353

Page
FOREWORD .............................................................................................................. v
TABLE OF CONTENTS .......................................................................................... ix
ABBREVIATIONS ................................................................................................. xiii
LIST OF TABLES ................................................................................................. xxii
LIST OF FIGURES ............................................................................................... xxv
SUMMARY .................................................................................................. .......xxxiii
ÖZET ......... .......................................................................................................... .xxxv
1. INTRODUCTION .................................................................................................. 1
1.1 Objectives of the Thesis ..................................................................................... 2
1.2 Outline of the Thesis .......................................................................................... 4
1.3 Brief History of Bricks, Mortar and Masonry Walls ......................................... 6
1.3.1 Bricks .......................................................................................................... 6
1.3.2 Mortar .......................................................................................................... 7
1.3.3 Masonry Walls ............................................................................................ 7
1.4 An Overview of Literature ................................................................................. 8
1.4.1 The characteristics of historical masonry .................................................... 8
1.4.2 The predictions of masonry compressive strength .................................... 15
1.4.3 The relationships proposed for masonry stress-strain ............................... 20
1.4.4 The relationships proposed for modulus of elasticity-compressive strength
of masonry .................................................................................................. 21
1.4.5 The shear strength components ................................................................. 23
2. GENERAL OUTLINE OF THE AKARETLER ROW HOUSES AND
DESCRIPTION OF ORIGINAL TEST MATERIALS ....................................... 27
2.1 General Outline of the Akaretler Row Houses ................................................ 27
2.2 Description of Original Test Materials ............................................................ 31
3. DETERMINATION OF MATERIAL CHARACTERISTICS OF MASONRY
CONSTITUENTS OF BRICKS AND MORTAR ................................................. 35
3.1 Mechanical Tests on Bricks ............................................................................. 36
3.1.1 Flexural tests on bricks.............................................................................. 36
3.1.1.1 Specimen preparation ......................................................................... 36
3.1.1.2 Test procedure .................................................................................... 37
3.1.1.3 Test results ......................................................................................... 38
3.1.2 Compression tests on bricks ...................................................................... 39
3.1.2.1 Specimen preparation ......................................................................... 39
3.1.2.2 Test procedure .................................................................................... 40
3.1.2.3 Test results ......................................................................................... 42
3.1.3 Compression tests on brick specimens in parallel to bed joint ................. 51
3.1.4 Compression tests on three-brick specimens ............................................ 53
3.1.4.1 Specimen preparation ......................................................................... 53
3.1.4.2 Test procedure .................................................................................... 54
3.1.4.3 Test results ......................................................................................... 55
x
3.1.5 Rebound hammer tests on the bricks ......................................................... 61
3.2 Mechanical Tests on Mortar ............................................................................. 63
3.2.1 Flexural tests on mortar ............................................................................. 63
3.2.1.1 Specimen preparation ......................................................................... 63
3.2.1.2 Test procedure .................................................................................... 64
3.2.1.3 Test results .......................................................................................... 65
3.2.2 Compression tests on mortar ..................................................................... 66
3.2.2.1 Specimen preparation ......................................................................... 66
3.2.2.2 Test procedure .................................................................................... 67
3.2.2.3 Test results .......................................................................................... 68
3.2.3 Rebound hammer tests on mortar joints .................................................... 73
3.3 Physical Tests on Bricks .................................................................................. 74
3.4 Chemical Tests on Brick and Mortar ............................................................... 77
3.5 Evaluation of the Tests ..................................................................................... 80
4. EXPERIMENTAL STUDIES ON ORIGINAL MASONRY - CORE,
WALLET AND IN-SITU WALL TESTS .............................................................. 87
4.1 Mechanical Tests on Cores .............................................................................. 88
4.1.1 Splitting tests on cores .............................................................................. 88
4.1.1.1 Specimen preparation ......................................................................... 88
4.1.1.2 Test procedure .................................................................................... 90
4.1.1.3 Test results .......................................................................................... 90
4.1.2 Compression tests on cores ....................................................................... 92
4.1.2.1 Specimen preparation ......................................................................... 92
4.1.2.2 Test procedure .................................................................................... 93
4.1.2.3 Test results .......................................................................................... 95
4.1.3 Shear tests on cores ................................................................................. 104
4.1.3.1 Specimen preparation ....................................................................... 104
4.1.3.2 Test procedure .................................................................................. 106
4.1.3.3 Test results ........................................................................................ 107
4.2 Compression Tests on Wallets ....................................................................... 115
4.2.1 Compression tests on wallets under monotonic loads ............................. 115
4.2.1.1 Specimen preparation ....................................................................... 115
4.2.1.2 Test procedure .................................................................................. 116
4.2.1.3 Test results ........................................................................................ 118
4.2.2 Compression tests on wallets under cyclic loads .................................... 122
4.2.2.1 Specimen preparation ....................................................................... 122
4.2.2.2 Test procedure .................................................................................. 122
4.2.2.3 Test results ........................................................................................ 122
4.3 In-situ Shear Tests on Walls ........................................................................... 133
4.3.1 In-situ shear tests on walls under monotonic loads ................................. 133
4.3.1.1 Specimen preparation ....................................................................... 133
4.3.1.2 Test results ........................................................................................ 135
4.3.2 In-situ shear tests on walls under cyclic loads ........................................ 149
4.3.2.1 Specimen preparation ....................................................................... 149
4.3.2.2 Test procedure .................................................................................. 149
4.3.2.3 Test results ........................................................................................ 150
4.4 Evaluation of Compression and Non-destructive Tests ................................. 153
4.5 Evaluation of Shear Tests ............................................................................... 155
5. TESTS ON PRISMS AND WALLS BUILT WITH HISTORICAL BRICKS
AND REPRODUCED MORTAR ......................................................................... 161
xi
5.1 The Reproduced Mortar ................................................................................. 162
5.2 Mechanical Tests on Reproduced Mortar ...................................................... 163
5.2.1 Flexural tests on reproduced mortar ........................................................ 163
5.2.1.1 Specimen preparation ....................................................................... 163
5.2.1.2 Test procedure .................................................................................. 164
5.2.1.3 Test results ....................................................................................... 164
5.2.2 Compression tests on reproduced mortar ................................................ 166
5.2.2.1 Specimen preparation ....................................................................... 166
5.2.2.2 Test procedure .................................................................................. 166
5.2.2.3 Test results ....................................................................................... 167
5.3 Evaluation of Reproduced Mortar Tests ........................................................ 172
5.4 Comparison of Original Mortar and Reproduced Mortar .............................. 175
5.5 Mechanical Tests on Masonry Prisms ........................................................... 176
5.5.1 Compression tests on masonry prisms under monotonic loads .............. 176
5.5.1.1 Specimen preparation ....................................................................... 176
5.5.1.2 Test procedure .................................................................................. 179
5.5.1.3 Test results ....................................................................................... 180
5.5.2 Compression tests on masonry prisms under cyclic loads ...................... 185
5.5.2.1 Specimen preparation ....................................................................... 185
5.5.2.2 Test procedure .................................................................................. 186
5.5.2.3 Test results ....................................................................................... 186
5.5.3 Shear tests on prisms (triplets) ................................................................ 196
5.5.3.1 Specimen preparation ....................................................................... 196
5.5.3.2 Test procedure .................................................................................. 198
5.5.3.3 Test results ....................................................................................... 200
5.6 Mechanical Tests on Walls ............................................................................ 208
5.6.1 Compression tests on walls ..................................................................... 208
5.6.1.1 Specimen preparation ....................................................................... 208
5.6.1.2 Test procedure .................................................................................. 211
5.6.1.3 Test results ....................................................................................... 213
5.6.2 Diagonal tension tests on walls ............................................................... 215
5.6.2.1 Specimen preparation ....................................................................... 215
5.6.2.2 Test procedure .................................................................................. 215
5.6.2.3 Test results ....................................................................................... 217
5.6.3 Shear tests on walls ................................................................................. 221
5.6.3.1 Specimen preparation ....................................................................... 221
5.6.3.2 Test procedure .................................................................................. 222
5.6.3.3 Test results ....................................................................................... 223
5.7 Evaluation of Compression Tests................................................................... 229
5.7.1 Relation between masonry prism and wall, brick unit and mortar ......... 229
5.7.2 Relation between Young's modulus and compressive strength of masonry
230
5.8 Evaluation of Shear Tests............................................................................... 232
6. OVERALL EVALUATION OF TEST RESULTS ......................................... 233
6.1 The Test Results of Masonry Components .................................................... 233
6.2 The Compression Test Results of the Masonry ............................................. 234
6.2.1 Comparison of core, wallet, prism, and wall test results ........................ 234
6.3 The Shear Test Results of the Masonry ......................................................... 239
6.4 Interaction Curves .......................................................................................... 242
6.5 Evaluation of the Test Results Considering Turkish Seismic Design Code .. 245
xii
6.5.1 Compression............................................................................................ 245
6.5.2 Shear........................................................................................................ 247
7. NUMERICAL ANALYSES .............................................................................. 251
7.1 The Material Models ...................................................................................... 251
7.1.1 The classical metal plasticity model........................................................ 252
7.2 Masonry Modeling ......................................................................................... 253
7.3 The Analyses of the Prisms under Compression............................................ 254
7.3.1 The establishment of the model .............................................................. 254
7.3.2 The results of the numerical analyses of the prisms................................ 256
7.4 The Analyses of the Masonry Walls under Compression .............................. 260
7.4.1 The establishment of the model .............................................................. 260
7.4.2 The results of the numerical analyses of the walls.................................. 262
7.5 The Numerical Analyses Results of the Masonry.......................................... 264
8. CONCLUSIONS................................................................................................. 269
REFERENCES....................................................................................................... 275
APPENDICES ........................................................................................................ 283
CURRICULUM VITA........................................................................................... 293
xiii
ABBREVIATIONS
a : A constant of compressive strain-compressive stress function
Amb : The area of a mortar bed joint
Amb,ul : The total initial area of upper and lower bed joints for in-situ wall
Ao : Initial cross section area of any specimen
ASTM : American Society for Testing Materials
Aw : The percentage of water absorption
Awdt : Area of wall specimen of diagonal tension test
b : A constant of compressive strain-compressive stress function
bb : Width of brick
bb,p : Width of brick tested in parallel to bed joint
BC : Brick tested under compression loads
BCh : Brick samples of chemical tests
BCp : Brick tested in parallel to bed joint under compression loads
BFT : Brick tested under flexural tension effects
bm : Width of mortar
BP : Brick specimen of physical tests
BR : Brick specimen of rebound hammer tests
brm : Width of reproduced mortar
bt : Width of triplet
btb : Width of three-brick
bw : Width of wall
bws : Width of in-situ wall for in-situ shear test
bwt : Width of wallet
C : Constant of Young's modulus-compressive strength function
CC : Core specimen of compression test
CoV : Coefficient of variation
CS : Core specimen of shear test
CST : Core specimen of splitting test
D : Constant of compressive stress at proportional limit-compressive
strength function
Dcx : Diameter of core in x direction
Dcy : Diameter of core in y direction
e : Constant (depending on the quality of workmanship) of Eq. (1.6)
E : Young’s modulus
Ebc : Young’s modulus of brick
Ebc,s : Secant modulus of brick at peak
Ecc,p : Young’s modulus of core
Ecc,s : Secant modulus of core
Emasc : Young’s modulus of masonry
Emc : Young’s modulus of mortar
Epc : Young’ modulus of prism
Epc,s : Secant modulus of prism
xiv
Epcall : Young’s modulus of prism when monotonic and cyclic tests are
evaluated together
Epcall,s : Secant modulus of prism when monotonic and cyclic tests are
evaluated together
Epcf : Young’s modulus of prism obtained from numerical analysis
Ermc : Young’s modulus of reproduced mortar
Ermc,a : Average Young’s modulus of reproduced mortar
Etbc : Young’s modulus of three-brick
Etbc,s : Secant modulus of three-brick
Euc : Young’s modulus of masonry unit
Ewc : Young’s modulus of wall
Ewcf : Young’s modulus of wall obtained from numerical analysis
Ewtc : Young’s modulus of wallet
f : Compressive strength
F : Constant of compressive stress at the visible crack-compressive
strength function
fbc : Compressive strength of brick
fbft : Flexural tensile strength of brick
fc : Compressive strength
fcc : Compressive strength of core
fcc,c : Characteristic compressive strength of concrete
fcft,c : Characteristic flexural tensile strength of concrete
FEMA 356 : Federal Emergency Management Agency
fft : Flexural tensile strength
fmasc : Compressive strength of masonry
f'masc : Specified compressive strength of masonry
fmasc,c : Characteristic compressive strength of masonry
fmc : Compressive strength of mortar
fmft : Flexural tensile strength of mortar
fpc : Compressive strength of prism
fpcall : Compressive strength of prism when monotonic and cyclic tests are
evaluated together
fpcf : Compressive strength of prism obtained from numerical analyses
frmc : Compressive strength of reproduced mortar
frmc,a : Average compressive strength of reproduced mortar
frmf : Flexural tensile strength of reproduced mortar
frmf,a : Average flexural tensile strength of reproduced mortar
ftbc : Compressive strength of three-brick
fuc : Compressive strength of unit
fuc,n : Normalized compressive strength of unit
fwc : Compressive strength of wall
fwcf : Compressive strength of wall obtained from numerical analyses
fwtc : Compressive strength of wallet
g : Gauge length
Gwdt : Shear modulus of wall obtained from diagonal tension test
hb : Height or thickness of brick
hb,p : Height of brick tested in parallel to bed joint
hc : Height of core
hm : Height of mortar
hrm : Height of reproduced mortar
xv
ht : Height of triplet
htb : Height or thickness of three-brick
hw : Height of wall
hws : Height of in-situ wall for in-situ shear test
hwt : Height of wallet
ID : No reliable or insufficient data
K : Constant coefficient of Eqs. (1.3) and (1.13)
k : Modulus of sub-grade reaction of soil
lb : Length of brick specimen
Lb : The distance between the centers of the supports for flexural tension
test of brick
lb,p : Length of brick specimen tested in parallel to bed joint
lc : Length of core specimen
lm : Length of mortar specimen
Lm : The distance between the centers of the supports for flexural tension
test of mortar specimen
lo : Initial gauge length
lrm : Length of reproduced mortar specimen
Lrm : The distance between the centers of the supports for flexural tension
test of reproduced mortar specimen
lt : Length of triplet specimen
ltb : Length of three-brick specimen
LVDT : Linear variable displacement transducer
lw : Length of wall specimen
lws : Length of in-situ wall specimen for in-situ shear test
lwt : Length of wallet specimen
MC : Mortar specimen tested under compression loads
MCF : Mortar specimen, which is obtained from the flexural test, tested
under compression loads
MCh : Mortar samples of chemical tests
n : Number of specimens tested
NA : Not available data
Nbr : Rebound number of brick
nd : Not detected
Nmr : Rebound number of mortar joints
p : Equivalent pressure stress
PC : Prism specimen tested under monotonic compression loads
Pc : The load applied to any specimen during compression test
PCC : Prism specimen tested under cyclic compression loads
PCF : Prism analyzed for compression loads
Pcst,m : Maximum load recorded during splitting test of core specimen
Pft : The load applied to any specimen during flexural tension test
Pft,m : The maximum load resisted by any specimen during flexural tension
test
Po : Porosity of any specimen
Pts : Shear load recorded during shear test of triplet specimen
Pwdt : Load recorded during diagonal tension test of wall specimen
Pws : Shear load recorded during in-situ shear test of wall specimen
R : Correlation coefficient
R2 : Coefficient of determination
xvi
Stdev : Standard deviation
Strg : Straingage
TBC : Three-brick tested under compression loads
Tmax : Maximum shear load recorded
tmb : Thickness of mortar bed joint
tmh : Thickness of mortar head joint
TS : Triplet specimen tested under shear loads
TSDC : Turkish Seismic Design Code
ucs,f : Relative horizontal displacement at shear strength level of core
specimen
ucs,n : Normalized relative horizontal displacement of core specimen
ucs,p : Relative horizontal displacement at proportional shear limit of core
specimen
uws,cr : Horizontal displacement at the first crack level of in-situ wall
specimen
uws,f : Horizontal displacement at shear strength level of in-situ wall
specimen
uws,p : Horizontal displacement at proportional limit of in-situ wall
specimen
WC : Wall tested under compression loads
WCF : Wall analysed under monotonic compression loads
Wd : Dry weight of brick
WDT : Wall specimen of diagonal tension test
Ws : Saturated weight of brick
WS : Wall specimen of in-situ monotonic shear test
WSC : Wall specimen of in-situ cyclic shear test
WSh : Wall tested under shear loads
WtC : Wallet specimen of monotonic compression test
WtCC : Wallet specimen of cyclic compression test
X : Predictor variable
Y : Criterion variable
εcc,cr : Compressive strain at the first crack level of core specimen
εcc,f : Compressive strain at compressive strength level of core specimen
εcc,n : Normalized compressive strain of core specimen under compression
loads
εcc,p : Compressive strain at proportional limit of core specimen
εen,n : Normalized compressive envelope strain
εf : Compressive strain at compressive strength
εmasc : Compressive strain of masonry
εn : Normalized compressive strain
εp : Compressive strain at proportional limit
εp,n : Normalized compressive plastic strain
εpc,f : Compressive strain at compressive strength level of prism specimen
εpc,n : Normalized compressive strain of prism specimen
εpc,p : Compressive strain at proportional limit level of prism specimen
εpcall,cr : Compressive strain at the first visible crack of prism specimen when
monotonic and cyclic tests are evaluated together
εpcall,f : Compressive strain at compressive strength level of prism specimen
when monotonic and cyclic tests are evaluated together
xvii
εpcall,n : Normalized compressive strain of prism specimen when monotonic
and cyclic tests are evaluated together
εpcall,p : Compressive strain at proportional limit of prism specimen when
monotonic and cyclic tests are evaluated together
εpcf,f : Compressive strain at compressive strength level of prism obtained
from numerical analysis
εpcf,p : Compressive strain at proportional limit of prism obtained from
numerical analysis
εrmc,n : Normalized compressive strain of reproduced mortar specimen
εv,bc : Compressive strain of brick under compression load
εv,bc,0.85f : Compressive strain at the 85 percent of compressive strength in the
descending branch for brick
εv,bc,f : Compressive strain at compressive strength level of brick
εv,bc,p : Compressive strain at proportional limit of brick
εv,mc,f : Compressive strain at compressive strength level of mortar
εv,mc,n : Normalized compressive strain of mortar
εv,mc,p : Compressive strain at proportional limit of mortar
εv,tbc,f : Compressive strain at compressive strength level of three-brick
under compression load
εv,tbc,n : Normalized compressive strain of three-brick under
compression load
εv,tbc,p : Compressive strain at proportional limit of three-brick under
compression load
εv,wtc,cr : Compressive strain at the first visible crack of wallet under
compression load
εv,wtc,f : Compressive strain of wallet under compression load
εv,wtc,p : Compressive strain at proportional limit of wallet under compression
load
εwc,f : Compressive strain at compressive strength level of wall
εwc,p : Compressive strain at proportional limit level of wall
εwcf,f : Compressive strain at compressive strength level of wall obtained
from numerical analysis
εwcf,p : Compressive strain at proportional limit of wall obtained from
numerical analysis
μ : Ductility
μbc : Ductility of brick under compression loads
μcc : Ductility of core under compression loads
μcs : Friction coefficient of core
μf : Friction coefficient
μmc : Ductility of mortar under compression loads
μpc : Ductility of prism under compression loads
μpcall : Ductility of prism when monotonic and cyclic tests are evaluated
together
μpcf : Ductility of prism obtained from numerical analysis
μrmc : Ductility of reproduced mortar under compression loads
μrmc,a : Average ductility of reproduced mortar under compression loads
μtbc : Ductility of three-brick under compression loads
μwc : Ductility of wall under compression loads
μwcf : Ductility of wall obtained from numerical analysis
μwdt : Ductility of wall obtained from diagonal tension test
xviii
μws : Friction coefficient of in-situ wall
μwtc : Ductility of wallet under compression loads
ρA,b : Bulk density of brick
ρR,b : Real density of brick
σ : Compressive (vertical) stress applied
σallowable,c : Allowable stress of the cores according to TSDC (2007)
σallowable,p : Allowable stress of the prisms according to TSDC (2007)
σallowable,w : Allowable stress of the walls according to TSDC (2007)
σallowable,wt : Allowable stress of the wallets according to TSDC (2007)
σbc : Compressive stress of brick
σbc,p : Compressive stress at proportional limit of brick
σcc,cr : Compressive stress at the first crack level of core
σcc,n : Normalized compressive stress of core
σcc,p : Compressive stress at proportional limit level of core
σcr : Compressive stress at the first crack
σmasc : Compressive stress of masonry
σmc,n : Normalized compressive stress of mortar
σmc,p : Compressive stress at proportional limit of mortar
σp : Compressive stress at proportional limit
σpc,n : Normalized compressive stress of prism
σpc,p : Compressive stress at proportional limit level of prism
σpcall,cr : Compressive stress at the first visible crack of prism when
monotonic and cyclic tests are evaluated together
σpcall,n : Normalized compressive stress of prism when monotonic and cyclic
tests are evaluated together
σpcall,p : Compressive stress at proportional limit of prism specimen when
monotonic and cyclic tests are evaluated together
σptf,p : Compressive stress at proportional limit of prism obtained from
numerical analysis
σrmc,n : Normalized compressive stress of reproduced mortar
σsa : Allowable bearing capacity of soil
σtbc,n : Normalized compressive stress of three-brick
σtbc,p : Compressive stress at proportional limit of three-brick
σwc,p : Compressive stress at proportional limit of wall
σwtc,cr : Compressive stress at the first visible crack of wallet
σwtc,p : Compressive stress at proportional limit of wallet
σwtf,p : Compressive stress at proportional limit of wall obtained from
numerical analysis
υmc : Poisson’s ratio of mortar
υuc : Poisson’s ratio of masonry unit or brick
υwtc : Poisson’s ratio of wallet specimen
n : Normalized compressive stress
 : Constant (shape factor) of Eq. (1.13) and conversion factors for of
brick compressive strength
 : Constant of Eq. (1.3)
 : Constant of Eq. (1.3)
 : Shear strength
cs,f : Shear strength of core
cs,n : Normalized shear stress of core
cs,o : Shear strength at zero nominal compression stress of core
xix
cs,p : Shear stress at proportional limit of core specimen
Hwdt : Extension in vertical direction of wall of diagonal tension test
o : Shear strength at zero nominal compression stress (shear bond
strength)
ts : Shear stress of triplet under shear loads
ts,f : Shear strength of triplet under shear loads
Vwdt : Shortening in vertical direction of wall of diagonal tension test
wdt : Shear strain of wall of diagonal tension test
wdt : Shear stress of wall of diagonal tension test
wdt,f : Shear strain at shear stress level of wall of diagonal tension test
wdt,f : Shear strength of wall of diagonal tension test
wdt,p : Shear strain at proportional level of wall of diagonal tension test
ws : Shear stress of in-situ wall
ws,cr : Shear stress at the first crack level of in-situ wall
ws,f : Shear strength of in-situ wall
ws,o : Shear strength at zero nominal compression stress of in-situ wall
ws,o : Shear strength at zero nominal vertical stress of in-situ wall
ws,p : Shear stress at proportional limit of in-situ wall
wsh,f : Shear strength of wall
xx
xxi
LIST OF TABLES
Page
Table 1.1 : Proposals for values in Eq. (1.17). .......................................................... 22
Table 1.2 : The compressive stress range for the calculation of modulus of elasticity.
................................................................................................................. 22
Table 1.3 : The values of shear strength components given in several codes. .......... 25
Table 1.4 : The values of shear strength components collected from the literature. . 26
Table 2.1 : The geometrical characteristics of the blocks. ........................................ 30
Table 2.2 : Soil properties of the historical houses. .................................................. 31
Table 2.3 : The range of sizes of the original bricks. ................................................ 33
Table 3.1 : The brick sizes for the flexural tension tests. .......................................... 37
Table 3.2 : The flexural tensile strengths of the bricks. ............................................ 38
Table 3.3 : The statistical parameters of the brick flexural tension test results. ....... 38
Table 3.4 : The brick sizes for the compression tests. .............................................. 40
Table 3.5 : The allowable values according to Chauvenet Criterion. ....................... 44
Table 3.6 : The results of the brick compression tests. ............................................. 46
Table 3.7 : The statistical parameters of the brick compression tests. ...................... 47
Table 3.8 : The comparison of Young’s moduli for the bricks (LVDTs and strain
gages). ..................................................................................................... 49
Table 3.9 : The brick specimen sizes for the compression tests. .............................. 52
Table 3.10 : The compressive strengths of the bricks tested in parallel to the bed
joint and anisotropy ratios. .................................................................... 52
Table 3.11 : The statistical parameters of the compression tests on the bricks tested
in parallel to the bed joint. .................................................................... 52
Table 3.12 : The three-brick specimen sizes for the compression tests. ................... 54
Table 3.13 : The results of the three-brick compression tests. .................................. 57
Table 3.14 : The statistical parameters of the three-brick compression tests. ........... 57
Table 3.15 : The rebound numbers of the narrow side of the bricks......................... 62
Table 3.16 : The rebound numbers of the wide sides of the bricks........................... 62
Table 3.17 : The statistical parameters of the brick rebound number. ...................... 63
Table 3.18 : The mortar specimen sizes for the flexural tension tests. ..................... 64
Table 3.19 : The results of the mortar flexural tension tests. .................................... 65
Table 3.20 : The statistical parameters of the mortar flexural tension tests. ............. 65
Table 3.21 : The mortar specimen sizes of the first group. ....................................... 67
Table 3.22 : The mortar specimen sizes of the second group. .................................. 67
Table 3.23 : The results of the mortar compression tests (the first group). .............. 70
Table 3.24 : The results of the mortar compression tests (the second group). .......... 70
Table 3.25 : The statistical parameters of the mortar compression tests (all
specimens). ........................................................................................... 70
Table 3.26 : The rebound numbers of the mortar joints. ........................................... 74
Table 3.27 : The dry and saturated weights of the bricks. ........................................ 75
Table 3.28 : The bulk and real densities of the bricks. .............................................. 76
Table 3.29 : The water absorption and real porosity values of the bricks................. 76
xxii
Table 3.30 : The oxide components of the brick samples. ........................................ 77
Table 3.31 : The oxide components of the mortar samples. ...................................... 78
Table 3.32 : The loss of ignition and organic matter of the brick samples. .............. 79
Table 3.33 : The loss of ignition and organic matter of the mortar samples. ............ 79
Table 3.34 : The mineralogical of the brick samples. ............................................... 79
Table 3.35 : The mineralogical of the mortar samples. ............................................. 79
Table 4.1 : The core sizes for the splitting tests. ....................................................... 90
Table 4.2 : The results of the core splitting tests. ...................................................... 91
Table 4.3 : The core sizes for compression tests. ...................................................... 94
Table 4.4 : The results of the core compression tests. ............................................... 96
Table 4.5 : The statistical parameters of the core compression tests. ....................... 97
Table 4.6 : The stress and strain values at the first cracks for several cores. .......... 102
Table 4.7 : The statistical parameters of the compressive stress and strain at the first
cracks. ................................................................................................... 102
Table 4.8 : The core specimen sizes of the shear tests. ........................................... 105
Table 4.9 : The results of the core shear tests. ........................................................ 109
Table 4.10 : The statistical parameters of the core shear tests. ............................... 110
Table 4.11 : The wallet sizes for the monotonic compression test. ......................... 116
Table 4.12 : The results of the wallet monotonic compression tests. ...................... 120
Table 4.13 : The statistical parameters of the wallet monotonic compression tests.
............................................................................................................. 120
Table 4.14 : The wallet sizes for the cyclic compression tests................................ 122
Table 4.15 : The results of the wallet compression tests (cyclic). ........................... 127
Table 4.16 : The statistical parameters of the wallet compression tests (cyclic). ... 127
Table 4.17 : The stress and strain values at the first cracks of the several wallets
under the compression loadings. ......................................................... 130
Table 4.18 : The statistical parameters concerning the first cracks of the several
wallets under the compression loadings. ............................................ 130
Table 4.19 : The statistical parameters of the wallet compression tests. ................. 131
Table 4.20 : The wall specimen sizes for the in-situ shear tests. ............................ 135
Table 4.21 : The results of the in-situ monotonic shear tests on the walls. ............. 138
Table 4.22 : The statistical parameters of the in-situ monotonic shear tests on the
walls. ................................................................................................... 139
Table 4.23 : The values of the shear stress and horizontal displacement at the first
cracks of the in-situ monotonic shear tests on the walls. .................... 146
Table 4.24 : The statistical parameters regarding the first cracks of the in-situ
monotonic shear test walls. ................................................................. 146
Table 4.25 : The wall specimen sizes for the in-situ cyclic shear tests. .................. 149
Table 4.26 : The results of the in-situ cyclic shear tests on the walls. .................... 152
Table 4.27 : The statistical parameters of the in-situ cyclic shear tests on the walls.
............................................................................................................. 152
Table 4.28 : The average compressive strengths and shear strength components of
the other buildings. .............................................................................. 157
Table 5.1 : The reproduced mortar specimen sizes of the flexural tension tests. .... 164
Table 5.2 : The results of the reproduced mortar flexural tension tests. ................. 165
Table 5.3 : The statistical parameters of the reproduced mortar flexural tension tests.
............................................................................................................... 166
Table 5.4 : The reproduced mortar specimen sizes for the compression tests. ....... 167
Table 5.5 : The results of the reproduced mortar compression tests. ...................... 168
xxiii
Table 5.6 : The statistical parameters of the reproduced mortar compression tests.
.............................................................................................................. 168
Table 5.7 : Young’s modulus and ductility of the reproduced mortar specimens at
the age of 210 days. .............................................................................. 171
Table 5.8 : The increment rate of compressive strength for the concrete and
reproduced mortar. ................................................................................ 172
Table 5.9 : The flexural and compressive strengths of several mortars compiled from
the literature. ......................................................................................... 174
Table 5.10 : The prism sizes for the monotonic compression tests......................... 178
Table 5.11 : The results of the monotonic compression tests on the prisms. .......... 181
Table 5.12 : The statistical parameters for the monotonic compression tests on the
prisms. ................................................................................................. 181
Table 5.13 : The correction factors for the prisms (ASTM C 1314-03b). .............. 182
Table 5.14 : The prism sizes for the cyclic compression tests. ............................... 185
Table 5.15 : The results of the cyclic compression tests on the masonry prisms.... 188
Table 5.16 : The statistical parameters of the cyclic compression results on the
masonry prisms. .................................................................................. 188
Table 5.17 : The stress and strain values at the first cracks of several prisms. ....... 193
Table 5.18 : The statistical parameters of the first cracks for several prisms. ........ 193
Table 5.19 : The statistical parameters for all compression tests on the prisms. .... 193
Table 5.20 : The triplet sizes for the shear tests. ..................................................... 197
Table 5.21 : The results of the triplet shear tests. .................................................... 201
Table 5.22 : The statistical evaluation of the compressive stresses recorded during
the triplet shear tests. .......................................................................... 204
Table 5.23 : The wall sizes for the compression tests. ............................................ 211
Table 5.24 : The size requirements of TS EN 1052-1 (2000). ................................ 211
Table 5.25 : The results of the wall compression tests............................................ 214
Table 5.26 : The statistical parameters of the wall compression tests. ................... 214
Table 5.27 : The wall sizes for the diagonal tension tests. ...................................... 216
Table 5.28 : The results of the wall diagonal tension tests. .................................... 220
Table 5.29 : The statistical parameters of the wall diagonal tension tests. ............. 220
Table 5.30 : The wall sizes of the shear tests. ......................................................... 222
Table 5.31 : The results of the wall shear tests. ...................................................... 225
Table 5.32 : The statistical evaluation of compressive stress recorded during the wall
shear tests. ........................................................................................... 225
Table 5.33 : The failure mechanisms of the walls (shear tests). ............................. 227
Table 5.34 : The compressive strengths of masonry prism or wall, brick and mortar.
............................................................................................................ 230
Table 6.1 : The average results of the flexural tensile tests. ................................... 233
Table 6.2 : The average results of the compression tests. ....................................... 233
Table 6.3 : The average results of the masonry compression tests. ........................ 234
Table 6.4 : The correction factors. .......................................................................... 235
Table 6.5 : The constants of the relations of the compression tests. ....................... 238
Table 6.6 : The shear strength components obtained from the tests. ...................... 239
Table 6.7 : The prediction of the friction coefficient for the other buildings. ........ 240
Table 6.8 : The predictions of the shear strength components for the laboratory tests.
.............................................................................................................. 241
Table 6.9 : The allowable shear components of the shear tests. ............................. 248
Table 7.1 : The masonry components of the prisms for the numerical analyses. ... 256
Table 7.2 : The results of the numerical analyses for the masonry prisms. ............ 259
xxiv
Table 7.3 : The masonry components of the walls for the numerical analyses. ...... 261
Table 7.4 : The results of the numerical analyses for the masonry walls. ............... 263
Table 7.5 : The comparison of the numerical analyses for the prisms and walls. ... 265
Table 7.6 : The comparison of the compressive strengths (numerical analysis and
Eurocode 6 prediction). ......................................................................... 266
Table A.1 : The conversion factors for the brick compressive strength .................. 283
Table A.2 : The normalized compressive strengths of the bricks ........................... 284
Table A.3 : The normalized compressive strengths of the bricks in parallel to the bed
joint. ...................................................................................................... 284
xxv
LIST OF FIGURES
Page
Figure 2.1 : The aerial view of the historical row houses. ........................................ 28
Figure 2.2 : The façade of Block B. .......................................................................... 28
Figure 2.3 : The original vaulted slabs of Block B. .................................................. 28
Figure 2.4 : The masonry walls and reinforced concrete slabs of Block C. ............. 29
Figure 2.5 : The mortar joints of the historical walls of the houses. ......................... 29
Figure 2.6 : The plan of Block B (entrance level). ................................................... 30
Figure 2.7 : The longitudinal section of Block B. ..................................................... 30
Figure 2.8 : A view of fill layer of Block C. ............................................................. 31
Figure 2.9 : The bricks collected from the historical row houses. ............................ 31
Figure 2.10 : The examples of the bricks collected from the historical row houses. 32
Figure 2.11 : The definition of sizes of the bricks. ................................................... 33
Figure 2.12 : Examples of warpages of the bricks. ................................................... 33
Figure 2.13 : The composition and colors of the bricks............................................ 34
Figure 2.14 : The visual appearance of original mortar. ........................................... 34
Figure 3.1 : The test setup with measurement system for the brick flexural tests. ... 37
Figure 3.2 : The failures of BFT-1 and BFT-4 bricks under flexural effect. ............ 39
Figure 3.3 : The test setup with measurement system for the brick compression. ... 41
Figure 3.4 : The ductility definition. ......................................................................... 43
Figure 3.5 : The compressive stress-compressive strain relationships for the bricks.
............................................................................................................... 45
Figure 3.6 : The Young’s modulus-compressive strength relationship for the bricks.
............................................................................................................... 48
Figure 3.7 : The compressive stress at proportional limit-compressive strength
relationship for the bricks. ..................................................................... 48
Figure 3.8 : The Young's modulus-secant modulus at peak relationship for the
bricks. .................................................................................................... 49
Figure 3.9 : The failure of BC-2 under compression loads. ...................................... 49
Figure 3.10 : The comparison of the compressive stress-compressive strain
relationships for the bricks (LVDTs and strain gages). ...................... 50
Figure 3.11 : The description of the brick specimens tested in parallel to bed joint. 51
Figure 3.12 : The failure of the bricks compressed in direction parallel to the bed
joint...................................................................................................... 53
Figure 3.13 : The description of the three-brick specimens. ..................................... 54
Figure 3.14 : The test setup with measurement system for the three-brick
compression tests................................................................................. 55
Figure 3.15 : The compressive stress-compressive strain relationships for the threebrick
specimens. .................................................................................. 56
Figure 3.16 : The normalized compressive stress-normalized compressive strain
relationship for the three-brick specimens. ......................................... 58
Figure 3.17 : The Young’s modulus-compressive strength relationship for the threebrick
specimens. .................................................................................. 59
xxvi
Figure 3.18 : The compressive stress at proportional limit-compressive strength
relationship for the three-brick specimens. ......................................... 60
Figure 3.19 : The failure mechanism of TBC-14. ..................................................... 60
Figure 3.20 : The Young's modulus-secant modulus at peak relationship for the
three-brick specimens. ......................................................................... 61
Figure 3.21 : The test setup with measurement system for the mortar flexural tests.
............................................................................................................. 64
Figure 3.22 : The failure of the mortar specimens under the flexural effect. ........... 66
Figure 3.23 : The test setup with measurement system for the mortar compression
tests. ..................................................................................................... 68
Figure 3.24 : The compressive stress-compressive strain relationships for the first
group of the mortar. ............................................................................. 69
Figure 3.25 : The compressive stress-compressive strain relationships for the second
group of the mortar. ............................................................................. 69
Figure 3.26 : The normalized compressive stress-normalized compressive strain
relationship for the mortar specimens. ................................................ 71
Figure 3.27 : The Young’s modulus-compressive strength relationship for all mortar
specimens. ........................................................................................... 72
Figure 3.28 : The compressive stress at proportional limit-compressive strength
relationship for the mortar specimens. ................................................ 72
Figure 3.29 : The Young's modulus-secant modulus at peak relationship for the
mortar specimens. ................................................................................ 73
Figure 3.30 : The average oxide components of the brick and mortar samples. ....... 78
Figure 3.31 : The average results of mineralogical analysis of the brick and mortar
samples. ............................................................................................... 80
Figure 4.1 : The drilling of the cores from the structural masonry walls of the
Akaretler Historical Row Houses. ......................................................... 89
Figure 4.2 : The preparation of the cores. ................................................................. 89
Figure 4.3 : The schematic view of the cores. ........................................................... 90
Figure 4.4 : The test configuration of the core splitting tests. ................................... 91
Figure 4.5 : The view of CST-2 core during and after the splitting test. .................. 91
Figure 4.6 : The view of CST-3 core during the splitting test. ................................. 92
Figure 4.7 : The capping of the cores for the compression tests. .............................. 92
Figure 4.8 : The schematic view of the cores tested under compression. ................. 93
Figure 4.9 : The setup with measurement system for the core compression tests. ... 93
Figure 4.10 : The compressive stress-compressive strain relationships for the cores.
............................................................................................................. 95
Figure 4.11 : The normalized compressive stress-normalized compressive strain
relationship for the cores. .................................................................... 97
Figure 4.12 : The Young’s modulus-compressive strength relationship for the cores.
............................................................................................................. 98
Figure 4.13 : The compressive stress at proportional limit and compressive strength
relationship for the cores. .................................................................... 99
Figure 4.14 : The Young's modulus and secant modulus at peak relationship for the
cores. .................................................................................................. 100
Figure 4.15 : The damage of CC-25 core during and after the compression test. .. 101
Figure 4.16 : The compressive stress at the first crack and compressive strength
relationship for several cores. ............................................................ 102
Figure 4.17 : The stress-to-strain ratio at the first crack and at the proportional limit
relationship for several cores. ............................................................ 103
xxvii
Figure 4.18 : The compressive stress-compressive strain relation for cores with
characteristic points. .......................................................................... 104
Figure 4.19 : The schematic view of the cores of the shear tests. ........................... 105
Figure 4.20 : The test setup with measurement system for the core shear tests. .... 106
Figure 4.21 : The shear stress-relative horizontal displacement relationships for the
cores under the compressive stress of 0.05 MPa. .............................. 108
Figure 4.22 : The shear stress-relative horizontal displacement relationships for the
cores under the compressive stress of 0.15 MPa. .............................. 108
Figure 4.23 : The shear stress-relative horizontal displacement relationships for the
cores under the compressive stress of 0.30 MPa. .............................. 109
Figure 4.24 : The shear strength-vertical stress relationship for the cores. ............ 111
Figure 4.25 : The normalized shear stress normalized relative horizontal
displacement relationship (linear) for the cores. ............................... 112
Figure 4.26 : The normalized shear stress-normalized relative horizontal
displacement relationship (parabolic). .............................................. 112
Figure 4.27 : The shear stress-to-relative horizontal displacement ratios at strength
level and at proportional limit relationship for the cores. ................. 113
Figure 4.28 : The damage development of CS-0.15-4. ........................................... 114
Figure 4.29 : The failure of CS-0.15-4.................................................................... 114
Figure 4.30 : The view of the several wallets of the compression test. .................. 115
Figure 4.31 : The testing machine of the Instron Satec 1000RD. ........................... 117
Figure 4.32 : The measurement system for the wallet compression tests. .............. 117
Figure 4.33 : The compressive stress-vertical strain relationships under monotonic
compression for the wallets (through LVDTs). ................................ 119
Figure 4.34 : The compressive stress-horizontal strain relationships under monotonic
compression for the wallets (through LVDTs). ................................ 119
Figure 4.35 : The Poisson’s ratio-vertical strain relationships for several wallets
under monotonic compression. ......................................................... 120
Figure 4.36 : The compressive stress-compressive strain relationships under cyclic
loadings for several wallets (through LVDTs). ................................. 124
Figure 4.37 : The envelope curves of the compressive stress-vertical strain under
cyclic loadings relationships of several wallets (through LVDTs). .. 124
Figure 4.38 : The compressive stress-horizontal strain relationships under cyclic
loadings for several wallets (through LVDTs). ................................. 125
Figure 4.39 : The envelope curves of the compressive stress-horizontal strain
relationships for several wallets under cyclic loads (LVDTs). ......... 125
Figure 4.40 : The relationships of the Poisson’s ratio-compressive strain of several
wallets under cyclic loads. ................................................................ 126
Figure 4.41 : The normalized compressive stress-normalized compressive strain
relationship for the wallets. ............................................................... 128
Figure 4.42 : The Young’s modulus-compressive strength relationship for several
wallets................................................................................................ 129
Figure 4.43 : The compressive stress at proportional limit-compressive strength
relationship for several wallets. ......................................................... 129
Figure 4.44 : The compressive stress at the first crack-compressive strength
relationship for several wallets. ......................................................... 131
Figure 4.45 : The development of failure of WtCC-5 wallet specimen. ................. 132
Figure 4.46 : The characteristic points on the compressive stress-compressive strain
curve for the wallets. ......................................................................... 133
Figure 4.47 : The images of the in-situ shear test site............................................. 134
xxviii
Figure 4.48 : The test setup with measurement system for the in-situ shear tests on
the walls. ............................................................................................ 135
Figure 4.49 : The shear stress-horizontal displacement relationships for the in-situ
walls at entrance. ............................................................................... 136
Figure 4.50 : The shear stress- horizontal displacement relationships for the in-situ
walls at the first story. ....................................................................... 137
Figure 4.51 : The shear stress- horizontal displacement relationships for the in-situ
walls at the second story. ................................................................... 137
Figure 4.52 : The pre-peak branches of the shear stress-horizontal displacement
curves for the entrance walls. ............................................................ 139
Figure 4.53 : The pre-peak branches of the shear stress-horizontal displacement
curves for the first story walls. .......................................................... 140
Figure 4.54 : The pre-peak branches of the shear stress-horizontal displacement
curves for the second story walls. ..................................................... 140
Figure 4.55 : The shear strength-vertical stress relationship for the in-situ walls. . 142
Figure 4.56 : The shear stress-to-horizontal displacement ratios at strength level and
at proportional limit relationship for the in-situ walls. ...................... 143
Figure 4.57 : The damage state of specimen WS-E-4 during the in-situ shear test. 145
Figure 4.58 : The view of the WS-E-4 wall after the in-situ shear test. .................. 146
Figure 4.59 : The shear stress at first crack-shear strength relationship for several insitu
walls. ........................................................................................... 147
Figure 4.60 : The shear stress-to-horizontal displacement ratio at the first crack level
and at strength level relationship for several in-situ walls. ............... 148
Figure 4.61 : The shear stress-horizontal displacement relation of in-situ walls with
characteristic points. .......................................................................... 148
Figure 4.62 : The shear stress-horizontal displacement relationships for the in-situ
walls under cyclic loadings. .............................................................. 151
Figure 4.63 : The envelope curves of shear stress-horizontal displacement the
relationships for the in-situ walls. ..................................................... 151
Figure 4.64 : The pre-peak branches of the shear stress-horizontal displacement
curves for the in-situ walls under cyclic loadings. ............................ 152
Figure 4.65 : The relationship between compressive strength of core and rebound
number of brick. ................................................................................ 154
Figure 4.66 : The comparison of shear strength-vertical stress values for the core and
in-situ tests. ........................................................................................ 156
Figure 4.67 : The shear strength-vertical stress relationships for the in-situ walls and
cores. .................................................................................................. 157
Figure 4.68 : The shear bond strength-compressive strength relationship for the
cores. .................................................................................................. 158
Figure 4.69 : The friction coefficient-compressive strength relationship for the cores.
........................................................................................................... 159
Figure 5.1 : The ingredients of the reproduced mortar. .......................................... 162
Figure 5.2 : The production steps of the reproduced mortar. .................................. 163
Figure 5.3 : The development of flexural tensile strength by age for the reproduced
mortar. ................................................................................................. 165
Figure 5.4 : The view of RMF-210-3 mortar specimen after the flexural test. ....... 166
Figure 5.5 : The development of compressive strength with age for the reproduced
mortar. ................................................................................................. 169
Figure 5.6 : The compressive stress-compressive strain relationships for reproduced
mortar at the age of 210 days. ............................................................. 170
xxix
Figure 5.7 : The normalized compressive stress-normalized compressive strain
relationship for reproduced mortar at the age of 210 days. ................. 170
Figure 5.8 : Young’s modulus-compressive strength relationship for the reproduced
mortar specimens at the age of 210 days. ............................................ 171
Figure 5.9 : The appearances of RMCF-28-1-A mortar specimen during and after the
compression test. ................................................................................. 171
Figure 5.10 : Flexural tensile and compressive strength relationship for the
reproduced mortar. ............................................................................ 173
Figure 5.11 : The comparison of the flexural tensile and compressive strengths of
the original mortar and reproduced mortar. ...................................... 175
Figure 5.12 : The construction steps of the prisms. ................................................ 177
Figure 5.13 : The view of the prisms. ..................................................................... 178
Figure 5.14 : The test setup and measurement system for the prism compression
tests. ................................................................................................... 179
Figure 5.15 : The compressive stress-compressive strain relationships for the prisms
tested under monotonic loadings (obtained from the LVDTs). ........ 181
Figure 5.16 : The normalized compressive stress-normalized compressive strain
relationship for the prisms tested under monotonic loading. ............ 183
Figure 5.17 : The Young’s modulus-compressive strength relationship for the prisms
under monotonic loading. .................................................................. 183
Figure 5.18 : The compressive stress at proportional limit-compressive strength
relationship for the prisms tested under monotonic loading. ............ 184
Figure 5.19 : The Young's modulus-secant modulus at peak relationship for the
prisms tested under monotonic compression. ................................... 185
Figure 5.20 : The cyclic compressive stress-compressive strain relationships for the
prisms. ............................................................................................... 187
Figure 5.21 : The envelope curves of cyclic compressive stress-compressive strain
relationships for the prisms. .............................................................. 187
Figure 5.22 : The normalized compressive stress-normalized strain relationship for
the prisms tested under monotonic and cyclic loading. .................... 189
Figure 5.23 : The Young’s modulus-compressive strength relationship for the prisms
tested under monotonic and cyclic loading. ...................................... 190
Figure 5.24 : The compressive stress at proportional limit-compressive strength
relationship for the prisms tested under monotonic and cyclic loading.
........................................................................................................... 191
Figure 5.25 : The Young's modulus-secant modulus at peak relationship for the
prisms tested under monotonic and cyclic loading. .......................... 192
Figure 5.26 : The compressive stress at the first crack level- compressive strength
relationship for several prisms. ......................................................... 192
Figure 5.27 : The plastic strain-envelope strain relationship for the prisms. .......... 194
Figure 5.28 : The stiffness degradation in the prisms under cyclic compression load.
........................................................................................................... 195
Figure 5.29 : The compressive stress-compressive strain relation of the prisms with
special points. .................................................................................... 195
Figure 5.30 : The construction steps of the triplets. ................................................ 197
Figure 5.31 : The description of the triplets. ........................................................... 198
Figure 5.32 : The test setup with measurement system for the triplet shear tests. .. 199
Figure 5.33 : The measurement system for the triplet shear tests. .......................... 200
Figure 5.34 : The shear stress-shear displacement relationships for the triplets
(through A1 and B1 LVDTs). ........................................................... 201
xxx
Figure 5.35 : The variation of compressive stress during the shear tests of TS-0.13-1,
TS-0.13-2, TS-0.25-1 and TS-0.25-2. ............................................... 203
Figure 5.36 : The variation of compressive stress during the shear tests of TS-0.50-2,
TS-0.50-4, TS-0.75-1, TS-0.75-2, TS-1.00-1 and TS-1.00-2. ........... 204
Figure 5.37 : The shear stress-shear strain relationships for TS-0.13-1, TS-0.13-2,
TS-0.25-1, TS-0.25-2, TS-0.50-2 and TS-0.50-4. ............................. 206
Figure 5.38 : The shear stress-shear strain relationships for TS-0.75-1, TS-0.75-2,
TS-1.00-1 and TS-1.00-2 (through Ac-Bc and Ad-Bd LVDTs). ...... 207
Figure 5.39 : The failure of the triplets under the compression stresses of 0.13, 0.25,
and 0.50 MPa. .................................................................................... 207
Figure 5.40 : The failure of TS-1.00-1 triplet. ........................................................ 208
Figure 5.41 : The shear strength-compressive stress relation for the triplets. ......... 208
Figure 5.42 : The construction steps of the walls. ................................................... 210
Figure 5.43 : The wall specimens. .......................................................................... 211
Figure 5.44 : The test setup with measurement system for the wall compression
tests. ................................................................................................... 213
Figure 5.45 : The compressive stress-compressive strain relationships for the walls.
........................................................................................................... 214
Figure 5.46 : The wall specimens for the diagonal tension tests. ........................... 216
Figure 5.47 : The test setup with measurement system for the wall diagonal tension
tests. ................................................................................................... 217
Figure 5.48 : The shear stress-horizontal strain and shear stress-vertical strain
relationships for the wall diagonal tension tests. ............................... 218
Figure 5.49 : The comparison of shear stress-vertical strain relations for the wall
diagonal tension tests. ........................................................................ 219
Figure 5.50 : The shear stress-shear strain relationships for the wall diagonal tension
tests. ................................................................................................... 219
Figure 5.51 : The failures of the walls (diagonal tension tests). ............................. 221
Figure 5.52 : The wall views of the shear tests. ...................................................... 222
Figure 5.53 : The test setup with measurement system for the wall shear tests. .... 223
Figure 5.54 : The shear stress-horizontal displacement relationships for the walls.
........................................................................................................... 224
Figure 5.55 : The variation of compressive stress during the shear tests of WSh-
0.13-1, WSh-0.13-2, WSh-0.25-1, WSh-0.25-2 and WSh-0.25-3. ... 226
Figure 5.56 : The variation of compressive stress during the shear tests of WSh-
0.50-1, WSh-0.50-2, WSh-0.75-1 and WSh-0.75-2. ......................... 227
Figure 5.57 : The failure of WSh-0.13-2 wall. ........................................................ 228
Figure 5.58 : The failure of WSh-0.50-2 wall. ........................................................ 228
Figure 5.59 : The shear strength-vertical stress relation for the walls. ................... 229
Figure 6.1 : The friction coefficient-shear bond strength relationship (the shear tests
in the laboratory). ................................................................................ 240
Figure 6.2 : The friction coefficient-shear bond strength relationship (the shear tests
of the Akaretler Row Houses and the other buildings in the laboratory).
............................................................................................................. 241
Figure 6.3 : The interaction between shear and compressive stresses of the cores. 243
Figure 6.4 : The interaction between shear and compressive stresses of the triplets.
............................................................................................................. 243
Figure 6.5 : The interaction between shear and compressive stresses of the walls. 244
Figure 7.1 : Modeling of masonry (Lourenço, 1994). ............................................. 254
Figure 7.2 : The description of the prism. ............................................................... 255
xxxi
Figure 7.3 : The finite element model of the prism and the element type used. ..... 256
Figure 7.4 : The compressive stress contours and deformed shapes of PCF-1. ...... 257
Figure 7.5 : The compressive stress-compressive strain relationships obtained from
the numerical analysis. ........................................................................ 259
Figure 7.6 : The comparison of compressive stress-compressive strain relationships
obtained from the tests and numerical analyses (prisms). ................... 260
Figure 7.7 : The finite element model of the wall. .................................................. 261
Figure 7.8 : The mesh of the wall. .......................................................................... 261
Figure 7.9 : The compressive stress contours and deformed shapes of WCF-1. .... 262
Figure 7.10 : The compressive stress-compressive strain relationships for the walls
(numerical analysis). ......................................................................... 263
Figure 7.11 : The comparison of compressive stress-compressive strain relationships
obtained from the tests and numerical analyses (walls). ................... 264
Figure 7.12 : The compressive stress-compressive strain relationships of the prisms
and the walls (numerical analysis). ................................................... 265
Figure B.1 : The relationships of compressive stress-vertical strain. ..................... 285
Figure B.2 : The relationships of compressive stress-vertical strain and their
envelope curves under cyclic loads. ................................................... 286
Figure C.1 : The compressive stress contours and deformed shapes of PCF-2. ..... 287
Figure C.2 : The compressive stress contours and deformed shapes of PCF-3. ..... 288
Figure C.3 : The compressive stress contours and deformed shapes of PCF-4. ..... 288
Figure C.4 : The compressive stress contours and deformed shapes of WCF-2. ... 289
Figure C.5 : The compressive stress contours and deformed shapes of WCF-3. ... 289
Figure C.6 : The compressive stress contours and deformed shapes of WCF-4. ... 290
xxxii
xxxiii
A COMPREHENSIVE EXPERIMENTAL RESEARCH ON THE BEHAVIOR
OF HISTORICAL BRICK MASONRY WALLS OF 19TH CENTURY
BUILDINGS
SUMMARY
Turkey has several historical structures ranging from city walls, bridges, palaces,
churches, mosques, underground cisterns and residental buildings, which are the
remainings of Roma, Byzantine, and Ottoman periods. The wish to protect them for
the future requires evaluating the present situation of these structures. One of the
necessary steps for the structural evaluation is to determine the in-situ material
characteristics. The related literature investigation indicates that the number of the
study on the material characterization of the historical structures in Turkey is very
limited. Consequently, in this study, it is aimed to carry out a comprehensive
experimental study on historical masonry samples, which were obtained from
historical Akaretler Row Houses built in 19th century in İstanbul. The row houses are
the first examples of row houses in Ottoman Empire. The load-bearing walls of the
row houses were constructed with solid clay brick laid in mortar joints. Since several
walls of the houses were to be removed according to the restoration design, a large
number of different types of specimens could be collected for laboratory tests and
considerable amount of in-situ destructive and non-destructive tests were carried out
as well. It is considered that the historical masonry knowledge obtained from these
experimental studies may be used in structural assessment and restoration works of
the other historical masonry row houses/buildings constructed in 19th century in
Turkey. Particularly, Beyoğlu (Pera) district in İstanbul has many historical brick
masonry buildings and row houses, which were built during the same period.
This study can be divided into four main parts. In the first part, the material
characterizations of the historical masonry componets (brick and mortar) are carried
out with mechanical, physical, and chemical tests. These tests were performed on the
brick and mortar samples, which were collected from the walls of the houses.
According to the test results and visual observations, it is concluded that the bricks
were simply produced in field kilns, and the binder of the mortar was hydrated lime
without brick powder. Additionally, the surface hardnesses of the in-place bricks and
mortar joints are measured.
In the second part, the structural behavior of the historical masonry samples (core
and wallet) extracted from the walls is investigated under tension, compression and
shear loads. According to this study, the wallet compressive strength may be taken as
about 60% of the core compressive strength. By evaluating the results of destructive
(core tests) and non-destructive tests (rebound hammer tests), an equation is
suggested to find compressive strength of masonry core specimens from the in-situ
rebound number of bricks. By testing the materials extracted from several 19th
century historical buildings, the relationships of shear bond strength-compressive
strength and friction coefficient-compressive strength are obtained for the core
specimens.
xxxiv
In the third part, tests on reproduced masonry (prism and wall), which were
constructed with historical bricks collected from the walls of the houses and
reproduced mortar, are given. Since the process of taking test specimens from
existing structures is destructive, the application of this process on the historical
structures is not generally allowed. Even if allowed, to take non-damaged specimens
and to take appropriate specimens in terms of number, size required for the test type,
and composition required for the simulation of in-place load-bearing masonry may
not be possible. In such cases, masonry properties may be identified through tests
performed on the reproduced specimens. In this study, the reproduced prisms
specimens are tested under monotonic/cyclic compression, and shear loads. The
cyclic characteristics of the prisms are defined through the envelope and plastic
strain relationships and stiffness degradation. The tests conducted on the reproduced
walls are compression, diagonal tension and shear tests. Using compressive strength,
shear strengths and corresponding compressive stresses, the interaction curves of
shear and compressive stresses are obtained. These curves indicate that Mohr-
Coulomb criterion should be used for cases under which the levels of compressive
stress are lower than 30-40% of corresponding compressive strengths.
In the last part, numerical analyses are performed on masonry prisms and walls under
compression loads. The main objective is to obtain the compressive stresscompressive
strain relations for comparing with the related relations obtained from
tests. It should be noted that in order to define the masonry components (brick and
mortar) in the numerical analyses, the experimental data of these components were
used.
Additionally, in the related parts of this thesis, in order to express the compression
test results quantitatively, the compressive strength and corresponding strain, the
compressive stress at proportional limit and corresponding strain, the Young’s
modulus and the ductility of each specimen are obtained. The relationships of
Young’s modulus-compressive strength, compressive stress at proportional limitcompressive
strength and Young’s modulus-secant modulus at peak are expressed
with various linear functions. Modeling the relationships between compressive stress
and compressive strain, parabolic functions are proposed based on the regression
analysis conducted on the corresponding test data. For evaluating the shear tests
quantitatively, the shear strength components (bond strength and friction of
coefficient), shear modulus, and shear stress-horizontal displacement relationships
are inferred from the test results. In the analysis and/or assessment of the existing
masonry structures, these functions can be utilized.
The obtained results are also evaluated according to Turkish Seismic Design Code
(TSDC) (2007) in a comparative manner. This evaluation may be concluded that the
specimens to be tested should be clearly described and that the knowledge on the
masonry buildings built with lime mortar should be provided by the code. The
statements related to Young’s modulus of the masonry wall should be clearer. The
design procedure of masonry structures and default values given in the code are
generally referred as the assessment procedure of the existing masonry structures.
This procedure may be developed taking into account the common characteristics of
the existing masonry structures.
xxxv
19. YY TARĠHĠ TUĞLA YIĞMA DUVARLARIN DAVRANIġI ÜZERĠNE
KAPSAMLI DENEYSEL BĠR ÇALIġMA
ÖZET
Ülkemiz, Roma, Bizans ve Osmanlı dönemine ait sur, köprü, saray, kilise, cami,
sarnıç ve sıraevler gibi birçok tarihi yapıya sahiptir. Bu yapıların gelecekte var
olmalarını sağlamaya yönelik çalışmaların önemli adımlarından biri, malzeme
özelliklerinin belirlenmesidir. Bu çalışma kapsamında yapılan literatür araştırması,
ülkemizdeki tarihi yapıların malzeme özelliklerinin belirlenmesini konu alan
deneysel çalışma sayısının yetersiz olduğunu göstermektedir. Bu nedenle, tarihi bir
yığma yapıdan alınan numuneler üzerinde kapsamlı bir deneysel çalışma yapılması
amaçlanmıştır. Numuneler, 19. yy'da inşaatı gerçekleştirilen tarihi Akaretler
Sıraevler Grubu'nun taşıyıcı duvarlarından alınmıştır. Osmanlı döneminin ilk
sıraevler örneği olan bu sıraevler grubunun taşıyıcı duvarları, dolu tuğla ve harçla
oluşturulmuştur. Bazı taşıyıcı duvarları, restorasyon projesi gereği kaldırıldığı için,
laboratuar deneyleri için farklı karakteristiklere sahip bir çok malzeme numunesi
alınabilmiş ve çok sayıda hasarlı ve hasarsız yerinde deney yapılabilmiştir. Elde
edilen sonuçların 19. yy. tuğla yığma yapıları için kullanılabileceği düşünülmektedir.
Tez çalışması, sekiz bölümden oluşmaktadır. Bölüm 1'de, çalışmanın amaçları ve
izlenen yol genel hatlarıyla verilmiştir. Ayrıca, tez konusu ile ilgili yapılan literatür
araştırması özetlenek verilmiştir.
Akaretler Sıraevler Grubu'nun tarihçesi, yapısal ve mimari özellikleri ve zemin
koşulları Bölüm 2'de anlatılmıştır. Ayrıca, sıraev grubunun taşıyıcı duvarlarını
oluşturan tuğla ve harcın doku, renk ve boyut gibi bazı fiziksel özellikleri de
verilmiştir.
Bilindiği üzere, kompozit bir malzeme olarak teşkil edilen yığma duvarların yapısal
karakteristikleri; duvar örgü tipi ve işçilik gibi parametrelerin yanında kendisini
oluşturan tuğla ve harcın karakteristiklerine de bağlıdır. Bu yüzden, sıraevler
grubunun taşıyıcı duvarlarından alınan tarihi tuğla ve harç örnekleri üzerinde
mekanik, fiziksel ve kimyasal deneyler gerçekleştirilmiştir. Söz konusu deneyler,
Bölüm 3'de verilmiştir. Tuğla ve harç örneklerinin eğilmede çekme ve basınç
dayanımları deneysel olarak belirlenmiştir. Belirlenen dayanımlar, malzemenin
kalitesi hakkında fikir sahibi olmak amacıyla günümüz malzemesi için ilgili
standartlarda verilen dayanım değerleri ile karşılaştırılmıştır. Bu karşılaştırma;
beklendiği üzere, tarihi malzemenin, günümüz malzemesine göre daha düşük
dayanımlı olduğunu göstermektedir. Günümüz yığma standartlarına göre; yığma
basınç dayanımı, tuğla ve/veya harç basınç dayanımlarına bağlı olarak verilen
ampirik bağıntılar kullanılarak hesaplanabilir. Bu nedenle, bu bölümde, deneysel
olarak saptanan, tuğla ve harç ortalama basınç dayanımları kullanılarak, yığma
malzemesinin basınç dayanımı tahmin edilmeye çalışılmıştır. Aynı zamanda,
Schmidt çekici ile tuğla ve harcın yüzey sertlikleri, yerinde ölçülmüştür.
xxxvi
Tuğla ve harç örnekleri üzerinde gerçekleştirilen deneyler, yığma duvar davranışını
tanımlamak açısından yeterli olmayabilir. Bu nedenle, Bölüm 4'de tarihi sıraevler
grubunun mevcut taşıyıcı duvarlarından çıkarılan karot ve duvar parçacıkları
üzerinde gerçekleştirilen laboratuar deneyleri ve yerinde gerçekleştirilen duvar
kayma deneyleri verilmiştir. Karot numuneler, iki tuğla parçası ve bunları birleştiren
bir yatay derzden oluşmaktadır. Duvar parçacıkları ise üç tuğla sırası, iki yatay derz
ve birkaç düşey derzden oluşmaktadır. Karotlar üzerinde yarma, basınç ve kayma
deneyleri gerçekleştirilmiştir. Duvar parçacıkları ise monotonik ve tekrarlı basınç
yüklemeleri altında test edilmişlerdir. Karot kayma deneyleri, kayma dayanımı
bileşenlerini (yapışma dayanımı ve sürtünme katsayısı) elde etmek amacıyla, üç
farklı eksenel basınç yük seviyesi (0.05, 0.15 ve 0.30 MPa) için yapılmıştır. Bu
bölümde verilen deneyler neticesinde, basınç ve kayma dayanımları, kırılma
biçimleri, yapışma dayanımı, sürtünme katsayısı ve basınç gerilmesi-düşey
şekildeğiştirme, Young modulü-basınç dayanımı, kayma gerilmesi-yatay
yerdeğiştirme gibi ilişkiler elde edilmiştir. Ayrıca, tarihi Akaretler Sıraevler Grubu
ile aynı dönemde yani 19. yy'da inşa edilen birkaç tarihi yapıdan alınan karot
numuneleri üzerinde de basınç ve kayma deneyleri yapılmıştır. Bu yapılara ait
ortalama basınç dayanımlarının birbirlerine yakın değerler aldıkları tespit edilmiştir.
Bu yapılara ait karot basınç ve kayma deney sonuçları, Akaretler Sıraevler Grubu
deney sonuçları ile beraber değerlendirilerek yapışma dayanımı-basınç dayanımı ve
sürtünme katsayısı-basınç dayanımı ilişkileri elde edilmiştir. Bu ilişkilere ek olarak,
karot basınç dayanımı ile tuğla yüzey sertliği değişkenleri arasında da bir ilişki
kurulabilmiştir.
Tarihi yapılardan yığma duvar numunesi almak veya yeterli büyüklükte ve sayıda
numune almak mümkün olmayabilir. Bu durumlarda, tarihi yığma malzemesinin
mekanik özelliklerini saptamak için deneyler, üretilmiş numuneler üzerinde
yapılabilir. Söz konusu numuneleri, incelenen yapıdan alınan tuğla ve orijinal harcın
mekanik özelliklerini karşılayan bir harçla üretmek gerekmektedir. Bu nedenle,
Bölüm 5, tarihi Akaretler Sıraevler Grubu'nun taşıyıcı duvarlarından toplanan tuğla
ve orijinal harcın mekanik özelliklerini karşılayan harçla üretilen prizma ve duvar
numuneler üzerinde gerçekleştirilen deneylere ayrılmıştır. Prizma numunelerin
imalatı, üç sıra tuğla ve iki yatay derz içerecek şekilde gerçekleştirilirken; duvar
numunelerin imalatı, beş sıra tuğla ve dört yatay derz ve bir sıra tuğla dört düşey derz
içerecek şekilde gerçekleştirilmiştir. Prizma ve duvar numuneleri üzerinde basınç ve
kayma deneyleri yapılmıştır. Deneyler neticesinde, ilgili dayanım ve deformasyon
özellikleri ve Bölüm 4'de elde edilen ilişkilere benzer ilişkiler elde edilmiştir.
Bölüm 6'da, tez kapsamında gerçekleştirilen ve ilgili bölümlerde sunulan deneysel
çalışmalar özetlenerek, karşılaştırmalı bir şekilde verilmiştir. Ayrıca; karot, prizma
ve duvar numuneleri için, basınç ve kayma gerilmeleri arasındaki etkileşimi gösteren
eğriler elde edilmiştir. Kayma dayanımı ve basınç gerilmesi arasındaki ilişki,
genellikle, Mohr-Coulomb varsayımına uygun olarak doğrusal bir ilişki olarak
tanımlanır. Ancak, elde edilen denklemler, doğrusal ilişkinin düşük basınç gerilme
seviyeleri için geçerli olduğunu göstermektedir. Elde edilen ortalama basınç ve
kayma dayanımı bileşenleri, Deprem Bölgelerinde Yapılacak Yapılar Hakkında
Yönetmelik (TSDC) (2007) hükümlerine göre değerlendirilmiştir.
xxxvii
Prizma ve duvarların basınç yükleri altındaki davranışları, doğrusal olmayan sonlu
eleman çözümlemeleri ile incelenmiş ve Bölüm 7'de verilmiştir. Sonlu eleman
çözümlemeleri için Abaqus (2009) sonlu eleman analiz programı kullanılmıştır.
Analizlerde, tuğla ve harcın mekanik özellikleri ve basınç gerilmesi-düşey
şekildeğiştirme ilişkileri için deneysel olarak elde edilen sonuçlar kullanılmıştır.
Sonlu elemanlar programı kullanılarak elde edilen prizma ve duvarın basınç
gerilmesi-düşey şekildeğiştirme ilişkileri, deneysel olarak elde edilen ilişkilerle
karşılaştırılmıştır. Sonlu eleman analizinin, numunelerin basınç dayanımlarını tahmin
etmede başarılı olduğu görülmüştür.
Bölüm 8'de çalışmanın sonuçlarına yer verilmiş ve daha sonra yapılması önerilen
çalışma konularına değinilmiştir.
xxxviii
1
1. INTRODUCTION
Masonry known as a traditional structural material is used for construction of
structures all over the world from ancient times to today. Almost until the 19th
century, when technological progresses took place, masonry constituted the main
material of constructions. This structural material is a composite material, which is
formed by laying units of stones and/or bricks generally bonded with mortar above
each other. As the properties of masonry constituents (stones, bricks, and mortar) are
connected with construction time and local conditions of region, such as raw material
features and workmanship techniques, the structural material characteristics show a
wide variety from one region to another.
Most of the historical structures, which are the architectural and cultural heritages of
the world, were constructed with masonry walls. The masonry walls of these
structures were used as both load-bearing elements and interior partitioning walls.
These walls are formed by using different types of unit (stone and/or brick) bonds
with/without mortar. As these structures are surviving examples of the progress of
civilization and the construction practice, they should be preserved for future
generations. The safety assessment and protection of these heritages against possible
seismic and/or aging effects require an extensive study including the steps of
geometrical and structural layout drawings, the investigation of masonry materials,
analysis of existing structural system and design of adequate strengthening
interventions if needed.
Although, recently, several studies on historical masonry structures have been going
on in the world, the number of these studies in Turkey is limited. As Turkey was host
to several civilizations such as Roman, Byzantine and Ottoman Empires during
history, the country have many invaluable historical masonry structures ranging from
city walls, bridges, caravansaries, churches, mosques and residential buildings.
2
The studies carried out in the world can be utilized for the assessment and
strengthening of the masonry structures in Turkey, but due to the specific local
properties and conditions, this kind of a direct adaptation cannot be sufficient to
evaluate the present state of the historical structures in Turkey. Consequently, the
investigation of structural characteristics of historical masonry materials and
structures in Turkey, which is located on a high seismic risk region, is a challenging
issue requiring extensive research.
1.1 Objectives of the Thesis
The subject of the thesis had initially been planned as the modeling of the seismic
behavior of the historical brick masonry structures. However, during the
investigation of the related literature, it was clearly found out that, the knowledge on
the material characteristics of the historical structures in Turkey is not sufficient and
that the number of the studies on the historical materials is very limited. Therefore,
the material characteristics required for the analyses of these structures are generally
based on rough assumptions. To base the material features on assumptions or to take
default values given in several codes during the analyses of these structures might
lead to non-realistic assessments.
As the knowledge on the traditional masonry walls and masonry constituents of the
historical structures in Turkey is scarce, the study intends to study the characteristics
of historical masonry walls and masonry constituents. The main purposes of the
study are:
 To make a contribution for the development of a systematic database on the
material characteristics of the historical masonry by conducting tests on bricks,
mortar and masonry specimens,
o To determine mechanical features of the masonry materials and
members under tension, compression and shear loads, which are vital
characteristics for the structural design and the assessment of the
existing structures,
o To determine physical and chemical properties of the masonry
constituents,
3
 To establish relationships between the material characteristics, such as the
relationships between the modulus of elasticity and compressive strength
based on the statistical assessment of the test data,
 To observe and define load-deformation or stress-strain behavior of the
masonry under monotonic or cyclic vertical and horizontal loads,
 To investigate the compliance of the default values and the relationships
given in several codes about the masonry mechanical characteristics, such as
default values of compressive or shear strength of masonry, or equations such
as about Young's modulus of masonry.
 To carry out different types of tests (destructive or non-destructive and in-situ
or laboratory) on specimens with different properties in terms of source
(original or reproduced), size (small or large) and composition (stacked bond
or running bond) for the evaluation of the obtained results in a comparative
manner,
 To conduct non-linear finite element analyses for the numerical prediction of
the behavior of the specimens tested under compression loads.
To achieve the objectives given above, an extensive experimental and theoretical
study was carried out on the brick masonry specimens, which were extracted from a
historical row house group. The historical row houses are the Akaretler Row Houses
constructed around 1875 in Istanbul. Since several structural walls of the houses
were to be removed according to the restoration design, a large number of different
types of specimens could be collected for laboratory tests and considerable amount
of in-situ destructive and non-destructive tests were carried out as well. It is
considered that the historical masonry knowledge obtained from these experimental
studies may be used in structural assessment and restoration works of the other
historical masonry row houses/buildings constructed in 19th century in Turkey.
Particularly, İstanbul has many historical brick masonry buildings and row houses
constructed in 19th century. The historical buildings located in Beyoğlu (Pera) and
the Akaretler Row Houses, İstanbul are among typical examples of this type
construction.
4
1.2 Outline of the Thesis
In this study, a large number of the destructive and non-destructive tests was carried
out in laboratory and in-situ on original specimens and on reproduced specimens
(under monotonic and/or cyclic compression, tension (flexural or splitting) and shear
loads. While original specimens are brick, mortar, core, and wallet, the reproduced
specimens are prism, triplet, and wall, which were built with original bricks and
reproduced mortar simulating original mortar mechanical characteristics. The test
results are supported through statistical analysis and numerical studies.
The thesis includes eight chapters. Chapter 1 addresses an introduction, a brief
history of masonry walls and their components and the objectives of the thesis, and
reports the related literature.
In Chapter 2, the historical Akaretler Row Houses were briefly explained in terms of
historical, geometrical, and structural characteristics. The structural materials of the
houses were also described in terms of their physical features.
Chapter 3 presents the identification of the material characteristics of the bricks and
mortar. The bricks and the mortar specimens taken from the walls of the houses were
tested under flexural tension and compression for assessment of their basic
mechanical characteristics. Besides studies on the physical and chemical aspects of
these specimens, several in-situ non-destructive tests were carried out for obtaining a
correlation between destructive and non-destructive test predictions.
The mechanical properties of masonry specimens were assessed by testing original
specimens extracted from the historical walls (cores and wallets) and by in-situ tests
as described in Chapter 4. While the core specimens consisted of two pieces of bricks
bonded with a bed joint, the wallet specimens were three brick high including two
bed joints and several head and longitudinal joints. For understanding the behavior of
the core specimens; splitting tension, compression, and shear tests were performed in
laboratory. The shear tests of the cores were conducted for three distinct precompression
levels to determine shear bond strength and friction coefficient as well
as the influence of the pre-compression level. The wallets were tested under
monotonic and cyclic compression loads. The in-situ shear tests of the walls were
also performed using either a monotonic or a cyclic loading pattern. These tests
provided important data on basic engineering parameters of historical masonry walls
5
as well as several relationships between these parameters and between the
mechanical characteristics obtained by different testing techniques.
Chapter 5 includes the experimental studies performed on the specimens produced
with the historical bricks, which were collected from the walls of the houses, and
reproduced cement-lime mortar is assumed to be representative of the original mortar
of the row houses. Two different types of the specimens were built: prisms and walls.
The prisms were composed of three bricks bonded with two bed joints in stacked
bond, and the walls were composed of five courses of the bricks bonded with four
bed joints and each course included four head joints in running bond. While the
prisms were tested under both monotonic and cyclic pattern of the compression and
shear loads, the walls were tested under monotonic increasing compression and shear
loads. In order to clarify the influence of the pre-compression levels on the shear
tests, the pre-compression levels adopted were 0.13, 0.25, 0.50, 0.75 and 1.00 MPa
for the prisms (triplets) and 0.13, 0.25, 0.50 and 0.75 MPa for the walls. The
outcomes of the tests were analyzed to define pre-peak and post-peak behavior, to
obtain relationships between certain mechanical characteristics of the walls, and to
describe failure patterns, as wells as to clarify the influences of the specimen and test
types on the test results. In addition, diagonal tension tests of the walls were also
performed.
Chapter 6 includes the evaluation and comparison of the test results, taking into
account specimen source (original and reproduced), specimen size (small and large),
bond type (stacked and running) and test type (monotonic and cyclic; laboratory and
in-situ). Additionally, the obtained results are also evaluated according to Turkish
Seismic Design Code (TSDC) (2007) in a comparative manner.
In Chapter 7, the results of the numerical analysis are presented with aim of
comparing to the experimental results of the compression tests on the prism and wall
specimens.
Additionally, it should be noted that the test results presented in each chapter are
evaluated taking into account the similar studies in literature.
In the last chapter, main conclusions derived from the experimental and numerical
studies are summarized. This chapter also includes recommendations for future
works.
6
1.3 Brief History of Bricks, Mortar and Masonry Walls
1.3.1 Bricks
In the first period of the civilization, adobe and bricks occurred in the regions where
the availability of stone and/or timber was not easy, like Mesopotamia (Kuban,
2002). These are the first prefabricated structural materials (Croci, 2000; Kuban,
2002). Some outstanding examples of brick masonry structures were constructed in
Mesopotamia, Iran and the Middle East (Kuban, 2002).
The knowledge obtained from the study of Hendry and Khalaf (2003) is presented as
follows. Sumerians developed the manufactured clay bricks using moulds as early as
3000 B.C. These bricks were generally dried in the sun, but fired clay bricks were
known and utilized for the construction of important buildings. King
Nebuchadnezzar’s Palace in Babylon in Iraq, which was built about 600 B.C., is a
later example of early brick construction on large scale.
While according to Croci (2000), bricks were firstly made from mud and dried in the
sun, but were later made from clay and burnt in the kilns to ~1000°C temperature;
according to Drysdale et al. (1994), the earliest bricks were made from mud or clay
and dried in the air or the sun. While according to Drysdale et al. (1994), clay bricks
have existed for at least 10.000 years and the idea of firing brick was discovered by
3000 B.C.; according to Beall (2000), sun-dried mud and clay bricks have been used
for ~10.000 years and ~5000 years respectively.
Although the color of the bricks is generally red, yellow and darker colors of bricks
are also present due to the processes of incomplete or over-burning, respectively
(Croci, 2000). While the bricks in the Roman period and the Middle Ages were
generally 30-40 mm and 25-40 mm high, respectively; those in the period of 14th-19th
centuries were generally approximately 55mm high (Croci, 2000). The sizes of some
Roman bricks were given as 225×450×75 mm, 225×225×75 mm, and 200×200×38
mm by Drysdale et al. (1994). Since the firing process was performed in uncontrolled
conditions, size and quality of the bricks showed large variations, (Drysdale et al.,
1994).
7
1.3.2 Mortar
The function of mortar in masonry structures is to transform a wall into an integral
unit by taking up irregularities in the masonry units, the errors in alignment and
providing necessary strength to resist loads as mentioned by Mulligan (1942),
Bayülke (1980) and Orton (1992).
The first production of bricks from soil caused the production of mortar from soil.
The first form of mortar was mud with/without wood pieces, and then an efficient
mortar was obtained by using lime, sand and water in the period of Roman Empire
(Kuban, 2002).
In the study of Croci (2000), the history of mortar was given as follows: Mortar
consists of binder, aggregates, and water. According to the types of binder and
aggregates used in mortar production, mortar can be classified as aerial lime mortars,
pozzolanic mortars, hydraulic lime mortars, cement mortars, and hybrid mortars. As
cement and hybrid mortars are currently available and used commonly, they are not
explained in the study. Aerial lime mortars, which include slaked lime, sand and
water, harden by carbon dioxide in the air. Pozzolanic mortars are composed of
slaked lime, natural or artificial pozzolana and water. These types of mortars were
extensively used in the Roman period. Artificial pozzolana is obtained generally by
crushing or pulverizing tiles or old bricks. Hydraulic lime mortars consist of lime
obtained by burned marl, sand, and water. This mortar, which can harden under
water in the absence of air, was a product of the 18th century.
According to Drysdale et al. (1994): Early mortars might have been formed with
clay, bitumen, or clay-straw mixtures. After Egyptian produced mortar including
calcined gypsum a few thousand years ago, the Greeks and Romans produced the
earliest types of concrete by adding lime, sand and crushed stone or brick. The fact
that lime mortars are not hardened under water caused the manufacturing of
pozzolanic cement by the Romans. This kind of mortar was composed of lime and
volcanic ash. The name "pozzolanic" comes from the name of the place where
volcanic ash was obtained.
1.3.3 Masonry Walls
The history of masonry walls are given in the study of Drysdale et al. (1994): The
walls of ancient times, which were solid and massive, were constructed with stone,
8
sun-dried brick or kiln-burnt brick. These walls were utilized to retain earth and to
enclose buildings. For example, the Roman masonry walls were arranged with
bonded bricks or with outer bricks and the space between the outer bricks was filled
with an early type of concrete. The earliest examples of masonry walls were built
with mud bricks laid in mortar. The thicknesses of the mortar joints varied between 1
to 40 mm. An example of this type wall construction is found at Ctesiphon in
Mesopotamia. The wall, which was a part of a palace, had a thickness of 5 m at the
bottom and a height of 34.4 m.
According to Croci (2000): The oldest walls, which were built with sun-dried bricks
and mortar (3000 B.C.), were revealed, the walls of Jericho. It is also stated that the
joint thicknesses of mortar were in ranges of 10-20 mm (the Roman period), 15-45
mm (the Middle Ages).
1.4 An Overview of Literature
Historical and recent/ordinary unreinforced existing masonry structures have been
studied by several researches. This section is arranged to present the knowledge on
the corresponding subjects collected from the literature. These subjects are the
material characterization of historical masonry, the relationships established to
predict masonry compressive strength based on the strengths of constituents (brick
and mortar), the relationships of compressive stress-compressive strain and Young’s
modulus-compressive strength and the shear strength characteristics such as the
friction coefficient and shear strength at zero vertical stress.
1.4.1 The characteristics of historical masonry
In this section, the material characteristics of several historical masonry structures
are compiled from the literature.
The studies of Postacıoğlu (1981), Güner (1984), and Akman et al. (1986): Masonry
prisms taken from a historical cistern and a historical building being used as a wine
storehouse in Istanbul and masonry prisms produced in laboratory were tested to
obtain average compressive strength, unit weight and water absorption rate. The
masonry prisms included bricks and khorasan mortar. While khorasan is the name of
burnt clay made from broken and crushed bricks, tiles and potteries, khorasan mortar
is the name of the mortar including khorasan, slaked lime, sometimes sand and
9
water. The average values of the compressive strengths, unit weights and water
absorption rates for the cistern and the wine storehouse were calculated as 4 and 7
MPa; 1360 and 1320 kg/m3; 32 and 33%, respectively. In addition to these values,
the shear strength of the masonry specimen taken from the cistern was determined as
0.7 N/mm2. The compressive strengths and moduli of elasticity of the produced
prisms, which were constructed with common bricks (burnt hand made bricks) and
khorasan mortar in laboratory, were in ranges of 3.2-4.8 MPa and 340-725 MPa. It
was determined that the curing conditions of the produced prisms, namely, dry and
humid, had no important effect on the mechanical properties.
The study of Baronio and Binda (1995): The historic mortar samples, which were
collected from the ruins of the Civic Tower collapsed in Pavia, 1989, were studied.
The results of chemical analyses of the mortar samples figured out that the type of
binder was lime putty and that aggregates had siliceous nature whose proportion
varied between 52-82%. Lime putty was explained with the existence of calcium
oxide between 6 and 12%. The binder-to-aggregate ratios were between 1:2 and 1:5.
While the porosity and the bulk density were estimated as 12-13%, 1862-1914
kg/m3, respectively, the compressive strengths and moduli of elasticity of mortar
samples having dimension of ~30x30x30 mm were 2.92-11.46 MPa and 268-1297
MPa, respectively. The mortar/masonry compressive strength ratio was found as 2.8.
The study of Ahunbay et al. (1997): Using non-destructive approaches, Hagia Sophia
was studied. Endoscopic investigations were utilized to detect the existing state of
the inner parts of the main piers carrying the dome that were built with khorasan
mortar and bricks. Chemical analysis results revealed that the piers were constructed
with good quality stones and hydrated lime mortar. In order to figure out the existing
state of several columns, ultrasonic tests were carried out and the test results
indicated the need of the observing these columns. Ten displacement transducers
were located on the existing cracks to monitor the evolution of the cracks for a period
of over one year. It was established that there was no significant changes in the
displacements and the slight changes in the crack openings were due to temperature
differences.
The study of Güleç and Tulun (1997): The mortar and plaster samples taken from
Roman Bath in Ankara, Tahtakale Bath and Esekapı Madrasah in Istanbul were
inspected. While Roman Bath was constructed in Roman, Tahtakale Bath and
10
Esekapı Madrasah were constructed in Ottoman periods. The mineralogical,
petrographical and physico-chemical tests of these samples were carried out to obtain
necessary data for choosing appropriate materials during restoration works. The
composition of the mortar samples of Roman Bath was determined as randomly
distributed lime binder, fine, and coarse aggregates. As the composition of the
samples was similar to that of the modern mortar, it was thought that the samples
might be taken from cement mortar, which was used in the past restoration of the
bath. The samples of Tahtakale Bath and Esekapı Madrasah consisted of uniformly
distributed binder, fine, and coarse aggregates. Due to the existence of brick pieces in
mortar, the mortar type of Tahtakale bath and Esekapı Madrasah was determined as
khorasan mortar.
The study of Karavezirogluo et al. (1997): The effect of the thickness and strength of
mortar on the compressive strength of masonry was researched experimentally for
historical constructions. Masonry prisms were built with mortar, which was similar
to the one taken from the historical constructions, and bricks, which had the physical
and geometrical characteristics of Roman and Byzantine period. The mortar
consisted of lime, natural pozzolana, sand, brick pieces, brick powder, and cement.
While the average compressive strength of the masonry was determined as 3.35
MPa; after 28 days, the average compressive strength of the bricks was determined
as 9.6 MPa. According to the test results, for the same joint thicknesses, the masonry
strength increased as the mortar strength increased. The deformation values of the
masonry with thick mortar joints were higher than that of the masonry with thin
mortar joints under compression loads.
The study of Papayianni (1997): The compressive strength, water absorption rate and
bulk density of the bricks and mortar, which were taken from the best known
historical structures remaining from Ottoman period (15th-16th centuries) in
Thessaloniki, Greece, were determined experimentally. The common size of the
bricks was 300x400x40 mm, and the color of the bricks was generally red-brown.
The reason of different tones of the color was low firing temperature. While the
compressive strengths of the bricks were generally in the range of 5-19 MPa, that of
the 10 % of the bricks were less than 5.0 MPa. The mortar samples were found to
have a high content of lime and a small percentage of siliceous material (8-17%).
Due to the high ratio of lime, the color of the mortar was generally grey-white. While
11
the compressive strength of the mortar samples was generally in the range of 1-6
MPa, that of the 48% of the samples were between1.5 and 2.2 MPa.
The study of Papayianni and Hatzitrifonos (1997): The characteristics of the brick
and mortar samples of Pazar Hamamı built in 15th century in Greece were given. The
compressive strengths of the bricks with 40-55 mm height and the mortar were 3-
11.5 and 1-6 MPa, respectively. While the mean values of apparent specific gravity
and water absorption rate of the bricks were 1550-1600 kg/m3 and 18-28%; those of
the mortars were 1400-1600 kg/m3 and 20-25%, respectively.
The study of Karaveziroglou-Weber et al. (1998): For deciding the quality and
properties of masonry materials to be used in the restoration of Archeiropoiteus
Church built in 5th century in Greece, several compressive tests were performed on
historical bricks and historical mortar taken from the church. The strengths of the
historical materials were obtained as 8-10 MPa for the bricks, whose average
apparent specific gravity was 1700 kg/m3, and 2.7 MPa for the mortars.
The study of Binda et al. (1999): The physical properties of the bricks and the
mortars, which were extracted from Basilica of S. Vitale of Byzantine period in
Ravenna, were investigated. The water absorption rate and apparent specific gravity
of the bricks were determined as 22% and 1543-1600 kg/m3, respectively; and the
average apparent specific gravity of the mortar was 1705 kg/m3.
The study of Baronio et al. (1999): The mortar samples of Tower of Pavia were
studied. The average compressive strength and Young’s modulus of the mortars were
identified as 4.5 and 700 MPa, respectively. The water absorption rate and porosity
were determined as 14 and 24%, respectively. The average values of the apparent
and real specific gravities were calculated as 1715 and 2252 kg/m3, respectively.
The study of Andersen et al. (1999): The mineralogy and texture of mortars sampled
from medieval churches in Denmark were inspected. The aggregates of the samples
were mainly quartz and feldspar. The aggregate/binder ratio was calculated as ~1:3.
The study of Croci (2000): The information on the compressive strengths of the
historical materials are given as 2-11 MPa for masonry, 15-30 MPa for good quality
bricks and 2.5 MPa for 1:3 to 1:5 lime:sand mortar.
The study of Binda et al. (2000): The material test methods of existing unreinforced
masonry structures were given and the difficulties in the assessment of the
12
information obtained from the non-destructive test methods were mentioned. Sonic,
radar and thermography tests are the most commonly used non-destructive tests. The
steps of diagnosis procedure of existing masonry structures were summarized as the
investigation of geometry and crack pattern, static and/or dynamic monitoring,
foundation and soil research, laboratory tests of existing materials to define chemical,
physical and mechanical characteristics, in-situ testing such as flat jack, rebound
hammer, penetration, pull-out and non-destructive tests. This procedure was applied
to the Bell Tower of Monza. Using flat jack tests, modulus of elasticity and Poisson's
ratio of several walls were found as 465-1380 MPa and 0.07-0.20, respectively.
According to the results of chemical and concerning analysis of mortar samples,
mortar included mainly putty lime and siliceous aggregates (~%65). As mortar was
weak, the mechanical tests of mortar samples could not be carried out. In terms of
color, there were two of type bricks, namely, brown and light red. While the brown
bricks had absorption by total immersion of 13%, compressive strength of 28-33
MPa and modulus of elasticity of 2050-5300 MPa; the light red ones had absorption
by total immersion of 18%, compressive strength of 4-12 MPa and modulus of
elasticity of 500-1330 MPa.
The study of Matovic and Milovanovic (2001): The physical characteristics of the
sandstone taken from the historical St. Marco Church in Belgrade, Serbia were given
as follows. The ranges of bulk density, density, absolute porosity, and water
absorption rate were determined as 1880-2420 kg/m3, 1940-2720 kg/m3, 7.4-27.9%
and 2.43-9.16%, respectively.
The study of Papayianni and Stefanidou (2001): The mortars of the several
constructions built in 2nd, 7th and 15th centuries were examined in terms of pore size
distribution and porosity. While the pore of these mortars was in the range of 1-
0.1μm, the open porosity was in the range of 20-40%.
The study of Binda and Saisi (2001): Non-destructive sonic tests (NDT) and minordestructive
flat-jack tests were utilized to characterize the current states of three
historical stone structures in Italy in terms of materials, material variations through
the structure, wall texture, the existing stress, modulus of elasticity, damage etc.
Compression and Brazilian tests were also conducted on the stone samples taken
from one of these structures. The ratio of the average tensile strength to the average
compressive strength was 0.11-0.12. It was explained that the evaluation of the
13
results obtained from sonic and flat-jack tests could lead to a relation between sonic
velocities and moduli of elasticity.
The study of Penelis (2002): The compressive strengths of mortars of Roman rotunda
(4th century), Christian rotunda (4-5th century), Hagia Sophia (8th century), Bey
Hamamı (15th century) and minaret of Roman rotunda (17th century) were given as
2.3, 3.7, 4.5, 1.18 and 1.25 MPa, respectively.
The study of Baronio et al. (2003): Chemical, physical, and mechanical tests were
conducted on the in-situ materials of the Noto Cathedral, Italy after the partial
collapse of the construction. Using chemical analysis, binder of the mortar samples
was determined as lime. The water absorption and the bulk density of the mortar
were estimated as 26-27% (in weight) and 1.36-1.37 kg/m3, respectively. The
compressive and tensile strengths, which were obtained from the splitting tests, of
the mortar samples varied between 0.4-0.9 MPa and 0.2-0.3 MPa, respectively.
According to the chemical and optical analysis results, the stones, namely,
calcarenite and giuggiolena, were different types of limestones. The physical
properties of the stone samples were determined: Bulk densities in dry and saturated
conditions are ~1500-1750 kg/m3, and ~1830-1990 kg/m3, respectively. While water
absorption is determined as 14-19%, initial rate of absorption (IRA) coefficient is
determined as ~2.4-6.8 kg/m2. The compressive strengths of the stone specimens
were also estimated in dry and saturated conditions. While the average compressive
strengths of the dry and saturated calcarenite specimens were determined as 18 and
11.6 MPa, respectively; those of the dry and saturated giuggiolena specimens were
determined as 5.3 and 5.1 MPa, respectively.
The study Arce et al. (2003): Several tests were conducted on the historical bricks of
12th-14th centuries, which were collected from the historical buildings in Toledo,
Spain, to figure out the composition of the bricks for utilizing in the restoration
works. The chemical analysis leaded to the result that the color tones of bricks
depended on the firing temperature, such as darker bricks were fired in high
temperature, and light brownish bricks were fired in low temperature. Water
absorption, water suction, density, and total porosity were estimated as 19-22%,
0.08-0.18%, 1510-1600 kg/m3, 32-43%, respectively. The average compressive
strength of the bricks was 33 MPa.
14
The study of Pohle and Jager (2003): The in-situ material properties of the historical
Frauenkirche Church in Dresden, Germany were evaluated by tests to determine the
appropriate materials to be used in the reconstruction of the structure. The materials
of stone, mortar, masonry and the metal components were collected from the ruins of
the historical structure. The compressive strength of the mortar including lime and
brick powder varied between 1.5-5.0 MPa.
The study of Güleç et al. (2005): The mortar characteristics of the land walls of
Constantinople at Yedikule in Istanbul were examined to obtain chemical and
physical properties and compressive strengths of the mortar samples. According to
the test results and observations, the mortar included lime as binder and crushed
bricks, river sand and limestone pieces as aggregates. It was determined that the
porosity of the samples varied between 22-38% and the compressive strengths varied
between 4-6 MPa.
The study of Felice (2006): While the average compressive strength of the bricks
taken from S. Luca vault of Bologna dating back to the 17th century in Italy was
determined as 21.8 MPa, the average compressive strength of the bricks taken from a
structure of 19th century was determined as 30 MPa. The size of the bricks of S. Luca
was 290x135x50 mm and the average compressive strength and Young’s modulus of
the mortars, for which lime was used as binder were 1.7 and 347 MPa.
The study of Krstevska et al. (2007): Seven mosques and two churches dating from
Ottoman period were inspected by in-situ and laboratory tests. The ambient vibration
tests were carried out to get natural frequencies, mode shapes, and damping
coefficients of these structures. While the natural frequencies of the minarets were
determined as 0.96-1.40 Hz, those of the others were determined as 2-6.6 Hz for
translational directions and 3.2-6.4 Hz for torsion. Surface hardnesses of the masonry
on site were measured by a P type Schmidt hammer. While the compressive
strengths of the stones were determined from core samples having a diameter of 50
mm in the range of 12.7-108.7 MPa, those of the bricks were determined as 6.8-14.7
MPa. In-situ compression stresses of the structures were measured using flat-jacks as
0.24-1.63 MPa.
The study of Palieraki et al. (2007): Radar one of the non-destructive methods and
boroscopy one of the minor-destructive methods were applied to three historical
monuments, which were built in Byzantine period, in Greece. These tests were
15
helpful for estimating the construction techniques, the presence and state of timber
ties, and the thickness of stones. However, it was mentioned that the use of radar
method in brick masonry structures is not suitable due to the effect of clay on emitted
and received signals and that very careful cleaning is required to observe the drilled
holes by boroscopy.
The study of Papayianni and Stefanidou (2007): The compressive strengths, porosity,
and specific densities of the mortars belonging to Roman, Byzantine, and Ottoman
periods are detailed. The compressive strengths, the values of porosity and specific
density were 2.50-5.50 MPa, 20-25% and 1650-1700 kg/m3 for Roman; 2.80-6.50
MPa, 17.5-22% and 1680-1720 kg/m3 for Byzantine; and 1.36-2.50 MPa, 18-27.5%
and 1500-1650 kg/m3 for Ottoman Periods, respectively.
As understood from the studies summarized above, the characteristics of the
historical masonry and masonry constituents vary in a wide range. Consequently, for
performing a realistic and reliable structural assessment of the historical masonry
structures, the determination of the masonry characteristics realistically is vitally
important.
1.4.2 The predictions of masonry compressive strength
The compressive strength of masonry depends on several factors such as unit and
mortar features, which are strengths, moduli of elasticity, Poisson's ratios, and
thicknesses, quality of workmanship, bonding type and adhesive strength between
unit and mortar joints. By considering some of these factors, several experimental
studies have been carried out by researchers to propose empirical relationships for
the estimation of the masonry compressive strength. In the following paragraphs,
several relationships proposed by codes and researches are presented.
Turkish Seismic Design Code (TSDC) (2007) recommends the determination of
masonry strength depending on unit compressive strength if masonry wall tests are
not performed, Eq. (1.1). According to the code, the wall tests can be performed on
the wall specimens built with unit and mortar to be used in the construction of loadbearing
walls. However, the size and bond type of masonry wall specimen is not
described in the code.
masc uc f  0.50 f (1.1)
16
In Eq. (1.1), masc f
is the compressive strength of masonry wall and uc f
is the
compressive strength of units. It should be noted that the value determined from Eq.
(1.1) is not allowable or specified stress. According to this code, 25% of this value is
taken as allowable compression stress.
Building Code Requirements and Specification for Masonry Structures (BCRSMS)
(2008) proposes two methods for the estimation of compressive strength of masonry.
The first method, called as unit strength method, is presented with the following
equation for clay brick masonry:
(400 ) '
masc uc f  A  Bf (1.2)
where A 1 (inspected masonry), B  0.2
for N type portland cement-lime mortar
(at 28 days strength of 5.2 MPa), B  0.25for S or M type portland cement-lime
mortar (compressive strengths of 17.2 and 12.4 MPa at age of 28 days, respectively),
uc f is the average compressive strength of units and '
masc f is the specified
compressive strength of masonry in psi. The second method, called as prism test
method, is based on the testing of prisms constructed with the same units and mortar
as those to be used in the structure. The method is standardized in ASTM C 1314-
03b (2003).
Eurocode 6, EN 1996-1-1 (2005) also provides two methods for the determination of
masonry strength. The first one is an equation depending on the strengths of unit and
mortar, the properties of unit and mortar (unit material, hole percentage in unit,
mortar joint thickness and existence of longitudinal joint).
 
masc c uc n mc f Kf f , ,  (1.3)
masc c f , is the characteristic compressive strength of masonry. Constant K takes the
value of 0.55 for masonry without longitudinal mortar joint and built with clay unit
of Group 1 and general purpose mortar. If masonry has a longitudinal joint through
all or a part of the length of the masonry; the value of the constant is multiplied by
0.8. The code classifies units into five groups according to raw material type and
hole percentage. If the ratio of the volume of all holes in a unit to its gross volume is
 25%, the unit is classified as Group 1. Masonry mortar without any special
characteristics in terms of joint thickness and of density is called general purpose
17
mortar.  , constants are equal to 0.7 and 0.3 for masonry formed with general
purpose mortar, respectively. uc n f , is the normalized mean compressive strength of
the units. The normalized compressive strength of the unit is the strength converted
to the air dried compressive strength of an equivalent 100 mm wide x 100 mm high
masonry unit. mc f is the mean compressive strength of the mortar. The coefficient of
variation of the strength of the masonry units should be not more than 25 % for using
Eq. (1.3). The second approach of EN 1996-1-1 (2005) is to test masonry specimens.
The test and production of the specimens are explained in TS EN 1052-1 (2000) and
EN 1996-1-1 (2005). It should be noted that all statements and limitations given in
EN 1996-1-1 (2005) are not detailed here and only statements/limitations related to
the materials tested in the thesis is mentioned.
Pande et al. (1994) suggests that the masonry compressive strength may be
calculated depending on the tensile strength of units. It is mentioned that this
suggestion is developed as an alternative of the equation provided by Eurocode 6. It
should be noted that the tensile strength is divided by stress factors (0.08) to find
the masonry compressive strength. These factors are given depending on dimensions
of masonry unit, Young’s moduli and Poisson's ratios of unit and mortar, and
thicknesses of mortar joints.
Tomazevic (2006) carried out several experimental studies on materials defined in
EN 1996-1-1 (2005), so as to detect the validity of Eq. (1.3). While the differences
between the results of the tests and of the equation are small for some cases, the
differences are large for the other cases.
Thomas (1971) reported that the compressive strength of brick masonry with
cement:lime:sand mortar can be taken as about the fourth root of the mortar strength,
and that the ratio of masonry compressive strength to brick compressive strength
varies between 0.2 and 0.4.
In the study of Akman (1978), three functions were reported for the predicting the
compressive strength of brick masonry walls:
masc uc f  0.25 f (1.4)
18
3
masc uc mc f  f f (1.5)
 
e
b
h
f f
f
w
w
uc mc
masc 



16 3
4 0.1
(1.6)
where w h is the height of wall, w b is the height of wall, and e is a factor depending on
the quality of workmanship. The factor takes the values of 10 (good quality), 0
(average quality), and -5 kgf/cm2 (poor quality).
According to Bayülke (1980), the compressive strength of masonry constructed with
common brick and 1:3 lime:sand or 1:2:8 cement:lime:sand mortar can be related to
the brick strength with Eq. (1.7). Additionally, it is mentioned that the ratio of
masonry compressive strength to brick unit compressive strength is in the range of
0.25-0.50.
masc uc f  0.27 f (1.7)
Hendry (1990) reported Eqs. (1.8)-(1.13) proposed before by several researchers to
predict masonry compressive strength. It should be noted that dimensions are in
MPa. The following relations are for solid brick masonry built with the
cement:lime:sand ratios of 1:1/4:3 or 1:1:6. The Eqs. (1.8) and (1.9) are for the
predictions of the characteristic compressive strengths of the 102.5 and 215 mm thick
walls, respectively. For the mean strengths, the constants of these equations are given
as 1.242 and 0.334, respectively.
0.531 0.281
, 1.017 masc c uc mc f  f f (1.8)
0.778 0.234
, 0.217 masc c uc mc f  f f
(1.9)
The mean compressive strength of the blockwork masonry ( ) masc f is predicted with
Eq. (1.10). It should be noted that this equation is for the specified mortar ingredients
(cement:lime:sand = 1:1:6).
0.67 0.33 0.9 masc uc mc f  f f (1.10)
19
According to other researcher, the mean compressive strength of masonry may be
obtained using Eq. (1.11).
0.66 0.33 0.83 masc uc mc f  f f (1.11)
According to Eq. (1.12), the masonry strength depends on the modulus of elasticity
of mortar rather than the mortar strength:
0.26
0.43
1000
97 . 0 




 mc
masc uc
E
f f (1.12)
where mc E is the modulus of elasticity of mortar.
The last proposal was expressed depending on type and size of unit and on masonry
with or without longitudinal mortar joints:
0.75 0.25
, ( ) masc c uc mc f  K f f (1.13)
In Eq. (1.13), K is 0.45 for masonry without longitudinal mortar joint and 0.35 for
masonry with longitudinal mortar joint.  is 1.0 for unit of 200 mm long x 200 mm
thick, 0.7 for brick of 65 mm long x 200 mm thick and 1.4 for block of 200 mm long
x 100 mm thick.
The equation reported by Karaveziroglou-Weber et al. (1998) includes the effect of
unit ( ) b h and mortar thicknesses ( ) mb t of brick and mortar:
   mc uc mc
b
mb
masc c f f f
h
t
f  








 


 


 1 0.83 0.4
,
(1.14)
The effects of mortar strength and mortar joint thickness on the masonry
compressive strength were also studied experimentally in the study of
Karaveziroglou-Weber et al. (1998). Although an equation is not proposed, the ratios
are given depending on the thicknesses of the mortar joints for the historical
structure. It is noted that the compressive strength of the structure might be taken as
30% of brick strength for the mortar joint thickness of 50 mm, as 42% of brick
strength for the mortar joint thickness of 40 mm and as 70% of brick strength for the
mortar joint thickness of 20 mm. It was also stated that the equations given by
20
Eurocode 6 ( K  0.60,  0.65and  0.25 ) underestimate the strength of old
masonry.
Lourenço and Pina-Henriques (2006) reported Eq. (1.15) which takes into account
the influences of the unit compressive strength and the elastic moduli ( , ) uc mc E E ,
Poisson's ratios ( , ) uc mc   and thicknesses ( , ) b mb h t of brick and mortar, respectively.
 
uc
m
mc
uc
u m
mc
uc
mc
uc
b
mb
masc f
E
E
E
E
E
E
h
t
f
 





 






 


 






 
1
1
1
(1.15)
1.4.3 The relationships proposed for masonry stress-strain
Using the test data of the brick masonry with an average masonry compressive
strength of 15 MPa (the compressive strengths of the masonry of 9-28 MPa), the
relationship of compressive stress-compressive strain is presented with a parabolic
function reported by Hendry (1990):
2
2
 


 



 


 



f
masc
f
masc
masc
masc
f 


 
(1.16)
In this equation, masc  and masc  are the compressive stress of masonry and the
corresponding compressive stress, masc f and f  are the compressive strength of
masonry and the corresponding compressive stress. This function is also obtained by
Kaushik et al. (2007) for clay brick masonry (the compressive strengths of the
masonry of 4.1-7.5 MPa). It should be also noted that the relationship is valid until
compressive stress drops to 90% of the corresponding compressive strength (Kaushik
et al., 2007).
Croci (2000) described a compressive stress-compressive strain relationship for
historical brick masonry, which has three parts as linear elastic branch, in-elastic and
descending branches. However, no function was given to express this relationship.
21
1.4.4 The relationships proposed for modulus of elasticity-compressive strength
of masonry
The modulus of elasticity in compression is determined by the compression tests of
masonry or by empirical formula depending on the masonry compressive strength.
Based on the compression test results, the modulus of elasticity ( ) masc E is generally
formulated as given in Eq. (1.17), where C is a constant coefficient.
masc masc E  Cf (1.17)
The values of C collected from codes and from the literature survey are given in
Table 1.1. As seen in this table, the value proposed by TSDC (2007) is smaller than
the values of FEMA 356 (2000), EN 1996-1-1 (2005) and (BCRSMS) (2008). While
FEMA 356 (2000) and (BCRSMS) (2008) propose prism test method detailed in
ASTM 1314-03b (2003); EN 1996-1-1 (2005) proposes the test method detailed in
EN 1052-1 (TS EN 1052-1) (2000).
For the experimental determination of modulus of elasticity, FEMA 356 (2000) and
BCRSMS (2008) defines the modulus as the slope of the chord between the points of
5 and 33% of masonry compressive strength on the stress-strain diagram obtained
from the prism test method. TS EN 1052-1 (2000) defines the modulus as the secant
modulus at the point of 1/3 times masonry compressive strength. It should be noted
that the only TS EN 1052-1 (2000) describes the gage length and locations for
measuring of compressive deformations.
The relationships between the elastic modulus and compressive strength were
expressed with different functions (Eqs. (1.18) and (1.19)) by Hendry (1990).
masc masc E  2116 f (1.18)
0.83 1180 masc masc E  f (1.19)
Hendry (1990) described the initial tangent modulus of the brick masonry with the
following expression:
f
masc
masc
f
E

 2
(1.20)
22
Modulus of elasticity of masonry can be estimated using Eq. (1.21), depending on
the thicknesses and the moduli of elasticity of unit and mortar joint, (Drysdale et al.,
1994; Lourenço and Pina-Henriques, 2006). Moduli of elasticity calculated by Eq.
(1.21) may deviate from those determined experimentally in the range of 6 to 30%,
(Lourenço and Pina-Henriques, 2006).
mc
b mb
b
uc
b mb
b
masc
E
h t
h
E
h t
h
E
( )
1
( )
1





(1.21)
Table 1.1 : Proposals for C values in Eq. (1.17).
Reference C Note
FEMA 356 (2000) 550 Default
EN 1996-1-1 (2005) 1000 If tests are not performed.
TSDC (2007) 200
BCRSMS (2008)
700*
800**
* for clay masonry
** for concrete masonry
Sahlin (1971) 400-1000
Bayülke (1980)
45*
97**
* for masonry with 1:3 (lime:sand)
** for masonry with 1:2:8
(cement:lime:sand)
UNDP/UNIDO PROJECT
RER/79/015 (1984)
500-3000
Drysdale et al. (1994)
210-1670
500-600 North American clay brick masonry
Kaushik et al. (1997)
300*
200**
550***
* for brick
** for mortar
*** for masonry
Tomazevic (2006) 200-2000
For masonry produced from bricks and
mortar defined in Eurocode 6
This literature investigation revealed that the values of the C constant vary from 45
to 3000. As seen in Table 1.2, the compressive stress range defined for the
calculation Young’s modulus varies between 5 to 60%. In addition, the locations of
the gage points and lengths are not clearly defined and standardized.
Table 1.2 : The compressive stress range for the calculation of modulus of elasticity.
Reference Range (%) Note
TS EN 1052-1 (2000) 33 Masonry wall
ASTM C 1314-03b (2003) 5-33 Masonry prism
Baronio and Binda (1995) 20-60 Historic mortar
Aprile et al. (2001) 40 Historic brick and representative mortar
Felice (2006) 40 Historic masonry (representative)
Gumaste et al. (2007) 25 Indian brick, mortar and masonry (1.2-13.6 MPa)
23
1.4.5 The shear strength components
The values of the components of shear strength are given depending on the raw
material and the hole ratio of units and on the composition, compressive strength and
bed joint thickness of mortar. The components of shear strength given in several
codes and the related literature are summarized in Table 1.3 and Table 1.4,
respectively.
FEMA 356 (2000) recommends the determination of the shear strength ( ) f  of the
existing masonry structures for assessment depending on the bond strength ( ) o  ,
which is obtained from the shear tests realized in accordance with ASTM C 1531-03
(2003), (Eqs. (1.22) and (1.23)).
 1.0
,
max  
mb ul
o A
T
(1.22)
  0.5  0.5 f o (1.23)
max T is the maximum shear load recorded during the shear test, mb ul A , is the total
initial area of upper and lower bed joints for in-situ wall tested and  is the vertical
stress of the in-situ wall tested. In these equations, while the coefficient of friction is
assumed as 1.0 for the calculation of the bond strength, the coefficient is assumed as
0.5 for the calculation of the shear strength.
According to ASTM C 1531-03 (2003), the friction coefficient ( ) f  is in a wide
range of 0.3-1.6. The average of this range is taken as 1.0 with a coefficient of
variation of 0.30. According to TS EN 1996-1-1 (2006), the characteristic values of
the friction coefficient is constant (0.4), the bond strength ones vary from 0.10 to
0.30 MPa depending on unit type and mortar compressive strength at 28 days. In
Table 1.3, mortar is symbolized with M and compressive strength at 28 days. The
bond strength of 0.10-0.25 MPa and the friction coefficient of 0.5 are proposed by
TSDC (2007) for both allowable stress design and assessment. Building code
requirements and specification for masonry structures, (BCRSMS) (2008) mention
only the friction coefficient of masonry built with autoclaved aerated concrete and
with horizontal joint thickness  1.5 mm. The friction coefficient is given as 0.48 for
allowable stress design and as 1.0 for strength design, BCRSMS (2008).
24
The shear strength components collected from the literature presented in Table 1.4
are detailed in this paragraph. Yorulmaz and Atan (1971) tested the masonry walls
constructed with three types of bricks and two types of mortars under compression
loads oriented at different directions to the bed joints. Atkinson et al. (1989)
conducted shear tests on the two groups of masonry specimens. While the first group
was constructed with clay bricks with a 33-MPa compressive strength and 1:2:9
mortar (volumetric ratio), the second group was constructed with clay bricks with a
64-MPa compressive strength and 1:1.5:4.5 mortar (volumetric ratio). Paulay and
Priestly (1992) mentioned that the values of o  and f  depend on the test method
and masonry type and the ranges of the values (Table 1.4) is for unreinforced
masonry. Valluzzi et al. (2002) tested the triplet specimens built with bricks and
mortar. The average compressive strengths of these bricks and mortars were
determined as 8.8 and 6.0 MPa, respectively.
The literature research showed that the shear strength components vary in a large
interval and the values of these components depend on the test method and masonry
type as mentioned Paulay and Priestly (1992). Consequently, to specify shear
strength components and failure modes of the historical masonry tested in this thesis,
the historical masonry under shear loads is required.
The relations (ratios, charts, equations, and default values) mentioned above are
generally based on the test results of the modern masonry. The modern masonry is
generally formed with cement or cement-lime mortar (not lime mortar), and higher
strength bricks with respect to the historical ones. Consequently, the verification of
these relations for old/historical masonry structures is required before they can be
used in the structural assessments of the historical structures. In order to fulfill this
verification, the test results obtained from this study and the relationships established
based on these test results were evaluated in a comparative manner with the
corresponding relationships and statements given in the literature inspected.
25
Table 1.3 : The values of shear strength components given in several codes.
Reference Unit characteristics Mortar characteristics o  (MPa) f  (MPa)
FEMA 356 (2000) Concrete or clay  0.69 1, 0.5
ASTM C 1531-03 (2003) Solid or ungrouted hollow made of clay or concrete 0.3-1.6
TS EN 1996-1-1 (2006)
Clay
M10-M20 0.30 0.4
M2.5-M9 0.20 0.4
M1-M2 0.10 0.4
Calcium silicate
M10-M20 0.20 0.4
M2.5-M9 0.15 0.4
M1-M2 0.10 0.4
Aggregate concrete M10-M20 0.20 0.4
Autoclaved aerated concrete M2.5-M9 0.15 0.4
Manufactured or dimensioned natural stone M1-M2 0.10 0.4
TSDC (2007)
Brick with vertical hole (the hole ratio  35)
Lime mortar
enhanced with cement
0.25 0.5
Brick with vertical hole (the hole ratio > 35) 0.12 0.5
Solid or common brick 0.15 0.5
Stone 0.10 0.5
Autoclaved aerated concrete Glue 0.15 0.5
Solid concrete block Cement 0.20 0.5
BCRSMS (2008) Autoclaved aerated concrete
Mortar joint thickness
 1.5 mm
1.0, 0.48
26
Table 1.4 : The values of shear strength components collected from the literature.
Reference o  (MPa) f 
Yorulmaz and Atan (1971) 0.15-0.53 0.21-0.69
Atkinson et al. (1989) 0.13-0.81 0.64-0.75
Paulay and Priestly (1992) 0.10-1.50 0.30-1.20
Drysdale et al. (1994) 0.24-0.69 0.60-1.00
Valluzzi et al. (2002) 0.66 1.36
27
2. GENERAL OUTLINE OF THE AKARETLER ROW HOUSES AND
DESCRIPTION OF ORIGINAL TEST MATERIALS
In this chapter, the general characteristics of the historical Akaretler Row Houses are
presented in terms of history, plans, elevations, and soil conditions. As the
experimental studies in the thesis were performed on the materials and wallets taken
from the houses, the material physical characteristics such as texture, color and size
are also detailed in the following paragraphs.
2.1 General Outline of the Akaretler Row Houses
The Akaretler Row Houses, which were constructed by the financial support of the
Ottoman court around 1875, are the first examples of row houses in the Ottoman
Empire. These houses designed by architect Salkis Balyan, were constructed either
for the workers of Dolmabahçe Palace (Kırşan, 1996; Erenoğlu, 1998) or for
providing income for Maçka Aziziye Mosque (Sağdıç, 1999; Çakmak, 2001). The
houses are one of the best examples of the Ottoman civil architecture. The neoclassical
façades and ornaments of these houses are among their impressive
properties. The houses are located in the south part of Istanbul, adjacent to the
Marmara Sea. This area is ranked as the second degree seismic zone in Turkish
Seismic Design Code (TSDC) (2007).
The houses, which are stepped parallel to the slope of the terrain, include six blocks
with totally 133 housing units. The aerial view of the houses including six blocks,
which are A, B, C, D, E and F Blocks with similar characteristics, and the façade of
B Block can be seen in Figure 2.1 and Figure 2.2, respectively. While Block A is
servicing to private companies as offices for a while, the restoration works of the
other blocks have just been completed. They are being used as residences, offices,
shops, a hotel and a museum.
28
Figure 2.1 : The aerial view of the historical row houses.
Figure 2.2 : The façade of Block B.
Although the original structural system of these blocks were constituted with solid
brick masonry walls and vaulted slabs made of bricks and steel as shown in Figure
2.3, several vaulted slabs were replaced with reinforced concrete slabs during the
past restoration activities, Figure 2.4.
Figure 2.3 : The original vaulted slabs of Block B.
A Block
B Block
C Block
D Block
F Block
E Block
A Block
29
Figure 2.4 : The masonry walls and reinforced concrete slabs of Block C.
Almost all masonry walls have cross bond (also known as English bond) brick
arrangements with approximately 10-35 mm thick bed and 10-25 mm thick head
mortar joints, Figure 2.5. The main construction materials are stone for foundations,
and brick for the walls of the basement and upper structure and relatively small span
arches over the entrance of corridors. While stairs were originally built with masonry
or wood, they were later reconstructed using reinforced concrete. The doors and
balustrades of balconies of these masonry houses were made of cast iron.
Figure 2.5 : The mortar joints of the historical walls of the houses.
Basic geometrical characteristics of the blocks excepting with A block are given in
Table 2.1, where B is basement and E is entrance story. The structures are almost
rectangular in plan and the plan of B block can be seen in Figure 2.6. The
longitudinal section of the block is illustrated in Figure 2.7.
Bed joint
Longitudinal
joint
Head joint
30
Table 2.1 : The geometrical characteristics of the blocks.
Block
Story
number
Story height
(m)
Area of entrance
floors (m2)
Thickness of inner
walls (cm)
Thickness of
outer walls (cm)
B E+2 3.15-5.00 1309 20-70 50-120
C E+2 3.60-4.15 1231 40-70 50-85
D E+2 3.40-5.30 1676 20-60 60-80
E B+E+2 3.00-4.25 906 20-100 60-100
F B+E+2 3.00-4.10 940 10-60 60-100
Figure 2.6 : The plan of Block B (entrance level).
Figure 2.7 : The longitudinal section of Block B.
The information on the soil conditions of the Akaretler Row Houses were taken from
the report of Çoban and Günay (2006). The soil type and properties were determined
by means of in-situ tests with six boreholes and laboratory tests conducted on
disturbed/undisturbed samples and cores. The standard penetration tests indicated an
artificial fill layer with a maximum thickness of 2700 mm, which comprises of
coarse gravel and soft organic soil, as shown in Figure 2.8. There is a layer of brown
colored, weak greywacke with discontinuities below the fill layer for A, B, C, and D
blocks. The greywacke consists of fine grained and greenish-grey claystone and
siltstone. The soil of E and F blocks was defined as alluvial, which is dense brown
colored with a loose mixture of sand, gravel and silt, and clay. According to the
results of the soil investigation, allowable bearing capacities ( ) s,a  , moduli of subgrade
reaction (k) , and local site classes of the blocks obtained at the end of site and
laboratory tests are given in Table 2.2.
31
Figure 2.8 : A view of fill layer of Block C.
Table 2.2 : Soil properties of the historical houses.
Block s,a  (kN/m2) k (kN/m3) Local site class
A, B, C, D 250 35000-50000 Z2
E, F 125 27900 Z3
2.2 Description of Original Test Materials
Several structural walls of the row houses were demolished according to the
restoration project of B, C, D, E, and F blocks. Thanks to this project, large number
of different types of materials (brick, mortar, core, and wall samples) could be
obtained for laboratory and in-situ tests. While core samples included two pieces of
bricks bonded with a bed mortar joint, wall samples included at least three rows of
bricks and several bed, head and longitudinal joints. Additionally, many bricks were
collected for the construction of the specimens of masonry walls and masonry prisms
(Figure 2.9).
Figure 2.9 : The bricks collected from the historical row houses.
32
The structural walls of the historical Akaretler Row Houses were constructed with
solid clay bricks laid in mortar. Sizes, colors, and marks that may be labels of
manufacturer firms of these bricks and the thicknesses of bed and head mortar joints
showed a large variation throughout the houses.
In order to give the size ranges of bricks, the bricks were grouped according to their
labels. These groups are shown in Figure 2.10 and each group was symbolized with a
number. The ranges of average dimensions of each group were calculated
considering at least five samples, Table 2.3. The definition of the sizes is illustrated
in Figure 2.11 and the surfaces symbolized with 1 and 2 in this figure are called as
header and stretcher, respectively. The width and length of a specimen were the
average of corresponding sizes measured on the upper and the lower sides of the
header and stretcher surfaces, and the height of a specimen was determined as the
average of dimensions measured on the middle of the four sides, by taking into
account the recommendations of TS EN 772-16 (2002). As shown in Table 2.3, the
sizes of the bricks were 222-261 mm in length ( ) b l , 109-126 mm in width ( ) b b and
54-68 mm in height ( ) b h .
Figure 2.10 : The examples of the bricks collected from the historical row houses.
As seen in Figure 2.10 and Table 2.3, the bricks were not identical in terms of colors,
textures, and shapes. The differences may have resulted from the non-uniform
processes of firing and molding as well as different composition. As shown in Figure
11
1 2 3
4 5 6
7 8 9
10
33
2.12, several bricks had warpages. The colors of the bricks varied from light red to
dark red, and there were cracks and sea-shells in some bricks, Figure 2.13. The
composition of aggregates of the bricks showed a large variation in terms of the
distribution and size, as also seen in Figure 2.13. Aggregates and lumps of lime,
which are constituents of mortar, could be detected by visual inspection of mortar
samples, Figure 2.14.
Figure 2.11 : The definition of sizes of the bricks.
Table 2.3 : The range of sizes of the original bricks.
Brick b l (mm) b b (mm) b h (mm)
1 253-258 120-125 58-62
2 245-261 121-126 55-59
3 249-255 123-125 56-64
4 227-240 109-116 60-68
5 235-239 109-113 55-63
6 222-229 115-118 63-65
7 227-235 113-116 64-68
8 243-251 117-123 60-66
9 230-235 110-112 54-56
10 225-236 109-123 55-61
11 234-245 115-120 58-63
Figure 2.12 : Examples of warpages of the bricks.
hb
bb
1 lb
1
2
2
34
Figure 2.13 : The composition and colors of the bricks.
Figure 2.14 : The visual appearance of original mortar.
35
3. DETERMINATION OF MATERIAL CHARACTERISTICS OF
MASONRY CONSTITUENTS OF BRICKS AND MORTAR
The first step in the investigation of in-place masonry is to determine the basic
properties of bricks and mortar. The knowledge of in-place material characteristics of
masonry constituents, bricks, and mortar is specified to assess the compliance of
material characteristics to the relevant codes, to collect material characteristics of
constructions of any period for developing a structural material database, to follow of
variation of any structural material during history, and to produce materials similar to
in-place materials for repairing/strengthening if needed, and to utilize for predicting
of the masonry characteristics.
As masonry is generally formed by units laid in bed and head mortar joints, the
mechanical characteristics of masonry as a composite material depend on material
characteristics of bricks and mortar in addition to quality of workmanship, type of
wall construction, the thicknesses of the mortar joints, bond type, and uniformity and
regularity of size and shape of brick as mentioned by Mulligan (1942); Bayülke
(1980) and Croci (2000). Consequently, in this chapter, material tests on the brick
and mortar samples, which were collected from the removed walls of the historical
Akaretler Row Houses, are presented. The tests were carried out at Structural and
Earthquake Engineering Laboratory, Structural Laboratory and Chemical Laboratory
at Istanbul Technical University. The objectives of this chapter are as follows:
 to determine the strength and deformability characteristics of the brick and
mortar samples, which were taken from the historical houses, by conducting
flexural tension and compression tests,
 To have an idea on the quality of the historical masonry and of its constituents
by comparing them with the material properties of the present day masonry
given in several codes,
 to predict the compressive strength of the masonry of the historical houses
based on the average compressive strength of the brick and mortar samples,
36
 to determine surface hardnesses of the bricks and bed mortar joints at in-situ
by a rebound hammer to predict the masonry compressive strength,
 to determine physical properties (bulk density, water absorption and porosity)
and chemical composition of masonry constituents with the objective of
having an idea regarding the production procedure of the bricks and mortar,
In the experimental studies of the thesis, due to the wish of reflecting the physical
properties of the in-place materials on the results of mechanical tests, drying and
cooling processes, such as given in ASTM C 67-05 (2005), were not applied on
specimens taken from the houses. Consequently, the specimens were tested with their
water content. Owing to the variability of in-place material properties throughout the
houses, the samples of tests were selected randomly to be representative of in-place
materials. It should be noted that unless otherwise specified, the tests of the brick and
wall specimens were conducted in a direction normal to bed joint considering the
bricks position in masonry walls.
3.1 Mechanical Tests on Bricks
3.1.1 Flexural tests on bricks
3.1.1.1 Specimen preparation
For determining the flexural strengths of the bricks, a total of eight full-size bricks
were tested after cleaning from the remaining of dirt, mortar and other foreign
materials with wire brush in accordance with ASTM C 67-05 (2005). Since the
surfaces of the bricks were not flat, the parts of the bricks, which were to be
contacted with the steel supports and loading steel rod, were capped with a cementtype
mortar. The specimens were symbolized by BFT, which was the first capital
letters of brick and flexural, and followed by an order number. The sizes of the bricks
of length ( ) b l , width ( ) b b and height ( ) b h are given in Table 3.1.
37
Table 3.1: The brick sizes for the flexural tension tests.
Specimen b l (mm) b b (mm) b h (mm)
BFT-1 235 125 67
BFT-2 236 116 61
BFT-3 237 123 58
BFT-4 234 118 61
BFT-5 239 117 63
BFT-6 234 116 63
BFT-7 229 118 65
BFT-8 259 112 60
3.1.1.2 Test procedure
Utilizing the test configuration given in Figure 3.1, the flexural strengths of the fullsize
bricks were estimated. The flexural tests were carried out by considering the
relevant provisions of ASTM C 67-05 (2005). The Amsler testing machine with a
capacity of 1000 kN was used. Each specimen was positioned on the two steel
supports as shown in Figure 3.1 and the distance between the centers of the supports
was equal to 200 mm( ) b L . The load was applied to the midspan of each specimen by
means of a steel rod whose length was grater than the width of the specimen. The
maximum load resisted by the specimen ( ) ft ,m P was determined from the readings of
a ring, which had a capacity of 5.0 kN, located on the rod. According to ASTM C 67-
05 (2005), the loading rate should be smaller than 8896 N/min.
Figure 3.1 : The test setup with measurement system for the brick flexural tests.
Upper plate
Ring
Rod
Specimen
Supports
Lb=200 mm
Pft
hb
bb
lb
Pft: the load applied during
the flexural tension test
38
3.1.1.3 Test results
The flexural tensile strengths of the specimens or moduli of rupture called by ASTM
C 67-05 (2005) were calculated using Eq. (3.1):
2
1.5 ,
b b
ft m b
bft b h
P L
f  (3.1)
where ( ) ft P is the maximum load obtained from the readings of the ring; b l , b b and
b h are the width, length and height of the specimens at the plane of failure, and ( ) b L
is the distance between the supports. The flexural tensile strengths of the bricks
( ) bft f are given in Table 3.2. The duration of the tests was generally less than 1 min.
Consequently, the loading rates applied to the specimens conformed to the proposal
of ASTM C 67-05 (2005) given above.
The average flexural tensile strength of bricks was 1.35 MPa with a standard
deviation of 0.31 MPa and with a coefficient of variation of 0.23, Table 3.3. In this
table, standard deviation and coefficient of variation were represented with Stdev and
CoV, respectively.
Table 3.2: The flexural tensile strengths of the bricks.
Specimen ft m P , (N) bft f (MPa)
BFT-1 2023.0 1.08
BFT-2 1453.0 1.01
BFT-3 2498.0 1.81
BFT-4 2450.5 1.67
BFT-5 2165.5 1.40
BFT-6 2165.5 1.41
BFT-7 1785.5 1.07
BFT-8 3163.0 --
Table 3.3 : The statistical parameters of the brick flexural tension test results.
Statistical parameter bft f
Minimum (MPa) 1.01
Maximum (MPa) 1.81
Average (MPa) 1.35
Stdev (MPa) 0.31
CoV 0.23
39
The failure of the bricks took place suddenly in a manner of brittle. While the failure
of BFT-2 and BFT-4 formed due to the formation of a crack inclined from midspan
to near one of the supports, the failure of other specimens formed due to an
approximately vertical crack at midspan. The failures of BFT-1 and BFT-4
specimens are illustrated in Figure 3.2.
Figure 3.2 : The failures of BFT-1 and BFT-4 bricks under flexural effect.
3.1.2 Compression tests on bricks
3.1.2.1 Specimen preparation
The preparation works of the brick specimens of the compression tests were as
follows: Firstly, the bricks collected were cleaned from the remaining of mortar and
dirt. Secondly, the broken bricks out of the bricks collected were selected for the
compression tests. Thirdly, the broken bricks were cut parallel to their widths in a
way to obtain straight specimens. Then, if required, the other uneven surfaces of the
specimens were straightened. Lastly, the upper and lower surfaces of the bricks were
capped with a type of cement mortar before compression tests to provide parallel
loading surfaces (ASTM C 67-05, 2005 and TS EN 772-1, 2002). The tests were
carried out at least 2 days after capping.
Additionally, small areas were formed with a type of epoxy-based glue on the
surfaces of the specimens to be attached strain gauges (epoxy layer of about 40×40
mm) as the surfaces were not smooth. After at least one day, these areas were rubbed
with pieces of emery paper.
The sizes of the specimens are given in Table 3.4. The heights of the bricks included
the thicknesses of the caps. The brick specimens were denoted by BC, which is the
initial capital letters of brick and compression, respectively, and the specimen
number. Generally, the lengths of almost all the specimens were larger than the
40
halves of the corresponding full-size bricks, the widths of the specimens were
smaller than ones and the heights of the specimens were approximately equal to the
corresponding size.
While ASTM C 67-05 (2005) recommends the use of half brick samples in
compression tests, which should have full height and width of units and a length
which is the half of length of units; TS EN 772-1 (2002) recommends the use of
bricks having at least a height of 40 mm after cutting for obtaining flat surfaces or if
the heights of bricks are smaller than 40 mm, the standard proposes the use of
composite specimens which are formed with bricks in stack bond without using any
type of mortar or glue between bricks. It is seen that the heights of the specimens
meet the proposal of ASTM C 67-05 (2005) and TS EN 772-1 (2002).
Table 3.4: The brick sizes for the compression tests.
Specimen b l (mm) b b (mm) b h (mm)
BC-1 160 100 50
BC-2 106 81 55
BC-3 125 105 45
BC-4 135 105 50
BC-5 135 107 55
BC-6 180 104 50
BC-7 120 99 52
BC-8 199 85 49
BC-9 145 76 48
BC-10 135 107 60
BC-11 130 126 79
BC-12 114 111 72
BC-13 120 119 71
BC-14 125 123 75
BC-15 127 123 72
BC-16 112 111 76
BC-17 118 112 76
BC-18 120 112 79
BC-19 119 108 70
BC-20 161 124 83
BC-21 120 109 78
BC-22 113 109 72
BC-23 114 111 72
BC-24 111 111 75
BC-25 125 118 72
3.1.2.2 Test procedure
A total of 25 brick specimens were tested under monotonic increasing compression
loads, using the Amsler testing machines with 1000 or 5000 kN load capacities in an
appropriate loading range. The testing machine and the test setup with measurement
41
systems can be seen in Figure 3.3. The upper loading plate of the testing machine has
a spherically seated hinge, which decreases the influence of non-parallel loading
surfaces of the specimens on the test results. A load cell with a capacity of 200 kN
and two linear variable displacement transducers (LVDT) with a capacity of 25 mm
(TML CDP-25) were utilized to record load applied and measure displacement
taking place, respectively. The data obtained from the load cell and LVDTs were
stored by means of a data logger. As shown in Figure 3.3, the LVDTs were
positioned on the opposite sides between the lower and upper plates of the test
machine. The specimens to be tested were put on the lower loading plates and the
centers of the specimens were coincided with those of the upper/lower loading plates.
Figure 3.3 : The test setup with measurement system for the brick compression.
In addition to the LVDTs, to observe the differences to be resulted from the distinct
measurement techniques and from locations where measurements taken, strain
gauges (TML PFL-30-11-3L) were used to obtain displacement readings. These
strain gauges (Strg) were located in a centered position on the surfaces of several
specimens and attached to the specimens using the epoxy-based glue. For each
specimen, a total of maximum eight strain gauges with a capacity of 30 mm were
fixed in vertical and horizontal directions (Figure 3.3).
The rate of load applied is described in ASTM C 67-05 (2005) and TS EN 772-1
(2002). According to these standards, the rate of load should be appropriate until half
of the expected maximum load; then, the maximum load should be reached in
LVDT1 LVDT2
Load cell
Upper plate
Lower plate
Specimen
Strg-1-2
Strg-3-4 Strg-7-8
Strg-5-6
42
between 1 and 2 min for ASTM C 67-05 (2005), at least in 1 min for TS EN 772-1
(2002). Although, the test machines used do not have any rate control system, by
taking into consideration the recommends of the standards mentioned, the rate of
loading was adjusted by manual.
3.1.2.3 Test results
The behavior of the brick specimens in pre-peak and post-peak regions under
compression loads is characterized with the relationships of compressive stressstrain,
the compressive strengths, the moduli of elasticity, the ductility values, and
the failure modes of the specimens. These material characteristics, which are defined
in the following paragraphs, were denoted by the corresponding symbol and
subscripts, which were the initial small letters of material and test type. For example,
bc f represents the strength of brick under compression loads.
Compressive stress ( ) bc  is defined as the ratio of the compressive load ( ) c P on the
specimen tested to the initial cross section area of the specimen ( ) o A Eq. (3.2).
Compressive strength ( ) bc f is taken as the maximum value of the compressive
stresses.
o
c
bc A
P
  (3.2)
Compressive strain ( ) v,bc  is the ratio of the average change in height of the specimen
tested, which were recorded by the LVDTs, to the initial height of the specimen or
the ratio of the average change in gauge length (l) to the initial gauge length ( ) o l .
o
v bc l
l
 , 
(3.3)
The modulus of elasticity ( ) bc E , which is one of the deformation parameters of the
material, is determined by taking into account the compliance of the material to
Hooke’s law. The modulus is termed as Young’s modulus according to ASTM E
111-04 (2004). Young’s modulus, which is determined using the least squares
method in this standard, is the slope of the linear branch of stress-strain relationship
below the proportional limit.
43
In this study, ductility ( ) bc  was calculated as the ratio of the compressive strain at
the 85 percent of strength level in the post-peak region ( ) v,bc,0.85 f  to the compressive
strain at strength level of the specimen ( ) v,bc, f  Eq. (3.4). In order to clarify the
expressions mentioned in the definition of ductility can be seen in Figure 3.4.
v bc f
v bc f
bc
, ,
, ,0.85


  (3.4)
Vertical strain
Compressive stress
fbc,f
0.85fbc,f
v,bc,f v,bc,0.85f
Figure 3.4 : The ductility definition.
The elimination of the results showing high deviations with respect to the
corresponding average value of the results was realized according to Chauvenet
Criterion (Akman, 1978). To apply the criterion to the test results, the steps followed
are: Firstly, the average and the standard deviation of the test results are determined.
Secondly, each test result is subtracted from the average. Thirdly, by taking the
absolute of the value obtained in the second step, this value is divided by the
standard deviation. Lastly, the value calculated in the third step is compared with the
corresponding allowable value given in Table 3.5 and if the value is larger than the
allowable value, the test result is eliminated. After the elimination, the values of the
average and standard deviation are calculated over again. As shown in this table, the
44
allowable values are given with a constant value, depending on the number of the
specimen tested. The criterion permits the use of linear interpolation. In this study,
the results of the test specimens eliminated according to this criterion are symbolized
with "--" in relevant tables.
Table 3.5 : The allowable values according to Chauvenet Criterion.
Specimen number Allowable value
2 1.15
3 1.38
4 1.54
5 1.65
6 1.73
7 1.80
10 1.96
15 2.13
25 2.33
50 2.57
100 2.81
300 3.14
500 3.29
1000 3.48
Using the readings of the load cell and of the LVDTs, the responses of the specimens
during the tests were characterized with the diagrams of compressive stress and
compressive strain, Figure 3.5. While the compressive stress was calculated using
Eq. ( 3.2), the compressive strain was estimated using Eq. ( 3.3). The compressive
strain was taken as the average compressive strains calculated from the two LVDTs
readings. The diagrams given in Figure 3.5 can be expressed with several mechanical
parameters thought to be display the behavior of the specimens. These parameters,
which were compressive strength ( ) bc f and corresponding strain ( ) v,bc, f ε ,
compressive stress at proportional limit ( ) bc, p σ and corresponding strain ( ) v,bc, p ε ,
Young’s modulus ( ) bc E and ductility ( ) bc μ , are tabulated, Table 3.6. This table
demonstrates that the parameters took values in a wide range. In order to express this
variation quantitatively, the statistical assessment of the parameters were realized,
Table 3.7. The average values obtained are 5.5 MPa with a CoV of 0.55 for the
compressive strengths, 5.2% with a CoV of 0.34 for the strains at the compressive
45
strength levels, 3.5 MPa with a CoV of 0.53 for the compressive stresses at the
proportional limits, 2.4% with a CoV of 0.39 for the strains at the proportional limits,
150 MPa with a CoV of 0.66 for Young’s moduli and 1.4 with a CoV of 0.15 for the
ductilities. The high deviation may be resulted from the differences between the test
specimen characteristics. As mentioned in Chapter 2, the bricks had distinct
characteristics observable visually, such as size, color, and aggregates size.
0
2
4
6
8
10
12
14
0.0 0.1 0.2 0.3
Compressive strain
Compressive stress (MPa)
Figure 3.5 : The compressive stress-compressive strain relationships for the bricks.
To investigate whether the variation in the value of a variable (strength and
deformation parameters) may be related with another variable(s), if related, to find
out the function expressing the relation between these variables or at least, to
understand the trend of the variation; correlation and simple regression analyses were
conducted on the test data. While regression analysis calibrates the unknown
coefficients of a prediction equation, correlation analysis assures a measure of
goodness of fit between the prediction equation and the data sample (Ayyub and
McCuen, 1996). The steps followed for these analyses were the graphical
presentation of the variables, the calculation of the correlation coefficient (R) , the
determination of the function expressing the relation between the variables and the
calculation of the coefficient of determination, ( ) 2 R . These steps were realized by
46
means of Microsoft Office Excel 2003. The following statements were collected
from the study of Ayyub and McCuen (1996): The graphical presentations of the
variables (Y and X ) are required to determine the relation characteristics
(linear/nonlinear and direct/indirect). The value of the correlation coefficient shows if
the regression provides accurate predictions or not. While the values of 1 and -1 of
R define a perfect relation, the value of zero indicates no relation. The coefficient of
determination is the percentage of the variance in the criterion variable (Y) that is
explained by the predictor variable (X ) .
Table 3.6 : The results of the brick compression tests.
Specimen bc f (MPa) v,bc, f  (%) bc, p  (MPa) v,bc, p  (%) bc E (MPa) bc 
BC-1 9.84 8.97 5.84 3.40 172 1.5
BC-2 5.70 4.77 4.52 2.50 181 2.0
BC-3 7.86 6.85 5.98 4.14 144 1.5
BC-4 5.52 6.74 2.39 1.78 135 1.5
BC-5 2.04 5.57 1.32 2.19 60 1.8
BC-6 2.01 8.62 1.14 3.32 34 1.4
BC-7 4.50 6.45 3.42 3.77 91 1.4
BC-8 7.04 7.13 5.73 4.11 140 1.4
BC-9 1.88 6.27 1.52 2.74 56 1.8
BC-10 3.94 3.17 3.13 1.60 196 --
BC-11 11.99 3.75 7.58 1.88 404 1.6
BC-12 3.87 4.21 2.32 1.99 117 1.5
BC-13 5.09 7.29 1.93 1.88 103 1.4
BC-14 4.65 5.55 2.91 2.20 132 1.3
BC-15 9.45 3.12 5.52 1.28 430 1.3
BC-16 9.30 2.83 5.90 1.20 -- 1.4
BC-17 3.24 4.42 2.11 1.48 142 1.3
BC-18 5.14 6.15 1.80 1.28 141 1.4
BC-19 5.26 6.07 4.14 4.18 99 1.4
BC-20 11.82 4.41 5.67 1.95 291 1.2
BC-21 4.50 4.41 3.72 3.22 115 1.2
BC-22 2.07 3.57 1.68 2.09 80 1.3
BC-23 1.78 3.17 1.34 1.83 73 1.2
BC-24 2.61 3.98 2.18 2.95 74 1.2
BC-25 5.33 3.45 4.07 2.11 193 1.2
47
Table 3.7 : The statistical parameters of the brick compression tests.
Statistical parameter Minimum Maximum Average Stdev CoV
bc f (MPa) 1.78 11.99 5.5 3.05 0.55
v,bc, f  (%) 2.83 8.97 5.2 1.76 0.34
bc, p  (MPa) 1.14 7.58 3.5 1.87 0.53
v,bc, p  (%) 1.20 4.18 2.4 0.94 0.39
bc E (MPa) 34 430 150 99 0.66
bc  1.2 2.0 1.4 0.21 0.15
The relations obtained from the brick compression tests were shown in Figure 3.6 for
Young’s modulus-compressive strength, in Figure 3.7 for compressive stress at
proportional limit and in Figure 3.8 for Young's modulus-secant modulus at peak
relationships. The expressions of these relationships are as follows:
bc bc E  28 f (3.5)
bc p bc 0.6 f ,   (3.6)
bc bc s
v bc f
bc
v bc p
bc p E E
f
,
, , , ,
, 1.4   1.4
 


 



 

(3.7)
In Eq. (3.7), bc s E , is the secant modulus at peak of the bricks.
Although 15 specimens were instrumented with the strain gauges in horizontal and
vertical directions, reliable data can be obtained from only five specimens. Figure
3.10 displays the compressive stress-compressive strain diagrams obtained from the
mean displacement readings of LVDTs and of strain gauges. The analysis of these
figures indicates that the strain gauges could not be able to work to end of the tests
and that the slopes of the curves of the strain gauges which were denoted with Strg
are steeper than those of the LVDTs, Table 3.8. This may be ascribed to the strain
gauge readings local, to the difference in the stiffnesses of the bricks and of the
epoxy layer on the surfaces of the specimens and to the thicknesses of the epoxy
layers. Generally, the ratio between Young’s moduli decreases as corresponding
compressive strength increases, Table 3.8.
48
Ebc = 28.071fbc
R2 = 0.678
0
100
200
300
400
500
0 2 4 6 8 10 12 14
Compressive strength (MPa)
Young's modulus (MPa)
Figure 3.6 : The Young’s modulus-compressive strength relationship for the bricks.
bc,p = 0.6245fbc
R2 = 0.818
0
2
4
6
8
0 2 4 6 8 10 12 14
Compressive strength (MPa)
Compressive stress at proportional limit (MPa)
Figure 3.7 : The compressive stress at proportional limit-compressive strength
relationship for the bricks.
49
Ebc= 1.3586Ebc,s
R2 = 0.9532
0
100
200
300
400
500
600
0 100 200 300 400
Secant modulus at peak (MPa)
Young's modulus (MPa)
Figure 3.8 : The Young's modulus-secant modulus at peak relationship for the
bricks.
Table 3.8 : The comparison of Young’s moduli for the bricks (LVDTs and strain
gages).
Specimen f bc (MPa) Ebc (MPa) bc strg E , (MPa) bc strg bc E E ,
BC-11 11.99 404 993 2.5
BC-12 3.87 117 430 3.7
BC-13 5.09 103 323 3.1
BC-14 4.65 132 429 3.3
BC-15 9.45 430 829 1.9
BC-22 2.07 80 260 3.3
During the tests, the formations and progresses of cracks or separations under the
loading were observed. Before failure, vertical cracks and/or face shell separations
initiated, while compression stresses were approaching the compressive strength and
afterwards the specimens generally failed by crushing as shown in Figure 3.9.
Figure 3.9 : The failure of BC-2 under compression loads.
50
Figure 3.10 : The comparison of the compressive stress-compressive strain
relationships for the bricks (LVDTs and strain gages).
0
2
4
6
8
10
12
0.00 0.05 0.10 0.15 0.20
Vertical strain
Compressive stress (MPa)
BC-12
BC-12-Strg
0
2
4
6
8
10
12
0.00 0.05 0.10 0.15 0.20
Vertical strain
Compressive stress (MPa)
BC-13
BC-13-Strg
0
2
4
6
8
10
12
0.00 0.05 0.10 0.15 0.20
Vertical strain
Compressive stress (MPa)
BC-14
BC-14-Strg
0
2
4
6
8
10
12
0.00 0.05 0.10 0.15 0.20
Vertical strain
Compressive stress (MPa)
BC-15
BC-15-Strg
0
2
4
6
8
10
12
0.00 0.05 0.10 0.15 0.20
Vertical strain
Compressive stress (MPa)
BC-22
BC-22-Strg
0
2
4
6
8
10
12
0.00 0.05 0.10 0.15 0.20
Vertical strain
Compressive stress (MPa)
BC-11
BC-11-Strg
Compressive strain Compressive strain
Compressive strain Compressive strain
Compressive strain Compressive strain
51
3.1.3 Compression tests on brick specimens in parallel to bed joint
Bricks can be display different material characteristics depending on loading
direction, namely anisotropy. This might be attributed to the production process of
the bricks; pressing them in one direction and non-uniformity in firing conditions. In
order to clarify this state for the bricks under the consideration, compression tests
were realized on the halves of five full-size bricks in the directions of parallel and
normal to bed joint.
The specimen preparation and test procedure adopted were the same as those of the
bricks specimens tested in normal to the bed joint, aforementioned. While the
specimens tested in the normal direction were BC-21, BC-22, BC-23, BC-24 and
BC-25, whose results were given above; the specimens tested in the parallel direction
were BCp-21, BCp-22, BCp-23, BCp-24 and BCp-25. It should be noted that two
halves of a brick were signed with same number and that while one half was tested in
the normal direction; the other was tested in the parallel direction. The geometrical
description of the specimens tested in the parallel direction and sizes are presented in
Figure 3.11 and Table 3.9, respectively.
Figure 3.11 : The description of the brick specimens tested in parallel to bed joint.
Loading in parallel
to bed joint
Loading in normal
to bed joint
hb,p
lb,p
bb,p
lb
hb
bb
52
Table 3.9: The brick specimen sizes for the compression tests.
Specimen b p l , (mm) b p b , (mm) b p h , (mm)
BCp-21 124 66 131
BCp-22 116 63 135
BCp-23 116 61 125
BCp-24 109 64 132
BCp-25 120 62 141
The compressive strengths of the brick specimens obtained and their statistical
assessment are tabulated in Table 3.10 and Table 3.11. These tables show the
influence of the loading direction on the strength. According to the test results, the
compressive strengths of the specimens tested in parallel to bed joint are smaller than
in normal to bed joint. As shown in Table 3.10, the values of anisotropy are in a
range of 1.1-2.4 with an average of 1.9.
Table 3.10 : The compressive strengths of the bricks tested in parallel to the bed
joint and anisotropy ratios.
Specimen bc p f , (MPa) Anisotropy
BCp-21 2.35 1.9
BCp-22 1.85 1.1
BCp-23 0.85 2.1
BCp-24 1.40 1.9
BCp-25 2.20 2.4
Table 3.11 : The statistical parameters of the compression tests on the bricks tested
in parallel to the bed joint.
Statistical parameter bc p f ,
Minimum (MPa) 0.85
Maximum (MPa) 2.35
Average (MPa) 1.7
Stdev (MPa) 0.61
CoV 0.36
Figure 3.12 illustrates the views of several specimens, which were subjected to loads
parallel to the bed joint, after the compression tests. Generally, conical type failures
were observed as a result of several face shell separations following cracks.
53
Figure 3.12 : The failure of the bricks compressed in direction parallel to the bed
joint.
3.1.4 Compression tests on three-brick specimens
3.1.4.1 Specimen preparation
In order to figure out the friction affects, which can result from the small heights of
the brick specimens, between the brick specimens and the lower/upper loading plates
on the compression test results, the compression tests of the three-brick specimens
were carried out.
The three-brick specimens were built with three bricks, which were cut to obtain flat
surfaces, bonded with a thin layer of a high strength cement-type mortar in stacked
bond, Figure 3.13. The thicknesses of bonding were kept thin. The upper and lower
surfaces of the specimens were also capped with the mortar to provide parallel
loading surfaces, (ASTM C67-05, 2005 and TS EN 772-1, 2002). The tests were
carried out at least 2 days after the capping.
Fifteen three-brick specimens were tested under compression loads. The specimens
were denoted by TBC, which indicates the initial capital letters of three, brick, and
compression, respectively, and a specimen number. The height-to-thickness ratios of
the specimens were between 1.3 and 1.7. The thickness is defined as the least lateral
dimension (width or length) of the specimen. The average area of surfaces
perpendicular to the direction of loading is calculated using the average sizes of the
three bricks while calculating the compressive strengths of each specimen, Table
3.12.
54
Figure 3.13 : The description of the three-brick specimens.
Table 3.12: The three-brick specimen sizes for the compression tests.
Specimen tb l (mm) tb b (mm) tb h (mm)
TBC-1 132 116 150
TBC-2 110 108 174
TBC-3 117 110 174
TBC-4 107 99 165
TBC-5 131 122 155
TBC-6 118 113 160
TBC-7 112 109 165
TBC-8 151 105 146
TBC-9 107 92 156
TBC-10 120 98 155
TBC-11 114 101 154
TBC-12 128 97 142
TBC-13 148 97 153
TBC-14 144 109 157
TBC-15 129 103 160
3.1.4.2 Test procedure
The test procedure adopted is the same as the procedure followed for the brick
specimens. The Amsler testing machine described above was used to apply the
ltb
htb
btb
Front view Side view
Brick
Brick
Brick
Cap
Cap
Brick
Brick
Brick
55
compression loads to the specimens. A load cell with a capacity of 200 kN and two
LVDTs with a capacity of 25 mm were used to obtain necessary data for the
characterization of the specimen behavior. The readings of the load cell and LVDTs
were stored by TDS 303 data logger. The test setup and measurement systems of the
compression tests on the three-brick specimens are illustrated in Figure 3.14. The
specimens were positioned on a plate above the load cell that was located on the
centre of the lower plate and the LVDTs were located between the upper and lower
plates.
Figure 3.14 : The test setup with measurement system for the three-brick
compression tests.
3.1.4.3 Test results
The graphical illustration of the test data is presented for each specimen in Figure
3.15 in terms of the compressive stress and compressive strain relationships. The
compressive stress was obtained by dividing the readings of the load cell to the initial
cross section area of the corresponding specimen, Eq. (3.2). Compressive strain was
calculated as the ratio of the average value of the readings of the LVDTs to the initial
height of the corresponding specimen, Eq. (3.3). These curves showed the responses
of the specimens to the compression loads applied during the tests. In order to
emphasize several mechanical parameters obtained from the graphical analyses,
Table 3.13 was arranged. The dispersion of the parameters describing the
compressive behavior is presented in Table 3.14. As seen in this table, the mean
values are 2.3 MPa with a CoV of 0.41 for the compressive strength ( ) tbc f , 1.7%
with a CoV of 0.16 for the compressive strain at compressive strength ( ) v,tbc, f  , 1.5
MPa with a CoV of 0.41 for the compressive stress at the proportional limit ( ) tbc, p  ,
LVDT1 LVDT2
Load cell
Upper plate
Lower plate
Specimen
56
0.8% with a CoV of 0.24 for the compressive strain at the limit ( ) v,tbc, p  , 192 MPa
with a CoV of 0.34 for Young’s modulus ( ) tbc E and 1.5 with a CoV of 0.10 for the
ductility ( ) tbc  .
Figure 3.15 : The compressive stress-compressive strain relationships for the threebrick
specimens.
The difference between the test results of the brick and the three-brick specimens
make clarify the effect of confinement, due to friction between the loading plates and
specimens, on the compression test results. Especially, the fact that the average
strength of the brick specimens (5.5 MPa) is greater than the strength of the threebrick
specimens (2.3 MPa) indicates the magnitude of the confinement influence.
Moreover, that the statistical parameters of the three-brick specimens exhibited less
deviation with respect to the values of the brick specimens may be one of the
indicators of the confinement taking place in the tests of the brick specimens.
0
1
2
3
4
0.00 0.05 0.10
Compressive strain
Compressive stress (MPa)
57
Table 3.13 : The results of the three-brick compression tests.
Specimen ftbc (MPa) v,tbc, f  (%) tbc, p  (MPa) v,tbc, p  (%) tbc E (MPa) tbc 
TBC-1 4.66 1.61 3.08 0.57 -- 1.5
TBC-2 1.40 1.39 1.09 0.79 138 1.8
TBC-3 1.45 1.59 0.89 0.46 191 1.4
TBC-4 3.28 1.60 2.29 0.71 322 1.5
TBC-5 -- 1.83 -- 0.91 -- 1.4
TBC-6 2.86 1.60 1.79 0.61 292 1.6
TBC-7 3.43 1.75 2.29 0.81 282 1.4
TBC-8 1.56 2.29 0.95 0.70 135 1.5
TBC-9 1.84 1.62 1.30 0.75 173 1.5
TBC-10 1.69 1.49 1.12 0.67 167 1.8
TBC-11 2.27 -- 1.37 1.16 118 1.4
TBC-12 1.90 1.53 1.30 0.65 199 1.8
TBC-13 1.80 1.94 1.14 0.58 196 1.4
TBC-14 1.89 2.26 1.39 1.02 136 1.5
TBC-15 1.90 1.89 1.44 0.95 152 1.5
Table 3.14 : The statistical parameters of the three-brick compression tests.
Statistical parameter Minimum Maximum Average Stdev CoV
tbc f (MPa) 1.40 4.70 2.3 0.94 0.41
v,tbc, f  (%) 1.39 2.29 1.7 0.27 0.16
tbc, p  (MPa) 0.89 3.08 1.5 0.62 0.41
v,tbc, p  (%) 0.46 1.16 0.8 0.19 0.24
tbc E (MPa) 118 322 192 66 0.34
tbc  1.4 1.8 1.5 0.15 0.10
Simple regression analyses were performed to establish relationships between the
parameters given in Table 3.13. The relationship between the variables of
compressive stress and compressive strain was introduced by normalizing
compressive stress and compressive strain data of each specimen with the
corresponding compressive strength and the compressive strain at the strength,
respectively. After normalizing the compressive stress and compressive strain values
of all specimens, the regression analysis was conducted on a range of the normalized
data points of 0.8 ,  tbc n  in the descending branch and 1.5 , ,  v tbc n  thought to be a
reliable data range to find out the mathematical expression of the relationship, Figure
58
3.16. The function of the three-brick specimens is obtained with a high coefficient of
determination of 0.99 2 R  , as follows:
tbc n v tbc n v,tbc,n
2
, , ,   0.96 1.96 (3.8)
The equation indicates that the form of the relationship between the variables of the
compressive stress and the compressive strain can be explained with the parabolic
function, Eq. (3.8).
tbc,n = -0.9601v,tbc,n
2 + 1.9654v,tbc,n
R2 = 0.9883
0.0
0.2
0.4
0.6
0.8
1.0
1.2
0.0 0.5 1.0 1.5
Normalized compressive strain
Normalized compressive stress
Figure 3.16 : The normalized compressive stress-normalized compressive strain
relationship for the three-brick specimens.
Whether Young’s modulus was associated with compressive strength was
investigated, Figure 3.17. The relation might be stated with the following linear
relation:
tbc tbc E  90 f (3.9)
59
Etbc = 90.768ftbc
R2 = 0.664
0
100
200
300
400
0 1 2 3 4 5
Compressive strength (MPa)
Young's modulus (MPa)
Figure 3.17 : The Young’s modulus-compressive strength relationship for the threebrick
specimens.
By conducting regression analysis on the pairs of compressive strength and
corresponding stress at proportional limit, the fact that the type of the relationship
between these variables is linear was figured out, Figure 3.18. The expression of the
relation can be given by the following equation:
tbc p tbc 0.7 f ,   (3.10)
This equation paraphrases that the behavior of the three-brick specimens exhibits a
non-linear characteristic under compressive stress larger than almost 70 percent of
the compressive strength.
To point out the non-linearity taking place between the end points of compressive
stress at proportional limit and compressive strength, these points and corresponding
strains were correlated with the following function as shown in Figure 3.20.
tbc s tbc
v tbc f
tbc
v tbc p
tbc p E E
f
1.7 1.7 ,
, , , ,
,  
 


 



 

(3.11)
60
tbc,p = 0.6689ftbc
R2 = 0.9721
0
1
2
3
4
0 1 2 3 4 5
Compressive strength (MPa)
Compressive stress at proportional limit (MPa)
Figure 3.18 : The compressive stress at proportional limit-compressive strength
relationship for the three-brick specimens.
If the right side of the function can be esteemed as Young’s modulus ( ) tbc E , it is seen
that Young’s modulus might be taken as about 1.7 times the secant modulus at
peak ( ) tbc,s E . It is possible to derive that there is about a reduction of 40% in the
secant modulus (at peak) with respect to Young's modulus.
The three-brick specimens generally failed in a conical break as the result of vertical
cracks and face shell separation, as shown in Figure 3.19.
Figure 3.19 : The failure mechanism of TBC-14.
61
Etbc = 1.6818Etbc,s
R2 = 0.924
0
100
200
300
400
500
600
0 100 200 300 400
Secant modulus at peak (MPa)
Young's modulus (MPa
Figure 3.20 : The Young's modulus-secant modulus at peak relationship for the
three-brick specimens.
3.1.5 Rebound hammer tests on the bricks
The rebound hammer test method was recommended by FEMA 356 (2000) as a
supplementary test for investigating the surface hardness of masonry during material
characterization of existing masonry structures. This method is a non-destructive
method and has been used in concrete for long years.
In order to measure the surface hardness of the bricks, rebound hammer was applied
on the faces of bricks on site after plaster was removed. Additionally, for examining
the possible difference between rebound hammer test results on narrow (100-130
mm) and wide sides (190-260 mm) of the bricks, the tests were done on both sides of
the bricks. For these objectives, a total of 32 bricks were tested by the hammer.
Size and average rebound numbers of each brick are presented in Table 3.15 and
Table 3.16 for narrow and wide sides of the bricks, respectively. The brick
specimens were symbolized with BR, which was the first capital letters of brick and
rebound, and the number of bricks tested. Average rebound number for each brick
( ) br N was calculated as the average of ten readings.
62
Table 3.15 : The rebound numbers of the narrow side of the bricks.
Specimen b l or b b (mm) b h br N
BR-1 100 52 37
BR-2 115 60 25
BR-3 120 55 36
BR-4 120 57 34
BR-5 120 60 32
BR-6 120 60 31
BR-7 120 60 39
BR-8 120 60 42
BR-9 120 68 24
BR-10 120 67 34
BR-11 120 60 24
BR-12 125 60 27
BR-13 125 60 31
BR-14 130 60 30
BR-15 130 60 38
Table 3.16 : The rebound numbers of the wide sides of the bricks.
Specimen b l or b b (mm) b h br N
BR-16 190 60 19
BR-17 195 60 21
BR-18 205 60 36
BR-19 220 45 28
BR-20 225 70 34
BR-21 225 65 23
BR-22 235 70 23
BR-23 235 60 21
BR-24 235 60 25
BR-25 240 55 25
BR-26 240 58 35
BR-27 240 55 31
BR-28 240 60 25
BR-29 240 60 29
BR-30 240 55 27
BR-31 250 55 27
BR-32 260 60 --
While the average rebound number for narrow sides, whose rebound numbers varied
between 24 and 42, was determined as 32 with a standard deviation of 5.61; the
63
average value for wide sides, whose rebound numbers varied between 19 and 36, was
determined as 27 with a standard deviation of 5.13, Table 3.17. The coefficients of
variations for the narrow and wide sides were estimated as about 0.2.
Table 3.17 : The statistical parameters of the brick rebound number.
Statistical parameter Minimum Maximum Average Stdev CoV
Narrow side 24 42 32 5.61 0.17
Wide side 19 36 27 5.13 0.19
According to the results, the ratio of the average number of narrow sides to that of
wide sides was determined as approximately 1.2. When the number results obtained
were evaluated as total, the average rebound number was equal to 30.
3.2 Mechanical Tests on Mortar
3.2.1 Flexural tests on mortar
3.2.1.1 Specimen preparation
Taking appropriate mortar specimens from bed and head joints was not easy since
the thicknesses of the bed and head joints varied 10-35 mm and 10-25 mm,
respectively. Consequently, the specimens were generally taken from longitudinal
mortar joints having higher thicknesses relative to the bed and head mortar joints.
After the specimens were cleaned, the surfaces of the specimens were cut to get flat
surfaces. Five mortar specimens were tested to obtain the flexural tensile strength of
in-place mortar. The number of the flexural test specimens was low compared with
the specimen numbers of the other tests. There were two reasons of the low specimen
number. First reason was that taking appropriate mortar specimens from walls was
not easy due to the reason mentioned above. The second reason was that some
specimens were destroyed while the specimens were being cut.
The specimens were symbolized by MF, which was the first capital letters of mortar
and flexural, and followed by the order number of the specimen. The sizes of the
specimens of width ( ) m b and height ( ) m h are given in Table 3.18. TS EN 1015-11
(2000) and ASTM C 348-02 (2002) propose the use of mortar specimens having
160x40x40 mm.
64
However, as shown in Table 3.18, the dimensions of the specimens did not provide
this proposal, as the thicknesses of the in-place mortar joint were not suitable to
obtain larger specimens. It should be also noted that these codes are for molded
hardened mortar specimens.
Table 3.18: The mortar specimen sizes for the flexural tension tests.
Specimen m b (mm) m h (mm)
MF-1 27 34
MF-2 47 35
MF-3 35 35
MF-4 29 29
MF-5 38 35
3.2.1.2 Test procedure
The flexural tensile strength of the mortar specimens were determined taking account
the procedure given in TS EN 1015-11 (2000) and ASTM C 348-02 (2002). Threepoint
bending test configuration shown in Figure 3.21 was utilized to estimate the
flexural strength of the mortar specimens.
Figure 3.21 : The test setup with measurement system for the mortar flexural tests.
Each specimen was located on the two steel supports whose distance between the
centers of them equal to 100 mm( ) m L in accordance with TS EN 1015-11 (2000). The
load applied was transmitted to the midspan of each specimen by means of a steel
rod whose length was grater than the width of the specimen. The peak load resisted
by the specimen ( ) ft ,m P was estimated from the readings of a ring on the rod.
Ring
Specimen
Upper plate
Supports
Rod
65
3.2.1.3 Test results
The peak loads recorded and the flexural tensile strengths ( ) mft f determined from Eq.
(3.1) are presented in Table 3.19.
Table 3.19: The results of the mortar flexural tension tests.
Specimen ft m P , (N) mft f (MPa)
MF-1 265.5 1.28
MF-2 403.3 1.05
MF-3 336.3 1.18
MF-4 218.0 1.34
MF-5 479.3 1.54
While the average flexural tensile strength of the specimens was calculated as 1.28
MPa, the standard deviation and the coefficient of variation of the specimens were
0.18 MPa and 0.14, respectively, as given in Table 3.20.
Table 3.20 : The statistical parameters of the mortar flexural tension tests.
Statistical parameter mft f
Minimum (MPa) 1.05
Maximum (MPa) 1.54
Average (MPa) 1.28
Stdev (MPa) 0.18
CoV 0.14
The statements concerning statistical assessment in ASTM C 348-02 (2002) might be
used to evaluate the dispersion of the flexural strengths obtained. The statement is
that the coefficient of variation of the test results of three prisms, which were molded
from a single batch of mortar and tested at the same age, is 8.4 and 5.1% for the
multilaboratory and singlelaboratory, respectively. The CoV of the tests conducted is
larger than the values of CoV given in ASTM C 348-02 (2002). However, taking into
account the fact that while the specimen of the code is molded from a single batch,
the test specimens of this study were samples of the historical mortars; the CoV of
the tests in the study (14%) might be deemed as satisfactory.
As expected, the vertical cracks formed close to the mid-span caused the failure of
mortar specimens, as shown in Figure 3.22. Due to the brittle nature of mortar the
66
formation of this failure developed quickly, namely in non-ductile manner, as in the
case of brick specimens.
Figure 3.22 : The failure of the mortar specimens under the flexural effect.
3.2.2 Compression tests on mortar
3.2.2.1 Specimen preparation
The compression tests of mortar were undertaken on two mortar prism groups. First
group was the remaining halves of the specimens used for the flexural tests.
According to TS EN 1015-11 (2000), ASTM C 348-02 (2002) and ASTM C 349-02
(2002), the portions obtained from the flexural tests of mortar specimens can be used
to determine compressive strength. Specimens of this group were not capped since
they were previously cut for the flexure tests to provide parallel opposite surfaces.
Second group of mortar specimens were prepared only for the compression tests. The
bearing surfaces of these specimens were capped with a thin layer of the highstrength
cement mortar.
The dimension of the first and second group is presented in Table 3.21 and Table
3.22 respectively. The first group specimens were denoted with MCF-numberA or
MCF-numberB which was the first capitals of mortar, compression, and flexural. The
two portions obtained from each flexural test specimen were signed with A and B
following the same number. The second group specimens were signed with MCnumber.
67
Table 3.21: The mortar specimen sizes of the first group.
Specimen m l (mm) m b (mm) m h (mm)
MCF-1A 40 27 34
MCF-1B 40 27 34
MCF-2A 40 43 35
MCF-2B 40 45 35
MCF-3A 40 29 35
MCF-3B 40 33 35
MCF-4A 40 29 29
MCF-4B 40 29 29
MCF-5A 40 35 35
MCF-5B 40 38 35
Table 3.22: The mortar specimen sizes of the second group.
Specimen m l (mm) m b (mm) m h (mm)
MC-1 40 35 35
MC-2 40 40 40
MC-3 30 40 40
MC-4 35 40 40
MC-5 35 40 46
MC-6 35 40 39
MC-7 40 40 46
MC-8 40 50 49
MC-9 40 40 48
MC-10 39 38 39
MC-11 39 40 39
MC-12 39 37 40
MC-13 39 43 36
MC-14 40 40 40
MC-15 37 35 42
MC-16 40 39 39
MC-17 37 35 35
MC-18 34 43 34
MC-19 35 38 35
MC-20 36 36 33
3.2.2.2 Test procedure
The test setup and measurement system of the compression test of the mortar
specimens are illustrated in Figure 3.23. The Amsler testing machines with 1000 or
68
5000 kN load capacities were used in an appropriate loading range. The
measurement system consisted of a load cell of 100 or 200 kN and two LVDTs of 10
or 25 mm located on the opposite corners of the lower plate.
Figure 3.23 : The test setup with measurement system for the mortar compression
tests.
3.2.2.3 Test results
In order to make clarify the behavior of mortar under compression loading, a total of
thirty mortar specimens were tested. In order to show the variation of stress and
corresponding strain occurring during the tests, Figure 3.24 and Figure 3.25 are
plotted for the first and second mortar specimen groups, respectively. The parameters
highlighted to illustrate the properties of the mortar quantitatively were the same as
those of the bricks, and are presented in Table 3.23 for the first group and in Table
3.24 for the second group. Table 3.25 includes the analyses related with the level of
scatter in the parameters. The average values obtained by evaluating all mortar
specimen results are 3.1 MPa with a CoV of 0.30 for the compressive strength ( ) mc f ,
2.3% with a CoV of 0.21 for the compressive strain at compressive strength ( ) v,mc, f  ,
1.8 MPa with a CoV of 0.28 for the compressive stress at the proportional limit
( ) mc, p  , 0.9% with a CoV of 0.29 for the compressive strain at the limit ( ) v,mc, p  ,
232 MPa with a CoV of 0.44 for Young’s modulus ( ) mc E and 1.9 with a CoV of 0.14
for the ductility ( ) mc  .
LVDT-1 LVDT-2
Loadcell
Upper plate
Lower plate
Specimen
69
0
1
2
3
4
5
0.00 0.05 0.10 0.15 0.20
Compressive strain
Compressive stress (MPa)
MCF-1A
MCF-2A
MCF-2B
MCF-3A
MCF-5A
MCF-5B
Figure 3.24 : The compressive stress-compressive strain relationships for the first
group of the mortar.
0
1
2
3
4
5
0.00 0.05 0.10 0.15 0.20
Compressive strain
Compressive stress (MPa)
Figure 3.25 : The compressive stress-compressive strain relationships for the second
group of the mortar.
70
Table 3.23 : The results of the mortar compression tests (the first group).
Specimen fmc (MPa) v,mc, f  (%) mc, p  (MPa) v,mc, p  (%) mc E (MPa) mc 
MCF-1A 2.13 1.44 1.67 0.95 175 --
MCF-1B ID ID ID ID ID ID
MCF-2A 2.62 2.70 1.34 0.60 222 1.5
MCF-2B 1.61 2.25 1.00 1.00 100 2.3
MCF-3A 1.90 2.36 1.12 0.83 134 1.5
MCF-3B ID ID ID ID ID ID
MCF-4A ID ID ID ID ID ID
MCF-4B ID ID ID ID ID ID
MCF-5A 2.43 2.33 1.64 0.93 177 1.6
MCF-5B 2.24 3.41 1.51 0.61 249 1.8
Table 3.24 : The results of the mortar compression tests (the second group).
Specimen mc f (MPa) v,mc, f  (%) mc, p  (MPa) v,mc, p  (%) mc E (MPa) mc 
MC-1 4.64 2.07 -- 0.86 373 2.3
MC-2 3.69 2.24 1.56 0.63 246 1.9
MC-3 4.75 2.44 2.50 0.71 350 2.2
MC-4 4.14 1.83 2.43 0.62 391 1.9
MC-5 3.29 2.31 2.00 1.14 175 1.8
MC-6 3.21 -- 2.57 -- 107 1.6
MC-7 4.38 1.66 2.13 0.60 353 1.8
MC-8 3.60 2.72 2.40 1.11 216 2.0
MC-9 2.88 2.15 1.88 1.24 151 1.8
MC-10 2.60 3.16 1.59 1.21 131 2.0
MC-11 2.30 2.45 1.42 1.09 131 1.9
MC-12 1.72 2.70 1.22 1.28 96 1.8
MC-13 3.70 1.69 1.75 0.58 304 2.0
MC-14 3.23 1.90 1.79 0.77 232 2.2
MC-15 3.73 1.86 2.27 0.57 399 2.0
MC-16 3.40 2.20 1.84 0.62 297 1.5
MC-17 1.66 2.43 1.26 1.21 104 2.3
MC-18 3.44 1.50 2.77 0.78 356 2.5
MC-19 4.00 1.94 1.81 0.49 369 1.9
MC-20 3.50 2.81 2.47 1.22 203 1.8
Table 3.25 : The statistical parameters of the mortar compression tests (all
specimens).
Statistical parameter Minimum Maximum Average Stdev CoV
mc f (MPa) 1.61 4.75 3.1 0.92 0.30
v,mc, f  (%) 1.44 3.41 2.3 0.49 0.21
mc, p  (MPa) 1.00 2.77 1.8 0.50 0.28
v,mc, p  (%) 0.49 1.28 0.9 0.26 0.29
mc E (MPa) 96 399 232 102 0.44
mc  1.5 2.5 1.9 0.27 0.14
71
By following the procedure given for the bricks to derive relations, the normalized
stress and normalized compressive strain is correlated with a parabolic function as
shown in Figure 3.26:
mc n v mc n v,mc,n
2
, , ,   0.85 1.86 (3.12)
where mc,n  is the normalized stress of the mortar and v,mc,n  is the normalized
compressive strain of the mortar.
mc,n = -0.8493v,mc,n
2 + 1.8583v,mc,n
R2 = 0.9742
0.0
0.2
0.4
0.6
0.8
1.0
1.2
0.0 0.5 1.0 1.5
Normalized compressive strain
Normalized compressive stress
Figure 3.26 : The normalized compressive stress-normalized compressive strain
relationship for the mortar specimens.
The relation of Young’s modulus and compressive strength, Figure 3.27 might be
expressed with the following linear function:
mc mc E  75 f (3.13)
According to Figure 3.28, the compressive stress might be taken as about 0.6 times
compressive strength:
mc p mc 0.6 f ,   (3.14)
72
Emc = 76.063fmc
R2 = 0.6605
0
100
200
300
400
500
0 2 4 6
Compressive strength (MPa)
Young's modulus (MPa)
Figure 3.27 : The Young’s modulus-compressive strength relationship for all mortar
specimens.
mc,p = 0.5885fmc
R2 = 0.5012
0
1
2
3
4
0 2 4 6
Compressive strength (MPa)
Compressive stress at proportional limit (MPa)
Figure 3.28 : The compressive stress at proportional limit-compressive strength
relationship for the mortar specimens.
73
Figure 3.29 demonstrates the correlation of stress-to-strain ratios at proportional limit
and strength level. The correlation might be formulated as follows:
mc mc s
v mc f
mc
v mc p
mc p E E
f
,
, , , ,
, 1.6   1.6
 


 



 

(3.15)
Emc = 1.5756Emc,s
R2 = 0.8364
0
100
200
300
400
500
0 100 200 300
Secant modulus at peak (MPa)
Young's modulus (MPa)
Figure 3.29 : The Young's modulus-secant modulus at peak relationship for the
mortar specimens.
3.2.3 Rebound hammer tests on mortar joints
Rebound hammer tests were carried out on sixty horizontal mortar joints of the inplace
masonry walls of the historical houses. The test results are grouped according
to the thickness of the mortar joints, Table 3.26. Average rebound number of each
joint was obtained by taking five readings. The average rebound number of the
mortar joints and the standard deviation are 14 and 0.71, respectively. As seen the
table, there is no effect of the thickness of mortar joints on the rebound number.
74
Table 3.26 : The rebound numbers of the mortar joints.
bm t (mm) mr N
14-18 14
19-23 15
24-28 14
29-33 13
34-38 14
3.3 Physical Tests on Bricks
The physical features of the bricks, which are bulk density, real density, water
absorption and porosity, were figured out by conducting laboratory tests. The
objective of these tests is to define the physical properties of the bricks which are
thought to be effective on the mechanical properties.
The physical tests of a total of 10 almost half-length specimens were realized. In
order to define the properties, the knowledge on the dry and saturated weights of
each specimen should be required. The dry weight and saturated weight was found
out in accordance with ASTM C 67-05 (2005) and TS EN 771-1 (2005), and
determination of the dry weight includes two successive processes which are called
as drying and cooling. The steps of the drying process are explained as follows. The
specimens were left on an oven at 105°C during a 24-hour period. In the end of this
period, the specimens were weighted, and these weights were recorded as 1 W . After
then, the specimens were left again on the oven during a 2-hour period. In the end of
this period, the specimens were weighted, and these weights were recorded as 2 W .
The differences between the corresponding weights were estimated for each
specimen and as the specimen, whose weight difference between the two periods was
smaller than 0.2% was accepted dry. To realize the cooling process, these specimens
were left in a room having 21°C during 4 hours. After then, the specimens were
weighted and these weights were recorded as d W , namely, dry weights. Then, the dry
specimens were submerged in cool water during 24 hours to estimate the saturated
weights of the specimens ( ) s W . The brick specimens and their related weights are
presented in Table 3.27. The brick specimens were symbolized with the first capitals
of brick and physical, namely, BP.
75
Table 3.27 : The dry and saturated weights of the bricks.
Specimen 1 W (g) 2 W (g)  


 

 
1
1 2 100
W
W W
d W (g) s W (g)
BP-1 1335.2 1335.2 0.0 1337.2 1610.9
BP-2 1745.7 1745.6 0.1 1747.6 2068.6
BP-3 2163.7 2163.2 0.2 2166.3 2499.3
BP-4 1783.3 1783 0.2 1785 2047.5
BP-5 1111.2 1110.9 0.3 1112.9 1332.4
BP-6 1134.8 1134.6 0.2 1136.4 1380.8
BP-7 1869.6 1869.2 0.2 1871.2 2170.7
BP-8 1614.1 1614.1 0.0 1616.1 1837.6
BP-9 1171.6 1171.2 0.3 1173.8 1415.6
BP-10 1158.6 1158.3 0.3 1160.6 1402.8
Bulk density or unit weight of a specimen is defined as the ratio of the dry specimen
weight to the exterior volume, including pores:
V
Wd
A b  ,  (3.16)
where A,b  is bulk density of brick, d W is the dry specimen weight and V is the
exterior volume of the specimen.
Real density or specific density is determined by dividing the dry specimen weight
by the volume without open pores.
The bulk and real densities of the specimens are given in Table 3.28. The bulk
(apparent) densities of the bricks varied between 1550 and 1950 kg/m3. The average
bulk density and its standard deviation were 1730 and 160 kg/m3 with a coefficient
of variation of 0.09, respectively. The real densities of bricks ( ) R,b  determined
using picnometer, varied between 2750 and 2970 kg/m3 with an average of 2860
kg/m3. The standard deviation and the coefficient of variation were calculated as 70
kg/m3 and 0.02, respectively.
Water absorption is defined as the percentage of water absorbed in weight by a dry
specimen.
w s d d A %  (W W )100/W (3.17)
76
where w A , is the percentage of water absorption, s W is the saturated weight of the
specimen after submersion in cold water and d W is the dry weight. According to the
test results given in Table 3.29, the water absorption of bricks varied between
approximately 14 and 22% with an average of 18%. The standard deviation and the
coefficient of variation were calculated as 3% and 0.16, respectively.
Table 3.28 : The bulk and real densities of the bricks.
Specimen V (cm3) A,b  (gr/cm3) R,b  (gr/cm3)
BP-1 780 1.71 2.87
BP-2 1019 1.72 2.88
BP-3 1116 1.94 2.97
BP-4 972 1.84 2.94
BP-5 720 1.55 2.87
BP-6 734 1.55 2.85
BP-7 1000 1.87 2.91
BP-8 828 1.95 2.75
BP-9 735 1.60 2.76
BP-10 729 1.59 2.84
Porosity (Po) is defined as the volume of open pores divided by exterior volume
including pores of the specimen. In this study, porosity of the brick specimens was
calculated using Eq. (3.18) and the results obtained can be seen in Table 3.29. The
porosity of the bricks was calculated in the range of 29-46% with an average of 40%.
The standard deviation and the coefficient of variation were calculated as 5% and
0.14, respectively.
 


 


 
R b
A b Po
,
, 1


(3.18)
Table 3.29 : The water absorption and real porosity values of the bricks.
Specimen w A (%) Po (%)
BP-1 20.5 40
BP-2 18.4 40
BP-3 15.4 35
BP-4 14.7 38
BP-5 19.7 46
BP-6 21.5 46
BP-7 16.0 36
BP-8 13.7 29
BP-9 20.6 42
BP-10 20.9 44
77
3.4 Chemical Tests on Brick and Mortar
In order to obtain chemical and mineralogical composition of the bricks and mortar
samples taken from the Akaretler Row Houses, experimental works were carried out
by Prof. Dr. Tülay Tulun and her team in Istanbul Technical University, Faculty of
Science and Letters, Chemistry Department. Chemical, calcimetry and mineralogical
analysis were realized on a total of 17 bed/horizontal mortar joints and 4 brick
samples, which were taken from the inner walls of the houses. The results of the
experimental works were presented as a part of paper, (İlki et al, 2007).
Chemical analysis was realized using wet chemistry, UV (ultra violet) visible
spectrophotometer and atomic absorption methods were used to determine silica and
other oxide components. The oxide components of the brick and mortar samples by
weight percentage were tabulated in Table 3.30 and Table 3.31, respectively, and the
average oxide components can be seen in Figure 3.30 to assess weight ratios with
respect to each other. The samples were symbolized with BCh (brick-chemical) and
MCh (mortar-chemical). The component not detected is symbolized with nd in the
related tables.
Table 3.30 : The oxide components of the brick samples.
Sample SiO2 CaO MgO Fe2O3 Al2O3 Na2O K2O CdO CoO NiO CuO ZnO
BCh-1 19.84 5.26 10.08 24.49 24.47 2.25 3.7 0.28 1.46 1.77 0.06 0.76
BCh-2 25.12 3.44 8.80 25.09 23.97 4.62 2.8 0.26 1.73 2.43 0.07 0.96
BCh-3 35.68 2.04 6.08 20.65 24.24 1.87 3.67 0.27 1.99 2.51 0.09 0.91
BCh-4 30.88 3.32 6.56 20.09 15.77 3.24 4.12 0.02 1.96 1.88 0.1 1.31
Calcimetry analysis was utilized to estimate loss of ignition and organic contents,
Table 3.32 for the brick samples and in Table 3.33 for the mortar samples.
Mineralogical analysis was arranged by utilizing XRD (X-ray diffraction) to detect
the existence of feldspar, quartz and calcite. The percentages of these components
were tabulated in Table 3.34 for the brick samples and in Table 3.35 for the mortar
samples. The composition of the brick and mortar samples is compared in terms of
the average percentages of feldspar and quartz, Figure 3.31.
78
Table 3.31 : The oxide components of the mortar samples.
Sample SiO2 CaO MgO Fe2O3 Al2O3 Na2O K2O CdO CoO NiO CuO ZnO
MCh-1 51.72 22.71 4.00 1.55 5.16 4.17 6.83 1.58 0.39 0.37 0.02 0.22
MCh-2 42.44 24.52 11.25 1.24 6.19 4.17 8.44 1.05 nd 0.39 0.04 0.37
MCh-3 42.84 18.39 3.77 2.16 4.17 3.19 9.64 1.36 nd 5.85 0.03 0.36
MCh-4 58.64 6.16 4.05 3.77 5.73 2.94 10.85 1.62 0.04 3.32 0.14 0.86
MCh-5 18.64 23.01 11.3 3.86 15.18 3.68 9.24 1.89 nd 7.86 0.15 0.43
MCh-6 20.12 19.74 12.8 1.05 2.98 2.94 8.44 2.15 0.01 4.86 0.32 1.05
MCh-7 44.32 10.87 6.55 1.72 3.60 2.70 9.24 2.42 0.06 2.24 0.1 0.63
MCh-8 35.08 23.05 11.20 4.28 7.37 2.21 8.03 2.62 0.41 3.41 0.23 0.59
MCh-9 54.84 6.27 3.54 1.4 1.02 5.39 4.02 2.84 0.53 1.19 0.16 0.51
MCh-10 66.48 4.18 6.42 1.85 4.70 6.74 4.02 3.07 0.9 0.81 0.11 2.04
MCh-11 48.24 13.21 3.20 3.40 5.51 8.36 4.02 3.27 0.8 0.96 0.1 0.55
MCh-12 58.08 5.88 0.81 4.64 5.39 6.20 4.02 1.15 0.44 1.71 0.06 0.59
MCh-13 62.48 9.58 1.94 4.58 5.39 6.74 8.03 0.18 1.00 1.32 0.03 0.90
MCh-14 38.72 11.85 4.80 6.65 6.74 8.63 8.03 0.19 0.86 1.7 0.05 1.89
MCh-15 53.04 9.62 6.72 3.42 6.96 5.66 8.03 0.21 0.83 3.51 0.07 0.73
MCh-16 13.92 14.26 3.36 18.87 19.69 3.49 2.07 0.2 1.09 1.75 0.07 1.06
MCh-17 17.60 12.22 2.24 28.66 29.44 3.24 3.53 0.26 1.42 0.28 0.05 0.79
0
10
20
30
40
50
SiO2
CaO
MgO
Fe2O3
Al2O3
Na2O
K2O
CdO
CoO
NiO
CuO
ZnO
Oxide components
Weight (%)
Brick
Mortar
Figure 3.30 : The average oxide components of the brick and mortar samples.
79
Table 3.32 : The loss of ignition and organic matter of the brick samples.
Sample Loss of Ignition (%) Organic Matter (%)
BCh-1 1.69 1.27
BCh-2 4.01 3.42
BCh-3 0.67 0.28
BCh-4 1.52 1.33
Table 3.33 : The loss of ignition and organic matter of the mortar samples.
Sample Loss of Ignition (%) Organic Matter (%)
MCh-1 4.02 3.16
MCh-2 9.47 8.89
MCh-3 2.60 1.93
MCh-4 2.64 1.74
MCh-5 2.62 2.05
MCh-6 4.87 4.30
MCh-7 4.40 4.01
MCh-8 3.23 2.89
MCh-9 2.16 4.07
MCh-10 2.70 1.90
MCh-11 3.74 2.83
MCh-12 2.62 2.08
MCh-13 3.38 2.74
MCh-14 8.69 8.25
MCh-15 3.39 1.51
MCh-16 6.72 4.85
MCh-17 2.82 2.18
Table 3.34 : The mineralogical of the brick samples.
Sample Feldspar Quartz Calcite
BCh-1 46.96 35.84 17.19
BCh-2 70.39 21.43 8.18
BCh-3 32.44 62.02 5.53
BCh-4 48.12 44.18 7.71
Table 3.35 : The mineralogical of the mortar samples.
Sample Feldspar Quartz Calcite
MCh-1 56.44 20.72 22.84
MCh-2 64.15 10.9 24.95
MCh-3 65.43 15.16 19.41
MCh-4 69.19 24.32 6.49
MCh-5 66.68 9.11 24.21
MCh-6 72.47 1.62 25.91
MCh-7 64.61 23.09 12.3
MCh-8 59.27 12.72 28.01
MCh-9 56.81 36.31 6.88
MCh-10 64.53 31 4.47
MCh-11 71.02 15.73 13.25
MCh-12 61.39 32.28 6.33
MCh-13 72.56 18.50 8.94
MCh-14 86.71 1.85 11.44
MCh-15 73.05 17.04 9.91
MCh-16 65.09 5.25 29.66
MCh-17 72.96 2.25 24.79
80
0
10
20
30
40
50
60
70
Feldspar
Quartz
Calcite
Mineralogical components
Weight (%)
Brick
Mortar
Figure 3.31 : The average results of mineralogical analysis of the brick and mortar
samples.
3.5 Evaluation of the Tests
The mechanical, physical, and chemical characteristics of brick and mortar form the
basic characteristics of masonry. These characteristics, especially compressive
strength, are also required to assess whether the requirements of the related codes are
provided.
Although the test results varied in a wide range, the average values of them give an
idea about the material characteristics. The average compressive strengths of the
bricks, the three-brick specimens and the mortar specimens were obtained as 5.5, 2.3
and 3.1 MPa.
To eliminate size effect on the compressive strengths of the bricks obtained from the
tests, TS EN 772-1 (2002) suggests that the compressive strengths should be
transformed to the normalized compressive strength, namely, to the strength of an
equivalent 100 mm cube unit, using conversion factors  . The conversion factors
taken from TS EN 772-1 (2002) are presented in Table A.1. TS EN 772-1 (2002)
permits linear interpolation between the dimensions given. The normalized strengths
of the specimens in this study are also presented in Table A.2 and Table A.3. The
81
average values of the normalized compressive strengths of the bricks tested in
normal and parallel to the bed joint were calculated as 4.5 and 2.8 MPa, respectively.
The anisotropy of the bricks is calculated as 1:1.9 from the compressive strengths. As
seen, the difference between the compressive strengths, which resulted from different
loading directions, is noticeable. This difference may be associated with the pressing
and firing processes of the bricks. The elimination of size effects on the anisotropy
ratios may be a suitable way to make a realistic comparison related with loading
directions, by using conversion factors given in Table A.1, since the heights of the
specimens subjected to loads in normal to the bed joint were smaller than in parallel
to the bed joint. The normalized compressive strengths of the bricks tested in parallel
to the bed joint are presented in Table A.3. While the average value of the
normalized compressive strengths of these bricks was determined as 2.1 MPa; that of
the other halves of these bricks tested in normal to the bed joint was determined as
2.8 MPa. Consequently, the bricks had an anisotropy of 1:1.3. This result indicates
the effect of specimen size on the compressive strength. As expected, the strength of
the brick tested in direction parallel to the bed joint is reduced with respect to that in
the direction normal to the bed joint.
Distinct failure types (crushing and conical types) were observed for the brick
specimens under compression effects, depending on the directions of loading. This
might be due to the anisotropic properties of the bricks and the difference between
the heights of the specimens in two directions. The average height of the specimens
compressed in parallel to the bed joint were about 2 times that in normal to the joint.
Comparing the strengths of the historical masonry constituents with those given in
the related codes can provide an insight into the assessment of these historical
materials. Thus, the information complied from the codes is given in the following
paragraphs:
The requirements of TSDC (2007) are that the minimum compressive strength of
masonry units should be 5.0 MPa and that mortars to be used in the masonry walls
should be cement-lime or cement mortar.
TS EN 771-1 (2005) gives a classification for the bricks used in Turkey. According
to this code, while the common solid bricks with the average and normalized average
compressive strengths of at least 5.0 and 4.0 MPa are defined as medium-strength
82
solid bricks; those with the average and normalized compressive strengths of at least
3 and 2.5 MPa are defined as low-strength solid bricks. The average size of common
solid brick is also given as 190x90x50 mm (lengthxwidthxheight).
EN 1998-1 (2004) requires the use of units with a normalized compressive strength
at least 5 MPa for normal direction to the bed joint and 2 MPa for parallel direction
to the bed joint (except in cases of low seismicity) and mortar with a compressive
strength at least 5 MPa for unreinforced masonry.
One of the mortar classification methods given in BCRSMS (2008) depends on the
average compressive strengths at 28 days. The method defines the compressive
strengths of cement-lime, cement and masonry cement mortars in the range of 2.4-
17.2 MPa. BCRSMS (2008) recommends the use of ASTM C 62-05 (2006) for the
requirements related with solid building bricks. Each brick should have a minimum
compressive strength of 8.6, 15.2 or 17.2 MPa depending on the specified resistance
level of the brick to cyclic freezing damage (ASTM C 62-05, 2006).
Considering the compressive strengths of the historical bricks (minimum values of
1.5 MPa for the bricks and 1.4 MPa for the three-brick specimens; average values of
5.5 MPa for the bricks and 2.3 MPa for the three-brick specimens), it is seen that
these bricks do not conform to the requirements of TSDC (2007), EN 1998-1 (2004)
and ASTM C 62-05 (2006), but, these bricks can be identified as low-strength bricks
according to the requirements given in TS EN 771-1 (2005).
The mortar tested under the study is thought to be lime mortar (this idea is detailed
below), but these codes do not consider the simple lime mortar. Consequently, to
compare the strengths of the mortar specimens with the mortars defined in the codes
may not be suitable.
While the average compressive strength of the individual bricks was 5.5 MPa, the
average strength of the three-brick specimens was 2.3 MPa. The strength of the
three-brick specimens was significantly lower than the strength of the individual
bricks since the confinement effect of loading plates due to friction forces was
reduced in case of the three-brick specimens. The adverse effect of inevitable
accidental flexure in case of the three-brick specimens and relatively larger specimen
size are other factors causing lower compressive strength.
83
Tensile strength can be calculated using several formulations as a function of
compressive strength. Therefore, it is worth to mention about several ratios obtained
from flexural and compressive tests. The flexural tensile strength of the bricks is
about 25 and 60% of the average compressive strength of the individual bricks and
the three-brick specimens, respectively. The ratio of 25% obtained for the bricks is
consistent with the range of 10-32% given by Drysdale et al. (1994) for bricks. The
flexural tensile strength of mortars varied between 42-65% of the corresponding
compressive strengths.
While the ratio of average compressive strength of single brick to mortar is
approximately 1.8 according to destructive laboratory tests, the ratio of the average
rebound hammer number of the bricks to that of the horizontal mortar joints is
approximately 2.1. Since these tests were carried out in different conditions, namely,
laboratory and in situ, these slightly different ratios should be acceptable. Therefore,
it is possible to conclude that rebound hammer tests on bricks and horizontal mortar
joints were consistent with the corresponding laboratory tests.
The high values of the water absorption rate (an average of 18%) and porosity (an
average of 40%) of the bricks may be the indicators of the low quality and low
strengths. According to Bayülke (1980), the compressive strength of brick generally
gradually decreases if the porosity of the brick is higher than 25%.
The chemical tests were evaluated by Prof. Dr. M. Süheyl Akman and the evaluation
was a part of the study of Ilki et al. (2007). This study gives the following
information: Considering the construction period of the building, one would think
that as binder Khorassani type mortar, which is obtained by addition of brick powder
to hydrated lime might have been used. In this case, hydrates of calcium silicates and
calcium aluminates should have been formed due to pozzolanic effect. However, in
XRD tests, these formations are not observed. Consequently, it is decided that the
binder is made without brick powder. However, in this case, the presence of adequate
conditions and sufficient aging should be discussed for the carbonation of Ca(OH)2.
According to the results of XRD tests, it might also be possible that Ca(OH)2
remained as hexagonal portlantide crystal without being carbonated. However,
considering the absence of portlantide, it is assumed that all hydrated lime was
transformed into calcite. The results of mechanical tests on mortar specimens support
the validity of this assumption. The high content of feldspar indicates that the sand
84
used in the mortars also contained eruptive rock particles. In addition, the sand
contains quartz and calcite from sedimentary origin. It is a high probability that same
type of sand was used during the production of bricks as non-plastic flushing
material. The significantly higher ratios of calcite in mortars with respect to bricks
prove that the increase in the ratio of calcite in mortars is due to carbonation of
hydrated lime, as well as the sand content. The origin of clay minerals of the bricks is
sodium feldspars. According to their appearance and color, it is clear that the bricks
are simply produced (common type) in field kilns and they contain components with
iron.
As a result, the low values of strength and Young’s modulus, high values of water
absorption rate and of porosity and the non-uniformity in color, texture and shapes
indicate that the bricks were burnt in the kilns, but not in uniform conditions and so,
these bricks can be called as common bricks. The test results displayed a high
scattering given with large coefficients of variation. This may be resulted from the
differences in the production procedures (such as firing temperature and pressing
method), and in the raw materials (such as aggregate type and size). These large
deviations may cause several vital problems in the determination of the historical
masonry material properties and in the decision regarding material selection for the
restoration/strengthening works, if sufficient amount of samples are not taken into
consideration.
To predict masonry compressive strength, depending on the average compressive
strength values of the bricks (5.5 MPa) and the mortars (3.1 MPa); the equations
given in Chapter 1, which are proposed by TSDC (2007), BCRSMS (2008) and EN
1996-1-1 (2005), were utilized. The results obtained are presented, respectively, as
follows:
f f f MPa masc uc masc  0.50   0.505.5  2.8 (3.19)
f A Bf psi MPa masc uc (400 ) 1 (400 0.2 798) 560 3.9 '        
(3.20)
f f MPa
f Kf f f MPa
masc masc c
masc c uc n mc masc c
1.2 1.2 2.2 2.6
0.55 (4.5) (3.1) 2.2
,
0.7 0.3
, , ,
     
       
(3.21)
85
The masonry compressive strengths based on the equations of TSDC (2007) and EN
1996-1-1 (2005) are close to each other, namely, 2.8 and 2.6 MPa. It should be noted
that the normalized brick strength (4.5 MPa) was used in the equation of EN 1996-1-
1 (2005). The value of 3.9 MPa, which is called as the specified compressive strength
of masonry, is the prediction obtained from the equation of BCRSMS (2008). In
order to transform it to corresponding compressive strength, the value (3.9 MPa)
should be multiplied by a partial factor larger than 1.0 should be used.
86
87
4. EXPERIMENTAL STUDIES ON ORIGINAL MASONRY - CORE,
WALLET AND IN-SITU WALL TESTS
In this chapter, the tests conducted on masonry specimens (cores and wallets), which
were taken from the masonry walls of the historical houses and in-situ wall tests are
described and the test results are evaluated.
Since masonry is a composite material consisting of units and mortar joints; the
knowledge on these components of masonry is useful, but many times, this
knowledge cannot be sufficient to define the behavior of masonry. Consequently, in
addition to the tests performed on individual original brick and mortar specimens,
compression tests on the cores and wallets, and shear tests on the cores and the insitu
walls were carried out as a part of the thesis. The fact that several walls of the
houses would be removed according to the restoration project enabled the extraction
of the cores and wallets from these walls as well as the realization of the in-situ shear
tests on these walls. While the core specimens comprised two pieces of bricks
bonded with a mortar bed joint, the wallets comprised three rows of bricks bonded
with two mortar bed joints and several mortar head joints. Testing zone of the in-situ
walls consisted of two brick rows, one bed joint and several head joints.
The main objectives of the test groups explained in this Chapter are:
 To obtain the behavior of the original masonry of the historical houses as a
composite material under various loading conditions,
 To understand the roles of the components on the masonry behavior,
 To extract correlations between strength and deformation parameters of each
test group,
 To relate the indicative parameters of material behavior obtained from different
test groups,
 To deduce the influence of the specimen composition and size (core and
wallet) on the test results,
88
With the intention of a comprehensive assessment of the mechanical behavior of the
original walls, the tests included splitting, compression and shear tests on the cores,
compression tests on the wallets, and shear tests on the in-situ walls.
The tests explained in this Chapter were executed at Istanbul Technical University,
Civil Engineering Faculty, Structural Material Laboratory with the exception of the
in-situ shear tests.
4.1 Mechanical Tests on Cores
4.1.1 Splitting tests on cores
4.1.1.1 Specimen preparation
The splitting tests were performed on the core specimens for the purpose of
determination of splitting tensile strength and failure mechanism of the historical
masonry.
The core specimens were randomly taken out from the structural masonry walls of
the historical houses using a core-drilling machine as shown in Figure 4.1. Before
drilling, the plaster on the surface of the walls was removed and the surface was
roughly cleaned. The nominal diameter and length of the cores were 95 mm and 110
mm, respectively. Each core specimen was tied with wires to reduce the possibility
of damage during transportation and other preparation works. After cores were taken
to the laboratory, the ends of each core were cut to obtain flat surfaces as shown in
Figure 4.2. Then, the wires were taken out and the cores were left in the laboratory
until the test day. The view of cores is illustrated in Figure 4.3.
Sizes of the specimens of the splitting tests are presented in Table 4.1. Each core
specimen was symbolized with the first letters of core, splitting and tensile, and an
order number. c h is the height of the core specimen, c l is the length of the specimen,
and mb t is the thickness of the mortar bed joint. As the thickness of the bed joint is
not constant through the specimen, mb t shows a range of the thicknesses measured,
Figure 4.3.
89
The height of each specimen was determined as the average of two diameters
perpendicular to each other, as there was a small difference between the two
diameters of cx D and cy D . The diameters of cx D and cy D are the average of the
corresponding two values measured at the ends. The length is the average value of
four measures. The two measures of four are the lengths of the lines connecting
crowns at each end and the other two are the side lengths of the specimen, which are
measured through the middle of the mortar bed joint.
Figure 4.1 : The drilling of the cores from the structural masonry walls of the
Akaretler Historical Row Houses.
Figure 4.2 : The preparation of the cores.
Direction of drilling
90
Figure 4.3 : The schematic view of the cores.
Table 4.1 : The core sizes for the splitting tests.
Specimen c h (mm) c l (mm) mb t (mm)
CST-1 101 82 32-40
CST-2 101 106 6-12
CST-3 98 84 20-25
CST-4 98 85 18-28
4.1.1.2 Test procedure
In order to find out the average splitting tensile strength of the core specimens, four
cores were tested using the test configuration shown in Figure 4.4. The tests of the
cores were carried out using the Instron testing machine with 100 kN load capacity.
Compression loads were applied to the top and bottom of each specimen using a pair
of steel rod bearings. The splitting tensile strength was calculated using the following
equation:
c c
cst m
cst h l
P
f

, 2
 (4.1)
cst f is the splitting tensile strength of each core, and cst m P , is the maximum load
applied during the test.
4.1.1.3 Test results
The maximum loads resisted and the splitting tensile strengths of the specimens are
given in Table 4.2. The average splitting tensile strength is calculated as 0.41 MPa
with a standard deviation of 0.10 MPa. The coefficient of variation is 0.25.
lc
a-a cross section
Dcy
Brick
Brick
Mortar Dcy
a
a
Dcx
Brick
Brick
tmb tmb
91
Figure 4.4 : The test configuration of the core splitting tests.
The failure of the specimens was due to formation of an approximately vertical crack
occurring along a line between the two rods, as shown in Figure 4.5. Since the failure
of CST-3 was not similar to typical failure mode of splitting tests, the strength of
CST-3 was not taken into consideration while computing the average strength,
Figure 4.6.
Table 4.2 : The results of the core splitting tests.
Specimen cst m P , (N) cst f (MPa) Stdev (MPa) CoV
CST-1 5430 0.42
0.10 0.25
CST-2 4330 --
CST-3 2080 --
CST-4 5195 0.40
Figure 4.5 : The view of CST-2 core during and after the splitting test.
ass
Rod
Rod
lc
hc
a-a cross section
Brick
Brick
a
a
Mortar
Brick
Brick
Pcst
Lower plate
Upper plate
92
Figure 4.6 : The view of CST-3 core during the splitting test.
4.1.2 Compression tests on cores
4.1.2.1 Specimen preparation
Preparation procedure of core specimens was similar to that of the cores subjected to
the splitting test. In addition, the upper and lower surfaces of the core specimens
were capped with a thin mortar layer as shown in Figure 4.7 and Figure 4.8 for
providing uniform distribution of the compression loads along the lengths of the
specimens. The dimensions of the cores and the average thicknesses of mortar bed
joints are presented in Table 4.3.
Figure 4.7 : The capping of the cores for the compression tests.
93
Figure 4.8 : The schematic view of the cores tested under compression.
4.1.2.2 Test procedure
A total of forty-five cores were tested using the Amsler testing machine with a load
capacity of 1000 kN, Figure 4.9. Displacement was applied to the specimens in a
monotonic pattern, and compression loads, which each specimen was subjected to,
was followed through the load indicator of the Amsler testing machine and recorded.
Displacements were measured by four LVDTs with a capacity of 25 mm located on
the lower loading plate, Figure 4.9. The data was collected using a data logger (TML
TDS-303). The direction of the loading was the same as the direction of gravity loads
on the existing walls.
Figure 4.9 : The setup with measurement system for the core compression tests.
LVDT 1-2 LVDT 3-4
Specimen
hc
lc
a
a
Mortar cap
Mortar cap
Mortar
Brick
Brick
Mortar cap
a-a cross section
Brick
Brick
Dcx
Mortar cap
94
Table 4.3 : The core sizes for compression tests.
Specimen cx D (mm) c l (mm) c h (mm) mb t (mm)
CC-1 89 89 101 25
CC-2 91 89 100 18
CC-3 92 90 96 22
CC-4 92 87 97 16
CC-5 89 87 103 14
CC-6 90 89 105 25
CC-7 91 92 103 16
CC-8 88 88 102 29
CC-9 89 80 104 25
CC-10 90 89 104 34
CC-11 91 88 106 19
CC-12 91 92 102 25
CC-13 91 86 109 23
CC-14 91 89 102 17
CC-15 92 91 102 19
CC-16 91 97 107 26
CC-17 86 88 104 19
CC-18 88 87 104 20
CC-19 90 80 108 30
CC-20 90 92 108 21
CC-21 89 80 104 26
CC-22 89 91 109 25
CC-23 90 91 105 29
CC-24 91 90 110 18
CC-25 89 90 104 22
CC-26 90 90 111 23
CC-27 89 87 105 24
CC-28 91 89 106 15
CC-29 92 90 106 15
CC-30 92 85 102 15
CC-31 88 87 101 24
CC-32 88 75 102 32
CC-33 84 86 99 16
CC-34 87 88 104 18
CC-35 83 88 106 25
CC-36 88 87 98 29
CC-37 86 87 106 22
CC-38 90 86 104 20
CC-39 89 88 106 21
CC-40 83 89 100 32
CC-41 90 89 110 26
CC-42 87 88 100 15
CC-43 88 87 103 17
CC-44 89 86 101 23
CC-45 88 86 103 23
95
4.1.2.3 Test results
The response of each specimen to compression is expressed through a curve showing
the variation of compressive stress and compressive strain (Figure 4.10).
Compressive strain was obtained by the average displacement readings of the four
LVDTs.
In order to present the obtained technical data quantitatively; the compressive
strength ( f cc ), the strain at the strength level ( cc, f  ), the stress at proportional limit,
( cc, p  ), the strain at proportional limit ( cc, p  ), Young’s modulus ( cc E ), and ductility
( cc  ) of each specimen are given in Table 4.4.
The statistical assessment of the technical data is presented in Table 4.5. As shown in
this table, while the ductility values displayed a lower scatter, the Young’s modulus
values displayed a larger scatter, with respect to the distribution of the other
mechanical parameters.
0
1
2
3
4
5
0.00 0.02 0.04 0.06 0.08 0.10 0.12
Compressive strain
Compressive stress (MPa)
Figure 4.10 : The compressive stress-compressive strain relationships for the cores.
96
Table 4.4 : The results of the core compression tests.
Specimen f cc (MPa) cc, f  (%) cc, p  (MPa) cc, p  (%) cc E (MPa) cc 
CC-1 4.25 2.12 2.65 0.89 298 1.4
CC-2 2.32 2.35 1.67 0.94 177 1.7
CC-3 1.99 3.43 1.03 1.13 91 1.3
CC-4 -- 2.21 -- 0.70 559 1.4
CC-5 2.42 -- 1.49 1.16 129 1.3
CC-6 3.12 2.63 2.56 1.28 200 1.5
CC-7 3.55 2.30 1.13 0.46 249 1.8
CC-8 3.02 2.24 1.23 0.66 185 1.8
CC-9 2.84 2.18 1.62 0.90 179 1.5
CC-10 2.75 1.57 1.37 0.45 308 1.8
CC-11 3.06 2.64 1.56 0.82 191 1.2
CC-12 2.89 1.71 2.39 1.25 192 1.2
CC-13 2.36 2.19 1.53 0.99 155 1.3
CC-14 3.48 1.84 2.04 0.72 283 1.4
CC-15 3.87 2.69 2.09 0.79 263 1.3
CC-16 3.85 1.77 2.38 0.69 346 1.4
CC-17 3.57 1.98 1.59 0.41 391 1.4
CC-18 3.62 2.22 1.70 0.60 281 1.3
CC-19 3.17 1.81 1.94 0.82 237 --
CC-20 2.78 2.48 1.63 1.12 145 1.3
CC-21 2.32 1.31 1.40 0.63 224 2.1
CC-22 2.16 2.09 1.17 0.67 176 1.5
CC-23 3.03 2.26 2.63 -- 144 1.4
CC-24 4.58 1.56 1.59 0.29 552 1.5
CC-25 4.24 1.72 1.94 0.62 314 2.0
CC-26 2.84 1.40 1.73 0.51 337 1.4
CC-27 4.26 2.67 2.52 1.21 208 1.5
CC-28 3.95 2.27 1.30 0.31 416 1.2
CC-29 -- -- 2.90 0.87 332 1.6
CC-30 3.90 1.37 2.75 0.76 362 1.1
CC-31 2.22 3.47 1.31 0.93 140 1.8
CC-32 4.62 2.48 2.27 0.35 -- 1.3
CC-33 2.63 2.74 1.59 1.01 157 1.4
CC-34 2.09 1.80 1.57 0.87 181 1.6
CC-35 3.01 1.79 1.85 0.87 213 1.5
CC-36 2.29 2.41 1.76 1.51 117 1.5
CC-37 1.87 2.20 1.27 1.09 117 1.5
CC-38 4.73 1.29 3.10 0.54 572 1.4
CC-39 2.62 1.88 1.79 0.74 240 1.9
CC-40 1.67 2.13 0.94 0.86 109 1.5
CC-41 4.18 1.54 2.50 0.66 381 1.4
CC-42 4.18 1.55 2.48 0.60 416 1.7
CC-43 2.42 2.04 1.44 0.89 162 1.6
CC-44 3.72 2.34 2.35 1.03 228 1.5
CC-45 4.10 2.96 2.97 1.12 264 1.5
97
Table 4.5 : The statistical parameters of the core compression tests.
Statistical parameter Minimum Maximum Average Stdev CoV
cc f (MPa) 1.67 4.73 3.2 0.84 0.26
cc, f  (%) 1.29 3.47 2.1 0.51 0.24
cc, p  (MPa) 0.94 3.1 1.9 0.57 0.30
cc, p  (%) 0.29 1.51 0.8 0.28 0.35
cc E (MPa) 91 572 255 120 0.47
cc  1.1 2.1 1.5 0.21 0.14
Based on a simple regression analysis conducted on the test data, the function
showing the relationship of the normalized compressive stress-normalized
compressive strain was obtained as given by Eq. (4.2) , (Figure 4.11):
cc n cc n cc,n
2
, ,   0.90 1.88 (4.2)
In Eq. (4.2), cc,n  is the normalized compressive stress and cc,n  is the normalized
compressive strain.
n,cc = -0.8951n,cc
2 + 1.8827n,cc
R2 = 0.9543
0.0
0.2
0.4
0.6
0.8
1.0
1.2
0.0 0.5 1.0 1.5
Normalized compressive strain
Normalized compressive stress
Figure 4.11 : The normalized compressive stress-normalized compressive strain
relationship for the cores.
98
Through Eq. (4.2), it may be possible to obtain the compressive stress-compressive
strain relationship of a masonry core when the compressive strength and
corresponding strain of the core are known.
The tendency is that the specimen having higher strength could be achieved higher
Young’s modulus is formulated with a linear function, Figure 4.12:
cc cc E  80 f (4.3)
This correlation may enable the estimation of the Young’s modulus of any masonry
core when its compressive strength is known.
Ecc = 80.327fcc
R2 = 0.5952
0
100
200
300
400
500
600
0 1 2 3 4 5 6
Compressive strength (MPa)
Young's modulus (MPa)
Figure 4.12 : The Young’s modulus-compressive strength relationship for the cores.
The compressive stress at the proportional limit, which is the end point of the
proportional relation between compressive stress and compressive strain, is usually
taken into account during the estimation of the Young’s modulus. According to the
test results, the ratios of the compressive strengths to the corresponding proportional
limit stresses had an average of 0.60 with a coefficient of variation 0.20, (Figure
4.13).
99
cc p cc 0.6 f ,  
(4.4)
cc,p = 0.5759fcc
R2 = 0.4222
0
1
2
3
4
5
0 1 2 3 4 5
Compressive strength (MPa)
Compressive stress at proportional limit (MPa)
Figure 4.13 : The compressive stress at proportional limit and compressive strength
relationship for the cores.
These values calculated indicate that the compressive stress at proportional limit can
be taken as the 60% of the compressive strength. In addition, taking into
consideration related compressive strains, an other correlation can be established as
given in Eq. (4.5), (Figure 4.14):
cc cc s
cc f
cc
cc p
cc p E E
f
,
, ,
,  1.6   1.6
 

(4.5)
In case the ratio in right left of this equation equal to Young’s modulus is accepted,
the strain at strength level ( ) cc, f  was found as 1.98%, substituting the relation given
in Eq. (4.3) for Young’s modulus to Eq. (4.5). This value of the strain close to the
average of the corresponding strains (2.10%).
100
Moreover, it is derived from Eq. (4.5) that as compressive stress increases, the values
of secant modulus ( ) cc,s E decreases in the remarkable amount. This means that the
increase in the compressive stress is smaller than the increase in the corresponding
vertical strain due to the crack/damage propagation.
Ecc = 1.5505Ecc,s
R2 = 0.8311
0
100
200
300
400
500
600
700
0 100 200 300 400
Secant modulus at peak (MPa)
Young's modulus (MPa)
Figure 4.14 : The Young's modulus and secant modulus at peak relationship for the
cores.
The formation of damage initiated through approximately vertical cracks at brick or
mortar phases. Then, as the initial cracks were developing, new cracks were formed
and some pieces of bricks and mortar spelled off due to widening and spreading of
the cracks. Due to further increase of the damage, the specimens failed. The damage
development and failure of CC-25 specimen are presented in Figure 4.15.
The pairs of compression stress and corresponding compressive strain at which the
visible first crack occurred are presented in Table 4.6 for several specimens. cc,cr  is
the compressive stress at first crack, and cc,cr  is the corresponding strain value. The
statistical parameters related with the stress and strain at the formation of the first
cracks are presented in
Table 4.7.
101
According to the values given in Table 4.6, the average ratio of stresses at first crack
to corresponding strengths is 0.70 with a coefficient of variation of 0.17. Using the
values in Table 4.6 and corresponding strengths, the existence of a correlation
between stresses at the first cracks and corresponding strengths is detected. The
function showing this correlation is shown by Eq. (4.6) and Figure 4.16. This
relationship confirms the average ratio of 0.70 mentioned before:
cc cr cc 0.7 f ,   (4.6)
According to this finding, the cracks can be seen visually just after exceeding the
proportional limit.
Figure 4.15 : The damage of CC-25 core during and after the compression test.
102
Table 4.6 : The stress and strain values at the first cracks for several cores.
Specimen cc,cr  (MPa) cc,cr  (%)
CC-1 3.09 1.09
CC-7 2.09 0.94
CC-8 2.52 --
CC-9 1.54 0.86
CC-15 2.27 0.88
CC-25 3.37 1.11
CC-26 1.98 0.61
CC-30 2.94 0.82
CC-31 1.63 1.23
CC-40 0.81 0.73
CC-45 3.44 1.36
Table 4.7 : The statistical parameters of the compressive stress and strain at the first
cracks.
Statistical parameter Minimum Maximum Average Stdev CoV
cc,cr  (MPa) 0.81 3.44 2.3 0.83 0.36
cc,cr  (%) 0.61 1.36 1.0 0.23 0.23
cc,cr = 0.7136fcc
R2 = 0.8018
0
1
2
3
4
5
0 1 2 3 4 5
Compressive strength (MPa)
Stress at first crack level (MPa)
Figure 4.16 : The compressive stress at the first crack and compressive strength
relationship for several cores.
103
The existence of a correlation between stress and strain values at the first crack and
at proportional limit is determined as shown by Eq. (4.7), (Figure 4.17):
cc p
cc p
cc cr
cc cr
,
,
,
, 0.96




 (4.7)
The stress-to-strain ratios in Eq. (4.7) can be taken into consideration as secant
modulus at the first crack and at the proportional limit. According to Eq. (4.7) , there
is a small reduction in stress-to-strain ratio in the transition from the proportional
limit level to the first crack level.
cc,cr/cc,cr = 0.964(cc,p/cc,p)
R2 = 0.9926
0
1
2
3
4
0 1 2 3 4
Stress-to-strain ratio at proportional limit (MPa / %)
Stress-to-strain ratio at first crack level (MPa / %)
Figure 4.17 : The stress-to-strain ratio at the first crack and at the proportional limit
relationship for several cores.
The general form of the relationship between the compressive stress and the
compressive strain is presented in Figure 4.18. Using the average values and the
relationships established above, the stresses at the first crack and at the proportional
limit with the corresponding strains are shown in Figure 4.18.
104
The remarkable outcomes of the compression tests of the cores can be summarized as
follows: While the first visible crack generally occurred at around 70% of the
compressive and at a compressive strain of 1%, the proportional limit stress was
obtained as around 60% of the compressive strength at a compressive strain of 0.8%.
The strain at compressive strength level can be considered as 2% as mentioned
above.
Figure 4.18 : The compressive stress-compressive strain relation for cores with
characteristic points.
4.1.3 Shear tests on cores
4.1.3.1 Specimen preparation
In addition to the specimen preparation steps described for the splitting and
compression tests of the cores, two lateral surfaces of the cores were also flattened
using a cement type mortar for uniform application and distribution of shear loads
and a uniform support arrangement, as shown in Figure 4.19. A total of fourteen
cores were tested under shear loads together with constant pre-compression loads
(vertical loads).
cc,f = 2
Vertical strain (%)
Compressive stress
fcc
cc,cr = 0.7fcc
cc,p = 0.6fcc
cc,p = 0.8 cc,cr = 1
105
The compression stress levels ( ) were selected as 0.05, 0.15 and 0.30 MPa to
examine the influence of the vertical stress on the shear strength. These values of
0.05, 0.15 and 0.30 MPa corresponds about the 1.6, 4.7 and 9.4% of the average core
compressive strength.
The sizes of the cores together with the thicknesses of the mortar bed joints are
presented in Table 4.8. The symbolization of the specimens, CS-0.05/0.15/0.30-
number, signed the first letters of core and shear, the level of the compressive stress
and an order number.
Figure 4.19 : The schematic view of the cores of the shear tests.
Table 4.8 : The core specimen sizes of the shear tests.
Specimen cx D (mm) c l (mm) c h (mm) mb t (mm)
CS-0.05-1 91 91 105 25
CS-0.05-2 90 90 104 20
CS-0.05-3 89 86 108 23
CS-0.05-4 91 88 105 22
CS-0.05-5 88 87 103 21
CS-0.05-6 88 89 100 21
CS-0.15-1 88 90 105 19
CS-0.15-2 91 90 111 20
CS-0.15-3 88 88 99 29
CS-0.15-4 89 85 99 26
CS-0.30-1 83 89 103 11
CS-0.30-2 91 92 101 21
CS-0.30-3 91 91 110 21
CS-0.30-4 88 87 103 26
hc
Mortar cap
Mortar cap
Mortar
a
a
Brick
Brick
Dcx
Support
surface
Shear loading
surface
Mortar
lc
a-a cross section
Brick
Brick
106
4.1.3.2 Test procedure
The instrumentation of the specimens included two LVDTs on each side to measure
horizontal displacement, a hydraulic jack to apply shear loads and a load cell to
measure the shear loads applied. The capacities of the LVDTs, the hydraulic jack and
the load cell were 10 mm, 200 kN and 200 kN, respectively. The compressive load
was applied using the Amsler testing machine with a capacity of 1000 kN and the
compressive load level was controlled through the load indicator on the test machine.
The shear load was manually applied using the hydraulic jack in a monotonic
manner. The TML TDS-303 data logger was used to collect the test data. The test
setup and the measurement system are described in Figure 4.20.
Rubber plates were placed on the surfaces of the specimens subjected to the shear
and compression loads, so as to solve the potential problems to might be resulting
from non-flat loading surfaces of specimens and to provide the uniform distribution
of the shear and compression loads. For making the movement of the slipping easier
under the compression loads, the grease lubrication was applied between the two
upper loading plates, Figure 4.20.
Figure 4.20 : The test setup with measurement system for the core shear tests.
Load cell
Hydraulic
jack
Lower plate
Specimen
LVDT 1-3
LVDT 2-4
Compression load
Shear load
LVDT 1-3
LVDT 2-4
Load indicator
107
4.1.3.3 Test results
The shear stress-horizontal relative displacement relationships are given in Figure
4.21, Figure 4.22 and Figure 4.23 for compressive stress levels of 0.05, 0.15 and 0.30
MPa, respectively.
While the shear stress was determined by the ratio of shear load to the initial area of
the mortar joint, the horizontal relative displacement was taken as the average of the
relative displacements measured on the two faces of the cores. The relative
displacement on each face was determined by subtracting the measurements of the
top LVDT from those of the bottom LVDT. Since the readings of the LVDTs of CS-
0.30-3 and CS-0.30-4 were not reliable, the diagrams of these cores could not be
plotted.
The shear strength of each core ( ) cs, f  , which was calculated considering the peak
shear load recorded during the test and corresponding displacement ( ) cs, f u are
illustrated in Table 4.9. The statistical evaluation of the test results are given in Table
4.10. As seen in this table, the average shear strengths were found as 0.38 MPa under
the compressive stress of 0.05 MPa, as 0.43 MPa under the stress of 0.15 MPa, and
as 0.48 MPa under the vertical stress of 0.30 MPa.
As the differences between the values of the compressive stresses are small and the
compressive stress levels are not high, the average shear strengths obtained for each
stress group were close to each other, and there was no a clear difference between the
shear stress-relative horizontal displacement relationships plotted for each
compressive stress group. However, as expected the average shear strength of the
cores under higher compressive stresses was higher.
108
0.0
0.2
0.4
0.6
0 2 4 6 8 10
Relative horizontal displacement (mm)
Shear stress (MPa)
CS-0.05-1
CS-0.05-2
CS-0.05-4
CS-0.05-5
CS-0.05-6
Figure 4.21 : The shear stress-relative horizontal displacement relationships for the
cores under the compressive stress of 0.05 MPa.
0.0
0.2
0.4
0.6
0 2 4 6 8 10
Relative horizontal displacement (mm)
Shear stress (MPa)
CS-0.15-1
CS-0.15-2
CS-0.15-3
CS-0.15-4
Figure 4.22 : The shear stress-relative horizontal displacement relationships for the
cores under the compressive stress of 0.15 MPa.
109
0.0
0.2
0.4
0.6
0 2 4 6 8 10
Relative horizontal displacement (mm)
Shear stress (MPa)
CS-0.30-1
CS-0.30-2
Figure 4.23 : The shear stress-relative horizontal displacement relationships for the
cores under the compressive stress of 0.30 MPa.
Table 4.9 : The results of the core shear tests.
Specimen σ (MPa) cs, f  (Map) cs f u , (mm)
CS-0.05-1 0.05 0.39 1.23
CS-0.05-2 0.05 0.40 1.10
CS-0.05-3 0.05 -- --
CS-0.05-4 0.05 0.40 0.99
CS-0.05-5 0.05 0.35 1.07
CS-0.05-6 0.05 0.35 0.86
CS-0.15-1 0.15 0.39 1.38
CS-0.15-2 0.15 0.35 0.56
CS-0.15-3 0.15 0.51 1.78
CS-0.15-4 0.15 0.46 1.00
CS-0.30-1 0.30 0.46 0.99
CS-0.30-2 0.30 0.41 1.49
CS-0.30-3 0.30 0.52 ID
CS-0.30-4 0.30 0.51 ID
110
Table 4.10 : The statistical parameters of the core shear tests.
Statistical parameter
σ = 0.05 MPa σ = 0.15 MPa σ = 0.30 MPa
cs, f 
(MPa)
cs f u ,
(mm)
cs, f 
(MPa)
cs f u ,
(mm)
cs, f 
(MPa)
cs f u ,
(mm)
Minimum 0.35 0.86 0.35 0.56 0.41 0.99
Maximum 0.40 1.23 0.51 1.38 0.52 1.49
Average 0.38 1.05 0.43 1.06 0.48 1.24
Stdev 0.03 0.14 0.07 0.37 0.05 0.35
CoV 0.08 0.13 0.16 0.35 0.10 0.28
Under low compressive stress levels, shear strength can be formulated with the
Mohr-Coulomb criterion, (Drysdale et al., 1994). According to this criterion, shear
strength ( ) is the sum of the shear strength at zero nominal compression stress;
namely, shear bond strength between mortar and unit ( o  ) and the shear frictional
strength resulting from compressive stress ( ) and friction between unit and mortar
( f  ):
    o f   (4.8)
By adopting the Mohr-Coulomb criterion and carrying out a regression analysis
between shear strengths ( cs, f  ) obtained from the core tests and the corresponding
compressive stresses, a linear relationship is obtained Eq. (4.9), (Figure 4.24).
 0.36 0.40 ,   cs f (4.9)
As seen in Eq. (4.9), the shear strength at zero nominal compression stress, ( cs,o  ),
and the friction coefficient, ( cs  ), of the cores are 0.36 MPa and 0.40, respectively.
A simple regression analysis is executed on the data of the shear stresses and
corresponding relative horizontal displacements of all vertical stress groups by
following a procedure like the case of core compression tests. The shear stresses and
relative horizontal displacements of each core are normalized with its shear strength
and corresponding relative displacement, respectively. The regression analysis ended
up with Eqs. (4.10) and (4.11), (Figure 4.25 and Figure 4.26). It should be noted that
only the data points up to peak stresses are taken into account, since at that point all
specimens suddenly lost a substantial part of their strength.
111
cs n cs n u , ,  1.08 (4.10)
cs n cs n u u
cs n ,
2
, 0.58 1.51
,
    (4.11)
In these equations, cs,n  is the normalized shear stress and cs n u , is the normalized
relative horizontal displacement. As seen the relationship between shear stress and
relative horizontal displacement might be characterized with a linear or a parabolic
function. While the predictions of Eq. (4.11) are slightly better with respect to Eq.
(4.10), Eq. (4.10) can be more preferable because of its simplicity without a
significant reduction in precision.
tcs,f = 0.3947s + 0.3642
R2 = 0.9868
0.0
0.2
0.4
0.6
0.0 0.1 0.2 0.3 0.4
Vertical stress (MPa)
Shear stress (MPa)
Figure 4.24 : The shear strength-vertical stress relationship for the cores.
112
cs,n = 1.0824ucs,n
R2 = 0.8369
0.0
0.3
0.6
0.9
1.2
0.0 0.3 0.6 0.9 1.2
Normalized relative horizontal displacement
Normalized shear stress
Figure 4.25 : The normalized shear stress normalized relative horizontal
displacement relationship (linear) for the cores.
cs,n = -0.5789ucs,n
2 + 1.5119ucs,n
R2 = 0.8914
0.0
0.3
0.6
0.9
1.2
0.0 0.3 0.6 0.9 1.2
Normalized relative horizontal displacement
Normalized shear stress
Figure 4.26 : The normalized shear stress-normalized relative horizontal
displacement relationship (parabolic).
113
cs,f/ucs,f = 0.7553(cs,p/ucs,p)
R2 = 0.7783
0.0
0.2
0.4
0.6
0.8
0.0 0.2 0.4 0.6 0.8 1.0
Stress-to-horizontal displacement ratio at proportional limit
(MPa /mm)
Stress-to-horizontal displacement ratio at strength level
(MPa / mm)
Figure 4.27 : The shear stress-to-relative horizontal displacement ratios at strength
level and at proportional limit relationship for the cores.
The formation of damage of each specimen started with a diagonal crack at mortar
joint including a relatively small part of brick close to the location where shear loads
were applied and/or with a crack along the interface of the brick-mortar joint. Then,
as the widths of the crack were opening, slipping along the mortar bed joint was
visible. As a result, the failure of the core specimens generally resulting due to the
diagonal shear crack formed in mortar with/without a part of brick and/or the
separation of the interface of the brick-mortar joint. The separation surfaces of the
specimens were generally not smooth.
The damage development with increasing relative horizontal displacement in the
post-peak region and the failure of the CS-0.15-4 specimen for each side are shown
in Figure 4.28 and Figure 4.29, respectively.
114
Figure 4.28 : The damage development of CS-0.15-4.
Figure 4.29 : The failure of CS-0.15-4.
cs = 0.19 MPa
ucs = 4 mm
cs = 0.28 MPa
ucs = 3.0 mm
cs = 0.28 MPa
ucs = 6 mm
115
4.2 Compression Tests on Wallets
4.2.1 Compression tests on wallets under monotonic loads
4.2.1.1 Specimen preparation
The aim of this study is to obtain more realistic information about the behavior of
masonry under compression loads with respect to the core tests. For this purpose,
small size original wallets were extracted from the masonry walls of the historical
houses. The thicknesses of the in-place walls did not allow extracting wallets with
flat sides. Consequently, it was necessary to cut the irregular parts of the walls for
obtaining regular prisms. The composition of the wallets were chosen to include
three rows of bricks, two bed joints and several head and longitudinal joints so as to
provide the simulation of the original wall behavior realistically. A sample of the
wallets is presented in Figure 4.30.
Figure 4.30 : The view of the several wallets of the compression test.
First row
*
*
*
*
*: head joint
First bed joint
Second bed joint
Third row
Second row
*
a a
hwt
lwt
a-a cross section
lwt
bwt
Bed joint
Head joint
Head joint
Bed joint
tmb
Head joint
tmh
Longitudinal joint
Head joint
116
The upper and lower surfaces of the wallets to be subjected to compression loads
were flattened with mortar cement. Geometrical dimensions and the ranges of the
bed and head joint thicknesses are given in Table 4.11.
It should be noted that since all surfaces of the specimens were cut, the faces of the
specimens were unclear, and so both longitudinal and head joints were called as head
joints in this study. Consequently, the head joint thicknesses given in Table 4.11 are
small relative to the real thicknesses of the head joints.
The wallet specimens tested under monotonic compression are symbolized with
WtC, which are the first and the last letters of wallet and the first letters of
compression followed by an order number. The specimens damaged during
preparation works are denoted with * in Table 4.11. The type of the damage was
generally the partial disruption of bond between a brick and a mortar bed or head
joint.
Table 4.11 : The wallet sizes for the monotonic compression test.
Specimen wt l (mm) wt b (mm) wt h (mm) mb t (mm) mh t (mm)
WtC-1* 278 230 253 12-36 9-24
WtC-2 244 197 245 14-27 14-20
WtC-3* 264 220 257 17-34 15-45
WtC-4* 281 232 237 15-27 14-25
WtC-5 231 196 253 18-40 11-36
WtC-6 213 216 252 20-33 10-22
WtC-7 214 244 269 22-38 7-40
WtC-8 241 231 270 10-31 6-44
WtC-9 245 200 289 7-38 9-27
WtC-10 238 187 265 8-30 4-38
*: The specimens damaged
4.2.1.2 Test procedure
The compression tests of the wallets were performed using the Instron Satec 1000RD
testing machine with a load capacity of 5000 kN, Figure 4.31. The closed-loop servocontrolled
testing machine has a capability of performing the test through load or
displacement control techniques. The loading procedure and loading rate was
adjusted using the Bluehill 2 software provided by the Instron.
117
After the pre-load of 5 kN was applied in load control, the test was performed in
displacement control with a rate of 0.3 mm/min, which was kept constant through the
test. Ten specimens were tested with this type of loading pattern. It should be noted
that the displacement control was done according to the shortening over all height of
the specimens. In addition to the measurement system, a maximum of 12 LVDTs
with a capacity of 25 mm were utilized for measuring displacements in vertical and
horizontal directions. The LVDTs were installed on the wallets using L-shaped
metals bonded on the specimen surfaces. Two LVDTs for vertical displacements and
one LVDT for horizontal displacements were used on each side of the wallets, Figure
4.32. The data measured by these LVDTs were stored by the TML TDS 303
datalogger. It was decided that the displacement data of the defected specimens
(denoted by * in Table 4.11) would not be reliable, only load was measured during
their tests. It should be noted that the surfaces and/or compositions of WtC-7 and
WtC-9 were not appropriate to install LVDTs in vertical-horizontal and horizontal
directions, respectively.
Figure 4.31 : The testing machine of the Instron Satec 1000RD.
Figure 4.32 : The measurement system for the wallet compression tests.
~lwt /4 ~lwt /4
LVDT 5-8
lwt
~lwt /2
~hwt /6
~hwt /6
~2hwt /3
LVDT 1-4
LVDT 9-12
: gage points
Upper loading plate
Lower loading plate
118
During tests, a pair of Teflon sheets was placed between the loading plates and the
upper/lower surfaces of the specimens, so as to reduce the friction effects resulting
from the confinement effect of the upper and lower loading plates. Furthermore, a
pair of rubber pieces was used between the loading plates and the upper/lower
surfaces of several test specimens when the surfaces of the wallets were not perfectly
flat.
4.2.1.3 Test results
Utilizing the measurements of the LVDTs and the Instron machine, the behaviors of
the specimens under compression loads were obtained. Because of the deformations
of the Teflon sheets and the rubber pieces, the displacement data measured by the
Instron machine can not be utilized directly. Yet, the data can be used to explain the
general form of the relationships of compressive stress-compressive strain until
failure, Figure B.1.
The compressive stress-vertical strain and compressive stress-horizontal strain
relationships derived using the measurements of the LVDTs are presented in Figure
4.33 and Figure 4.34, respectively. In this figures, the vertical and horizontal strains
are the average strains calculated using the measurements of the LVDTs in vertical
and horizontal directions, respectively. Figure 4.35 presents the variation of
Poisson’s ratio, which is defined as the ratio of the horizontal strain to the vertical
strain under the vertical load. The mechanical properties and the statistical
parameters of these properties are given in Table 4.12 and Table 4.13, respectively.
As mentioned above, the displacement data of the defected specimens denoted with *
in Table 4.12 is not taken into consideration. Since the surfaces and/or compositions
of WtC-7 and WtC-9 were not appropriate to install LVDTs in vertical-horizontal
and horizontal direction, respectively; Poisson’s ratios of these wallets could not
been acquired, and the corresponding cells in Table 4.12 were signed with NA (not
available). Furthermore, when damage increased, the measurements of the LVDTs
become unreliable, such case are shown by ID in Table 4.12.
119
0
1
2
3
0.000 0.005 0.010 0.015
Compressive strain
Compressive stress (MPa)
WtC-2
WtC-5
WtC-9
WtC-10
Figure 4.33 : The compressive stress-vertical strain relationships under monotonic
compression for the wallets (through LVDTs).
0
1
2
3
0.00 0.02 0.04 0.06
Horizontal strain
Compressive stress (MPa)
WtC-2
WtC-5
WtC-10
Figure 4.34 : The compressive stress-horizontal strain relationships under monotonic
compression for the wallets (through LVDTs).
120
0
1
2
3
0.00 0.01 0.02 0.03
Compressive strain
Poisson's ratio
WtC-2
WtC-5
WtC-10
Figure 4.35 : The Poisson’s ratio-vertical strain relationships for several wallets
under monotonic compression.
Table 4.12 : The results of the wallet monotonic compression tests.
Specimen f wtc (Map) v,wtc, f  (%) wtc, p  (MPa) v,wtc, p  (%) wtc E (MPa) wtc 
WtC-1* 1.16 * * * * *
WtC-2 1.14 0.94 0.66 0.26 256 0.28
WtC-3* 0.95 * * * * *
WtC-4* 1.20 * * * * *
WtC-5 1.48 ID 0.74 0.34 222 0.28
WtC-6 2.18 ID ID ID ID ID
WtC-7 2.38 NA NA NA NA NA
WtC-8 1.84 ID ID ID ID ID
WtC-9 2.48 0.85 1.29 0.23 548 NA
WtC-10 1.75 0.65 1.18 0.33 357 0.21
ID: No reliable or insufficient data
NA: Not available data
Table 4.13 : The statistical parameters of the wallet monotonic compression tests.
Statistical parameter Minimum Maximum Average Stdev CoV
wtc f (Map) 1.14 2.48 1.9 0.49 0.26
v,wtc, f  (%) 0.65 0.94 0.8 0.15 0.19
wtc, p  (MPa) 0.66 1.29 1.0 0.31 0.31
v,wtc, p  (%) 0.23 0.34 0.3 0.05 0.17
wtc E (MPa) 222 548 346 146 0.42
wtc  0.21 0.28 0.26 0.04 0.15
121
The analysis of the stresses and the strains recorded during the tests ensured the
derivation of the following results: The forms of the curves of stress-vertical strain
and stress-horizontal strain relations were similar, namely, linear part up to
proportional limit and non-linear part from proportional limit to peak. The functions
to be used for the expression of the stress-strain relationships would be derived
below, based on the data of the monotonic and cyclic tests due to the number of the
monotonic tests limited. While the mean value of the strengths, wtc f (1.14-2.48 MPa)
was 1.9 MPa with excepting of the strengths of the damaged specimens, the mean
value of all specimens was 1.7 MPa. The vertical strains at corresponding strenghs,
v,wtc, f  varied between 0.65 to 0.94% with an average of 0.8%. While the mean value
of the stresses at related proportional limits, wtc, p  between 1.29-0.66 MPa was 1.0
Map; the corresponding average strain, v,wtc, p  (0.23-0.34%) was 0.3%. Similarly, to
the result obtained before, it is understood that Young’s modulus and proportional
limit stress took higher values with higher strength. The possible correlations would
be detected below, taking into account the values of the cyclic tests. While the mean
value of Young’s moduli, wtc E (222-548 MPa) was 346 MPa, the mean value of
Poisson’s ratio, wtc  (0.21-0.28) was 0.26. When the curves showing the relations of
Poisson ratio’s-vertical strain were inspected, the fact that each curve generally
composed of four successive distinct parts; Poisson ratio’s is constant in the first
part, increasing in the second part, almost constant (a very slight variation) in the
third part and increasing rapidly in the last part, is ascertained. A reason of Poisson’s
ratio almost constant in the third part might be due to the horizontal strains taking
place out of the measurement area which were the horizontal LVDTs located on. The
last part in which Poisson ratio’s is increasing rapidly represent the formation of
failure. As shown in Figure 4.35, Poisson ratio’s can take values of larger than 1. As
the post-peak parts of the stress-strain could not be achieved, the ductilities of the
specimens could not be calculated.
The existence of non-linear behavior at pre-peak region can be figure out through the
comparison of secant moduli at proportional limit and at peak. The secant moduli at
peak were about 47-75% of those at corresponding proportional limit. This means
that there was a reduction of about 25-53% on the secant moduli, namely, the
increment rate of strain was larger than that of stress at pre-peak region.
122
4.2.2 Compression tests on wallets under cyclic loads
4.2.2.1 Specimen preparation
The wallet specimens subjected to cyclic compression loads were prepared according
to the steps addressed for the wallet specimens tested under the monotonic
compression. The wallets denoted with WtCC are described in terms of dimensions
in Table 4.14. WtCC represents the first and the last letters of wallet and the first
letters of compression and cyclic followed by an order number.
Table 4.14 : The wallet sizes for the cyclic compression tests.
Specimen wtc l (mm) wtc b (mm) wtc h (mm) mb t (mm) mh t (mm)
WtCC-1 255 189 245 10-28 9-26
WtCC-2* 165 173 277 14-26 6-34
WtCC-3 225 143 279 14-30 15-43
WtCC-4 206 161 273 10-37 4-36
WtCC-5 207 213 270 10-30 7-39
*: The specimens damaged
4.2.2.2 Test procedure
The test machine, the software, and the measurement system utilized for the cyclic
tests were the same as those utilized for the monotonic compression tests of the
wallets. Similar to the test procedure of the monotonic tests of the wallets, the cyclic
tests were carried out with the control mode of the vertical strain with a rate of 0.3
%/min kept constant throughout the tests, and each specimen was subjected to a preload
of 5 kN with a rate of 10 kN/min before the test.
The tests included loading and reloading-unloading cycles. The unloading branches
were started at the vertical strain values of 0.15, 0.5, 1, 1.5, 2, 2.5, 3, 3.5, 4, 4.5 and
5% and the reloading branches were started at 0 kN. It should be noted that the
vertical strain was the ratio of the shortening to all height of specimen, namely the
vertical strain measured by the Instron testing machine.
4.2.2.3 Test results
The compressive stress-vertical strain and compressive stress-horizontal strain
relationships are plotted using the data measured by the external LVDTs, Figure 4.36
and Figure 4.38, respectively. The envelope curves of these cyclic tests are illustrated
123
in Figure 4.37 and Figure 4.39. It should be noted that the vertical strains in Figure
4.36 are not equal to the vertical strains defined while explaining the loading pattern
since the curves were plotted using the readings of the external LVDTs, not those of
the Instron. Because of no ensuring good sticking between the L-shaped metals,
which were used for positioning the LVDTs, and specimen surface and/or taking
place crack/damage around the metals during the tests, the LVDTs of WtCC-1 and
WtCC-5 wallets could not be work until the end of the tests. Consequently, the postpeak
branches of the curves of WtCC-1 and WtCC-5 wallets could not be acquired.
The analysis of the loading-unloading branches of the curves related to vertical strain
show that the slopes of the loading-unloading branches are higher than those of the
branches remaining between the loading-unloading branches are, Figure 4.36. This
fact might be attributed to permanent strain. It is understood from the diagrams of the
compressive stress-horizontal strain that when compression load is unloaded, a small
part of horizontal strain is back, namely, a large part of the strain is permanent with
respect to the vertical strain. In the post-peak regimes, unloading and re-loading
branches of several cycles might not cut each other. This means that the strains in
horizontal direction are irrecoverable in significant amount, when the loads are
unloaded.
Figure 4.40 shows Poisson’s ratio variation with vertical strain during each test.
These diagrams were based on the envelope curves. These ratios take constant values
until corresponding proportional limits, and after then, the ratios were taking larger
values.
To have an idea on the full trend of the response of the wallets until failure, by
utilizing the data of the Instron testing machine, the diagrams and their envelope
curves were drawn, Figure B.2.
124
0
1
2
3
0.00 0.01 0.02 0.03
Compressive strain
Compressive stress (MPa)
WtCC-1
WtCC-3
WtCC-4
WtCC-5
Figure 4.36 : The compressive stress-compressive strain relationships under cyclic
loadings for several wallets (through LVDTs).
0
1
2
3
0.00 0.01 0.02 0.03
Compressive strain
Compressive stress (MPa)
WtCC-1
WtCC-3
WtCC-4
WtCC-5
Figure 4.37 : The envelope curves of the compressive stress-vertical strain under
cyclic loadings relationships of several wallets (through LVDTs).
125
0
1
2
3
0.00 0.02 0.04 0.06
Horizontal strain
Compressive stress (MPa)
WtCC-1
WtCC-3
WtCC-4
WtCC-5
Figure 4.38 : The compressive stress-horizontal strain relationships under cyclic
loadings for several wallets (through LVDTs).
0
1
2
3
0.00 0.02 0.04 0.06
Horizontal strain
Compressive stress (MPa)
WtCC-1-Envelope
WtCC-3-Envelope
WtCC-4-Envelope
WtCC-5-Envelope
Figure 4.39 : The envelope curves of the compressive stress-horizontal strain
relationships for several wallets under cyclic loads (LVDTs).
126
0
1
2
3
4
0.00 0.01 0.02 0.03
Compressive strain
Poisson's ratio
WtCC-1
WtCC-3
WtCC-4
WtCC-5
Figure 4.40 : The relationships of the Poisson’s ratio-compressive strain of several
wallets under cyclic loads.
In order to express the cyclic behavior of the wallets, several quantitative values
derived from the cyclic tests are presented in Table 4.15. Using the readings of the
LVDTs placed on the wallets and the load readings of the Instron, these quantitative
results, which are compressive strength ( fwtc ), vertical strain ( v,wtc, f  ) at strength,
compressive stress at proportional limit ( wtc, p  ) and corresponding vertical strain
( v,wtc, p  ), Young’s modulus ( wtc E ), ductility ( wtc  ) and Poisson’s ratio ( wtc  ), were
computed. The ranges of the test results were 1.26-2.20 MPa for the compressive
strengths with an average of 1.8 MPa (in case of taking into account the damaged
specimen result, the average of the compressive strength was 1.7 MPa), 0.85-0.95%
for the vertical strains at corresponding strengths with an average of 0.9%, 0.68-1.31
MPa for compressive stresses at corresponding proportional limits with an average of
1.1 MPa, 0.19-0.34% for the vertical strains at corresponding proportional limits with
an average of 0.3%, 332-442 MPa for Young’s moduli with an average of 370 MPa,
1.4-1.8 for ductilities with an average of 1.6, and 0.26-0.31 for Poisson’s ratios with
an average of 0.29, Table 4.15 and Table 4.16.
127
The cells indicated with ID showed that since no reliable data was obtained, the
related values could not be determined, Table 4.15. According to Table 4.16 showing
the statistical analysis of the results, vertical strain at proportional limit displayed a
higher deviation with respect to the others.
Non-linear behavior occurring in ascending branches might be clarified by
comparing secant moduli at peaks with the moduli at proportional limits, namely,
Young’s moduli. Consequently, it is detected that there is a range of reduction of 40-
58% on the moduli. This means that the linear relationships between compressive
stress and strain transform into the non-linear ones after the proportional limits.
Table 4.15 : The results of the wallet compression tests (cyclic).
Specimen WtCC-1 WtCC-2* WtCC-3 WtCC-4 WtCC-5
wtc f (Map) 2.20 1.58 1.75 1.26 1.91
v,wtc, f  (%) ID * 0.91 0.85 0.95
wtc, p  (MPa) 1.31 * 1.10 0.68 1.14
v,wtc, p  (%) 0.30 * 0.31 0.19 0.34
wtc E (MPa) 442 * 350 356 332
wtc  ID * 1.4 1.8 ID
wtc  0.31 * 0.29 0.29 0.26
Table 4.16 : The statistical parameters of the wallet compression tests (cyclic).
Statistical parameter Minimum Maximum Average Stdev CoV
wtc f (Map) 1.26 2.20 1.8 0.39 0.22
v,wtc, f  (%) 0.85 0.95 0.9 0.05 0.06
wtc, p  (MPa) 0.68 1.31 1.1 0.27 0.25
v,wtc, p  (%) 0.19 0.34 0.3 0.07 0.23
wtc E (MPa) 332 442 370 49 0.13
wtc  1.4 1.8 1.6 0.29 0.18
wtc  0.26 0.31 0.29 0.02 0.07
By evaluating the monotonic and cyclic test results together, the relationships of
compressive stress-vertical strain by adopting normalized procedure defined as
above, of Young’s modulus-compressive strength and compressive stress at
proportional limit-compressive stress were acquired, Figure 4.41, Figure 4.42 and
Figure 4.43, respectively. While the relationship of the stress-strain might be
128
expressed with a parabolic function, the others might be expressed with linear
functions:
wtc n wtc n wtc,n
2
, ,   0.94 1.96 (4.12)
wtc wtc E  200 f (4.13)
wtc p wtc 0.6 f ,   (4.14)
Thanks to Eq. (4.13), if required, Young’s modulus might be estimated by using the
knowledge of the compressive strength. According to Eq. (4.14), the linear relation
between the variables of compressive stress and strain continue to about 60% of
corresponding strength, and then, a non-linear relation accompanying cracks and/or
damage is developing.
wtc,n = -0.9398wtc,n
2 + 1.9612wtc,n
R2 = 0.9729
0.0
0.2
0.4
0.6
0.8
1.0
1.2
0.0 0.5 1.0 1.5
Normalized compressive strain
Normalized compressive stress
Figure 4.41 : The normalized compressive stress-normalized compressive strain
relationship for the wallets.
129
Ewtc = 204.01fwtc
R2 = 0.6989
0
100
200
300
400
500
600
0 1 2 3
Compressive strength (MPa)
Young's modulus (MPa)
Figure 4.42 : The Young’s modulus-compressive strength relationship for several
wallets.
wtc,p = 0.5787fwtc
R2 = 0.8571
0
1
2
0 1 2 3
Compressive strength (MPa)
Compressive stress at proportional limit (MPa)
Figure 4.43 : The compressive stress at proportional limit-compressive strength
relationship for several wallets.
130
The observations made related to the first visible cracks during the several tests are
presented in Table 4.17 in terms of the corresponding values of the stress and strain.
The scatter on the stress and strain values ( wtc,cr  , wtc,cr  ) is expressed with the values
of minimum, maximum, average, standard deviation and coefficient of variation
(Table 4.18).
Table 4.17 : The stress and strain values at the first cracks of the several wallets
under the compression loadings.
Specimen wtc,cr  (MPa) wtc,cr  (%)
WtC-1* 0.80 *
WtC-2 0.90 0.48
WtC-3* 0.65 *
WtC-4* 0.50 *
WtC-7 1.55 NA
WtC-8 1.10 ID
WtCC-3 1.28 0.39
WtCC-4 1.02 0.35
WtCC-5 1.35 0.40
Table 4.18 : The statistical parameters concerning the first cracks of the several
wallets under the compression loadings.
Statistical parameter Minimum Maximum Average Stdev CoV
wtc,cr  (MPa) 0.50 1.55 1.0 0.34 0.34
wtc,cr  (%) 0.35 0.48 0.4 0.05 0.13
By analyzing the relationships between the compressive stresses at the first cracks
wtc,cr  and corresponding compressive strengths, it is clarified that there was a linear
relation between both variables, Figure 4.44. The relation might be represented by
the following equation:
wtc cr wtc 0.7 f ,   (4.15)
Using this expression given by Eq. (4.15), it is possible to estimate compressive
stress induced the formation of the first crack. The fact that the visible first crack
takes place at about 70% of corresponding compressive strength might be derived
from Eq. (4.15).
131
wtc,cr = 0.6659fwtc
R2 = 0.8066
0
1
2
0 1 2 3
Compressive strength (MPa)
Compressive stress at the first crack (MPa)
Figure 4.44 : The compressive stress at the first crack-compressive strength
relationship for several wallets.
Table 4.19 : The statistical parameters of the wallet compression tests.
Statistical parameter Minimum Maximum Average Stdev CoV
wtc f (Map) 1.14 2.48 1.9 0.44 0.23
v,wtc, f  (%) 0.65 0.95 0.9 0.11 0.12
wtc, p  (MPa) 0.66 1.31 1.0 0.27 0.27
v,wtc, p  (%) 0.19 0.34 0.3 0.06 0.20
wtc E (MPa) 222 548 358 102 0.28
wtc  1.4 1.8 1.6 0.28 0.18
wtc  0.21 0.31 0.3 0.03 0.10
The observation of damage occurring at a surface of WtCC-5 during and after the
test is illustrated with about stress and strain values, Figure 4.46. The failure
mechanisms of the wallets were generally characterized with about vertical cracks
occurring in bricks and bed joints and separation between head joints and bricks.
Generally, the first crack was in about vertical direction occurring in mortar or brick.
New vertical cracks appear and propagate to the other bricks and mortar joints. The
formation of damage was not suddenly. After peak, the specimen divided into
slender parts separated through cracks.
132
Figure 4.45 : The development of failure of WtCC-5 wallet specimen.
The responses of the specimens to the compression loads can be highlighted with the
illustration given in Figure 4.49. According to this figure, the compressive stress at
A
B
D
C
1
C
3
3
A
3
B
3
D
4
B D
5
B
4
A
4
B
D
2
B
4
C
5
A
5
D
5
C
0
1
2
0.000 0.005 0.010 0.015 0.020
Vertical strain
Compressive stress (MPa)
1
2
3
4
5
2
A
133
proportional limit is about 0.6 times compressive strength and the first visible crack
takes place at about 0.7 times compressive strength.
Compressive strain (%)
Compressive stress
fwtc
wtc,cr = 0.7fwtc
wtc,p = 0.6fwtc
wtc,p = 0.3 wtc,cr = 0.4 wtc,f = 0.9
Figure 4.46 : The characteristic points on the compressive stress-compressive strain
curve for the wallets.
4.3 In-situ Shear Tests on Walls
4.3.1 In-situ shear tests on walls under monotonic loads
4.3.1.1 Specimen preparation
For determining the in-situ shear strength of masonry walls along the mortar bed
joints and comparing the obtained results with the results of the shear tests on the
cores in laboratory, in-situ destructive shear tests were carried out under monotonic
and cyclic loads. The steps to be followed for the preparation specimens of the in-situ
tests are as follows: Firstly, two sides of each wall part to be imposed to shear
loading were removed in a way that the wall part included two brick rows and three
mortar bed joints, Figure 4.47. These sides were named as first space and second
space.
134
While the first space was formed to locate shear loading system, the second space
was formed to permit the movement of the specimen in the horizontal direction and
to position the displacement measuring system. The nominal lengths of the first and
second spaces were 400 mm and 700 mm, respectively. Secondly, to prevent risks
due to excessive loading; the thickness of the wall part to be tested was reduced.
Thirdly, the surfaces to be in touch with loading system (a, b as shown in Figure
4.47) were smoothened with a cement type-mortar to provide uniform and parallel
loading surfaces. Lastly, the test side was cleaned from the remaining of bricks and
mortar. The sizes of the wall parts to be loaded (wall specimens) and the number of
stories where the tests were performed are presented in Table 4.20. The wall
specimens tested under monotonic shear (WS-E/F/S-number) were symbolized with
the first letters of wall, shear, story name (Entrance, First, Second) where the test was
carried out and an order number.
Figure 4.47 : The images of the in-situ shear test site.
The test arrangement with the loading and measurement systems are illustrated in
Figure 4.48. The loads were applied to a part of the specimen left between upper and
lower mortar bed joints, Figure 4.48. The load was manually applied using a
hydraulic jack and the load applied was measured with a 500 kN-capacity loadcell.
The horizontal displacements were measured by two LVDTs of 25 mm capacity.
Utilizing the TDS-303 data logger, the load and displacement measurements were
recorded throughout the tests. In addition to the measurement and loading devices,
700 mm
hws
lws
400 mm bws
Specimen
First space
a
lws
700 mm
hws b
Second space
400 mm
135
two U-shaped steel plates and rubber pieces were used to distribute the applied load
uniformly.
Table 4.20 : The wall specimen sizes for the in-situ shear tests.
Specimen Story ws l (mm) ws b (mm) ws h (mm) mb t (mm)
WS-E-1 Entrance 233 298 170 25-30
WS-E-2 Entrance 273 620 207 19-23
WS-E-3 Entrance 273 246 210 24-28
WS-E-4 Entrance 261 283 195 15-25
WS-F-1 First 255 310 200 17-30
WS-F-2 First 354 253 215 23-36
WS-S-1 Second 259 264 175 19-25
WS-S-2 Second 202 242 215 25-40
WS-S-3 Second 278 419 188 18-25
WS-S-4 Second 308 277 190 21-27
Figure 4.48 : The test setup with measurement system for the in-situ shear tests on
the walls.
4.3.1.2 Test results
The wall specimens tested were subjected to gravity loads as well as shear loads. The
behavior of the specimens was characterized with the shear stress-horizontal
displacement curves. The shear stress was calculated as the ratio of the shear load to
the total initial area of the upper and lower mortar joints, Eq. (4.16), and the
horizontal displacement was calculated as the average of displacements obtained
from the two LVDTs.
Upper mortar bed joint
Lower mortar bed joint
Shear load
Specimen
Upper mortar bed joint
Lower mortar bed joint
Loadcell Hydraulic jack
LVDT 1-2
136
mb ul
ws
ws A
P
,
 
(4.16)
where ws  is the shear stress, ws P is the shear load and mb ul A , is the total initial area
of the upper and lower bed joints. The shear stress-horizontal displacement curves of
the specimens are given in Figure 4.49 for entrance story tests, in Figure 4.50 for first
story tests and in Figure 4.51 for second story tests. As shown in Figure 4.50 and
Figure 4.51, the post-peak branches of WS-F-1 and WS-S-2 are shorter than those of
the other specimens. The reason of this is the severe damage forming around the
LVDTs. Consequently, the unreliable data due to the formation of this damage was
not used.
0.0
0.2
0.4
0.6
0.8
1.0
0 10 20 30 40 50
Horizontal displacement (mm)
Shear stress (MPa)
WS-E-1
WS-E-2
WS-E-3
WS-E-4
Figure 4.49: The shear stress-horizontal displacement relationships for the in-situ
walls at entrance.
The behavior of the walls under shear loads was characterized with the calculation of
the shear strength ( ws, f  ) and horizontal displacement at shear strength level ( ws f u , ),
the shear stress at proportional limit ( ws, p  ) and corresponding horizontal
displacement ( ws p u , ) of each specimen, Table 4.21. The distribution of these
parameters were statistically investigated for each story as shown in Table 4.22.
137
0.0
0.2
0.4
0.6
0.8
1.0
0 10 20 30 40 50
Horizontal displacement (mm)
Shear stress (MPa)
WS-F-1
WS-F-2
Figure 4.50: The shear stress- horizontal displacement relationships for the in-situ
walls at the first story.
0.0
0.2
0.4
0.6
0.8
1.0
0 10 20 30 40 50
Horizontal displacement (mm)
Shear stress (MPa)
WS-S-1
WS-S-2
WS-S-3
WS-S-4
Figure 4.51: The shear stress- horizontal displacement relationships for the in-situ
walls at the second story.
138
While the average shear strengths were calculated as 0.70 MPa with a standard
deviation of 0.12 MPa for the entrance story, as 0.61 MPa with a standard deviation
of 0.01 MPa for the first story and as 0.52 MPa with a standard deviation of 0.09
MPa for the second story; the average horizontal displacements at corresponding
shear strength levels were calculated as 2.02 mm with a standard deviation of 0.48
mm for entrance story, as 1.80 mm with a standard deviation of 0.06 mm for the first
story and as 2.33 mm with a standard deviation of 0.38 mm for the second story. The
average vertical stress levels of the walls were estimated as 0.07 MPa for the
entrance story walls, 0.18 MPa for the first story walls and 0.25 MPa for the second
story walls. As expected, the average values of the shear strengths are higher for
increasing vertical stresses.
Table 4.21 : The results of the in-situ monotonic shear tests on the walls.
Specimen Story ws, f  (MPa) ws f u , (mm) ws, p  (MPa) ws p u , (mm)
WS-E-1 Entrance 0.57 2.10 0.14 0.06
WS-E-2 Entrance 0.75 2.46 -- --
WS-E-3 Entrance 0.85 -- 0.18 0.06
WS-E-4 Entrance 0.63 1.50 0.19 0.05
WS-F-1 First 0.61 1.84 0.19 0.08
WS-F-2 First 0.60 1.75 0.20 0.07
WS-S-1 Second 0.65 2.81 0.21 0.13
WS-S-2 Second 0.44 2.37 0.19 0.12
WS-S-3 Second 0.51 1.91 0.16 0.09
WS-S-4 Second 0.48 2.21 0.20 0.11
The non-linear behavior is illustrated in Figure 4.52 for entrance walls, in Figure
4.53 for the first story walls and in Figure 4.54 for the second walls by focusing on
the pre-peak and around peak regions. It can be seen that the ascending branches of
the curves display three distinct characteristics: Firstly, until the shear stress of 0.20
MPa, there is a linear trend. Secondly, non-linear behavior development generally
occurs at a shear stress of ~0.20 MPa. Lastly, the rate of increment of horizontal
displacement is significantly larger than that of shear stress after the horizontal
displacement of 0.5 mm.
139
Table 4.22 : The statistical parameters of the in-situ monotonic shear tests on the
walls.
Story
Statistical
parameter
Minimum Maximum Average Stdev CoV
Entrance
ws, f  (MPa) 0.57 0.85 0.70 0.12 0.17
ws f u , (mm) 1.50 2.46 2.02 0.48 0.24
ws, p  (Map) 0.14 0.19 0.17 0.03 0.18
ws p u , (mm) 0.05 0.06 0.06 0.01 0.17
First
ws, f  (MPa) 0.60 0.61 0.61 0.01 0.02
ws f u , (mm) 1.75 1.84 1.80 0.06 0.03
ws, p  (MPa) 0.19 0.20 0.20 0.01 0.05
ws p u , (mm) 0.07 0.08 0.08 0.01 0.13
Second
ws, f  (MPa) 0.44 0.65 0.52 0.09 0.17
ws f u , (mm) 1.91 2.81 2.33 0.38 0.16
ws, p  (MPa) 0.16 0.21 0.19 0.02 0.11
ws p u , (mm) 0.09 0.13 0.11 0.02 0.18
0.0
0.2
0.4
0.6
0.8
1.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0
Horizontal displacement (mm)
Shear stress (MPa)
WS-E-1
WS-E-2
WS-E-3
WS-E-4
Figure 4.52: The pre-peak branches of the shear stress-horizontal displacement
curves for the entrance walls.
140
0.0
0.2
0.4
0.6
0.8
1.0
0.0 0.5 1.0 1.5 2.0
Horizontal displacement (mm)
Shear stress (MPa)
WS-F-1
WS-F-2
Figure 4.53: The pre-peak branches of the shear stress-horizontal displacement
curves for the first story walls.
0.0
0.2
0.4
0.6
0.8
1.0
0.0 0.5 1.0 1.5 2.0 2.5
Horizontal displacement (mm)
Shear stress (MPa)
WS-S-1
WS-S-2
WS-S-3
WS-S-4
Figure 4.54: The pre-peak branches of the shear stress-horizontal displacement
curves for the second story walls.
141
For determining the shear strength of walls having different vertical stress
magnitudes, the knowledge on the components of shear strength ( o  , f  ) is required.
These components are determined using the Mohr-Coulomb criterion based on a
linear relation between shear strength and vertical stress, Eq. (4.8). These
components are obtained using the average shear strengths obtained from in-situ
monotonic shear tests of the walls and corresponding vertical stresses, Figure 4.55:
 0.45 0.98 ,   ws f (4.17)
As shown in Eq. (4.17), the shear strength at zero nominal vertical stress, ( ws,o  ), and
the friction coefficient, ( ws  ), obtained from the in-situ monotonic shear stress were
0.45 MPa and 0.98, respectively.
The vertical stress level at a wall part tested is not exactly equal to the vertical stress
level of the wall. According to ASTM C 1531-03 (2003) , the vertical stress of the
unit tested can be taken as 1.7 times vertical stress found at any wall by means of
flatjacks. Consequently, by increasing the vertical stresses given above for the stories
in accordance of ASTM C 1531-03 (2003), ws,o  and ws  is calculated as 0.45 MPa
and 0.58, respectively. These values reflect the influence of the vertical stress on the
components of shear strength. As seen, the friction coefficient computed using
increasing vertical stresses is smaller than the coefficient computed using the vertical
stresses of the story walls. This outcome shows that it is important to compute
vertical stress correctly to make a realistic estimation of shear components.
The proportional limit of shear stress and corresponding horizontal displacement are
obtained as the ends of the linear branch of the shear stress-horizontal curves. The
average shear stresses at corresponding proportional limits were obtained as 0.17
MPa with a standard deviation of 0.03 MPa for entrance story, 0.20 MPa with a
standard deviation of 0.01 MPa for the first story, and 0.19 MPa with a standard
deviation of 0.02 MPa for the second story.
The corresponding horizontal displacements were obtained as 0.06 mm with a
standard deviation of 0.01 mm for entrance story, 0.08 mm with a standard deviation
of 0.01 mm for the first story, and 0.11 mm with a standard deviation of 0.02 mm for
the second story.
142
As seen, the average shear stresses at proportional limits and corresponding
horizontal displacements do not depend on the story where the tests were performed;
namely, gravity (compression) load levels. The scatters in the horizontal
displacements were larger with respect to the shear stresses (Table 4.22). This may
be resulting from the differences of the bonding conditions between bricks and
horizontal/vertical mortar joints, non-standard workmanship and brick/mortar
properties and the measurement of the displacements by LVDTs positioned on
different locations. ASTM C 1197-04 (2004) reports that the results of the tests
conducted on the old brick masonry have CoVs as great as 24%. Therefore, the
variation in these tests may be considered to be within the acceptable limits for old
masonry (ASTM C 1197-04, 2004). As shown in Table 4.22, the COVs of the test
results conform to the limit of 24%.
ws,f = 0.9838 + 0.446
R2 = 0.9838
0.0
0.2
0.4
0.6
0.8
0.0 0.1 0.2 0.3
Vertical stress (MPa)
Shear strength (MPa)
Figure 4.55: The shear strength-vertical stress relationship for the in-situ walls.
No correlation between shear strength and corresponding shear stress at proportional
limit was detected. The shear stress at proportional limit varied between 0.14-0.21
MPa.
143
Since there was no correlation between the shear stresses at proportional limits and
the corresponding vertical stress levels; it was decided that all shear stresses at
proportional limits may be evaluated together to find an average value of shear stress
at proportional limit for the wall specimens. While the average of the shear stress
was calculated as 0.18 MPa with a standard deviation of 0.02 MPa; the average of
the corresponding horizontal displacements varying between 0.05-0.13 mm was
calculated as 0.09 mm with a standard deviation of 0.03 mm.
A linear relationship between the ratio of the shear stress at proportional limit to the
corresponding horizontal displacement and the ratio of shear strength to the
corresponding horizontal displacement could be established (Figure 4.56):
ws p
ws p
ws f
ws f
u u ,
,
,
, 0.12
 
 (4.18)
ws,f/uws,f = 0.1221(ws,p/uws,p)
R2 = 0.819
0.0
0.1
0.2
0.3
0.4
0.5
0 1 2 3 4
Stress-to-horizontal displacement ratio at proportional limit
(MPa /mm)
Stress-to-horizontal displacement ratio at strength level
(MPa / mm)
Figure 4.56: The shear stress-to-horizontal displacement ratios at strength level and
at proportional limit relationship for the in-situ walls.
The damage evolution of the specimens tested under in-situ shear loads is shown for
a typical case in Figure 4.57. As shown in Figure 4.57, the formation of damage in
144
the walls started with thin diagonal cracks on the remaining layer of plaster covering
the upper and/or lower mortar bed joints. Then, as the pieces of the plaster layer were
falling; thin diagonal and/or thin horizontal/vertical cracks formed on the mortar
joints, and thin vertical cracks were formed on bricks. After that, new cracks
developed in the middle mortar joint and bricks as the extension of the former cracks.
As the cracks were widening, the partial separation at the interfaces of bricks and
upper/lower bed joints could be visible. Lastly, while the separation was clear and
the pieces of mortar were falling, the slip between bricks and the upper/lower bed
joints could be seen. Consequently, the specimens failed owing to the slip resulting
from the separation of the brick-upper/lower bed joints, to several vertical cracks
occurring at bricks, to several cracks at different directions at the mortar joints, and
to separations between bricks and mortar head joints. The appearances of specimen
WS-E-4 after the test are illustrated in Figure 4.58.
The pairs of shear stress, ( ) ws,cr  , and corresponding horizontal displacement,
(u ) ws,cr , causing the visible first cracks are among parameters necessary for defining
pre-peak behavior under shear loadings. In Table 4.23, these values are presented for
several specimens. The distribution of the first crack values was evaluated by
considering all story tests together; namely, without making any difference between
stories, Table 4.24. While the average shear stress and its standard deviation at the
first crack levels were 0.47 and 0.08 MPa; the corresponding average horizontal
displacement and its standard deviation were 0.66 and 0.18 mm, respectively.
145
Figure 4.57: The damage state of specimen WS-E-4 during the in-situ shear test.
ws = 0.57 MPa
uws = 0.60 mm
A
B
A D C
ws = 0.48 MPa
uws = 6.14 mm
A C
C
ws = 0.34 MPa
uws = 11.45 mm A
Lower mortar bed joint
Upper mortar bed
joint
Slip
A
ws = 0.27 MPa
uws = 19.62 mm
C
ws = 0.19 MPa
uws = 29.82 mm
A
146
Figure 4.58: The view of the WS-E-4 wall after the in-situ shear test.
Table 4.23 : The values of the shear stress and horizontal displacement at the first
cracks of the in-situ monotonic shear tests on the walls.
Specimen Story ws,cr  (MPa) ws cr u , (mm)
WS-E-4 Entrance 0.57 0.60
WS-F-2 First 0.50 0.47
WS-S-1 Second 0.52 0.80
WS-S-2 Second 0.39 0.70
WS-S-3 Second 0.39 0.46
WS-S-4 Second 0.42 0.92
Table 4.24 : The statistical parameters regarding the first cracks of the in-situ
monotonic shear test walls.
Specimen Minimum Maximum Average Stdev CoV
ws,cr  (MPa) 0.39 0.57 0.47 0.08 0.17
ws cr u , (mm) 0.46 0.92 0.66 0.18 0.27
The shear stresses at the visible first cracks are correlated with corresponding shear
strengths as given in Eq. (4.19), (Figure 4.59):
ws,cr ws, f   0.84 (4.19)
B D C B
D A
Slip Slip
A
A
B
D C
147
ws,cr = 0.8421ws,f
R2 = 0.8416
0.00
0.25
0.50
0.75
0.00 0.25 0.50 0.75 1.00
Shear strength (MPa)
Shear stress at the first crack level (MPa)
Figure 4.59: The shear stress at first crack-shear strength relationship for several insitu
walls.
A linear trend between shear stresses and corresponding horizontal displacements
initiating crack and shear strengths and corresponding horizontal displacements was
established, Figure 4.60, and this trend was formulated as follows:
ws f
ws f
ws cr
ws cr
u u ,
,
,
, 2.7
 
 (4.20)
For a better description of the specimen behavior, the main mechanical behavioral
characteristics of the in-situ wall specimens under shear loads are shown on a
representative shear stress-horizontal displacement diagram, Figure 4.61.
148
ws,cr/uws,cr = 2.6867(ws,f/uws,f)
R2 = 0.6787
0.0
0.5
1.0
1.5
0.00 0.25 0.50 0.75
Stress-to-horizontal displacement ratio at strength level
(MPa /mm)
Stress-to-horizontal displacement ratio at the first crack level
(MPa / mm)
Figure 4.60: The shear stress-to-horizontal displacement ratio at the first crack level
and at strength level relationship for several in-situ walls.
Figure 4.61: The shear stress-horizontal displacement relation of in-situ walls with
characteristic points.
0.0
0.2
0.4
0.6
0.8
0 2 4 6
Horizontal displacement (mm)
Shear stress (MPa)
uws,f
ws,f
ws,cr = 0.84ws,f
ws,p = 0.18
uws,p = 0.09 uws,cr = 0.31uws,f
0.0
0.2
0.4
0.6
0.8
0 2 4 6 8 101214161820222426283032343638404244464850
149
4.3.2 In-situ shear tests on walls under cyclic loads
4.3.2.1 Specimen preparation
For making realistic predictions related to the responses of the walls to seismic loads,
in-situ cyclic shear tests were performed on the walls. The test specimens were
prepared as explained above. The sizes of the wall specimens are given in Table
4.25. The wall specimens of the tests were symbolized with the first letters of wall,
shear, cyclic, second (story name) and an order number; namely, WSC-S-number.
All cyclic tests were performed in the second stories.
Table 4.25 : The wall specimen sizes for the in-situ cyclic shear tests.
Specimen Story ws l (mm) ws b (mm) ws h (mm) mb t (mm)
WSC-S-1 Second 243 240 191 21-28
WSC-S-2 Second 242 230 180 10-30
WSC-S-3 Second 356 254 204 15-40
WSC-S-4 Second 206 262 190 20-25
4.3.2.2 Test procedure
The cyclic shear tests were conducted through several loading and unloading cycles
until the occurrence of failure. The loading procedures of the cyclic tests consisted of
two phases, namely, load-controlled and displacement-controlled phases. The test
was controlled by load levels until around the shear strength (pre-peak load) and the
cycles were generally performed at the multipliers of 10 kN such as 10, 20, 30 kN.
Afterwards, displacement-controlled loading regime was used at post-peak
displacements and the selected amplitudes of the displacement were generally
arranged by several folds of the horizontal displacement at corresponding strength
( ws f u , ); namely the displacement at strength was increased by the multipliers of 2, 3,
4, 5, 7 and 11. Each cycle was repeated three times for each selected load or
horizontal displacement target.
As mentioned above for the in-situ monotonic shear tests, the test procedure used in
this test group is similar to the Method B given in ASTM C 1531-03 (2003). The
only difference is the loading of a part of masonry wall instead of only one brick.
The test setup can be seen in Figure 4.48.
150
4.3.2.3 Test results
The behavior of the specimens under cyclic shear loads is explained with the shear
stress-horizontal displacement curves. The shear stress was calculated as the ratio of
the shear load to the total initial area of the upper and lower mortar joints, Eq. (4.16),
and the horizontal displacement was determined as the average of displacements
measured by two LVDTs. The cyclic shear stress-horizontal displacement curves are
given in Figure 4.62. The curve of specimen WSC-S-1 could not obtained, as the
displacement data of the specimen was not reliable owing to the technical problems.
In order to compare the cyclic curves with the curves of the monotonic tests, the
envelope curves of the specimens are plotted together with the curves of the
specimens tested under monotonic shear stresses, Figure 4.63. As shown in this
figure, the strengths of the specimens tested under monotonic loads are higher than
the strengths of ones tested under cyclic loads. This may be due to strength and
stiffness degradations resulting from the cyclic loads. To show the non-linear
characteristics in the pre-peak regions, this region for each specimen is illustrated in
Figure 4.64. As shown in Figure 4.64, although the non-linear behavior taking place
under the cyclic loads share the similar characteristics with the monotonic tests, the
limit of linear relation (proportional limit) is lower than that of the monotonic tests.
The parameters reflecting material responses to cyclic shear loads, which are the
shear strength ( ws, f  ) and corresponding horizontal displacement ( ws f u , ), the shear
stress at proportional limit ( ws, p  ) and corresponding horizontal displacement ( ws p u , ),
the statistical assessment of these parameters are presented in Table 4.26 and Table
4.27, respectively. The average shear strength and average horizontal displacement
were 0.39 MPa with a standard deviation of 0.05 MPa and 1.84 mm with a standard
deviation of 0.72 mm, respectively. The average shear stress at proportional limit and
corresponding horizontal displacement were 0.13 MPa with a standard deviation of
0.03 MPa and 0.08 mm with a standard deviation of 0.01 mm, respectively.
The crack formation initiated at the shear stress and corresponding displacements of
0.18 MPa and 0.12 mm for WSC-S-2 specimen and at the shear stress and
corresponding displacement of 0.19 MPa and 0.5 mm for WSC-S-4 specimen.
151
It is seen that the values of the strength, the proportional limit and the first crack are
smaller than the corresponding values obtained from the in-situ monotonic shear
tests. This shows the influences of the loading pattern (monotonic or cyclic) on the
shear test results.
0.0
0.2
0.4
0.6
0.8
1.0
0 10 20 30 40 50
Horizontal displacement (mm)
Shear stress (MPa)
WSC-S-2
WSC-S-3
WSC-S-4
Figure 4.62: The shear stress-horizontal displacement relationships for the in-situ
walls under cyclic loadings.
0.0
0.2
0.4
0.6
0.8
1.0
0 10 20 30 40 50
Horizontal displacement (mm)
Shear stress (MPa)
WS-S-1
WS-S-2
WS-S-3
WS-S-4
WSC-S-2
WSC-S-3
WSC-S-4
Figure 4.63: The envelope curves of shear stress-horizontal displacement the
relationships for the in-situ walls.
152
0.0
0.2
0.4
0.6
0.0 0.5 1.0 1.5 2.0 2.5
Horizontal displacement (mm)
Shear stress (MPa)
WSC-S-2
WSC-S-3
WSC-S-4
Figure 4.64: The pre-peak branches of the shear stress-horizontal displacement
curves for the in-situ walls under cyclic loadings.
Table 4.26 : The results of the in-situ cyclic shear tests on the walls.
Specimen ws, f  (MPa) ws f u , (mm) ws, p  (MPa) ws p u , (mm)
WSC-S-1 0.41 ID ID ID
WSC-S-2 0.35 1.04 0.12 0.07
WSC-S-3 0.45 2.06 0.16 0.08
WSC-S-4 0.34 2.42 0.11 --
Table 4.27 : The statistical parameters of the in-situ cyclic shear tests on the walls.
Statistical parameter ws, f  (MPa) ws f u , (mm) ws, p  (MPa) ws p u , (mm)
Minimum 0.34 1.04 0.11 0.07
Maximum 0.45 2.42 0.16 0.08
Average 0.39 1.84 0.13 0.08
Stdev 0.05 0.72 0.03 0.01
CoV 0.13 0.39 0.23 0.13
153
4.4 Evaluation of Compression and Non-destructive Tests
In this section, the compression tests results of cores and wallets are evaluated in a
comparative manner as well as several explanations on non-destructive test results.
While the average compressive strength of the cores was 3.2 MPa, which of the
wallets was 1.9 MPa. This difference might be attributed to the size effect. While the
average size of the cores was 89x89x104 mm( ) c c c D l  h , the average size of the
wallets was 234x202x262 mm ( ) wt wt wt l b  h . The height of the wallets was about
2.5 times that of the cores. In addition to the size parameter, the fact that the
compositions of the cores and wallets were different might be taken as an other
parameter. The average Young’s moduli of the cores were about 70 percent of that of
the wallets. These differences may be resulting from the different gage lengths and
positions of LVDTs. While the changes in the heights of the cores were measured
over all height, those of wallets were measured about in the middle area. The mean
values of the ductility obtained for the cores and the wallets were close to each other.
The comparison of the core and wallet tests shows that the average strength obtained
from the cores should be reduced with a correction factor, which might be taken as
about 0.6 according to the tests conducted in this study. However, it should be noted
that the constant of 0.6 is obtained for the specimens in this thesis. To extract core
specimens from the masonry wall is easier than to extract wallet specimens and as
the cores have small sizes, to take large number core specimens is possible with
respect to the wallets,. Consequently, for determining average compressive strength
of any masonry wall, the compression tests may be carried out on core specimens
instead of wallet tests. However, the size effects on the results of the cores should be
eliminated.
In order to establish a relationship between rebound number of the bricks and
compressive strength of the cores, three other buildings belonging to the same
construction period as the Akaretler Row Houses were studied experimentally. These
buildings were Haci Sayid Han, Sari Apartment and Eminonu Apartment. The
average compressive strengths of the cores and the in-situ average rebound numbers
of the bricks were determined as 2.1 MPa and 28 for Haci Sayid Han, 3.9 MPa and
31 for Sari Apartment and 3.3 MPa and 30 for Eminonu Apartment, respectively.
Therefore evaluating these destructive and non-destructive test results with the
154
findings obtained for the Akaretler Row Houses, a linear relationship Eq. (4.21) is
proposed for predicting the compressive strength of masonry cores as a function of
rebound number. As seen in Figure 4.65, the proposed relationship is quite strong
with a R2 value of 0.996.
 0.60 14.57 cc br f N (4.21)
In Eq. (4.21), br N is the average rebound number of bricks and cc f is the average
compressive strength of cores. It should be noted that this equation is obtained for
low strength brick and mortar tested in this study. Therefore, the validity of this
relationship, when the mortar and brick strengths are significantly different, should
be checked. The obtained relationship follows a similar trend with the linear
relationship proposed by Brencich and Sterpi (2006), which was based on the test
data of masonry with compressive strengths of cores (150 mm diameter) in a range
of 5.3-7.6 MPa.
 0.713 17.4 cc br f N (4.22)
Akaretler Row
Houses
Eminonu
Apartment
Haci Sayid Han
Sari Apartment
fcc = 0.5947Nbr - 14.568
R2 = 0.9956
2
3
4
25 30 35
Rebound number of brick
Compressive strength of core (MPa)
Figure 4.65: The relationship between compressive strength of core and rebound
number of brick.
155
4.5 Evaluation of Shear Tests
In order to understand the behavior of masonry walls under shear effects, in-situ
shear tests of the walls and shear tests of the core specimens at laboratory were
carried out. While the in-situ test specimens were composed of two row bricks, two
mortar bed joints and several mortar head joints, the core specimens were composed
of two parts of bricks and one mortar bed joint.
As expected, one of the conclusions to be inferred from both shear tests is that there
is a clear trend for the wall specimens under the high vertical stresses to display
higher values of shear strength.
The average shear strengths and corresponding horizontal displacements of the core
tests were obtained as 0.38 MPa and 1.05 mm for the vertical stress of 0.05 MPa, as
0.43 MPa and 1.06 mm for the vertical stress of 0.15 MPa and as 0.48 MPa and 1.24
mm for the vertical stress of 0.30 MPa, Figure 4.66. The results of the in-situ tests
were obtained as 0.52 MPa and 2.33 mm for the vertical stress of 0.07 MPa, as 0.61
MPa and 1.80 mm for the vertical stress of 0.18 MPa, and as 0.70 MPa and 2.02 mm
for the vertical stress of 0.25 MPa, Figure 4.66. It is understood from the average
values that although the vertical stresses are close to each other for both test types,
the shear strengths and horizontal displacements of the cores were lower than those
of the in-situ walls. This difference may be attributed to several factors such as
material variability and possible damages occurred during extraction, transportation,
and/or preparation processes of the core specimens. In addition to these factors, the
differences between the test techniques and between the specimen composition and
sizes (while in-situ walls included two mortar bed joints and several head joints, the
cores included one bed joint and no head joint) might have lead to the attained
different values.
The components of shear strength ( o  and f  ) were obtained from both shear tests,
Figure 4.67. The fact that the components of the in-situ shear tests (  0.45 o  MPa
and  0.98 f  ) were higher than those of the core tests (  0.36 o  MPa and
0.40 f   ) , Figure 4.67, may be resulting from the factors given above. However,
when taking into account the statement about the estimation of the vertical stresses of
in-situ walls in ASTM C 1531-03 (2003), the differences between the coefficients of
156
friction, which were obtained from the core and in-situ shear tests, decreases. The
characteristic values of o  and f  are calculated by multiplying the corresponding
average values obtained from the shear tests by 0.80 in accordance with TS EN
1052-3 (2004). According to this statement, while the characteristic values of o  and
f  obtained from the is-situ tests were 0.36 MPa and 0.78; those obtained from the
core tests were 0.29 MPa and 0.32, respectively.
The number of the cyclic in-situ shear tests is not sufficient to derive relationships
showing the effect of monotonic and cyclic loads. However, the comparison of the
cyclic and monotonic tests carried out on the second stories shows that there is
strength degradation due to cyclic load pattern.
0.0
0.2
0.4
0.6
0.8
0.05 0.07 0.15 0.18 0.30 0.25
Vertical stress (MPa)
Shear strength (MPa)
Core tests
In-situ tests
Figure 4.66: The comparison of shear strength-vertical stress values for the core and
in-situ tests.
157
tcs,f = 0.3947s + 0.3642
R2 = 0.9868
ws,f = 0.9838 + 0.446
R2 = 0.9838
0.0
0.2
0.4
0.6
0.8
0.0 0.1 0.2 0.3 0.4
Vertical stress (MPa)
Shear strength (MPa)
Core tests
In-situ tests
Figure 4.67: The shear strength-vertical stress relationships for the in-situ walls and
cores.
In order to compare the shear components (shear bond strength and friction
coefficient) of the cores, three other buildings belonging to the same construction
period as the Akaretler Row Houses were studied experimentally. These buildings
were Haci Sayid Han, Haydarpasa Hospital and Ozturk Apartment. The average
compressive strengths, shear bond strengths and the friction coefficients of the cores
tested under similar conditions are presented in Table 4.28. As seen in this table,
while the average compressive strengths of the cores taken from these building walls
were in the range of 2.1-3.8 MPa; the shear bond strengths and the friction
coefficients were in ranges of 0.22-0.48 MPa and of 0.27-1.18, respectively.
Table 4.28 : The average compressive strengths and shear strength components of
the other buildings.
Building cc f (MPa) o  (MPa) f 
Haci Sayid Han 2.1 0.22 1.18
Haydarpasa Hospital 2.9 0.30 0.68
Ozturk Apartment 3.8 0.48 0.27
Akaretler Row Houses 3.2 0.36 0.40
158
It was derived from the compression and shear test results of the cores that the
compressive strengths might be correlated with the shear bond strengths and with the
friction coefficients as shown in Figure 4.68 and Figure 4.69, respectively. These
correlations can be expressed with the following exponential functions:
fcc
cs o e 0.5
,   0.08
(4.23)
fcc
cs e 0.9 7.9   
(4.24)
As shown in these figures, while the shear bond strengths increase with increasing
values of the compressive strength, the friction coefficients decrease.
cs,o= 0.082e0.4611fcc
R2 = 0.9932
0.0
0.2
0.4
0.6
2.0 2.5 3.0 3.5 4.0
Compressive strength of core (MPa)
Shear bond strength of core (MPa)
Haci Sayid Han
Haydarpasa Hospital
Akaretler Row Houses
Ozturk Apartment
Figure 4.68: The shear bond strength-compressive strength relationship for the
cores.
159
cs = 7.9312e-0.8941fcc
R2 = 0.9712
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
2.0 2.5 3.0 3.5 4.0
Compressive strength of core (MPa)
Friction coefficient of core
Haci Sayid Han
Haydarpasa Hospital
Akaretler Row Houses
Ozturk Apartment
Figure 4.69: The friction coefficient-compressive strength relationship for the cores.
160
161
5. TESTS ON PRISMS AND WALLS BUILT WITH HISTORICAL BRICKS
AND REPRODUCED MORTAR
Since the process of taking test specimens from existing structures is destructive, the
application of this process on the historical structures is not generally allowed. Even
if allowed, to take non-damaged specimens and to take appropriate specimens in
terms of number, size required for the test type, and composition required for the
simulation of in-place load-bearing masonry may not be possible. In such cases,
masonry properties may be identified through tests performed on the reproduced
specimens.
Consequently, this chapter of the thesis is assigned to compression and shear tests on
the specimens constructed with the historical bricks and reproduced mortar. The
historical bricks were collected from the load-bearing walls of the Akaretler Row
Houses. Reproduced mortar was arranged in a way to conform the mechanical
characteristics of the in-place mortar.
Two distinct specimen types (prism and wall) in terms of size and composition were
built in accordance with the related statements of FEMA 356 (2000), EN 1996-1-1
(2005) and BCRSMS (2008).
This chapter studies the subjects summarized below:
 The mechanical characteristics of the reproduced masonry specimens (prism and
wall), which were constructed with the original bricks and hybrid mortar produced,
which had similar mechanical features to the in-place mortar, under compression
and shear loadings,
 The differences between the test results of the reproduced specimens (prism and
wall) and of the original specimens (core, wallet and in-situ wall)
 The effects of test type, specimen size and composition on the test results,
The prism and wall tests given in this Chapter were conducted at Istanbul Technical
University, Civil Engineering Faculty, Structural Material Laboratory, and Structural
& Earthquake Engineering Laboratory, respectively.
162
5.1 The Reproduced Mortar
Masonry prism and masonry wall specimens for compression and shear tests were
constructed with original (extracted or historical) bricks and reproduced mortar,
which was produced in a way to conform to mechanical characteristics of original
mortar.
Although original mortar included only lime as binder, both cement and lime were
used for reproduced mortar as the process of hardening of lime mortar takes more
time than that of cement or hybrid mortar. Cement, lime, and sand used for mortar
production are the commercial productions of Lafarge, Paksan, and BASF firms,
respectively, Figure 5.1. While cement used was Portland cement, whose
compressive strength at 28 days of age is 42.5 MPa; lime was hydrated lime and sand
was the filler of epoxy based structural repair mortar.
In order to find appropriate proportion of cement, lime, sand, and water, a series of
trial mixtures was made. The appropriate proportion of cement, lime, sand, and water
was found as 01:02:15:2.9 in weight, respectively. Reproduced mortar production
steps are presented in Figure 5.2. Firstly, sand, lime, and cement were put in a plastic
vessel; secondly, these ingredients were mixed and lastly, the mass was mixed by a
workman until obtaining a homogeneous mixture by adding water.
Figure 5.1 : The ingredients of the reproduced mortar.
In order to follow the variation of mechanical characteristics of reproduced mortar by
time, flexural and compression tests were performed at the ages of 7, 14, 28, 90 and
210 days following the production of reproduced mortar.
163
The flexural tests were performed on prisms of mortar having nominal dimensions of
40x40x160 mm. The compression tests were carried on mortar specimens obtained
from the halves of the flexural test specimens. The mortar specimen preparation, and
flexural and compression tests were carried out in accordance with provisions of TS
EN 1015-11 (2000).
Figure 5.2 : The production steps of the reproduced mortar.
5.2 Mechanical Tests on Reproduced Mortar
5.2.1 Flexural tests on reproduced mortar
5.2.1.1 Specimen preparation
Mortar test specimens were produced randomly from mortar mixtures produced for
the construction of masonry prisms, triplets, and walls. Steel molds having nominal
dimensions of 40×40×160 mm were used for the specimens. Before the molds were
filled with mortar; the molds were cleaned and oiled for reducing mortar sticking to
molds. Mortar was put on each mold in two layers. After the first layer was
compacted by shaking on a small table, the second layer was put and again
compacted by vibration. The upper surface of the mortar in the mold was smoothed
by a trowel. The mortar specimens were left in the molds for 5 days. Then, the
specimens were extracted from the molds and left in a room with a temperature of
21-22 °C until test days.
164
The specimens were symbolized by RMF, which was composed of the first capital
letters of the words of reproduced, mortar and flexural, and two numbers, which
indicate specimen age at testing date and specimen order. The sizes of the specimens
of length ( rm l ), width ( rm b ), and height ( rm h ) are given in Table 5.1.
Table 5.1 : The reproduced mortar specimen sizes of the flexural tension tests.
Specimen rm l (mm) rm b (mm) rm h (mm)
RMF-7-1 160 40 39
RMF-7-2 160 40 40
RMF-7-3 160 39 39
RMF-14-1 160 40 40
RMF-14-2 160 41 40
RMF-14-3 160 40 39
RMF-28-1 160 40 39
RMF-28-2 160 40 40
RMF-28-3 160 40 39
RMF-90-1 161 40 39
RMF-90-2 162 39 39
RMF-90-3 161 39 39
RMF-210-1 162 40 39
RMF-210-2 163 40 39
RMF-210-3 160 39 40
5.2.1.2 Test procedure
The flexural test procedure based on a three-point bending test configuration was
utilized to estimate the flexural strength of the mortar specimens, ( ) rmf f (Figure
3.21). The distance between the centers of the supports was equal to 100 mm ( rm L ).
The peak load resisted by the specimen was calculated from the readings of a ring,
whose capacity was 5.0 kN, located on the loading rod.
5.2.1.3 Test results
The flexural tensile strength ( ) rmf f determined using Eq. (3.1) for each specimen
and average flexural tensile strength ( ) rmf ,a f , for each test age are presented in Table
5.2. In order to observe the variability of the flexural tensile strength of reproduced
mortar by age, the relationship of flexural tensile strength and age is presented in
Figure 5.3.
165
As shown in Figure 5.3, the increment rate of strength slows down as age increases
and after the age of 90 days, the flexural strength may be considered as 0.85 MPa.
Table 5.2 : The results of the reproduced mortar flexural tension tests.
Specimen Age f rmf (MPa) rmf a f , (MPa)
RMF-7-1 7 0.61
RMF-7-2 7 -- 0.6
RMF-7-3 7 0.59
RMF-14-1 14 0.71
RMF-14-2 14 ID 0.7
RMF-14-3 14 0.66
RMF-28-1 28 ID
RMF-28-2 28 0.94 0.8
RMF-28-3 28 0.66
RMF-90-1 90 --
RMF-90-2 90 0.93 0.9
RMF-90-3 90 0.93
RMF-210-1 210 0.77
RMF-210-2 210 0.77 0.8
RMF-210-3 210 0.84
0.0
0.2
0.4
0.6
0.8
1.0
0 50 100 150 200 250
Age (day)
Flexural tensile strength (MPa)
Figure 5.3 : The development of flexural tensile strength by age for the reproduced
mortar.
166
The statistical evaluation of the flexural test results was carried out and tabulated in
Table 5.3 for each age group. All mortar specimens failed due to an approximately
vertical crack occurring at midspan as shown in Figure 5.4.
Table 5.3 : The statistical parameters of the reproduced mortar flexural tension tests.
Age (day) n Stdev (MPa) CoV
7 3 0.01 0.02
14 3 0.04 0.06
28 3 0.20 0.25
90 3 0.00 0.00
210 3 0.04 0.05
Figure 5.4 : The view of RMF-210-3 mortar specimen after the flexural test.
5.2.2 Compression tests on reproduced mortar
5.2.2.1 Specimen preparation
In order to obtain an average compressive strength of the reproduced mortar for
comparing with the strength of the original mortar and to observe development of the
strength of reproduced mortar with time, a total of 30 mortar specimens at different
age were tested under compression loads. According to TS EN 1015-11 (2000), the
compression tests were performed on the halves of the mortar specimens obtained
from the flexural tests. The sizes of the specimens are given in Table 5.4. The
specimens were symbolized by the first capitals of reproduced, mortar, compression
and flexural, and A or B letters which were used for showing each half of a
specimen.
5.2.2.2 Test procedure
Compression tests on reproduced mortar were conducted by adopting the test
procedure given for original mortar (TS EN 1015-11, 2000), (Figure 3.23).
167
Table 5.4 : The reproduced mortar specimen sizes for the compression tests.
Specimen Age (day) rm l (mm) rm b (mm) rm h (mm)
RMCF-7-1-A 7 40 40 39
RMCF-7-1-B 7 40 40 39
RMCF-7-2-A 7 40 40 40
RMCF-7-2-B 7 40 40 40
RMCF-7-3-A 7 39 40 39
RMCF-7-3-B 7 39 40 39
RMCF-14-1-A 14 40 40 40
RMCF-14-1-B 14 41 40 40
RMCF-14-2-A 14 41 40 40
RMCF-14-2-B 14 40 40 40
RMCF-14-3-A 14 40 40 39
RMCF-14-3-B 14 40 40 39
RMCF-28-1-A 28 40 40 39
RMCF-28-1-B 28 40 40 39
RMCF-28-2-A 28 40 40 40
RMCF-28-2-B 28 40 40 40
RMCF-28-3-A 28 40 40 39
RMCF-28-3-B 28 40 40 39
RMCF-90-1-A 90 40 40 39
RMCF-90-1-B 90 40 62 39
RMCF-90-2-A 90 41 40 39
RMCF-90-2-B 90 40 62 39
RMCF-90-3-A 90 39 39 39
RMCF-90-3-B 90 39 39 39
RMCF-210-1-A 210 40 40 39
RMCF-210-1-B 210 40 40 39
RMCF-210-2-A 210 40 40 39
RMCF-210-2-B 210 40 40 39
RMCF-210-3-A 210 40 39 40
RMCF-210-3-B 210 40 39 40
5.2.2.3 Test results
Compressive strength of each mortar specimen and average compressive strength of
each age group are given in Table 5.5.
The average compressive strengths of the specimens at ages of 7, 14, 28, 90 and 210
days were 1.2, 1.6, 2.5, 3.1 and 2.9 MPa, respectively. The values of the standard
deviation (Stdev) and coefficient of variation (CoV) are presented in Table 5.6.
There is no statement concerning limit of standard deviation or coefficient of
variation in TS EN 1015-11 (2000), but these values are believed to be acceptable.
A result to be drawn from these strength values was that the compressive strength of
reproduced mortar generally increased with time, but increment rate reduce by time,
Figure 5.5.
168
Table 5.5 : The results of the reproduced mortar compression tests.
Specimen Age (day) f rmc (MPa) rmc a f , (MPa)
RMCF-7-1-A 7 1.13
1.13
1.2
RMCF-7-1-B 7 1.13
RMCF-7-2-A 7 1.19
1.19
RMCF-7-2-B 7 --
RMCF-7-3-A 7 1.22
1.22
RMCF-7-3-B 7 1.22
RMCF-14-1-A 14 1.63
1.66
1.6
RMCF-14-1-B 14 1.69
RMCF-14-2-A 14 1.59
1.62
RMCF-14-2-B 14 1.65
RMCF-14-3-A 14 1.56
1.53
RMCF-14-3-B 14 1.50
RMCF-28-1-A 28 2.04
1.98
2.5
RMCF-28-1-B 28 1.92
RMCF-28-2-A 28 2.66
2.74
RMCF-28-2-B 28 2.82
RMCF-28-3-A 28 2.66
2.64
RMCF-28-3-B 28 2.61
RMCF-90-1-A 90 --
3.06
3.1
RMCF-90-1-B 90 3.06
RMCF-90-2-A 90 2.95
3.24
RMCF-90-2-B 90 3.53
RMCF-90-3-A 90 3.01
3.08
RMCF-90-3-B 90 3.14
RMCF-210-1-A 210 2.90
2.89
2.9
RMCF-210-1-B 210 2.88
RMCF-210-2-A 210 3.15
3.11
RMCF-210-2-B 210 3.07
RMCF-210-3-A 210 2.50
2.50
RMCF-210-3-B 210 --
Table 5.6 : The statistical parameters of the reproduced mortar compression tests.
Age n Stdev (MPa) CoV
7 3 0.05 0.04
14 3 0.07 0.04
28 3 0.37 0.15
90 3 0.23 0.07
210 3 0.25 0.09
During the compression tests of reproduced mortar specimens at age of 210 days,
compression loads and changes over all height of each specimen by increasing loads
were recorded. Utilizing these readings, the stress and strain relation was plotted for
each specimen as shown in Figure 5.6.
169
By normalizing stress and strain values of each specimen with its strength and its
strain at strength level and by evaluating for these normalized values of all
reproduced mortar specimens at the age of 210 days, the relation of normalized
stresses and normalized strains was expressed with a parabolic function:
rmc ,n
2
rmc ,n rmc ,n   0.87 1.88 (5.1)
rmc,n  is the normalized compressive stress of the reproduced mortar and rmc,n  is the
normalized compressive strain of reproduced mortar.
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
0 50 100 150 200 250
Age (day)
Compressive strength (MPa)
Figure 5.5 : The development of compressive strength with age for the reproduced
mortar.
An increase in compressive strength generally leads to an increase in Young’s
modulus. As shown Figure 5.8, the relationship between compressive strength and
Young's modulus can be represented by a linear function;
rmc rmc E  77 f (5.2)
The specimens failed after formation of approximately vertical cracks. The
appearances of RMCF-28-1-A specimen during and after the test are given in Figure
5.9.
170
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
0.00 0.01 0.02 0.03 0.04 0.05 0.06
Compressive strain
Compressive strength (MPa)
RMCF-210-1-A
RMCF-210-1-B
RMCF-210-2-A
RMCF-210-2-B
RMCF-210-3-A
Figure 5.6 : The compressive stress-compressive strain relationships for reproduced
mortar at the age of 210 days.
rmc,n = -0.8691rmc,n
2 + 1.8761rmc,n
R2 = 0.9965
0.0
0.3
0.6
0.9
1.2
0.0 0.3 0.6 0.9 1.2 1.5
Normalized compressive strain
Normalized compressive stress
Figure 5.7 : The normalized compressive stress-normalized compressive strain
relationship for reproduced mortar at the age of 210 days.
171
Table 5.7 : Young’s modulus and ductility of the reproduced mortar specimens at
the age of 210 days.
Specimen Age (day) Ermc (MPa) rmc a E , (MPa) rmc  rmc,a 
RMCF-210-1-A 210 230
224
1.4
1.5
RMCF-210-1-B 210 221 1.6
RMCF-210-2-A 210 257 1.6
RMCF-210-2-B 210 221 1.4
RMCF-210-3-A 210 193 1.6
RMCF-210-3-B 210 -- 1.4
Ermc = 77.396frmc
R2 = 0.7695
150
200
250
300
2.0 2.5 3.0 3.5
Compressive strength (MPa)
Young's modulus (MPa)
Figure 5.8 : Young’s modulus-compressive strength relationship for the reproduced
mortar specimens at the age of 210 days.
Figure 5.9 : The appearances of RMCF-28-1-A mortar specimen during and after the
compression test.
172
5.3 Evaluation of Reproduced Mortar Tests
In this study, the ratios of flexural strength to compressive strength for reproduced
mortar were determined as 0.50 at 7 days, 0.44 at 14 days, 0.32 at 28 days, 0.29 at 90
days, and 0.28 at 210 days. By evaluating these strength values, the existence of any
correlation between these strengths was investigated. The correlation obtained is
shown in Figure 5.10. The function of the correlation may be used to calculate
flexural strength of hybrid (cement-lime) mortar:
0.5 0.51 rmf rmc f  f (5.3)
The flexural tensile strength of concrete can be calculated from the following
equation (TS 500, 2000):
0.5
, , 2 0.35 cft c cc c f   f (5.4)
where cft c f , is the characteristic flexural tensile strength of concrete, and cc c f , is the
characteristic compressive strength of concrete. The value of 2.0 in Eq. (5.4) is used
to transform direct tensile strength to flexural tensile strength, (TS 500, 2000). As
shown in Eqs. (5.3) and (5.4), the multiplier of 0.70 for concrete is larger than that of
0.51 for reproduced mortar.
It was observed that the compressive strength development rate of the reproduced
mortar with time was generally less than that of concrete, (Table 5.8). This state may
be resulting from the usage of lime in the reproduced mortar and from the differences
between sizes of the reproduced mortar in this study and concrete specimens. It
should be noted that the increment rate is calculated as the ratio of the strength at a
specified test day to the strength at the age of 28 days.
Table 5.8 : The increment rate of compressive strength for the concrete and
reproduced mortar.
Age (day)
Concrete
(Berktay, 1995)
The reproduced
mortar in this study
7 0.65 0.48
14 0.64
28 1.00 1.00
90 1.20 1.24
210 1.35 1.16
173
The compressive and flexural strengths of mortar depend on several parameters such
as the type, size and distribution of aggregate grain, workmanship and water-binder
ratio (Bayülke, 1980).
The flexural and compressive strengths of several mortars, which are complied from
the literature, are summarized in Table 5.9. In the study of Bayülke (1980); the 28
days compressive strengths of two mortar mixtures including the ingredients of
cement:lime:sand of 1:1:6 and 1:1.5:8 by weight were given as 2.8-7 MPa and 3.5
MPa, respectively.
frmf = 0.5093(frmc)0.5
R2 = 0.8182
0.0
0.5
1.0
0.0 0.5 1.0 1.5 2.0
Compressive strength0.5 (MPa)
Flexural strength (MPa)
Figure 5.10 : Flexural tensile and compressive strength relationship for the
reproduced mortar.
Baronio et al. (1999) determined the flexural and compressive strengths with time for
different mortar types having a ratio of binder to aggregate of 1/3 and including two
type of hydrated lime as binder, Table 5.9. The results showed the influence of lime
type on flexural and compressive strength. The flexural strengths obtained for two
types of limes were not equal to each other. The flexural strengths and compressive
strengths at the age of 28 days were determined as 0.36-0.45 MPa and 0.73-1.06
MPa, respectively.
174
The pairs of flexural-compressive strengths at 28 days were 1.16-4.70 MPa, 0.96-
3.71 MPa, 1.07-3.30 MPa for a mortar mixture having the volumetric ratio of
cement:lime:sand of 1:1.5:8 (Yorulmaz and Atan,1971), Table 5.9. In order to
understand the validity of Eq. (5.3), the flexural tensile strengths accompanying
compressive strengths in Table 5.9 are calculated using Eq. (5.3). As seen in these
tables, while ratios of the flexural tensile strengths (test/prediction) are close to 1.0
for the first two studies, these ratios are generally higher than 1.0 for the last studies
(Url-1, 2006). This may be due to evaluation of all mortar specimens together
without taking into account the ages of the specimens.
Table 5.9 : The flexural and compressive strengths of several mortars compiled from
the literature.
Study
Mixture ratio
(C:L:S)
Age
(day)
rmf f
(MPa)
rmc f
(MPa)
rmf rmc f f
Prediction
(Eq. (5.3)) Test/Prediction
rmf f (MPa)
Yorulmaz
and Atan
(1971)
1:1.5:8
28 1.16 4.70 0.25 1.11 1.0
28 0.96 3.71 0.26 0.98 1.0
28 1.07 3.30 0.32 0.93 1.2
Baronio
et al. (1999)
1(C+L):3
(Volume)
(the first lime
type)
28 0.36 0.73 0.49 0.44 0.8
90 0.51 1.39 0.37 0.60 0.8
180 0.67 2.33 0.29 0.78 0.9
360 0.76 2.42 0.31 0.79 1.0
720 0.78 2.44 0.32 0.80 1.0
1(C+L):3
(Volume)
(the second
lime type)
28 0.45 1.06 0.42 0.53 0.9
90 0.60 2.14 0.28 0.75 0.8
180 0.80 2.65 0.30 0.83 1.0
360 0.89 2.91 0.31 0.87 1.0
Tests and
research –
Lime
mortar tests
(Url-1,
2006)
1:1:6
(Volume)
7 2.05 5.02 0.41 1.14 1.8
28 1.95 7.70 0.25 1.42 1.4
180 2.10 8.10 0.26 1.45 1.4
360 2.20 8.70 0.25 1.50 1.5
720 2.20 8.50 0.26 1.49 1.5
1:2:9
(Volume)
7 1.65 4.96 0.33 1.14 1.5
28 1.55 5.56 0.28 1.20 1.3
180 1.50 5.75 0.26 1.22 1.2
360 1.70 6.05 0.28 1.25 1.4
720 1.75 5.95 0.29 1.24 1.4
175
5.4 Comparison of Original Mortar and Reproduced Mortar
Flexural and compression tests were performed on both original and reproduced
mortar specimens. While the average flexural tensile and average compressive
strengths of reproduced mortar were 0.9-0.8 MPa and 3.1-2.9 MPa at the age of 90-
210 days, those of original mortar were 1.3 and 3.2 MPa, respectively, Figure 5.11.
While the average modulus of elasticity and ductility of reproduced mortar at the age
of 210 days were 224 MPa and 1.5, those of original mortar were 232 MPa and 1.9,
respectively. These results show that the average flexural and compressive strengths
of the reproduced mortar were 65 and 94% of the corresponding values of the
original mortar, respectively. The average compressive strength of the reproduced
mortar was close to that of the original mortar.
As the values of the strengths required were small and so these values were sensitive
to small amount of itself ingredients, a small error made in amount of any ingredient
of the reproduced mortar might have been lead to the deviation of flexural strength.
However, when comparing mechanical parameters of the reproduced and original
mortar (compressive and flexural tensile strengths, Young's modulus and ductility), it
is concluded that the reproduced mortar can sufficiently represent the original
mortar.
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
Reproduced mortar
at 90 days
Reproduced mortar
at 210 days
Original mortar
Strength (MPa)
Flexural tensile
Compressive
Figure 5.11 : The comparison of the flexural tensile and compressive strengths of
the original mortar and reproduced mortar.
176
While the ratio of average flexural strength to average compressive strength for
original mortar was 0.41, the ratios for reproduced mortar were determined as 0.50 at
7 days, 0.44 at 14 days, 0.32 at 28 days, 0.29 at 90 days, and 0.28 at 210 days.
The functions showing the relation of modulus of elasticity and compressive strength
for original and reproduced mortars were expressed in equations of similar form and
their constants were determined as 76 and 77, respectively. The parabolic equations
of normalized stresses and normalized strains obtained for the original mortar and the
reproduced mortar have also similar coefficients.
While the flexural test specimens of the original and reproduced mortar failed due to
a crack at midspan, the compression test specimens failed due to spreading vertical
cracks. Consequently, the similarity of various mechanical characteristics and failure
modes confirm that the reproduced mortar can sufficiently represent the original
mortar.
5.5 Mechanical Tests on Masonry Prisms
5.5.1 Compression tests on masonry prisms under monotonic loads
5.5.1.1 Specimen preparation
As mentioned above, FEMA 356 (2000) recommends the use of prism specimens for
the determination of masonry characteristics of existing masonry structures under
compression loads as an alternative of the testing of the prisms extracted from the
existing walls. The test method is detailed in ASTM C 1314-03b (2003).
The prism specimens were constructed with the historical bricks collected and
reproduced mortar, according to ASTM C 1314-03b (2003). The bricks cleaned from
the remaining of the original mortar were chosen randomly to represent the historical
walls of the houses. The construction process of the prisms is illustrated in Figure
5.12. The prisms contained three bricks in stack bond with two mortar bed joints. It
should be noted that the mortar used for the construction of the prisms was the
reproduced mortar.
177
The prisms were generally constructed with bricks whose lengths were reduced and
the nominal thicknesses of the joints were 20 mm. This value is considered to
represent the average bed joint thickness of the historical masonry houses. As the
selection of the bricks was random, the heights of the bricks in prisms were varying.
The bricks were put into a vessel of water for a few minutes before the construction
for reducing the water absorption rate of the bricks from the mortar. If this was not
done, the bricks might absorb a portion of the water in mortar leading to low quality
bonding between unit and mortar. As shown in Figure 5.12, high strength cementbased
mortar was used for smoothing the lower and upper surfaces of each prism and
providing parallel loading surfaces. Wooden laths were used for providing constant
joint thickness along each joint.
However, as the surfaces of the bricks were not parallel and not uniform, the joint
thicknesses, which were formed with the reproduced mortar, varied in a range. After
capping the upper surface of the specimen, the wooden laths were extracted from the
mortar joints, and the empty spaces of the joints were filled with mortar.
Figure 5.12 : The construction steps of the prisms.
The geometrical definition and composition of the prisms are given in Figure 5.13.
The length ( p l ), width ( p b ), height ( p h ) and the range of the mortar joint thicknesses
( mb t ) of each specimen can be seen in Table 5.10. The specimens were denoted with
Cement mortar
for capping
Capping
Reproduced mortar
for mortar joint
178
PC, which was the capital letters of prism and compression. While the length and
width dimensions were determined as the mean value of the four measurements at
the top and bottom edges of the corresponding two surfaces, and the height was
determined as the mean value of the four measurements taken at the middle of all
surfaces. The requirements of ASTM C 1314-03b (2003) are to build the prisms with
full-size or reduced length bricks, to ensure a minimum length of 100 mm for each
prism, to construct the prisms at least two units high and to ensure height-tothickness
( p p h t ) ratios of 1.3 and 5.0. The thickness of each specimen is defined as
the minimum value of length and width. The prisms of this study were constructed
with the reduced length bricks and they were three bricks high. The minimum length
of the prisms was 138 mm and the height-to-thickness ratios varied from 1.9 to 2.1.
Consequently, it is seen that the prisms conformed to the requirements given in
ASTM C 1314-03b (2003).
Figure 5.13 : The view of the prisms.
Table 5.10 : The prism sizes for the monotonic compression tests.
Specimen p l (mm) p b (mm) p h (mm) mb t (mm)
PC-1 143 127 252 22-29
PC-2 143 121 235 15-21
PC-3 138 130 252 20-28
PC-4 162 128 248 19-33
PC-5 162 116 236 20-30
PC-6 181 117 241 14-31
PC-7 140 118 249 20-35
Front view Side view
lp bp
hp
tmb
Cap
Cap
Bed joint
Bed joint
Brick
Brick
Brick
179
5.5.1.2 Test procedure
The monotonic compression tests of the prisms were carried out using the Instron
testing machine and the Bluehill program detailed above. The tests were
displacement controlled with a constant rate of 0.3 mm/min. A preload of 5 kN was
applied to each specimen. In addition to the internal displacement measurement
systems of the test machine, external LVDTs were attached on each surface to
measure the changes in the gage lengths, Figure 5.14. The displacement measuring
points were not described in ASTM C 1314-03b. Consequently, to eliminate the
effects of friction and of crushing of capping on the displacements, gage points were
arranged as shown in Figure 5.14. The measurements of the LVDTs were collected
and stored by a TDS 303 datalogger. Compression loads, which were measured by
the load cell of the Instron device, accompanying the LVDT measurements were
manually recorded. To minimize friction effects, a pair of Teflon sheets was
positioned between the specimen and the loading plates.
Figure 5.14 : The test setup and measurement system for the prism compression
tests.
2hp/3-5hp/7
hp/6-hp/7
hp/6-hp/7
lp/2 lp/2
LVDT 1-2
bp/2 bp/2
LVDT 3-4
Front view Side view
Teflon
sheet
Teflon sheet
LVDT 1-2
LVDT 3-4
Upper loading plate
180
5.5.1.3 Test results
The test results presented here are based on the averages of the measurements
external LVDTs. The displacement measurements of the test machine were only
utilized to see the complete stress-strain behavior as the measurements included not
only the deformation of the specimen but also that of the Teflon sheets and the
deformation of the caps as well.
The compressive stress versus compressive strain relationships of the prisms are
illustrated in Figure 5.15. The parameters obtained from these curves with the
intention of quantitative explanation of the behavior can be seen in Table 5.11. The
statistical analyses of these parameters were explained with the values of minimum,
maximum, mean, standard deviation and coefficient of variation, Table 5.12.
The average values of compressive strengths ( pc f ), strains at corresponding
compressive strengths ( pc, f  ), compressive stresses at proportional limits ( pc, p  ),
and corresponding strains ( pc, p  ), Young’s moduli ( pc E ), and ductilities ( pc  ), were
calculated as 2.3 MPa, 1.0%, 1.0 MPa, 0.3%, 373 MPa and 1.4, respectively.
According to ASTM C 1314-03b (2003), the compressive strengths obtained from
the tests should be corrected to eliminate size effects. These correction factors take
values depending on p p h t ratios, Table 5.13. As shown in this table, the
compressive strengths of the prisms need to be converted to the compressive strength
of a prism with a height-to-thickness ratio of 2.0. As the corresponding ratios of the
prisms tested in this study were in the range of 1.9-2.1, the corrected average
compressive strengths of the prisms were obtained as equal to the average
compressive strength obtained from the tests (2.3 MPa).
181
0
1
2
3
4
5
0.00 0.01 0.02 0.03 0.04
Compressive strain
Compressive stress (MPa)
PC-1
PC-2
PC-3
PC-4
PC-5
PC-6
PC-7
Figure 5.15 : The compressive stress-compressive strain relationships for the prisms
tested under monotonic loadings (obtained from the LVDTs).
Table 5.11 : The results of the monotonic compression tests on the prisms.
Specimen f pc (MPa) pc, f  (%) pc, p  (MPa) pc, p  (%) pc E (MPa) pc 
PC-1 3.27 1.25 1.24 0.21 577 1.7
PC-2 2.44 1.09 1.11 0.26 435 ID
PC-3 1.67 0.83 0.89 0.26 349 1.5
PC-4 3.56 1.04 1.46 0.28 522 1.5
PC-5 1.73 0.95 0.7 0.28 246 1.2
PC-6 2.18 1.06 1.15 0.41 283 1.1
PC-7 1.32 1.12 0.64 0.32 197 1.3
Table 5.12 : The statistical parameters for the monotonic compression tests on the
prisms.
Statistical
parameter
Minimum Maximum Average Stdev CoV
pc f (MPa) 1.32 3.56 2.3 0.84 0.37
pc, f  (%) 0.83 1.25 1.0 0.13 0.13
pc, p  (MPa) 0.64 1.46 1.0 0.30 0.30
pc, p  (%) 0.21 0.41 0.3 0.06 0.20
pc E (MPa) 197 577 373 143 0.38
pc  1.1 1.7 1.4 0.22 0.16
182
Table 5.13 : The correction factors for the prisms (ASTM C 1314-03b).
p p h t 1.3 1.5 2.0 2.5 3.0 4.0 5.0
Correction factor 0.75 0.86 1.00 1.04 1.07 1.15 1.22
For suggesting several simplifications in the prediction of the masonry
characteristics, a simple regression analyses are conducted on the test data. The
outcomes obtained are presented in the following paragraphs.
By normalizing compressive stress and compressive strain of all prisms, a general
form of the relation between stress and strain is derived, Figure 5.16. This form can
be represented with a parabolic function with a significantly high value of 0.99 2 R  :
pc,n
2
pc,n pc,n   0.89 1.87 (5.5)
In Eq. (5.5), pc,n  is the normalized compressive stress of the prisms and pc,n  is the
normalized compressive strain. As seen in Figure 5.15, Young’s modulus and
compressive stress at proportional limit increase as compressive strength increases. It
is seen in Figure 5.17 that the relationship between Young’s modulus and
compressive strength can be formulated, with a high value of 2 R (0.84):
pc pc E 160 f (5.6)
To figure out the onset of nonlinearity in the stress-strain relationships, the
compressive stresses at the corresponding proportional limits are specified. Figure
5.18 shows that there is a correlation between the compressive strength and
compressive stress at the proportional limit, which can be expressed with the
following linear function ( 0.78 2 R  ):
pc p pc 0.44 f ,   (5.7)
The constant value (0.44) means that the nonlinearity behavior takes place at about
40% of the compressive strength in the ascending branch. This value is larger than
the value of 33% given in ASTM C 1314-03b (2003) to estimate modulus of
elasticity.
183
pc,n = -0.8855pc,n
2 + 1.8693pc,n
R2 = 0.9884
0.0
0.2
0.4
0.6
0.8
1.0
1.2
0.0 0.5 1.0 1.5
Normalized compressive strain
Normalized compressive stress
Figure 5.16 : The normalized compressive stress-normalized compressive strain
relationship for the prisms tested under monotonic loading.
Epc = 161.62fpc
R2 = 0.8422
0
100
200
300
400
500
600
0 1 2 3 4 5
Compressive strength (MPa)
Young's modulus (MPa)
Figure 5.17 : The Young’s modulus-compressive strength relationship for the prisms
under monotonic loading.
184
pc,p = 0.435fpc
R2 = 0.7775
0
1
2
0 1 2 3 4 5
Compressive strength (MPa)
Compressive stress at proportional limit (MPa)
Figure 5.18 : The compressive stress at proportional limit-compressive strength
relationship for the prisms tested under monotonic loading.
The analysis of stress-to-strain ratios at proportional limits and corresponding peaks
may further explain nonlinear behavior in the pre-peak part and whether there is a
relation between Young’s moduli and stress-to-strain ratios at the peaks (called as
secant modulus, pc,s E ). These ratios at the peaks are taken as secant moduli at the
corresponding peaks. As seen in Figure 5.19, the ratios at the proportional limits are
correlated with those at corresponding strengths. This correlation can be expressed
by Eq. (5.8):
pc pc,s
pc,f
pc
pc,p
pc,p E 1.7E
f
1.7  




(5.8)
This equation indicates that about 1.7 times of secant modulus at peak may be
considered as Young’s modulus and due to the development of the nonlinear
behavior; there is a reduction of about 40% in the secant moduli at the peak with
respect to that at the proportional limit.
185
Epc = 1.7143Epc,s
R2 = 0.7188
0
1
2
3
4
5
6
0 1 2 3 4 5
Secant modulus at peak (MPa)
Young's modulus (MPa)
Figure 5.19 : The Young's modulus-secant modulus at peak relationship for the
prisms tested under monotonic compression.
5.5.2 Compression tests on masonry prisms under cyclic loads
5.5.2.1 Specimen preparation
In order to understand the responses of the masonry prisms against cyclic
compression loads, seven prisms were tested under cyclic compression loads. The
test specimens are described in terms of dimensions and bed joint thicknesses in
Table 5.14. PCC is used to symbolize the prism specimens of the cyclic compression
tests. The specimens signed with * were partially defected specimens at the interface
between brick and bed joint during the test preparation phase.
Table 5.14 : The prism sizes for the cyclic compression tests.
Specimen p l (mm) p b (mm) p h (mm) mb t (mm)
PCC-1 138 129 245 20-30
PCC-2 136 124 247 18-26
PCC-3 132 128 248 19-24
PCC-4 174 121 250 18-22
PCC-5 201 119 251 22-33
PCC-6* 174 119 245 21-27
PCC-7* 132 105 239 18-22
186
Following the related statements of ASTM C 1314-03b (2003), the prisms were built
three-bricks high with a minimum length of 132 mm and a height-to-thickness ratio
of 1.9-2.3.
5.5.2.2 Test procedure
The cyclic tests were conducted using the Instron test machine and the Bluehill
software. The test procedure adopted was the same as the cyclic loading procedure
applied on the original wallet tests. A total of four LVDTs were used for recording
the shortenings during the tests, Figure 5.14.
5.5.2.3 Test results
Using the displacement readings of the LVDTs and the load readings of the Instron
device through Bluehill software, the cyclic compressive stress-compressive strain
relations and the envelope curves of the relations, which are reflecting cyclic
behavior of the prisms, are plotted in Figure 5.20 and Figure 5.21, respectively.
Mechanical parameters obtained from these relationships and the statistical
assessments of these parameters are presented in Table 5.15 and Table 5.16,
respectively.
The general known trend between strength and ductility (inverse proportion) is
displayed by the cyclic prisms. As compressive strength increases, ductility takes
generally decreasing value. For instance, while the ductility of PCC-2 is 1.1, that of
PCC-4 is 1.9, Table 5.15. Although the parameters identified exhibit large
deviations, specially strength, stress at proportional limit and Young’s modulus, to
give mean values might be useful to have a better understanding about the material.
Without consideration of the defected specimens, the mean values are 2.7 MPa for
cyclic compression strength, 1.0% for corresponding strain, 1.1 MPa for stress at
proportional limit, 0.3% for corresponding strain, 360 MPa for Young’s modulus and
1.5 for ductility. The strength, deformation and failure characteristics obtained from
monotonic and cyclic compression tests are quite similar. From this similarity, it may
be concluded that the monotonic stress-strain relationship can be used as an envelope
for the cyclic stress-strain relationship. The comparison of the behavior of the prisms
under monotonic and cyclic compression loads is detailed in the study of Ispir and
Ilki (2010).
187
0
1
2
3
4
5
0.00 0.01 0.02 0.03 0.04
Compressive strain
Compressive stress (MPa)
PCC-1
PCC-2
PCC-3
PCC-4
Figure 5.20 : The cyclic compressive stress-compressive strain relationships for the
prisms.
0
1
2
3
4
5
0.00 0.01 0.02 0.03 0.04
Compressive strain
Compressive stress (MPa)
PCC-1
PCC-2
PCC-3
PCC-4
Figure 5.21 : The envelope curves of cyclic compressive stress-compressive strain
relationships for the prisms.
188
Table 5.15 : The results of the cyclic compression tests on the masonry prisms.
Specimen f pc (MPa) pc, f  (%) pc, p  (MPa) pc, p  (%) pc E (MPa) pc 
PCC-1 3.54 1.03 1.25 0.27 472 1.5
PCC-2 4.16 0.74 1.59 0.22 -- 1.1
PCC-3 2.50 1.26 0.85 0.24 356 1.5
PCC-4 1.59 1.11 0.74 0.29 252 1.9
PCC-5 1.82 ID ID ID ID ID
PCC-6* 1.74 * * * * *
PCC-7* 1.75 * * * * *
Table 5.16 : The statistical parameters of the cyclic compression results on the
masonry prisms.
Statistical
parameter
Minimum Maximum Average Stdev CoV
pc f (MPa) 1.59 4.16 2.7 1.16 0.43
pc, f  (%) 0.74 1.26 1.0 0.22 0.22
pc, p  (MPa) 0.74 1.59 1.1 0.39 0.35
pc, p  (%) 0.22 0.29 0.3 0.03 0.10
pc E (MPa) 252 472 360 110 0.31
pc  1.1 1.9 1.5 0.33 0.22
Consequently, by analysis of both the monotonic and cyclic test data, Figure 5.22,
the following formulation can be proposed to represent normalized compressive
stress-normalized compressive strain relation of the masonry prisms:
pcall n pcall n pcall,n
2
, ,   0.87 1.85 (5.9)
In Eq. (5.9), pcall,n  is the normalized compressive stress of all prisms tested under
monotonic or cyclic loadings and pcall,n  is the normalized compressive strain of the
prisms. The symbol of “all” is used when monotonic and cyclic tests are evaluated
together. As seen, the constants of Eqs. (5.5) and (5.9) are similar. This shows the
known connection between the envelopes of the cyclic tests and the curves of the
monotonic tests.
189
pcall,n = -0.8669pcall,n
2 + 1.8514pcall,n
R2 = 0.9874
0.0
0.2
0.4
0.6
0.8
1.0
1.2
0.0 0.5 1.0 1.5
Normalized compressive strain
Normalized compressive stress
Figure 5.22 : The normalized compressive stress-normalized strain relationship for
the prisms tested under monotonic and cyclic loading.
The fact that Young’s moduli or the initial slopes of compressive stress-compressive
strain curves take increasing values as the compressive strengths increase can be
observed better than the other tests, Figure 5.20, Figure 5.21 and Table 5.15. When
the Young’s moduli and corresponding strengths of all prisms are taken into account
together, a linear function is obtained for the relationship of Young's modulusstrength
of the brick masonry prisms, (Figure 5.23):
pcall pcall E 150 f (5.10)
The relationship between compressive stress at proportional limit and compressive
strength is presented in Figure 5.24. This figure displays that compressive stress at
proportional limit is about 40% of compressive strength and then, nonlinear behavior
starts to develop in stress-strain relation after this proportional limit point. This
outcome is similar to that obtained from the monotonic tests, Eq. (5.7).
190
Epcall = 153.46fpcall
R2 = 0.8143
0
100
200
300
400
500
600
0 1 2 3 4 5
Compressive strength (MPa)
Young's modulus (MPa)
Figure 5.23 : The Young’s modulus-compressive strength relationship for the prisms
tested under monotonic and cyclic loading.
As shown in Figure 5.25, the statistical evaluation of secant modulus at peak
(E ) pcall,s and Young's modulus leads to a linear equation:
pcall pcall s
pcall f
pcall
pcall p
pcall p E E
f
,
, ,
,  1.5   1.5
 

(5.11)
where pcall, p  is the compressive stress at proportional limit and pcall, p  is the
corresponding compressive strain of the prisms tested under the monotonic and
cyclic loading.
If Eqs. (5.8) and (5.11) are compared, it can be seen that there is about a difference of
20% in the constants of the equations. This may be evaluated as a large difference
with respect to the other differences between the monotonic tests and all monotonic
and cyclic tests. This might be resulting from the stiffness degradation due to cyclic
loading pattern.
191
pcall,p = 0.4033fpcall
R2 = 0.7718
0
1
2
0 1 2 3 4 5
Compressive strength (MPa)
Compressive stress at proportional limit (MPa)
Figure 5.24 : The compressive stress at proportional limit-compressive strength
relationship for the prisms tested under monotonic and cyclic loading.
The stresses causing the first visible cracks were recorded for several specimens. An
overview of the first crack parameters and their statistical parameters are shown in
Table 5.17 and Table 5.18, respectively. The average values of the stress causing the
first crack and corresponding compressive strain are 1.5 MPa and 0.5%, respectively.
The stress and corresponding strength were plotted in Figure 5.26 to figure out the
relation between them:
pcall cr pcall 0.6 f ,   (5.12)
where pcall,cr  is compressive stress at the first crack and pcall f is compressive
strength of the monotonic and cyclic tests and the symbol of “all” is used when
monotonic and cyclic tests are evaluated together. According to this equation, it is
possible to note that compressive stress at the first visible crack level is about 60% of
the corresponding strength.
Using the results of the monotonic and cyclic tests conducted on 14 prisms, the
variation intervals, means, and deviations of the test results to outline the complete
behavior of these prisms are given in Figure 5.19. As seen, the compressive strengths
and the Young’s moduli exhibit higher deviations with respect to the others.
192
Epcall = 1.508Epcall,s
R2 = 0.7236
0
2
4
6
8
10
0 1 2 3 4 5 6
Secant modulus at peak (MPa)
Young's modulus (MPa)
Figure 5.25 : The Young's modulus-secant modulus at peak relationship for the
prisms tested under monotonic and cyclic loading.
pcall,cr= 0.6421fpcall,cr
R2 = 0.8287
0
1
2
3
0 1 2 3 4 5
Compressive strength (MPa)
Compressive stress at the first crack (MPa)
Figure 5.26 : The compressive stress at the first crack level- compressive strength
relationship for several prisms.
193
Table 5.17 : The stress and strain values at the first cracks of several prisms.
Specimen PC-1 PC-2 PC-5 PC-6 PCC-3 PCC-4
pcall,cr  (MPa) 2.2 1.3 1.1 1.6 1.5 1.1
pcall,cr  (%) 0.54 0.36 0.51 0.64 0.47 0.47
Table 5.18 : The statistical parameters of the first cracks for several prisms.
Specimen Minimum Maximum Average Stdev CoV
pcall,cr  (MPa) 1.1 2.2 1.5 0.41 0.27
pcall,cr  (%) 0.36 0.64 0.5 0.09 0.18
Table 5.19 : The statistical parameters for all compression tests on the prisms.
The cyclic characteristics of the masonry prisms can be defined through the envelope
and plastic strain relationships and stiffness degradation. While plastic strain is the
residual strain at zero stress after an unloading; envelope strain is the strain on the
envelope curve where unloading started. Although, the number of tests is not
sufficient to propose a general relationship between plastic and envelope strains, for
having an idea on this relationship, a regression analysis is carried out, (Figure 5.27),
which ended up with Eq. (5.13).
p n en n en,n
2
, ,   0.10  0.65 (5.13)
In this equation; p,n  and en,n  are the normalized plastic and envelope strains,
respectively. The envelope and plastic strains are normalized with respect to strain at
peak stress. Naraine and Sinha (1991) have also expressed the normalized plasticnormalized
envelope relation with a similar parabolic function. However, the
constants of the function were 0.27 and 0.17, respectively. These differences may be
Statistical parameter Minimum Maximum Average Stdev CoV
pcall f (MPa) 1.32 4.16 2.5 0.94 0.38
pcall, f  (%) 0.74 1.26 1.0 0.16 0.16
pcall, p  (MPa) 0.64 1.59 1.1 0.32 0.29
pcall,p  (%) 0.21 0.41 0.3 0.05 0.17
pcall E (MPa) 197 577 369 128 0.35
pcall  1.1 1.9 1.4 0.26 0.19
194
resulting from the different characteristics of the masonry specimens in terms of
strength, composition, and size that the equations are based on.
p,n = 0.1032en,n
2 + 0.6516en,n
R2 = 0.9913
0
1
2
3
4
0 1 2 3 4
Envelope strain/Strain at peak
Plastic strain/Strain at peak
Figure 5.27 : The plastic strain-envelope strain relationship for the prisms.
For determining the cyclic stiffness degradation, slopes of unloading-reloading
branches were calculated as the slope of a straight line plotted between intersection
point of unloading-reloading and zero-stress point. The variations of the stiffness
degradation with compressive strain are plotted in Figure 5.28. In this figure, the
prisms were denoted with the average compressive strength. The slope values
indicate that there is a stiffness degradation taking place especially in post-peak
branches. It should be noted that while the degraded stiffness and strength were
calculated as the percentage of the peak values; compressive strain was normalized
with peak strain.
Utilizing the correlations derived above and the mean values of the parameters, the
behavior of the prisms under compression loads is quantitatively explained by means
of proportional limit, first visible crack, and peak values as shown in Figure 5.29.
According to the tests on the prisms, the prisms subjected to compression loads show
a linear behavior until about 40% of compressive strength, and then show a nonlinear
behavior. In addition, the first visible crack occurs at about 60% of compressive
strength.
195
0
20
40
60
80
100
120
0 1 2 3
Normalized strain
Normalized stiffness
PCC-1
PCC-2
PCC-3
PCC-4
Figure 5.28 : The stiffness degradation in the prisms under cyclic compression load.
Compressive strain (%)
Compressive stress
fpc
pc,cr = 0.6fpc
pc,p = 0.4fpc
pc,p = 0.3 pc,cr = 0.5 pc,f = 1
Figure 5.29 : The compressive stress-compressive strain relation of the prisms with
special points.
196
5.5.3 Shear tests on prisms (triplets)
5.5.3.1 Specimen preparation
With the aim of the determination of shear strength along mortar bed joint and of the
shear strength components, the triplets similar to the prisms described above were
tested under shear and pre-compression loads by taking into account the
requirements of TS EN 1052-3 (2004).
The requirements of TS EN 1052-3 (2004) are as follows: The bed joint thickness
should be 8-15 mm for mortar representing old mortar. If the lengths and heights of
the bricks, which are used for triplets, provide the conditions of  300 b l and
 200 b h mm, the length and height of each triplet should be equal to those of the
bricks used for the construction of the triplet.
The production process of the triplets is illustrated in Figure 5.30. This process was
the same as that of the prisms built for the compression tests. In addition, to support
each triplet and to apply shear load, spaces indicated with 1 and 2 were capped with
the high-strength cement mortar, as shown in Figure 5.30 and Figure 5.31, few days
after completing the production of the triplets.
The triplet specimens are schematically shown in Figure 5.31, in terms of
composition and geometrical aspect. t l , t b and t h are length, width and height of
each triplet specimen. The size and the range of the bed joint thicknesses of each
triplet are shown in Table 5.20. The triplet specimens were denoted with TS and one
of the following of 0.13, 0.25, 0.50 or 0.75 showing the pre-compression level
applied to the triplets in MPa. The evaluation of Table 5.20 leads to the following
results: Since the triplets were built with full-size bricks, the triplets conform to the
related statement of TS EN 1052-3 (2004) in terms of specimen size. However, the
joint thickness range of the triplets were in 10-18 mm and this is slightly larger than
the upper limit of 15 mm given by TS EN 1052-3 (2004).
It may be noted the differences between the prisms and the triplets are as follows:
While the prisms were formed with reduced length bricks, the triplets were formed
with full-size bricks. While the nominal thicknesses of the bed joints in the prisms
were selected as 20 mm for representing the walls of the historical row houses by
197
following ASTM C 1314-03b (2003), those of the joints in the triplets were selected
as 15 mm by following the related requirement of TS EN 1052-3 (2004).
Table 5.20 : The triplet sizes for the shear tests.
Specimen t l (mm) t b (mm) t h (mm) mb t (mm)
TS-0.13-1 236 120 233 14-18
TS-0.13-2 243 117 232 14-18
TS-0.25-1 215 120 241 14-18
TS-0.25-2 235 116 231 14-17
TS-0.50-1 240 115 237 16-17
TS-0.50-2 256 111 237 10-14
TS-0.50-3 224 122 251 15-17
TS-0.50-4 220 118 237 12-17
TS-0.75-1 231 118 230 13-16
TS-0.75-2 242 121 235 13-16
TS-1.00-1 235 121 234 12-15
TS-1.00-2 236 125 236 12-15
Cement mortar
for capping
Cap for compression
load
Reproduced mortar
for bed joints
Support Support
Cap for shear load Cap for shear load
Bed joint
Bricks in water
1
1
2
1 1
2 2
Figure 5.30 : The construction steps of the triplets.
198
Figure 5.31 : The description of the triplets.
5.5.3.2 Test procedure
The test setup used for the shear tests is given in Figure 5.32. Since the shear tests
were performed for the pre-compression levels, these tests were carried out in two
steps. The first step is to apply the pre-compression load by a hydraulic jack with a
20-kN capacity in an appropriate rate. The second step is to apply shear load until
failure by means of the Instron testing machine with a 5000 kN capacity. The second
step, displacement controlled shear loading, was carried out under a constant
displacement rate of 0.1 mm/min until failure. It should be noted that the
displacement mentioned is the displacement measured by the in-built transducer of
the Instron.
In order to take into account the influence of compression load on shear behavior, the
shear tests were performed for five pre-compression levels of 0.13, 0.25, 0.50, 0.75
and 1.00 MPa. The suggestion given by TS EN 1052-3 (2004) related to the
magnitude of the compression stress is 0.1, 0.3 and 0.5 MPa for bricks with
compressive strengths lower than 10 MPa. So as to understand the effect of the
compression stress magnitude on the failure mode of the triplets, the shear tests were
also carried out under the larger compression stress of 0.75 and 1.00 MPa with
respect to the stresses proposed by TS EN 1052-3 (2004). The five compression
levels of 0.13, 0.25, 0.50, 0.75 and 1.00 MPa correspond to about 5, 10, 20, 30 and
40 percent of the average compressive strength of the triplets (2.5 MPa),
lt bt
Front view
a a-a cross section
a
p
Brick
tmb
Brick Brick Brick
a
2
1 1
ht
199
respectively. A steel frame was constructed and used to apply the compression load
as shown in Figure 5.32.
To place the hydraulic jack and load cell for measuring compression load, two
sockets were arranged. Thanks to these sockets, the centers of the jack, the load cell
and each specimen could be intersected precisely and arranged with respect to the
center of the upper/lower loading plates so that the rotations of the specimen in two
perpendicular planes could be avoided. In order to provide the uniform distribution
of the loads applied, steel plates with a thickness of 30-35 mm were used, Figure
5.32. The compression and shear loads, which the specimens were subjected to, were
measured by two load cells of 10 and 20 kN, respectively. The load cell used for the
shear load was placed between the upper loading plate and the steel plate. A total of
eight LVDTs of 10 mm were used in the positions given in Figure 5.33 for
measuring displacements parallel to the direction of the pre-compression load during
the pre-compression loading, for measuring the slip of the middle brick due to the
shear load.
Figure 5.32 : The test setup with measurement system for the triplet shear tests.
Frame for precompression load
Load cell for
precompression
load
Hydraulic jack for
precompression
Load cell for
shear load
Compression
load
Shear load
Steel plate
Cap or support made with high
strength cement mortar
Instron
200
Figure 5.33 : The measurement system for the triplet shear tests.
5.5.3.3 Test results
The results of the triplets under shear and pre-compression loads are presented in the
following paragraphs.
The shear stress of each triplet was calculated by dividing the shear load recorded by
the total area of the mortar bed joints:
mb
ts
ts A
P
2
  (5.14)
In Eq. (5.14), ts  is the shear stress, ts P is the shear load, and mb A is the area of a
mortar bed joint.
Substituting peak shear load resisted by each triplet specimen into Eq. (5.14), the
shear strength of each triplet was computed, Table 5.21. The average shear strengths
estimated were 0.30 MPa at 0.13 MPa compression stress level, 0.40 MPa at 0.25
MPa compression stress level, 0.53 MPa at 0.50 MPa compression stress level, 0.56
MPa at 0.75 MPa compression stress level and 0.83 MPa at 1.00 MPa compression
lp/6 2lp/3 lp/6
Ad Ac
Bd Bc
LVDT A1-B1
LVDT Ac-Bc
LVDT Ad-Bd
201
stress level. These mean values confirm that the shear strength takes larger value for
increasing compression stress.
Table 5.21 : The results of the triplet shear tests.
Specimen  (MPa) ts, f  (MPa) Average ts, f  (MPa)
TS-0.13-1 0.13 0.30
0.30
TS-0.13-2 0.13 0.30
TS-0.25-1 0.25 0.38
0.40
TS-0.25-2 0.25 0.42
TS-0.50-1 0.50 0.51
0.53
TS-0.50-2 0.50 0.56
TS-0.50-3 0.50 --
TS-0.50-4 0.50 0.53
TS-0.75-1 0.75 0.58
0.56
TS-0.75-2 0.75 0.53
TS-1.00-1 1.00 0.79
0.83
TS-1.00-2 1.00 0.86
The shear stress-shear displacement relationships of the triplets are shown in Figure
5.34. It should be noted that the shear displacement is calculated as the average of the
readings of A1 and B1 LVDTs shown in Figure 5.33. As seen in this figure, A1 and
B1 LVDTs are used to measure the slip of the middle brick.
0.0
0.2
0.4
0.6
0.8
1.0
0 1 2 3 4 5
Shear displacement (mm)
Shear stress (MPa)
TS-0.13-1
TS-0.13-2
TS-0.25-1
TS-0.25-2
TS-0.50-2
TS-0.50-4
TS-0.75-1
TS-0.75-2
TS-1.00-1
TS-1.00-2
Figure 5.34 : The shear stress-shear displacement relationships for the triplets
(through A1 and B1 LVDTs).
The variations of compressive stresses during triplet shear tests are plotted in Figure
5.35 and Figure 5.36. Due to the limited sensitivity of the hydraulic jack, keeping the
202
specified compression stress level constant during the shear test could not be
captured. Consequently, two limits (Limit 1 and Limit 2) were defined for keeping
the compressive stress within these limits. Limit 1 and Limit 2 were 1.1 and 0.9
times the specified compression stress level, respectively. Figure 5.35 and Figure
5.36 show that the keeping the compressive stress within these limits could be
succeeded for the higher specified compressive stress level. To have an idea about
the variation of the compressive stress applied during each shear test, the
compressive stresses recorded were evaluated statistically, Table 5.22. As can be
seen in this table, the mean values of the compressive stresses applied during the
shear tests were close to the corresponding specified compressive stresses.
The average shear strains on both sides of middle bricks subjected to shear loads
were obtained by dividing the readings of Ac, Ad, Bc, and Bd LVDTs by related
gage lengths, Figure 5.33. Taking average values of shear strains on each side, the
relationships of the shear stress and the shear strains are plotted in Figure 5.37 and
Figure 5.38. The shear strains of TS-0.25-1 and TS-1.00-1 triplets could be obtained
for one side, as the related readings of other side were not reliable. The average
strains on both sides of triplets were generally not similar, as shown in Figure 5.37.
This is attributed to the different damage evolution at the mortar bed joints and
bricks resulting from the differences in the material properties such as pores, water
absorption rate, and possible defects in the bricks, and poor workmanship of mortar
causing non-homogenous distribution of mortar quality.
Comparing Figure 5.37 and Figure 5.38, it can be seen that the curves displayed
different characteristics. While the curves of each triplet in Figure 5.37 are not
generally similar to each other, those in Figure 5.38 are similar. The curves in Figure
5.38 are more stable with respect to those in Figure 5.37. This may be attributed to
the magnitude of the compression stresses applied and thus, to the differences in the
failure modes detailed below. The compression stress levels of the specimens in
Figure 5.37 (0.13, 0.25, and 0.50 MPa) are lower than those in Figure 5.38 (0.75 and
1.00 MPa). The triplets with the compression stresses of 0.13, 0.25 and 0.50 MPa
and the triplets with the compression stresses of 0.75 and 1.00 MPa exhibited
different failure mechanisms as shown in Figure 5.39 and Figure 5.40, respectively.
While the triplets tested under the compression stresses of 0.13, 0.25 and 0.50 MPa
203
failed due to mortar joint slip, those of 0.75 and 1.00 MPa failed due to cracks in the
middle bricks.
Figure 5.35 : The variation of compressive stress during the shear tests of TS-0.13-1,
TS-0.13-2, TS-0.25-1 and TS-0.25-2.
The typical failure development of the triplets with the compression stresses of 0.13,
0.25 and 0.50 MPa is as follows: The damage initiated around the peak load with
generally diagonal cracks in one or two mortar joints. While these cracks developed
slowly, new cracks formed in diagonal or horizontal directions on the mortar joints
and separations occurred on the interface between the middle brick and one or two
mortar joints. Then, these damages caused the separation of brick and the mortar
joint over the height of the specimen. It should be noted that the number of cracks
was limited and these cracks were not spreading all through the mortar joints. As
shown in Figure 5.39, the separation along mortar joint, namely, joint slip took place
in one or two sides of the middle bricks. On the other hand, the triplets tested under
the compression stresses of 0.75 and 1.00 MPa failed due to cracks in the bricks and
detachment of fragments of the bricks, Figure 5.40. Since the increased compressive
stress provided higher shear capacity, the joint slip did not occur and the specimens
failed because of the damages on the bricks.
0.0
0.1
0.2
00:00:00 00:28:48 00:57:36 01:26:24 01:55:12
Time (h:m:s)
Compressive stress (MPa)
Test
0.13 MPa
Limit 1
Limit 2
TS-0.13-1
0.0
0.1
0.2
00:00:00 00:28:48 00:57:36 01:26:24 01:55:12
Time (h:m:s)
Compressive stress (MPa)
Test
0.13 MPa
Limit 1
Limit 2
TS-0.13-2
0.0
0.1
0.2
0.3
0.4
00:00:00 00:28:48 00:57:36 01:26:24 01:55:12
Time (h:m:s)
Compressive stress (MPa)
Test
0.25 MPa
Limit 1
Limit 2
TS-0.25-2
0.0
0.1
0.2
0.3
0.4
00:00:00 00:28:48 00:57:36 01:26:24 01:55:12
Time (h:m:s)
Compressive stress (MPa)
Test
0.25 MPa
Limit1
Limit2
TS-0.25-1
204
Figure 5.36 : The variation of compressive stress during the shear tests of TS-0.50-2,
TS-0.50-4, TS-0.75-1, TS-0.75-2, TS-1.00-1 and TS-1.00-2.
Table 5.22 : The statistical evaluation of the compressive stresses recorded during
the triplet shear tests.
Specimen Specified  (MPa) Average  (MPa) Stdev (MPa) CoV
TS-0.13-1 0.13 0.13 0.01 0.08
TS-0.13-2 0.13 0.14 0.01 0.07
TS-0.25-1 0.25 0.24 0.01 0.04
TS-0.25-2 0.25 0.24 0.03 0.13
TS-0.50-1 0.50 0.49 0.02 0.04
TS-0.50-2 0.50 0.51 0.02 0.04
TS-0.50-4 0.50 0.48 0.01 0.02
TS-0.75-1 0.75 0.73 0.01 0.01
TS-0.75-2 0.75 0.76 0.01 0.01
TS-1.00-1 1.00 0.99 0.01 0.01
TS-1.00-2 1.00 0.99 0.00 0.00
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
00:00:00 00:28:48 00:57:36 01:26:24 01:55:12
Time (h:m:s)
Compressive stress (MPa)
Test
0.75 MPa
Limit 1
Limit 2
TS-0.75-1
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
00:00:00 00:28:48 00:57:36 01:26:24 01:55:12
Time (h:m:s)
Compressive stress (MPa)
Test
0.75 MPa
Limit 1
Limit 2
TS-0.75-2
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
1.2
00:00:00 00:28:48 00:57:36 01:26:24 01:55:12
Time (h:m:s)
Compressive stress (MPa)
Test
1.00 MPa
Limit 1
Limit 2
TS-1.00-1
0.0
0.1
0.2
0.3
0.4
0.5
0.6
00:00:00 00:28:48 00:57:36 01:26:24 01:55:12
Time (h:m:s)
Compressive stress (MPa)
Test
0.50 MPa
Limit 1
Limit 2
TS-0.50-2
0.0
0.1
0.2
0.3
0.4
0.5
0.6
00:00:00 00:28:48 00:57:36 01:26:24 01:55:12
Time (h:m:s)
Compressive stress (MPa)
Test
0.5 MPa
Limit 1
Limit 2
TS-0.50-4
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
1.2
00:00:00 00:28:48 00:57:36 01:26:24 01:55:12
Time (h:m:s)
Compressive stress (MPa)
Test
1.00 MPa
Limit 1
Limit 2
TS-1.00-2
205
Adopting Mohr-Coulomb law, the friction coefficient and shear bond strength are
obtained from the results of the shear tests on the triplets under the compression
stresses of 0.13, 0.25 and 0.50 MPa, Figure 5.41. Each point in this figure is obtained
as an average of test results of two to three test specimens. The shear strength based
on the triplet tests ( ts, f  ), can be expressed as the following:
 0.23 0.61 ,   ts f (5.15)
As shown in Eq. (5.15), the shear strength at zero nominal vertical stress, ( ts,o  ), and
the friction coefficient, ( ts  ), obtained from the shear tests on the triplets were 0.23
MPa and 0.61, respectively.
As the main failure pattern was not slip under the compressive stresses of 0.75 and
1.00 MPa, the test results obtained under these compression stresses were not used in
the prediction of the friction coefficient and shear bond strength. This indicates that
these compressive stresses are outside of the application limits of the Mohr-Coulomb
law. Drysdale et al. (1994) reported that the Mohr-Coulomb law is valid at low levels
of compression and this law is not used in case of failure modes other than slip along
the mortar joints.
206
Figure 5.37 : The shear stress-shear strain relationships for TS-0.13-1, TS-0.13-2,
TS-0.25-1, TS-0.25-2, TS-0.50-2 and TS-0.50-4.
0.0
0.2
0.4
0.000 0.025 0.050 0.075 0.100 0.125 0.150
Shear strain
Shear stress (MPa)
Ac-Bc
Ad-Bd
TS-0.13-2
0.0
0.2
0.4
0.000 0.025 0.050 0.075 0.100 0.125 0.150
Shear strain
Shear stress (MPa)
Ac-Bc
Ad-Bd
TS-0.13-1
0.0
0.2
0.4
0.6
0.000 0.025 0.050
Shear strain
Shear stress (MPa)
Ac-Bc
Ad-Bd
TS-0.25-2
0.0
0.2
0.4
0.6
0.000 0.025 0.050
Shear strain
Shear stress (MPa)
Ac-Bc
Ad-Bd
TS-0.50-2
0.0
0.2
0.4
0.6
0.000 0.025 0.050
Shear strain
Shear stress (MPa)
Ac-Bc
Ad-Bd
TS-0.50-4
0.0
0.2
0.4
0.6
0.000 0.025 0.050
Shear strain
Shear stress (MPa)
Ac-Bc
TS-0.25-1
0.0
0.2
0.4
0.6
0.0000 0.0025
Ac-Bc
0.0
0.2
0.4
0.0000 0.0025
Ac-Bc
Ad-Bd
0.0
0.2
0.4
0.0000 0.0025 0.0050
Ac-Bc
Ad-Bd
0.0
0.2
0.4
0.6
0.0000 0.0025 0.0050
Ac-Bc
Ad-Bd
0.0
0.2
0.4
0.6
0.0000 0.0025 0.0050
Ac-Bc
Ad-Bd
0.0
0.2
0.4
0.6
0.0000 0.0025 0.0050
Ac-Bc
Ad-Bd
207
Figure 5.38 : The shear stress-shear strain relationships for TS-0.75-1, TS-0.75-2,
TS-1.00-1 and TS-1.00-2 (through Ac-Bc and Ad-Bd LVDTs).
Figure 5.39 : The failure of the triplets under the compression stresses of 0.13, 0.25,
and 0.50 MPa.
0.0
0.2
0.4
0.6
0.8
0.000 0.025 0.050
Shear strain
Shear stress (MPa)
Ac-Bc
Ad-Bd
TS-0.75-2
0.0
0.2
0.4
0.6
0.8
1.0
0.0000 0.0025 0.0050 0.0075 0.0100
Shear strain
Shear stress (MPa)
Ad-Bd
TS-1.00-1
0.0
0.2
0.4
0.6
0.8
0.000 0.025 0.050
Shear strain
Shear stress (MPa)
Ac-Bc
Ad-Bd
TS-0.75-1
0.0
0.2
0.4
0.6
0.8
1.0
0.0000 0.0025 0.0050 0.0075 0.0100
Shear strain
Shear stress (MPa)
Ac-Bc
Ad-Bd
TS-1.00-2
TS-0.13-1 TS-0.13-2
208
Figure 5.40 : The failure of TS-1.00-1 triplet.
ts,f = 0.6062 + 0.2322
R2 = 0.9845
0.0
0.2
0.4
0.6
0.8
0.0 0.1 0.2 0.3 0.4 0.5 0.6
Compressive stress (MPa)
Shear strength (MPa)
Figure 5.41 : The shear strength-compressive stress relation for the triplets.
5.6 Mechanical Tests on Walls
5.6.1 Compression tests on walls
5.6.1.1 Specimen preparation
For experimentally determining the compressive strength of masonry through the
tests of larger specimens with respect to the others, namely, core, wallet and prism
209
specimens, three wall specimens were produced taking into account the related
proposals of TS EN 1052-1 (2000).
The walls were constructed with the historical bricks collected from the walls of the
houses and the reproduced mortar explained above. The bricks were selected
randomly, namely, without paying attention to the colors and sizes of the bricks.
Figure 5.42 presents the construction steps of the walls. As these steps were the same
as those of the prisms, only few remarkable points of the production are given here
as follows: The walls were produced on a platform consisting of a palette and a
plywood plate for providing a leveled surface for the construction and for facilitating
the transportation of specimens from production area to the test setup. Before the
construction of walls, grease oil was applied to the plywood plate to avoid the
adhesion of the specimen to the plywood plate. The upper and lower surfaces of the
specimens were capped with a high-strength cement mortar during the production of
the specimens.
The bonding type, the bed ( ) mb t and head ( ) mh t joint thicknesses of the walls were
selected in a way to represent a typical wall of the historical houses. The nominal bed
and head joint thicknesses of the walls in running bond were 20 mm and 10 mm,
respectively, (Figure 5.43). However, due to surface roughnesses and differences in
sizes, it was not possible to obtain a constant thickness for bed and head joints. Each
specimen was composed of five brick courses (twenty-two and a half bricks), four
bed joints and twenty head joints. The average weight of the specimens was
calculated as about 110 kg by taking the densities of 1860 kg/m3 for brick and of
2000 kg/m3 for mortar.
The geometrical aspects of the specimens were arranged taking into account TS EN
1052-1 (2000) requirements, the wall types of the historical houses, the capacity of
the loading frame to be used for the testing of the specimens and the number of the
bricks collected, Table 5.23. The wall specimens were symbolized with WC, which
is the first capital letters of wall and compression. The requirements of TS EN 1052-
1 (2000) regarding the sizes of the specimens to be built are presented in Table 5.24
for the length ( ) b l and height ( ) b h of the bricks providing the conditions of  300 b l
and 150 b h mm.
210
Considering Figure 5.43 and the sizes of the specimens given in Table 5.23 with the
average bricks length and height of ~120 and ~65 mm, respectively, it can be seen
that the wall specimens conformed to the requirements related to length ( ) w l ,
width ( ) w b and height ( ) w h but did not conform to w w h  3b and w w h  l , Table 5.24.
Figure 5.42 : The construction steps of the walls.
Palette and
plywood plate Greased plywood plate Lower cap First row
Mortar for
capping
Ingredients of masonry
mortar (reproduced)
Masonry mortar
(reproduced)
Head joints of
the first row
Bricks soaked in water
The leveling of the first
bed joint
The arrangement of
the head joint
thickness
Upper cap
211
Figure 5.43 : The wall specimens.
Table 5.23 : The wall sizes for the compression tests.
Specimen w l (mm) w b (mm) w h (mm) mb t (mm) mh t (mm)
WC-1 628 252 431 21-40 16-40
WC-2 636 214 440 18-31 13-34
WC-3 625 198 437 17-37 18-37
Table 5.24 : The size requirements of TS EN 1052-1 (2000).
w l (mm) w b (mm) w h (mm)
 2lb b  b
b  5h ; w  3b ;
w  15b ; w  l
5.6.1.2 Test procedure
The compression tests of the walls were performed utilizing a loading frame as
shown in Figure 5.44. The compression load was applied to each specimen by means
of a hydraulic jack of 500 kN capacity in an appropriate rate until the formation of
lw
Front view Side view
hw
bw
tmb tmb
tmh
hb
lb
bed joint
head joint
212
failure. Before each test, a preload of 20 kN was applied to the specimen. The load
was transferred to the specimen through an I-section steel profile (220 mm) and a
steel plate with 30 mm thick located on top of the specimens.
The measurement system was composed of a load cell of 1000 kN capacity, which
was located between I-section profile and the upper beam of the loading frame, and
LVDTs of 10 mm capacity, which were instrumented to measure the displacements
taking place in vertical and horizontal directions. In order to collect the
displacements occurring in different levels and to compare with each other, the
LVDTs in horizontal direction were positioned in three levels and those in vertical
direction were positioned in two levels, Figure 5.44. The readings of the load cell and
the LVDTs were stored by a TDS 303 data logger.
In order to fix the LVDTs into the specimen, steel anchors with 8 mm diameter were
used. The steps followed for the fixing the LVDTs are as follows: Firstly, anchor
holes were opened by a drill into each specimen in a depth of 60-70 mm and
secondly, the holes were cleaned from the remaining of bricks by air pressure (a
compressor). Then, 60-70 mm of the anchors was smeared on an epoxy-based
adhesive and lastly, the anchors were screwed in the holes.
213
Figure 5.44 : The test setup with measurement system for the wall compression
tests.
5.6.1.3 Test results
The compressive stress-compressive strain relationships are presented in Figure 5.45.
The compressive strains are taken as the average value of the strains obtained from
LVDTs. A summary of the results of the tests and their statistical assessment are
given in Table 5.25 and Table 5.26, respectively. The average values of compressive
strengths ( wc f ), strains at corresponding compressive strengths ( wc, f  ), compressive
stresses at proportional limits ( wc, p  ), and corresponding strains ( wc, p  ), Young’s
moduli ( wc E ), and ductilities ( wc  ), were calculated as 1.8 MPa, 1.4%, 0.9 MPa,
0.3%, 341 MPa and 1.7, respectively. The number of the wall tests is not sufficient to
derive relationships, but several ratios may be given. The ratios of Young's modulus
to compressive strength are varied between about 140-220. The compressive stresses
at corresponding proportional limits are about 40-60% of the corresponding
compressive strengths.
Hydraulic jack
Load cell
I-shaped profile
Data
logger
Switch box
Wall
Loading frame
Front side
2hw/5
LVDT 3-4
LVDT 1-2
214
0
1
2
3
0.00 0.01 0.02 0.03 0.04
Compressive strain
Compressive stress (MPa)
WC-1
WC-2
WC-3
Figure 5.45 : The compressive stress-compressive strain relationships for the walls.
Table 5.25 : The results of the wall compression tests.
Specimen f wc (MPa) wc, f  (%) wc, p  (MPa) wc, p  (%) wc E (MPa) wc 
WC-1 2.08 1.71 0.86 0.19 454 2.0
WC-2 1.79 1.67 1.08 0.43 253 1.4
WC-3 1.50 0.74 0.81 0.26 316 ID
Table 5.26 : The statistical parameters of the wall compression tests.
Statistical
parameter wc f (MPa) wc, f  (%) wc, p  (MPa) wc, p  (%) wc E (MPa) wc 
Minimum 1.50 0.74 0.81 0.19 253 1.4
Maximum 2.08 1.71 1.08 0.43 454 2.0
Average 1.8 1.4 0.9 0.3 341 1.7
Stdev 0.29 0.55 0.14 0.12 103 0.42
CoV 0.16 0.39 0.16 0.40 0.30 0.25
215
5.6.2 Diagonal tension tests on walls
5.6.2.1 Specimen preparation
With the purpose of assessing the effect of the test type on the shear behavior,
particularly, shear strength, in addition to the other shear tests, diagonal shear tests
are performed on wall specimens.
The walls were constructed according to the production procedure described above
for the wall specimens of the compression and shear tests and with the same
materials, by following the requirements of ASTM E 519-02 (2003). These walls
were composed of five rows bonded with four bed joints, Figure 5.46. A row
included three and a half bricks and three head joints. The nominal bed and head
joint thickness were 20 and 10 mm, respectively. The place of the half brick at the
opposite edges of the upper and the lower rows were left empty so that caps were
formed with high-strength cement mortar for applying load, Figure 5.46. The sizes
and the joint thicknesses of the walls are presented in Table 5.27. The specimens are
symbolized with the first letters of wall, diagonal, and tension, namely, WDT.
ASTM E 519-02 (2003) recommends the test of specimen with a size of 1.2x1.2 m.
However, as the available test equipments in the laboratory are appropriate for
specimen with a maximum height of about 500 mm and the number of the bricks
collected was limited, the nominal size of the specimens produced was decided as
400x400mm.
5.6.2.2 Test procedure
The tests were carried out using the Amsler testing machine with a load capacity of
5000 kN, Figure 5.47. Each specimen was subjected to diagonal compression load on
the opposite corners, namely, the caps, and the load was applied continuously to
failure. To measure the displacements between the lower and upper loading plates, a
total of 8 LVDTs were placed on the corners of the loading plates. The displacements
in the middle zone of each specimen were also measured in vertical and horizontal
directions by means of four LVDTs of 25 mm capacity, Figure 5.47. To fix the four
LVDTs into the specimen, a total of eight anchors with 8 mm diameters were used
by following the fixing procedure mentioned above for the compression tests of the
wall specimens. The readings of the LVDTs were collected and stored by a TML
216
TDS 303 data logger. The loads were observed from the built-in load indicator of the
test machine and these loads were recorded by hand.
Figure 5.46 : The wall specimens for the diagonal tension tests.
Table 5.27 : The wall sizes for the diagonal tension tests.
Specimen w l (mm) w b (mm) w h (mm) mb t (mm) mh t (mm)
WDT-1 442 250 385 12-28 15-19
WDT-2 441 255 387 17-30 15-28
WDT-3 392 236 328 11-32 10-30
Front view
Side view
bw
lw
tmb
tmh
hw
Cap
Cap
217
Figure 5.47 : The test setup with measurement system for the wall diagonal tension
tests.
5.6.2.3 Test results
In order to give a better description of the behavior of the specimens tested under the
diagonal tension loads, the average shear stresses and strains, and failure modes are
presented in the following paragraphs.
The shear stresses and shear strains at each load level are calculated by the equations
given in ASTM E 519-02 (2003):
wdt
wdt
wdt A
0.707P
  (5.16)
g
V Hwdt wdt
wdt
  
  (5.17)
In Eq. (5.16), wdt  is the shear stress, wdt P is the applied load, and wdt A is the
specimen area calculated as follows:
w
w w
wdt b
l h
A
2
(  )

(5.18)
Upper plate
Lower plate
Specimen
LVDT 1-4
LVDT 2-3
LVDT 5-6-7-8
LVDT 9-10-11-12
dw
dw/2
dw/2
Pwdt
Pwdt
LVDT 1-2
LVDT 3-4
218
In Eq. (5.17),  wdt is the shear strain, Vwdt is the shortening in vertical direction,
wdt H is the extension in horizontal direction, and g is the gage length. It should be
noted that this equation is for the cases of the equal gage lengths for the shortening in
vertical direction and the extension in horizontal direction. In this study, as there are
small differences between the corresponding gage lengths in vertical and horizontal
direction, shear strain was calculated as the sum of the vertical and horizontal strains
after they are calculated in vertical and horizontal directions separately. The shear
stress-average vertical and horizontal relationships are illustrated in Figure 5.48. It
should be noted that all strains in Figure 5.48 are average strains measured in the
middle zones of the specimens. The vertical strains, which were obtained from the
displacements measured by the LVDTs positioned at the upper and lower loading
plates, are presented in Figure 5.49 together with the average vertical strains
measured in the middle zone. As seen, the vertical strains measured in different
gages of these strains are quite similar. It should be noted that vertical strains
averaged on all height of the specimens are denoted with "all". Using the vertical and
horizontal strains, shear strains are calculated and the relationships of shear stressshear
strain are presented in Figure 5.50. These curves illustrate that while the curve
forms of WDT-1 and WDT-3 specimens are similar, that of WDT-2 specimen is
different. The curves of WDT-1 and WDT-3 specimens followed a steeper trend with
respect to one of WDT-2 specimen in the descending parts. This may be attributed to
the shear strength values.
0.00
0.05
0.10
0.15
0.20
-0.030 -0.025 -0.020 -0.015 -0.010 -0.005 0.000 0.005 0.010
Shear stress (MPa)
WDT-1-Vmid
WDT-2-Vmid
WDT-3-Vmid
WDT-1-Hmid
WDT-2-Hmid
WDT-3-Hmid
Horizontal strain Vertical strain
Figure 5.48 : The shear stress-horizontal strain and shear stress-vertical strain
relationships for the wall diagonal tension tests.
219
0.00
0.05
0.10
0.15
0.20
0.000 0.005 0.010
Vertical strain
Shear stress (MPa)
WDT-1-Vall
WDT-1-Vmid
WDT-2-Vall
WDT-2-Vmid
Figure 5.49 : The comparison of shear stress-vertical strain relations for the wall
diagonal tension tests.
0.00
0.05
0.10
0.15
0.20
0.000 0.005 0.010 0.015 0.020
Shear strain
Shear stress (MPa)
WDT-1
WDT-2
WDT-3
Figure 5.50 : The shear stress-shear strain relationships for the wall diagonal tension
tests.
220
The analysis of the curves given above provided the knowledge of shear strength
( wdt, f  ), shear strain at strength ( wdt, f  ), shear stress ( wdt, p  ) and shear strain ( wdt, p  )
at proportional limit, shear modulus ( wdt G ) and ductility ( wdt  ), Table 5.28.
Although the specimen number is low, the statistical analysis of the results is given
in Table 5.29 so as to have an idea on the scatter of the results. While the average
shear strength is calculated as 0.13 MPa with a standard deviation of 0.03 MPa, the
average shear modulus is calculated as 197 MPa with a standard deviation of 52
MPa. The shear strain at shear strength displayed a larger variation with respect to
the other characteristics. The ductility of WDT-2 is not calculated because the shear
strain at the 85 percent of the shear strength in the descending part could not be
obtained.
Table 5.28 : The results of the wall diagonal tension tests.
Specimen wdt, f  (MPa) wdt, f  (%) wdt, p  (MPa) wdt, p  (%) wdt G (MPa) wdt 
WDT-1 0.16 0.26 ID ID ID 2.3
WDT-2 0.10 -- 0.05 0.03 160 ID
WDT-3 0.13 0.12 0.07 0.03 233 1.7
Table 5.29 : The statistical parameters of the wall diagonal tension tests.
Statistical
parameter wdt, f  (MPa) wdt, f  (%) wdt, p  (MPa) wdt, p  (%) wdt G (MPa) wdt 
Average 0.13 0.19 0.06 0.03 197 2.0
Stdev 0.03 0.10 0.01 0.00 52 0.4
CoV 0.23 0.53 0.17 0.00 0.26 0.20
The first visible cracks occurred at the values of the shear stress close to the shear
strength at about 90-95% of the shear strength. These values of the shear stress were
about 0.15 MPa with a shear strain of %0.04 for WDT-1, 0.09 MPa with a shear
strain of %0.22 for WDT-2 and 0.12 MPa with a shear strain of %0.08 for WDT-3.
during the diagonal tension tests, two typical failure modes were observed: WDT-1
and WDT-2 specimens failed due to a diagonal crack taking place parallel to the
direction of the applied load. The crack, which followed bricks and bed/head joints in
a stepped pattern, caused the splitting of the specimens into two parts, Figure 5.51.
WDT-3 specimen failed due to a slip type of failure along a bed joint and a row of
bricks in a straight line, Figure 5.51. The interfaces between bricks and mortar joints,
where crack/slip occurred are considered to be the weaker planes of the interfaces.
221
Figure 5.51 : The failures of the walls (diagonal tension tests).
Comparing the obtained results with the observed failure modes, it is seen that the
specimens exhibit significantly different behavior under diagonal tension load. This
may be attributed to the differences of the quality of the bricks and mortar.
5.6.3 Shear tests on walls
5.6.3.1 Specimen preparation
This part includes the experimental determination of the behaviors of masonry walls
under shear forces in their own plane. To investigate the influence of precompression
level on the behaviors of the masonry shear walls, the shear loads were
accompanied by the pre-compression levels of 0.13, 0.25, 0.50 and 0.75 MPa. The
masonry shear walls were produced in accordance with the steps detailed above for
the walls of the compression tests. Additionally, two locations with about 25 mm
thick were formed with a high-strength mortar to apply shear loads and to arrange a
support (Figure 5.52). The specimens are introduced in terms of sizes and the ranges
of the thicknesses of bed/head joints in Table 5.30. In this table, the wall specimens
were signed with the first letter of wall, the first two letters of shear and the specified
level of pre-compression (for example WSh-0.13-1).
WDT-3
WDT-2 WDT-2
WDT-3 WDT-3
WDT-2
222
Figure 5.52 : The wall views of the shear tests.
Table 5.30 : The wall sizes of the shear tests.
Specimen w l (mm) w b (mm) w h (mm) mb t (mm) mh t (mm)
WSh-0.13-1 626 253 432 18-35 15-35
WSh-0.13-2 639 253 435 20-35 17-40
WSh-0.25-1 586 234 417 20-40 15-40
WSh-0.25-2 634 252 435 25-40 17-40
WSh-0.25-3 626 217 431 20-40 10-35
WSh-0.50-1 611 247 437 16-40 14-43
WSh-0.50-2 627 252 428 20-40 12-38
WSh-0.75-1 597 235 427 24-39 18-45
WSh-0.75-2 601 237 438 20-40 16-60
5.6.3.2 Test procedure
The walls were tested using the loading frame utilized for the compression tests on
the walls (Figure 5.44). The test of each wall specimen was performed in two stages.
In the first stage, the pre-compression specified was applied. In the second stage, by
keeping the pre-compression level in a specified range (between Limit 1 and Limit
2), shear load was applied in a monotonic manner until the specimen failed. The test
setup and measurement systems consisted of the rigid loading frame, I-shaped steel
beam, steel plate, two hydraulic jacks, two load cells and LVDTs, (Figure 5.54). The
I-shaped steel beam and steel plate were used to spread the compression loads. To
provide slip of each specimen under the compression loads, the location between
steel plates were greased. The hydraulic jacks with capacities of 200 and 1000 kN
were used for applying shear and compression loads, respectively. In order to
measure the compression and shear loads, the load cells of 200 and 1000 kN were
used. The LVDTs of 10 and 25 mm were utilized to gage vertical and horizontal
displacements.
support
for shear load
application
223
Figure 5.53 : The test setup with measurement system for the wall shear tests.
5.6.3.3 Test results
In order to present the shear behaviors of the walls, the relationships of the shear
stress and the horizontal displacement are given in Figure 5.54. The shear stress was
computed by dividing the shear load to the initial cross-sectional area. The horizontal
displacement was computed as the average displacement of the readings measured by
LVDTs of 1-4. As seen in Figure 5.54, the relationships were grouped according to
the pre-compression magnitudes. As a technical problem, which were related to the
hydraulic jack used for applying shear loads, took place during the tests of the
specimens of WSh-0.25-1 and WSh-0.25-2, the relationships obtained did not show
all responses of them.
It is also seen that there is a proportional trend between the pre-compression levels
and shear strengths (Figure 5.54). The shear strengths take larger values for
increasing pre-compression loads. Additionally, the specimens under lower precompression
loads display behaviors that are more ductile. However, the specimens
under the pre-compression level of 0.75 MPa display distinct characteristics in terms
I-shaped profile
Loading frame
Wall
Hydraulic jack
Load cell
for compression
load
Steel plate
Hydraulic jack Load cell
for shear load
LVDT 1-2 LVDT 3-4
Compression
load
Shear load
Support
224
of the form of the relationships between the shear stress and horizontal displacement,
especially in post-peak region.
Figure 5.54 : The shear stress-horizontal displacement relationships for the walls.
The variations of compression loads during each test are given in Figure 5.55 and
Figure 5.56. The average value of the compression loads recorded during each test
and the shear strength obtained are presented in Table 5.31. Additionally, the
compression loads applied to each specimen were evaluated statistically, Table 5.32.
It should be noted that WSh-0.75-1 specimen was subjected to an average
compression load of 0.67 MPa (Table 5.31). Consequently, in the estimations, the
value of 0.67 MPa was taken into account.
0.0
0.2
0.4
0.6
0 10 20 30 40 50
Horizontal displacement (mm)
Shear stress (MPa)
WSh-0.13-1
WSh-0.13-2
0.0
0.2
0.4
0.6
0 10 20 30 40 50
Horizontal displacement (mm)
Shear stress (MPa)
WSh-0.25-1
WSh-0.25-2
WSh-0.25-3
0.0
0.2
0.4
0.6
0 10 20 30 40 50
Horizontal displacement (mm)
Shear stress (MPa)
WSh-0.50-1
WSh-0.50-2
0.0
0.2
0.4
0.6
0 10 20 30 40 50
Horizontal displacement (mm)
Shear stress (MPa)
WSh-0.75-1
WSh-0.75-2
225
Table 5.31 : The results of the wall shear tests.
Specimen Average (MPa) ws, f  (MPa)
WSh-0.13-1 0.13 0.20
WSh-0.13-2 0.13 0.18
WSh-0.25-1 0.25 0.24
WSh-0.25-2 0.26 0.33
WSh-0.25-3 0.25 0.23
WSh-0.50-1 0.49 0.46
WSh-0.50-2 0.50 0.57
WSh-0.75-1 0.67 0.47
WSh-0.75-2 0.74 0.56
Table 5.32 : The statistical evaluation of compressive stress recorded during the wall
shear tests.
Specimen Specified  (MPa) Average  (MPa) Stdev (MPa) CoV
WSh-0.13-1 0.13 0.13 0.01 0.08
WSh-0.13-2 0.13 0.13 0.01 0.08
WSh-0.25-1 0.25 0.25 0.01 0.04
WSh-0.25-2 0.25 0.26 0.02 0.08
WSh-0.25-3 0.25 0.25 0.01 0.04
WSh-0.50-1 0.50 0.49 0.02 0.04
WSh-0.50-2 0.50 0.50 0.01 0.02
WSh-0.75-1 0.75 0.67 0.02 0.03
WSh-0.75-2 0.75 0.74 0.03 0.04
The specimens exhibited two distinct failure mechanisms under the shear loads:
slipping along a bed mortar joint or a diagonal crack following bed and head joints
(Table 5.33). The failure modes of WSh-0.13-2 and WSh-0.50-2 are illustrated in
Figure 5.57 and Figure 5.58, respectively. As known, if the tests were carried out
under the cyclic shear loads, the specimens would fail due to x-shaped diagonal
cracks.
226
Figure 5.55 : The variation of compressive stress during the shear tests of WSh-
0.13-1, WSh-0.13-2, WSh-0.25-1, WSh-0.25-2 and WSh-0.25-3.
Taking into account the failures of the specimens, the bond strength and the friction
coefficient are estimated as about 0.08 MPa and 0.98, respectively, from the
specimens failed due to slip, (Figure 5.59). Consequently, the shear strength based on
the wall tests ( wsh, f  ), can be expressed as the following:
 0.08 0.98 ,   wsh f (5.19)
0.0
0.1
0.2
00:00:00 00:36:00 01:12:00 01:48:00
Time (h:m:s)
Compressive stress (MPa)
Test
0.13 MPa
Limit 1
Limit 2
WSh-0.13-1
0.0
0.1
0.2
00:00:00 00:36:00 01:12:00 01:48:00
Time (h:m:s)
Compressive stress (MPa)
Test
0.13 MPa
Limit 1
Limit 2
WSh-0.13-2
0.0
0.1
0.2
0.3
0.4
00:00:00 00:21:36 00:43:12 01:04:48
Time (h:m:s)
Compressive stress (MPa)
Test
0.25 MPa
Limit 1
Limit 2
WSh-0.25-1
0.0
0.1
0.2
0.3
0.4
00:00:00 00:21:36 00:43:12 01:04:48
Time (h:m:s)
Compressive stress (MPa)
Test
0.25 MPa
Limit 1
Limit 2
WSh-0.25-2
0.0
0.1
0.2
0.3
0.4
00:00:00 00:21:36 00:43:12 01:04:48
Time (h:m:s)
Compressive stress (MPa)
Test
0.25 MPa
Limit 1
Limit 2
WSh-0.25-3
227
Figure 5.56 : The variation of compressive stress during the shear tests of WSh-
0.50-1, WSh-0.50-2, WSh-0.75-1 and WSh-0.75-2.
Table 5.33 : The failure mechanisms of the walls (shear tests).
Specimen Failure mechanism
WSh-0.13-1 Slip along a bed mortar joint
WSh-0.13-2 Diagonal crack following bed and head joints
WSh-0.25-1 Diagonal crack following bed and head joints
WSh-0.25-2 Slip along a bed mortar joint
WSh-0.25-3 Diagonal crack following bed and head joints
WSh-0.50-1 Diagonal crack following bed and head joints
WSh-0.50-2 Slip along a bed mortar joint
WSh-0.75-1 Diagonal crack following bed and head joints
WSh-0.75-2 Diagonal crack following bed and head joints
0.0
0.1
0.2
0.3
0.4
0.5
0.6
00:00:00 00:28:48 00:57:36 01:26:24
Time (h:m:s)
Compressive stress (MPa)
Test
0.50 MPa
Limit 1
Limit 2
WSh-0.50-1
0.0
0.1
0.2
0.3
0.4
0.5
0.6
00:00:00 00:28:48 00:57:36 01:26:24
Time (h:m:s)
Compressive stress (MPa)
Test
0.50 MPa
Limit 1
Limit 2
WSh-0.50-2
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
00:00:00 00:14:24 00:28:48 00:43:12
Time (h:m:s)
Compressive stress (MPa)
Test
0.75 MPa
Limit 1
Limit 2
WSh-0.75-1
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
00:00:00 00:14:24 00:28:48 00:43:12
Time (h:m:s)
Compressive stress (MPa)
Test
0.75 MPa
Limit 1
Limit 2
WSh-0.75-2
228
Figure 5.57 : The failure of WSh-0.13-2 wall.
Figure 5.58 : The failure of WSh-0.50-2 wall.
C D
B
A
C
B D
A
A B
C D
B
A
A D
B C
Slip surface
229
wsh,f = 0.9844 + 0.0846
R2 = 0.9915
0.0
0.2
0.4
0.6
0.8
0.0 0.2 0.4 0.6 0.8
Vertical stress (MPa)
Shear strength (MPa)
Figure 5.59 : The shear strength-vertical stress relation for the walls.
5.7 Evaluation of Compression Tests
5.7.1 Relation between masonry prism and wall, brick unit and mortar
According to the results obtained from the tests conducted in this thesis, the average
compressive strengths of the masonry prism (2.5 MPa) and wall (1.8 MPa)
specimens are about 45 and 33% that of the brick specimens (5.5 MPa). For further
elaboration of this issue, the average prism strength with respect to the average brick
strength, the studies of Sarangapani et al. (2005) and Gumaste et al. (2007) can be
used since the compressive strengths of prisms and its materials are in a comparable
level with the materials considered in this thesis, Table 5.34. It is depicted from the
study of Sarangapani et al. (2005) and Gumaste et al. (2007) that the values of
compressive strength ratios between prisms to bricks are 0.20-0.33 and 0.22-0.32,
respectively. The corresponding ratios are 0.41-0.66 and 0.20-0.36 in the studies of
Lourenço and Pina-Henriques (2006) and Kaushik et al. (2007), respectively. It
should be noted that the ratio increases with increasing mortar strength while keeping
brick strength constant, according to the results of these studies.
230
Table 5.34 : The compressive strengths of masonry prism or wall, brick and mortar.
Reference
Compressive strengths (MPa)
Prism or Wall Brick Mortar
Sarangapani et al. (2005) 0.6-1.4 3.2-4.3 4.1-10.6
Lourenço and Pina-Henriques (2006) 11-17.8 26.9 3.2-95
Gumaste et al. (2007) 1.3-1.8 5.7 0.86-6.6
Kaushik et al. (2007) 4.1-7.5 20.8 3.1-20.6
This study
1.3-4.2 (prism)
1.5-2.1 (wall)
1.8-12.0 1.6-4.8
5.7.2 Relation between Young's modulus and compressive strength of masonry
A linear relationship can be obtained to predict Young’s modulus from
corresponding compressive strength of the masonry prisms tested in this study.
According to this relation, Young’s modulus can be taken as 150 times the
compressive strength of the prism. Moduli of elasticity obtained in these prisms
(197-577 MPa) are close to those determined experimentally by Gumaste et al.
(2007) for varying between 365-508 MPa. It should be noted that the material
characteristics of the specimens tested by Gumaste et al. (2007) are given in Table
5.34. While the values of the prism moduli reported by Lourenço and Pina-Henriques
(2006) of 1110, 2210 and 2920 MPa are larger than the values of 197-577 MPa.
However, the ratios of Young’s modulus to prism strength are 101,152 and 164,
respectively. These ratios are close to the constant of 150, Eq. (5.10). While the value
of 150 is smaller than the constants given by different codes (Table 1.1) and close to
the value of 200 given by TSDC (2007), the constant of the prisms is close to the
lower limits of the ranges given by Hendry (1990) and by Tomazevic (2006) and the
constant is larger than those given by Bayülke (1980). The largest value in Table 1.1
(3000) is 20 times higher than the value obtained for the prisms. However, it is clear
that this coefficient may lead to extremely high elasticity modulus values. The
difference between the constant obtained from the prism tests in this study and the
constants reported from the codes (Table 1.1) may be resulting from the higher
strengths of unit and mortar considered in these codes. These codes do not target
historical constructions, but are basically for modern masonry construction.
According to the ratios of Young's modulus to compressive strength, which are
obtained in this study, it may be stated that the constant of 1000 given by EN 1996-1-
11 might be not suitable for masonry built with high strength masonry components,
231
Table 5.34. Consequently, the use of related constants given in modern codes may
lead to overestimation of the modulus.
Depending on the thicknesses and the moduli of elasticity of unit and mortar bed
joint, modulus of elasticity of masonry may be estimated using Eq. (1.22).
Substituting average Young’s moduli of the bricks (150 MPa), and the reproduced
mortar (224 MPa), and the average thicknesses of bricks (58 mm) and of bed mortar
joints (22 mm) into this equation, Young’s modulus of prism is calculated as 165
MPa. This value is about 45 percent of the value determined experimentally (369
MPa). This difference may be attributed to the variability in sizes and mechanical
properties of the masonry constituents.
The proportional limit used for the estimation of Young’s modulus is determined as
about 40 percent of the compressive strength for the prisms. This percent is close to
33 percent proposed by ASTM C1314-03b (2003) and TS EN 1052-1 (2000), Table
1.2. However, it should be mentioned that while ASTM C1314-03b (2003) proposes
the estimation of Young’s modulus of the masonry prisms in the range of 5-33% as
chord modulus; TS EN 1052-1 (2000) proposes the estimation of Young’s modulus
of the masonry walls at 33% as secant modulus. The obtained value of 40% (Eq.
(5.10)) is in good agreement with the value used in the studies of Aprile et al. (2001)
and Felice (2006), but is smaller than the value of 60% (Baranio and Binda, 1995)
and is larger than the value of 25% (Gumaste et al., 2007).
As shown Table 1.1 and Table 1.2, the values of the modulus show variability in a
wide range. These wide scatter is due to the differences in the properties of materials
tested, in the definition of the modulus used (tangent, secant or chord), in the
dimensions and bond type of the specimens tested, and in the locations of gage
points. This subject is detailed in the study of Ispir et al. (2008).
According to the results of the prism tests, the strains at corresponding strengths took
the values of 0.8-1.3%. These values are in a good agreement with the values of 0.5-
0.9% given by Gumaste et al. (2007), but they are larger than the values of  0.4%
derived from the tests of Kaushik et al. (2007). The differences in material properties
between Kaushik et al. (2007) (Table 5.34) and the present studies are believed to
lead to the different values of the strains at peak strengths. For the prediction of the
232
strain at strength, an empirical formula is proposed by Kaushik et al. (2007) taking
into account the influence of mortar:
 


 


 


 


,  0.25 0.7
0.27
pc
pc
mc
pc f E
f
f
 (5.20)
Substituting the values of  3.1 mc f ,  2.5 pc f and  369 pc E MPa, which were
experimentally determined in the prism tests of the present study, into Eq. (5.20), the
strain corresponding to peak stress is calculated as 0.8%. The strain of 0.8% is
approximate to the lowest value in the corresponding strains determined
experimentally, but the 0.8% strain is smaller than the average value of 1.0%.
5.8 Evaluation of Shear Tests
In order to obtain shear strength components, three types of the tests (shear test on
prism (triplet), diagonal tension test on wall and shear test on wall) were carried out
on the reproduced masonry specimens. The shear bond strengths and friction
coefficients are obtained as 0.23 MPa and 0.61 through the shear tests on the triplets,
and as 0.08 MPa and 0.98 through the shear tests on the walls, respectively. It should
also be noted that the shear bond strength is also obtained as 0.13 MPa through the
diagonal tension tests of the walls. As seen, the shear bond strengths and the friction
coefficients obtained from these tests, which are proposed by the codes, are different.
Consequently, to eliminate the influences of several parameters such as specimen
size and test type on the shear test results, correction factors should be defined.
233
6. OVERALL EVALUATION OF TEST RESULTS
This chapter presents a comparative evaluation of all test results obtained for the
historical masonry of the Akaretler Row Houses. In addition, the interaction curves
obtained based on the test results and the assessments in comparison with TSDC
(2007) are presented.
6.1 The Test Results of Masonry Components
For a better comprehension of the behavior of the historical masonry, the basic
properties of masonry components, such as the bricks and the mortar are tabulated in
Table 6.1 and Table 6.2, respectively. It should be noted that the values in
parenthesis are the corresponding coefficients of variation. These tables clearly
reveal that the scatters in the results are high and that the values of the flexural
tensile ( ) ft f and compressive ( ) c f strengths and modulus of elasticity (E) indicate
the poor quality of the materials. In these tables, CoV and f  are the coefficient of
variation and the strain at compressive strength.
Table 6.1 : The average results of the flexural tensile tests.
Specimen Brick Mortar Reproduced mortar
Number 8 5 6
ft f (MPa) 1.35 1.28 0.85
CoV 0.23 0.14 0.09
Table 6.2 : The average results of the compression tests.
Specimen Brick Three-brick Mortar Reproduced mortar
Specimen number 25 15 30 12
c f (MPa) 5.5 (0.55) 2.3 (0.41) 3.1 (0.30) 3.0 (0.09)
f  (%) 5.2 (0.34) 1.7 (0.16) 2.3 (0.21)
E (MPa) 150 (0.66) 192 (0.34) 232 (0.44) 224
234
6.2 The Compression Test Results of the Masonry
6.2.1 Comparison of core, wallet, prism, and wall test results
The four specimen groups are detailed in terms of their composition and geometrical
aspects, as follows: While the prisms and walls were built with the historical bricks
taken from the in-place walls of the houses and the reproduced mortar, the cores and
the wallets were taken from the in-place walls of the houses. While the prisms were
composed of three bricks and two bed joints in stacked bond, the walls was
composed of five brick courses, four bed joints and twenty head joints in running
bond. While the cores were composed of two brick parts and one mortar bed joint in
stacked bond, the wallets were composed of several bricks bonded with two bed and
several head and longitudinal joints in running bond. While the wallets in the
direction of width included more than ~1/2 brick as there were longitudinal joints
along the widths of the wallets, the other specimen groups included one brick with
reduced/non-reduced length or width.
To figure out the differences and similarities between the test results of the four
specimen groups, the average values of parameters characterizing their mechanical
behaviors are compared, (Table 6.3). In this table, the corresponding coefficients of
variation are also given in parenthesis. As seen in Table 6.3, the scatters of the
mechanical parameters (compressive strength ( ) c f and corresponding stress ( ) f 
and, compressive stress ( ) p  and strain ( ) p  at proportional limit, and ductility
() ) are remarkably high. Young’s modulus (E) shows the highest scattering in
these parameters.
Table 6.3 : The average results of the masonry compression tests.
Specimen Core Wallet Prism Wall
Number 45 15 14 3
c f (MPa) 3.2 (0.26) 1.9 (0.23) 2.5 (0.38) 1.8 (0.16)
f  (%) 2.1 (0.24) 0.9 (0.12) 1.0 (0.16) 1.4 (0.39)
p  (MPa) 1.9 (0.30) 1.0 (0.27) 1.1 (0.29) 0.9 (0.16)
p  (%) 0.8 (0.35) 0.3 (0.20) 0.3 (0.17) 0.3 (0.40)
E (MPa) 255 (0.47) 358 (0.28) 403 (0.41) 341 (0.30)
 1.5 (0.14) 1.6 (0.18) 1.4 (0.19) 1.7 (0.25)
235
The prism and wall tests are standardized tests given in ASTM C 1314-03b (2003)
and TS EN 1052-1 (2000), respectively. Consequently, the mechanical parameters of
the core and wallet specimens should be transformed with respect to the prism or
wall specimens, using correction factors. The correction factors obtained from the
tests results of this study are given in Table 6.4. For example, the compressive
strength obtained from the core tests should be multiplied by "0.8". It should be
noted that the correction factors given in Table 6.4 were based on the prism tests, as
the production and testing of the prism specimens are easier than those of the wall
specimens.
Table 6.4 : The correction factors.
Specimen Core Wallet Prism Wall
c f (MPa) 0.8 1.3 1.0 1.4
f  (%) 0.5 1.1 1.0 0.7
p  (MPa) 0.6 1.1 1.0 1.2
p  (%) 0.4 1.0 1.0 1.0
E (MPa) 1.6 1.1 1.0 1.2
 0.9 0.9 1.0 0.8
As seen in Table 6.3 and Table 6.4, the test results of the cores are different from the
test results of the wallets, prisms, and walls. While the differences between the
compressive strengths may be explained with size effect, the differences between the
other parameters may be explained with size effect and distinct gage lengths and
LVDTs' positions. It is considered that the difference between compressive strengths
may be handled by using correction factors, and the difference between the other
parameters may be made up by using appropriate gage lengths and positions.
However, it should be noted that the gage lengths and positions for the prism
compression tests are not defined in ASTM C 1314-03b (2003). It should be
mentioned that although the prism and wall tests are standardized tests, the obtained
compressive strengths from these tests are different (Table 6.3 and Table 6.4). These
difference may be attributed to the dissimilarities in the bond type (stacked or
running) and size. As expected, the existence of running bond and large size result in
low compressive strength. It is considered that the difference between the prismwallet
compressive strengths and between the prism-wall strengths may be resulting
from bond type and size/bond type effects, respectively.
236
The comparison of the parameters of the wallets and the prisms in Table 6.3 and
Table 6.4 reveals clearly that the differences between the parameters of these
specimen groups may be ignored, with exception of the compressive strengths. The
average compressive strength of the prisms is larger than that of the wallets. This
might be attributed to the possible damages occurring during the process of the
extraction, transportation, and cutting of the wallets, and the differences between the
prisms and wallets in terms of the bond type and size. These damages, which are
generally no visible, might be micro cracks occurred at the interfaces between bricks
and head/longitudinal joints or small separations at these interfaces. According to the
studies of the thesis, the average compressive strength of the wallets is ~75 % of the
compressive strengths of the prisms.
The masonry compressive strength can be predicted using the equations of TSDC
(2007), BCRSMS (2008) and EN 1996-1-1 (2005). These predictions were
calculated in Chapter 3. The masonry compressive strengths based on the equations
of TSDC (2007) and EN 1996-1-1 (2005) are close to each other, namely, 2.8 and
2.6 MPa. While these values (2.8 and 2.6 MPa) are in good agreement with the
average compressive strength of the prisms, the value obtained from the equation of
BCRSMS (2008) (3.9 MPa) are higher than the prism compressive strength. It
should be noted that the prediction of EN 1996-1-1 (2005) (2.6 MPa) are higher than
the average compressive strengths of the walls (1.8 MPa) determined in accordance
to TS EN 1052-1 (2000), which is the document of the test method given by EN
1996-1-1 (2005).
Based on the regression analyses conducted on the data of each compression test
group, several correlations can be established. The correlations between compressive
strength-compressive strain, Young’s modulus-compressive strength, compressive
stress at proportional limit-compressive strength, and compressive stress at the first
visible crack-compressive strength relations are outlined below.
The compressive stress-compressive strain relationship can be expressed with
parabolic functions with a and b constants for each test group:
n n n   a  b 2 (6.1)
237
These constants are given in Table 6.5. As seen, the constants of a and b took
similar values and the values of 2 R were high (close to 1.0). Hendry (1990) and
Kaushik et al. (2007) report the constants of a and b as -2.0 and 1.0 for masonry,
respectively.
The function to be used for the correlation between Young’s modulus and
compressive strength variables is linear as follows:
E  Cfc (6.2)
While the C values of the wallets and prisms were close to each other, the values of
the cores were quite different. This may be resulting from the differences in the gage
lengths. As seen in Table 6.3, the strains of the cores are higher than those of the
wallets/prisms are. As the strains of the cores were obtained from the displacements
measured the upper and the lower loading plates, which leads higher deformation
values), the average Young’s modulus of the cores was lower than those of the
wallets and the prisms.
The relationships of compressive stress at proportional limit-compressive strength
and compressive stress at the first visible crack-compressive strength can be
expressed the following functions:
p c   Df (6.3)
cr c   Ff (6.4)
p  and cr  are the compressive stress at proportional limit and the first visible
crack, respectively. D and F constants of Eqs. (6.3) and (6.4) are presented in
Table 6.5 for each specimen group. As mentioned in Chapter 1, according to the
literature investigation given in Table 1.2, the end points of the stress range defined
on compression stress-compressive strain curve for the calculation Young’s modulus
varies between 25-60%. The end point on the curve can be thought as Din Eq. (6.2).
As seen in Table 6.5, the end points of the stress range obtained in this thesis (40-
60%) are within the range of 25-60% in the literature. When the empirical
formulations obtained from the regression analysis conducted on the test data of the
prisms and the wallets are evaluated in a comparative manner, that the formulations
238
suggested for same variables are similar is shown. However, it should be addressed
that while the proportional compressive stress of the prisms is found as about 40% of
the compressive strength, that of the wallets is found as about 60% of the
compressive strength. This may be due to the difference in the places of the LVDTs
attached to the specimens. While the LVDTs of the wallets are not situated in the
center of the specimens, those of the prisms are situated in their centers.
Table 6.5 : The constants of the relations of the compression tests.
Constant Core Wallet Prism
a ; b -0.90;1.88 (0.95) -0.94;1.96 (0.97) -0.85;1.84 (0.99)
C 80 (0.59) 200 (0.70) 160 (0.87)
D 0.6 (0.42) 0.6 (0.86) 0.4 (0.77)
F 0.7 (0.80) 0.7 (0.81) 0.6 (0.72)
The assessment and comparisons made above may be lead to that the core and wallet
specimens for the compression tests may be used as alternatives of the standardized
specimens of the prisms and walls. However, as mentioned above, considering these
differences in the test results, the need of use of correction factors reveals, especially
for the core test results. The correction factors should be specified in a way to reflect
the effects of size, bond type, and possible damages occurring during extraction and
preparation of the cores and wallets. There are also two conclusions to be
highlighted. Firstly, the similarity between compression test results of the prism and
wallet specimens indicates the influences of specimen size and gage positions/
lengths on the test results. Secondly, if the extractions of original masonry specimens
from the in-situ walls are possible, testing these original masonry specimens may be
preferable to testing prisms or walls constructed with original bricks and
representative mortar simulating original mortar. As the mechanical characteristics
of the original mortar are generally low, to produce a simulating mortar may not be
easy. Additionally, to simulate the workmanship quality of the original masonry,
which is one of the parameter affecting the masonry wall behavior, may not be
possible.
239
6.3 The Shear Test Results of the Masonry
The shear bond strengths ( ) o  and the friction coefficients ( ) f  , which are
calculated adopting Mohr-Coulomb criterion, can be seen in Table 6.6. The analyses
of this table lead to the derivation of the following statements.
According to the table, the shear component values of the in-situ wall tests are
largest. While the ratio of the largest value of the shear bond strengths (0.04-0.23
MPa) to the smallest one is 5.75; the corresponding ratio for the friction coefficients
(0.20-0.50) is 2.5. Consequently, it is questionable which shear components used in
the structural analysis and/or assessment should be used. As seen in Table 6.6, the
values of the shear components depend on the test type. It should be noted that with
the exception of the core tests, the other tests were proposed by various codes.
The bond strengths of the original specimens (core and in-situ wall) are higher than
those of the reproduced specimens (prism and wall) and as the sizes of the specimens
increase, the bond strengths decrease for the laboratory tests (core, prism and wall).
As shown in Figure 6.1, there is an inverse relation between o  and f  values for
the tests performed in the laboratory. This relationship may be expressed with Eq.
(6.5) :
e o f
  3.2 1.3   (6.5)
Table 6.6 : The shear strength components obtained from the tests.
Shear strength
component
Core In-situ wall Prism (triplet)
Wall (diagonal
tension test)
Wall
(shear test)
Specimen
number
14 14 12 3 9
o  (MPa) 0.36 0.45 0.23 0.13 0.08
f  0.40 0.98 0.61 0.98
In order to detect the validity of Eq. (6.5) for the other buildings, the values
predicted by the equation for the other buildings are compared with the experimental
results, Table 6.7. The shear strength components of the tests in the laboratory
(Table 6.6) and of the other buildings (Table 6.7) are also plotted in Figure 6.2.
240
As shown, the inverse trend between the friction coefficients and the shear bond
strengths also exists in the case of other study of the same period (Eq. (6.6)).
e o f
  3.5 1.6   (6.6)
f = 1.2682e-3.1993o
R2 = 0.9999
0.0
0.2
0.4
0.6
0.8
1.0
1.2
0.0 0.1 0.2 0.3 0.4
Shear bond strength (MPa)
Friction coefficient
Wall
Prism
Core
Figure 6.1 : The friction coefficient-shear bond strength relationship (the shear tests
in the laboratory).
Table 6.7 : The prediction of the friction coefficient for the other buildings.
Building cc f
(MPa)
Experimental Prediction (Eq. (6.5))
o  (MPa) f  f 
Haci Sayid Han 2.1 0.22 1.18 0.64
Haydarpasa
Hospital
2.9 0.30 0.68 0.50
Ozturk Apartment 3.8 0.48 0.27 0.28
In Chapter 4, Eqs. (4.24) and (4.25) are proposed to estimate the shear strength
components of the cores, depending on the compressive strength. These equations
were obtained from the core test results of the Akaretler Row Houses and the other
buildings. These equations might be used to predict the shear strength components of
the specimens tested (prism and wall) in the laboratory, Table 6.8. The test results of
the prisms are compatible with the corresponding predictions in respect to those of
241
the walls. It is considered that while the test boundary conditions, bond types, and
sizes of the prisms were similar to those of the cores, those of the walls were almost
different.
f = 1.6364e-3.542o
R2 = 0.7695
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
0.0 0.1 0.2 0.3 0.4 0.5 0.6
Shear bond strength (MPa)
Friction coefficient
Figure 6.2 : The friction coefficient-shear bond strength relationship (the shear tests
of the Akaretler Row Houses and the other buildings in the laboratory).
Table 6.8 : The predictions of the shear strength components for the laboratory tests.
Shear strength component Prism (triplet) Wall (diagonal tension test) Wall (shear test)
o  (MPa)
Test 0.23 0.13 0.08
Prediction 0.28 0.20 0.20
f 
Test 0.61 0.98
Prediction 0.83 1.56
242
6.4 Interaction Curves
To understand how to develop shear stress and normal (tension and compression)
stress interaction curves, the test results are plotted in normal and shear stresses
plane. Simple regression analysis is conducted between shear strengths and
corresponding compressive stresses data to obtain the corresponding interaction
curves. Figure 6.3, Figure 6.4 and Figure 6.5 illustrate the interaction curves
obtained for the cores, the triplets, and the walls, respectively. While the average
strengths obtained from the tensile, compression and shear tests are utilized to plot
the interaction curve of the cores; the average strengths obtained from the
compression and shear tests are utilized to plot the curves of the others. As shown in
these figures, the curves of the cores, the triplets (prisms), and the walls can be
represented with parabolic functions, respectively:
0.30 0.88 0.29 2
,        cs f (6.7)
0.42 1.00 0.16 2
,        ts f (6.8)
0.66 1.17 0.04 2
,        wsh f (6.9)
By substituting   0 into Eqs. (6.7), (6.8) and (6.9), the related shear strengths in
the absence of normal stress are predicted as 0.29, 0.16 and 0.04 MPa for the cores,
the triplets and the walls, respectively. When these values are compared with the
related nominal bond strengths obtained from the linear function (Mohr-Coulomb
criterion), it is seen that these values are lower than the nominal ones, (0.36, 0.23,
and 0.08 MPa for the cores, the triplets, and the walls, respectively). The ratios of the
shear bond strengths obtained from the above equations to the corresponding
nominal ones are 0.81 for the cores, 0.70 for the triplets, and 0.50 for the walls. As
seen, as the sizes of the specimens increase, the differences between shear bond
strengths obtained from the interaction curves and from Mohr-Coulomb criterion
increase.
243
cs,f = -0.30342 + 0.8809 + 0.2871
R2 = 0.9768
0.0
0.2
0.4
0.6
0.8
1.0
-1 0 1 2 3 4
Compressive stress (MPa)
Shear strength (MPa)
Figure 6.3 : The interaction between shear and compressive stresses of the cores.
ts,f = -0.42272 + 0.9952 + 0.1594
R2 = 0.9423
0.0
0.2
0.4
0.6
0.8
1.0
0 1 2 3 4
Compressive stress (MPa)
Shear strength (MPa)
Figure 6.4 : The interaction between shear and compressive stresses of the triplets.
244
wsh,f = -0.66032 + 1.166 + 0.0398
R2 = 0.9677
0.0
0.2
0.4
0.6
0.8
1.0
0 1 2 3 4
Compressive stress (MPa)
Shear strength (MPa)
Figure 6.5 : The interaction between shear and compressive stresses of the walls.
The points intersecting the horizontal axes of these graphs correspond to tensile and
compressive strengths. Consequently, equalizing the right sides of Eq. (6.8) and (6.9)
to zero, the tensile strengths and may be predicted approximately as 0.15 MPa for
the triplets and 0.03 MPa for the walls, respectively. The ratios of the tensile strength
to compressive strength are obtained as ~6 % for the triplets and ~2% for the walls.
It should be noted this ratio is ~8 % for the cores. These differences in these ratios
may be resulting from the differences in the test types, specimen sizes, and specimen
configurations (stacked or running bond).
It is possible to observe the influence of the normal stress level on the failure modes.
According to the results of the tests conducted in the frame of the thesis, while at
low values of the compression stress, the specimens failed due to joint slip, at high
values of that, the specimens failed due to cracks/separations occurring in bricks.
The functions of the interaction curves and the differences in the failure modes
indicate that the relationship of shear strength-normal stress can be modeled using
Mohr-Coulomb criterion at low magnitudes of the normal stress, but the use of the
criterion at high magnitudes of the normal stress is not appropriate. The limit values
of the low compressive stresses for the triplets and walls may be calculated as about
30 and 40% of the corresponding compressive strengths, respectively.
245
The shear test results of the cores and triplets and corresponding relationships
obtained, as well, tensile strength predictions are closer each other with respect to
those of the walls. Consequently, to determine shear characteristics of a historical
wall, the core specimens may be used instead of the triplet specimens standardized in
EN 1996-1-1 (2005) and TS EN 1052-3 (2004). While the cores include original
mortar of the in-situ wall and have original workmanship characteristics of the insitu
wall and small sizes, the prisms should be constructed with a reproduced mortar
compatible with the original mortar and have large sizes. As the mechanical
characteristics of historical mortar are generally low, to reproduce a representative
mortar for the prisms may be hard and time consuming. Additionally, if the
thicknesses of the original mortar joints of the in-situ wall are small, the extraction
of original mortar specimens may be not possible for determining the mechanical
characteristics of the original mortar.
6.5 Evaluation of the Test Results Considering Turkish Seismic Design Code
6.5.1 Compression
In order to determine the compressive strength of a masonry wall or the allowable
compressive stress of the masonry wall, TSDC (2007) proposes four methods.
The first is to test masonry specimens built with the same material as the masonry
structure. The allowable compressive strength is defined as 25% of the average
strength obtained from the tests. However, the specimen characteristics are not
described in terms of size, bond type, and mortar joint. As these characteristics have
effects on test results, there is a need to define these characteristics exactly to obtain
comparable test results.
In the second method, the value of the allowable stress is taken from a table prepared
depending on mortar class (compressive strength) and compressive strengths of
units. It should be noted that lime mortar is not included in the table. Consequently,
to use this table is questionable for the historical buildings with lime mortar joints.
246
In the third method, while the 50 percent of the average compressive strengths of the
units determined experimentally is taken as the compressive strength of masonry
wall, the 25 percent of the compressive strength of masonry wall is taken as the
allowable compressive stress of the wall. In the last method, the allowable stresses
are given depending on unit and mortar types.
According to the first method, the allowable compressive stresses are calculated
from the average compressive strengths of the cores, the wallets, the prisms and the
walls, respectively, as follows:
0.25 3.2 0.8 ,    allowablec  MPa (6.10)
0.25 1.9 0.5 ,    allowablewt  MPa (6.11)
0.25 2.5 0.6 ,    allowablep  MPa (6.12)
0.25 1.8 0.5 ,    allowablew  MPa (6.13)
In these equations, allowable,c  is the allowable stress computed from the average
strength of the cores, allowable,wt  is the allowable stress computed from the average
strength of the wallets, allowable, p  is the allowable stress computed from the average
strength of the prisms, and allowable,w  is the allowable stress computed from the
average strength of the walls. According to TSDC (2007), using the related tests
performed in this thesis, the allowable compressive stresses are calculated as in the
range of 0.5-0.8 MPa.
According to the second method, as the average compressive strengths of the
individual bricks and the mortars were 5.5 and 3.1 MPa, the allowable compressive
stress of the masonry can be taken as 0.4 or 0.5 MPa. It should be noted that whether
linear interpolation is permitted is not mentioned by TSDC (2007).
Using the third methods, the compressive strength of masonry wall, and the
allowable compressive stress of the wall are determined as 0.505.5  2.8 and
0.252.8  0.7MPa, respectively.
The table formed for the last method does not include the lime mortar. However, the
allowable compressive stress of the masonry wall, which is built with solid or
247
common brick and cement-lime mortar, is given as 0.8 MPa. This value may be used
to make comparison. With the exception of lime mortar and specimen definition, the
conclusion to be drawn from this comparison is that the values proposed in TSDC
(2007) for the compression are in agreement with the test results. The allowable
compressive stress of the historical walls can be taken as at least 0.4 or 0.5 MPa with
a safety factor.
6.5.2 Shear
In order to determine the shear strength or allowable shear stress of masonry wall, no
test method is suggested by TSDC (2007). The allowable shear stress is calculated
by the following linear equation.
    f o f   (6.14)
In Eq. (6.14), f  is the allowable shear stress, o  is the allowable shear bond
strength, f  is the friction coefficient, and  is the existing vertical stress on the
wall.
While the values of the allowable shear bond strength are given depending on the
unit and mortar properties, the friction coefficient is given as 0.5 by TSDC (2007). It
should be noted that the case of lime mortar, which is common in case of historical
structures, is not included. Nevertheless, the allowable compressive stress of the
masonry wall, which is built with solid or common brick and cement-lime mortar, is
given as 0.15 MPa. This value (0.15 MPa) is taken into account for making
comparisons with the shear test results.
In order to compare the shear components, which were obtained from the tests
performed in this thesis, with the allowable shear components given by TSDC
(2007), the shear components obtained from the tests should be transformed to the
corresponding allowable ones. However, as no constant coefficient for this
transformation is suggested by the code, the constant of 0.25 given for compressive
strength transformation is adopted in this study to make comparison. The allowable
components calculated by adopting the constant of 0.25 are given in Table 6.9.
248
Table 6.9 : The allowable shear components of the shear tests.
Shear
component
Core
In-situ
wall
Prism
(triplet test)
Wall (diagonal
tension test)
Wall (shear
test)
Experimental
Allowable
o  (MPa) 0.09 0.11 0.06 0.03 0.02
f  0.10 0.25 0.15 0.25
As seen in this table, the allowable shear components calculated from the tests
(  0.02  0.11 o  MPa and  0.10  0.25) f  are lower than the corresponding
allowable components (  0.15 o  MPa and  0.5) f  given by TSDC (2007). This
might be resulting from using a low transformation constant of 0.25.
The comparison of the test results and TSDC (2007) and suggestions based on the
comparison are also outlined as follows:
 The masonry specimen, which is suggested by TSDC (2007) for the
determination of the compressive strength of the masonry wall, should be
defined exactly in terms of size, bond type, and composition such as the
number of bed, head, and longitudinal mortar joints.
 This study and the studies in the literature show that the mechanical
characteristics of the historical/ordinary masonry exhibit large variations.
Consequently, it is considered that the number of three masonry specimens
suggested by TSDC (2007) is not enough to obtain an average compressive
strength.
 In the table titled with "allowable compressive stress for masonry wall
depending on unit and bond types (corresponding compressive strengths)"
given in TSDC (2007), whether linear interpolation is permitted should be
clarified.
 The values of the C constant used for Young's modulus (Table 6.5), which
were determined based on the compression test data, may be comparable with
the value of 200 given in TSDC (2007). However, the literature shows that the
values of the C constant vary in a wide range depending on unit and mortar
characteristics (Table 1.1). Consequently, it is considered that to suggest only
one C value of 200 is not appropriate, and the different C values may be
recommended depending on masonry components and masonry
249
characteristics. Additionally, it should be noted that as the statement given for
the estimation of the modulus in TSDC (2007) is not clear, this statement
should be clarified; and that the determination of Young's modulus based on
the experimental data is not mentioned.
 As the use of lime mortar is common in the historical/ordinary masonry
structures, the case with only lime mortar should be considered and the
corresponding default values should be given.
 In the code, the experimental determination of shear strength components is
not mentioned. The code should include experimental identification of the
shear components.
 The use of the equation (Mohr-Coulomb) given for the calculation of shear
strength should be limited in case of high vertical stress. The limit may be
given as a specified percentage of masonry compressive strength. According
to this study, the limits are about 30-40% of the masonry compressive
strength.
 Although, a constant is defined for transforming compressive strength to
allowable compressive stress, no constant is defined for corresponding shear
strength transformation. Consequently, to use the shear components
determined experimentally in structural assessment to be carried out according
to TSDC (2007), a corresponding constant should be given.
 The equation given in EN 1996-1-1 (2005) (Eq. (1.3)) may also be suggested
by TSDC (2007) to predict masonry compressive strength based on the
compressive strengths and characteristics of masonry components, since the
compressive strengths obtained from tests are in good agreement with the
predictions made using the equation.
250
251
7. NUMERICAL ANALYSES
In the present chapter, three dimensional finite element analyses were carried out
using commercial finite element program Abaqus v6.9-1 (2009). The analyses were
performed on masonry prisms and walls under compression loads. The main
objective is to obtain the masonry behavior from the numerical analyses, based on
the experimental data of the masonry components.
In order to realize the numerical analyses of the masonry prism and walls under
compression loads, a micro-modeling approach was adopted. Consequently, the
components of masonry (the bricks and the mortar joints) were modeled separately
with individual material characteristics. Additionally, it should be noted that the bond
between the bricks and the mortar joints were accepted to be perfect so as to make
the numerical model simple. The material characteristics of the bricks and mortar,
which are needed for micro-modeling, were obtained from the experimental studies
explained in Chapter 3. To predict the behavior of the masonry (the prisms and
walls), the plastic behavior of the bricks and the mortar joints were modeled using
the classical metal plasticity models defined in Abaqus (2009). It should be noted
that the model is used in conjunction with the elastic material model.
The purposes of this chapter studies are:
 To understand the differences between the test and numerical analysis results in
terms of compressive stress-strain relation and strength,
 To assess the numerical analysis results of the prisms and the walls in a
comparative manner,
 To evaluate the analyze results of the masonry taking into account the test results
of the masonry components.
7.1 The Material Models
This section is arranged to present a brief introduction of the material models used in
this study. The knowledge presented here was complied from Abaqus 6.9
252
Documentation (Abaqus, 2009). While many material under low strain or stress
magnitudes exhibit linear elastic behaviors; the materials under higher stress (and
strain) magnitudes exhibit nonlinear, inelastic behavior, which is referred to as
plasticity. The plasticity models in Abaqus (2009) are generally based on incremental
theories. These theories are generally described in terms of a yield surface, a flow
rule and evolution laws defining the hardening. The yield surface can be used to
determine whether the material responds purely elastically under a particular state of
stress. The flow rule prescribes the inelastic deformation if the material point is no
longer responding purely elastically. The evolution laws of the hardening define the
way in which the yield and/or flow definitions change as inelastic deformation
occurs. The plasticity models defined are generally used with the linear elastic model
for describing the elastic part of the behavior. The isotropic elastic properties are
defined with Young’s modulus and Poisson’s ratio.
In this study, as the classical metal plasticity models was used, this model is
introduced in the following sub-headings. The detailed information is given in
Abaqus Documentation (Abaqus, 2009).
7.1.1 The classical metal plasticity model
The classical metal plasticity models are generally utilized for crash analyses, metal
forming and general collapse studies. In Abaqus (2009), these models use Mises or
Hill yield surfaces with associated plastic flow. While the Mises yield surface is used
to define isotropic yielding, the Hill yield surface is used to define anisotropic. In the
classical metal plasticity models, perfect plasticity or isotropic hardening can be
used. In isotropic hardening, the yield surface changes size uniformly in all
directions and so the yield stress increases (or decreases) in all stress directions while
plastic straining takes place. It is mentioned that the isotropic hardening model is
useful for case involving gross plastic straining. The isotropic hardening can be given
as a tabular function of plastic strain. The yield stress remains constant for plastic
strains exceeding the last value given as tabular data.
In this study, the uniaxial compression stress-strain relationships of the bricks and the
reproduced mortar were available experimentally and these relationships were based
on the gross plastic straining. Consequently, it was considered that a classical metal
plasticity with isotropic hardening model could be utilized to model the masonry
253
prisms and walls under the compression loading. The model was used in conjunction
with the linear elastic material model.
7.2 Masonry Modeling
As masonry is formed by the units (bricks and/or stones) with or without mortar, the
masonry is a composite material (Figure 7.1a). In order to simulate the masonry,
different modeling approaches have been used. The following information about the
modeling of masonry were given by Lourenço (1994) and Romano (2005). Masonry
can be modeled with individual components in micro-modeling or as a composite in
macro-modeling.
In detailed micro-modeling, the units and the mortar joints are represented by
continuum models and the unit-mortar interface is represented by discontinuous
elements (Figure 7.1b). In simplified micro-modeling, the units are extended through
the joints and are represented by discontinuous models (Figure 7.1c). In macromodeling;
units, mortar joints and the interfaces of unit-mortar are taken into account
as an isotropic or anisotropic homogeneous continuum (Figure 7.1d). Consequently,
the material differences between the units and the mortar joints are neglected.
As mentioned in the study of Prakash and Alagusundaramoorthy (2008), while
micro-modeling is appropriate for the analyses of small structures or structural
elements where the unit-mortar joint interaction is taken into account; macromodeling
is appropriate for those of large structures. It should be stated that micromodeling
of the large structures is difficult due to numerical convergence and
computational cost.
254
Figure 7.1 : Modeling of masonry (Lourenço, 1994).
7.3 The Analyses of the Prisms under Compression
7.3.1 The establishment of the model
The analyses were conducted on a prism of 155×122×245 mm ( ) p p p l b h with
two bed joints of 25 mm, Figure 7.3. These sizes were selected in a way to
characterize the corresponding average values of the prisms tested. The three
dimensional models of the prisms were formed with eight-noded C3D8 type
hexahedra elements with an approximate element size of 25 mm, Figure 7.3. These
elements have nodes only at their corners and are called linear elements (first-order
elements). The upper and lower boundary conditions of the prisms were assumed to
be fixed in x- and y-directions in terms of displacements. A perfect bond between
bricks and mortar joints was assumed as mentioned above.
The analyses were carried out by applying the incremental displacements to the
upper and lower surfaces of the prism in a stepwise manner.
255
As the material characteristics of the bricks and mortars indicated a high variability,
which were used for the manufacturing the prisms, the analyses were performed for
several compositions of bricks and mortar in order to provide a better insight into the
comparison between numerical and test results of the prisms, Table 7.1. The prisms
analyzed were signed with PCF-number, the first letters of prism, compression, and
finite.
In the finite element program, by adopting a micro-modeling approach, the material
characteristics of the brick and mortar were defined in a way to describe linear and
non-linear behavior. The linear elastic behavior was described with Young’s
modulus and Poisson’s ratio. While the values of Young’s modulus were obtained
from the compression tests, The Poisson’s ratios of the brick and mortar were
accepted as 0.25 and 0.20, respectively. Multi-linear stress-strain relationships
determined experimentally are used to model the non-linear behavior of each
constituent of the masonry. Abaqus (2009) call this method as the classical metal
plasticity with isotropic hardening. The isotropic hardening was represented with
compressive stress (yield stress)-plastic strain values in tabular form, which were
obtained from the compression tests of the masonry components.
Figure 7.2 : The description of the prism.
Mortar
Brick
Brick
Brick
Mortar
z
x
z
y
lp bp
tmb
hp
z
x
y
256
Figure 7.3 : The finite element model of the prism and the element type used.
Table 7.1 : The masonry components of the prisms for the numerical analyses.
Analysis Brick Reproduced mortar
PCF-1 TBC-9 RMCF-210-3-A
PCF-2 TBC-4 RMCF-210-1-B
PCF-3 TBC-1 RMCF-210-1-B
PCF-4 TBC-13 RMCF-210-1-B
7.3.2 The results of the numerical analyses of the prisms
The results obtained from the numerical analyses were explained with compressive
stress contours, deformed shapes, and compressive stress-compressive strain
relationships. While the compressive stress contours and deformed shapes of PCF-1
prism are presented Figure 7.4; for the other prisms are given in Appendix C.
Additionally, the maximum and minimum values of the stresses are given. In Abaqus
(2009), while the compressive stress is indicated with negative sign, the tensile stress
is indicated with positive sign.
The compressive stress-vertical strain relationships of the prisms are illustrated in
Figure 7.5 in conjunction with those of the masonry components for the two models.
In these relationships, compressive stress was calculated as the ratio of the average of
the compression force acting on upper and lower loading surfaces to the crosssectional
area. By taking into account the LVDT positions on the prisms of the
compression tests; the compressive strains derived from the numerical analysis were
taken as the average value of the strains computed between A1 to A2 and A3 to A4,
z Solid element C3D8
x
y
257
using a gauge length (about 180 mm), Figure 7.5. The strains between A1-A2 and
A3-A4 were calculated using the displacements of twelve and eight nodes,
respectively.
Figure 7.4 : The compressive stress contours and deformed shapes of PCF-1.
The analysis of the relationships in Figure 7.5 show that the relationships of the
prisms fall between those of the corresponding masonry components. However, it
can be seen that the post-peak regions obtained from the numerical analyses do not
conform to those of the corresponding masonry components. This may be due to that
the material model adopted in the numerical analyses is inappropriate to simulate the
post-peak behavior. Additionally, the paths are that the relationships of the prisms
follow are close to those of the bricks with respect to those of the mortar. This may
be resulting from the ratio of the bricks and the mortar joints formed the prism. The
percentages of the bricks and the mortar joints in the prism are about 80 and 20%,
respectively, with respect to the prism height. Additionally, in empirical equations in
literature, which are used to predict masonry compressive strength based on the
compressive strengths of masonry components, the contribution of the bricks to the
masonry strength is generally given higher than that of mortar.
z
x
z
y
z
x
A1
A2 z
y
A3
A4
Max stress = -4.606E-3 MPa
Min stress = -1.665 MPa
258
The quantitative parameters obtained from these figures are presented in Table 7.2.
In this table, f pcf and pcf , f  are the compressive strength and corresponding strain;
pcf , p  and pcf , p  are the compressive stress and strain at the proportional limit, pcf E
is Young's modulus and pcf  is the ductility.
The compressive stress-compressive strain relationships obtained from the tests and
the numerical analyses of the prisms are illustrated in Figure 7.6. As mentioned
above, since which masonry components used for the construction of each prism
were not known, it is not possible to make a direct comparison between the results of
the prism tests and numerical analyses. There is a similar trend between the results of
the tests and the analyses in terms of the relationship forms and compressive strength
values. However, the compressive strains at the strengths of the analyses are larger
than those of the tests are. The values of these strains of the tests and analyses are
0.7-1.3% and 1.4-1.9%, respectively. This may be resulting from the perfect bond
assumption made for the numerical analyses and the geometrical differences between
the prisms tested and analyzed. It may be expected that higher compressive strains at
strengths may occur for the prisms where the bond are perfect. The perfect bond
assumption is not appropriate to represent the real bond between bricks and mortar
joints. In cases tested, the bond had several imperfections due to workmanship and
the characteristics of brick and mortar. While all analyses were conducted on a prism
of 155×122×245 mm with two bed joints of 25 mm, there are differences between
the sizes and the mortar thicknesses of the prisms tested. As the bricks of the prisms
were selected randomly, the bricks in a tested prism might be having distinct material
characteristics. However, in the analyses, each prism is constituted with one brick
type.
259
Figure 7.5 : The compressive stress-compressive strain relationships obtained from
the numerical analysis.
Table 7.2 : The results of the numerical analyses for the masonry prisms.
Specimen pcf f
(MPa)
pcf , f 
(%)
pcf , p 
(MPa)
pcf , p 
(%)
pcf E
(MPa)
pcf 
PCF-1 1.5 1.40 0.74 0.47 158 1.1
PCF-2 3.2 1.83 1.45 0.48 300 1.0
PCF-3 4.1 1.91 2.05 0.50 410 1.1
PCF-4 1.9 1.85 1.08 0.52 206 1.1
0
1
2
3
4
5
0.00 0.01 0.02 0.03 0.04 0.05
Compressive strain
Compressive stress (MPa)
PCF-2
TBC-4
RMCF-210-1-B
0
1
2
3
4
5
0.00 0.01 0.02 0.03 0.04 0.05
Compressive strain
Compressive stress (MPa)
PCF-3
TBC-1
RMCF-210-1-B
0
1
2
3
4
5
0.00 0.01 0.02 0.03 0.04 0.05
Compressive strain
Compressive stress (MPa)
PCF-4
TBC-13
RMCF-210-1-B
0
1
2
3
4
5
0.00 0.01 0.02 0.03 0.04 0.05
Compressive strain
Compressive stress (MPa)
PCF-1
TBC-2
RMCF-210-3-A
260
0
1
2
3
4
5
0.00 0.01 0.02 0.03 0.04
Compressive strain
Compressive stress (MPa)
PCF
Test
Figure 7.6 : The comparison of compressive stress-compressive strain relationships
obtained from the tests and numerical analyses (prisms).
7.4 The Analyses of the Masonry Walls under Compression
7.4.1 The establishment of the model
The numerical analyses were conducted on an a wall of 633×225×435 mm
( ) w w w l ×b × h with five bed joints and four head joints for each brick row in running
bond, Figure 7.7. The thicknesses of bed and head joints are taken as 30 and 25 mm,
respectively. These sizes were selected in a way to characterize the corresponding
average values of the walls tested. The three dimensional models of the walls were
formed with eight-noded C3D8 type hexahedra elements with an approximate
element size of 30 mm, Figure 7.8. The upper and lower boundary conditions of the
walls were assumed to be fixed in x- and z-directions in terms of displacements.
Additionally, a perfect bond between bricks, and bed and head mortar joints was
assumed.
As the material characteristics of the bricks and mortars indicated a high variability,
which were used for the manufacturing the walls, the analyses were performed for
several compositions of bricks and mortar, Table 7.3. The walls analyzed were
signed with WCF-number, the first letters of wall, compression, and finite.
261
Figure 7.7 : The finite element model of the wall.
Table 7.3 : The masonry components of the walls for the numerical analyses.
Analysis Brick Reproduced mortar
WCF-1 TBC-9 RMCF-210-3-A
WCF-2 TBC-4 RMCF-210-1-B
WCF-3 TBC-1 RMCF-210-1-B
WCF-4 TBC-13 RMCF-210-1-B
Figure 7.8 : The mesh of the wall.
z
x
y
y
x
y
z
z
x
bw
hw
tmb
tmh
z
x
y
y
lw
262
7.4.2 The results of the numerical analyses of the walls
The compressive stress contours and deformed shapes are illustrated in Figure 7.9 for
WCF-1 and in Appendix C for the other walls.
Figure 7.9 : The compressive stress contours and deformed shapes of WCF-1.
The relationships of the compressive stress-compressive strain obtained from the
numerical analyses are given in Figure 7.10. Additionally, the mechanical parameters
derived from these figures are presented in Table 7.4. In this table, fwcf and wcf , f  are
the compressive strength and corresponding strain; wcf , p  and wcf , p  are the
compressive stress and strain at the proportional limit, wcf E is Young's modulus and
wcf  is the ductility. By taking into account the LVDT positions on the walls of the
compression tests; the compressive strains derived from the numerical analysis were
taken as the average value of the strains computed between A1 to A2 and A3 to A4,
using a gauge length (about 185 mm), Figure 7.9. The strains between A1-A2 and
A3-A4 were calculated using the displacements of eight nodes. Similar to the results
of the prisms, the compressive stress-compressive strain relationships of the walls are
between those of the masonry components with generally exception of the post-peak
regions.
y
x
A1
A2
A3
A4
Max stress = -1.795E-1 MPa
Min stress = -3.453 MPa
y
x
263
Figure 7.10 : The compressive stress-compressive strain relationships for the walls
(numerical analysis).
Table 7.4 : The results of the numerical analyses for the masonry walls.
Specimen wcf f
(MPa)
wcf , f 
(%)
wcf , p 
(MPa)
wcf , p 
(%)
wcf E
(MPa)
wcf 
WCF-1 1.7 1.40 0.70 0.42 166 1.5
WCF-2 3.0 1.82 1.53 0.55 279 1.2
WCF-3 3.8 2.06 2.09 0.62 338 1.1
WCF-4 2.0 1.80 1.03 0.49 209
The compressive stress-compressive strain relationships obtained from the tests and
the numerical analyses of the walls are illustrated in Figure 7.11. Since which
masonry components used for the construction of each wall were not known, it is not
0
1
2
3
4
5
0.00 0.01 0.02 0.03 0.04 0.05
Compressive strain
Compressive stress (MPa)
WCF-1
TBC-2
RMCF-210-3-A
0
1
2
3
4
5
0.00 0.01 0.02 0.03 0.04 0.05
Compressive strain
Compressive stress (MPa)
WCF-2
TBC-4
RMCF-210-1-B
0
1
2
3
4
5
0.00 0.01 0.02 0.03 0.04 0.05
Compressive strain
Compressive stress (MPa)
WCF-3
TBC-1
RMCF-210-1-B
0
1
2
3
4
5
0.00 0.01 0.02 0.03 0.04 0.05
Compressive strain
Compressive stress (MPa)
WCF-4
TBC-13
RMCF-210-1-B
264
possible to make a direct comparison between the results of the wall tests and the
numerical analyses.
0
1
2
3
4
0.00 0.01 0.02 0.03 0.04
Compressive strain
Compressive stress (MPa)
WCF
Test
Figure 7.11 : The comparison of compressive stress-compressive strain relationships
obtained from the tests and numerical analyses (walls).
7.5 The Numerical Analyses Results of the Masonry
As mentioned above, the elastic properties and compressive stress-strain
characteristics of brick and joints required for the analyses were obtained by
conducting experiments on brick and mortar specimens. Using this experimental data
of the masonry components, the behavior of masonry prisms and walls were obtained
numerically (Figure 7.12). It should be noted that the prism and wall specimens
analyzed for the same masonry component characteristics are symbolized with the
same number. In order to derive the differences between the numerical results of the
prism and wall with the same number, the ratios between corresponding mechanical
characteristics of the prisms to the walls are calculated (Table 7.5). As shown in
Table 7.5, these ratios are varied between 0.8 and 1.2. According to these ratios,
using the numerical analysis results, it is not possible to derive general
statements/relations between the behaviors of the prisms and walls. This may be due
265
to that the real compositions of the prisms and the walls may not be simulated with
the assumption of the perfect bond between the bricks and joints used in the
numerical analyses. It should be noted that the difference between the compositions
of the prisms and walls are the head joints.
0
1
2
3
4
5
0.00 0.01 0.02 0.03 0.04 0.05
Compressive strain
Compressive stress (MPa)
WCF-1
PCF-1
WCF-2
PCF-2
WCF-3
PCF-3
WCF-4
PCF-4
Figure 7.12 : The compressive stress-compressive strain relationships of the prisms
and the walls (numerical analysis).
Table 7.5 : The comparison of the numerical analyses for the prisms and walls.
Ratio fwcf wcf , f  wcf , p  wcf , p  wcf E wcf 
PCF-1/WCF-1 0.9 1.0 1.1 1.1 1.0 0.7
PCF-2/WCF-2 1.1 1.0 0.9 0.9 1.1 0.8
PCF-3/WCF-3 1.1 0.9 1.0 0.8 1.2 1.0
PCF-4/WCF-4 1.0 1.0 1.0 1.1 1.0
The compressive strength of the masonry can be predicted using Eurocode 6
equation (Eq. (1.3)), which is based on the compressive strengths of the bricks and
mortar determined experimentally. The prediction of the compressive strength is
illustrated for PCF-1 or WCF-1 in Eq. (7.1). In Table 7.6, the predictions made for
the all specimens of the numerical analyses are compared with the corresponding
numerical values. As shown in this table, the predictions of Eurocode 6 are about 26-
266
30% less than the results of the numerical analyses performed. This result indicates
that the numerical modeling adopted in this study should be improved in a way to
include the characteristics of the interfaces between the bricks and bed/head mortar
joints.
f f MPa
f Kf f f MPa
masc masc c
masc c uc n mc masc c
1.2 1.2 0.92 1.1
0.55 (1.4) (2.5) 0.92
,
0.7 0.3
, , ,
     
      
(7.1)
Table 7.6 : The comparison of the compressive strengths (numerical analysis and
Eurocode 6 prediction).
Specimen
Compressive strength (MPa) Ratio
Brick Mortar Analysis Prediction Eq. (1.3) (Prediction/Analysis)
PCF-1 (WCF-1) 1.4 2.5 1.5 (1.7) 1.1 0.73 (0.65)
PCF-2 (WCF-2) 3.3 2.9 3.2 (3.0) 2.1 0.66 (0.70)
PCF-3 (WCF-3) 4.7 2.9 4.1 (3.8) 2.7 0.66 (0.71)
PCF-4 (WCF-4) 1.8 2.9 1.9 (2.0) 1.4 0.74 (0.70)
To assess the contribution of the compressive strengths of the masonry components
to the masonry compressive strength, Eq. (1.3) can be used. In the equation, this
contribution is expressed with 0.7 0.3
uc,n mc f f . The contribution is calculated for each
specimen, (Eqs. (7.2), (7.3), (7.4) and (7.5)).
PCF-1 or WCF-1
f f MPa uc n mc (1.4) (2.5) 1.7 0.7 0.3 0.7 0.3
,   
(7.2)
PCF-2 or WCF-2
f f MPa uc n mc (3.3) (2.9) 3.2 0.7 0.3 0.7 0.3
,   
(7.3)
PCF-3 or WCF-3
f f MPa uc n mc (4.7) (2.9) 4.0 0.7 0.3 0.7 0.3
,   
(7.4)
PCF-4 or WCF-4
f f MPa uc n mc (1.8) (2.9) 2.1 0.7 0.3 0.7 0.3
,   
(7.5)
267
As seen, the contribution of the masonry components to the masonry compressive
strength is quite close to the corresponding compressive strength determined
numerically (Table 7.6). This indicates that the numerical analyses performed
overestimate the masonry compressive strengths, as this model does not include the
characteristics of the interfaces.
268
269
8. CONCLUSIONS
In this thesis, an experimental research is conducted on the historical masonry and
masonry constituents, which were taken from the historical Akaretler Row Houses
constructed around 1875. The historical masonry specimens are characterized,
particularly, in terms of mechanical behavior. It is considered that the mechanical
characteristics obtained from this study can be used for the historical brick masonry
structures built in 19th century. The knowledge on the mechanical characteristics of
the historical masonry structures are needed to understand the evolution of the
masonry materials by time, to find out the differences between the masonry
materials of several locations, to evaluate of the current state of the historical
structures and to produce similar materials for repair/restoration applications, if
needed. The conclusions drawn from this research are presented in this chapter.
Evaluation of the physical, chemical, and mechanical test results of the bricks and
the mortar revealed that the bricks are common type of bricks and that the binder of
the mortar is lime. The flexural tensile and compressive strengths of the bricks and
of the mortars are obtained. The ratios of the average flexural tensile strength to the
average compressive strength are calculated for the bricks and the mortars.
Therefore, when the compressive strengths of bricks and mortar are known, it is
possible to estimate values of the flexural tensile strengths using the obtained ratios
between tensile and compressive strengths. In addition, using the average
compressive strengths of the historical masonry components, the quality of these
components are assessed according to the related codes.
Original masonry specimens, which were taken from the historical walls (the cores
and the wallets), are tested under compression loads. The differences between the
test results of these specimen groups (core and wallet) were thought to be resulting
from size and composition differences. According to this study, the wallet
compressive strength may be taken as about 60% of the core compressive strength.
270
Using the average values of in-situ brick surface hardnesses and core compressive
strengths, which were obtained from the tests performed for the Akaretler Row
Houses and for several other structures belonging to the same construction period, a
linear relationship is proposed for predicting the average compressive strength of
masonry cores as a function of the average rebound number of bricks. Using this
equation, to estimate the average compressive strength of cores is possible without
carrying out core tests. As the rebound hammer test is non-destructive, the use of this
test method in the historical structures is appropriate.
Reproduced masonry specimens, which are prisms (ASTM C 1314-03b, 2003) and
walls (TS EN 1052-1, 2000), are tested under compression loads. The main objective
is to compare the test results of the two specimen groups built by taking into account
the proposals of the standards of ASTM C 1314-03b (2003) and TS EN 1052-1
(2000). The results of the reproduced specimens are also evaluated in a comparative
manner with the original specimens (core and wallet). It should be noted that the
other results of the prisms and wallets are close to each other. The average
compressive strengths of the wallets and walls are almost equal to each other. By
taking into account the similarities between the test results of the wallets and the
prisms, it may be concluded that the wallet specimens can be used if the original
wallet specimens can be taken from the in-situ walls.
The core and wallet specimens used to figure out masonry compressive strength are
not defined in the related codes, but the prism, and wall specimens are defined, and
the test methods are described in the corresponding codes. However, it is seen from
the results of the tests realized in this study that the average compressive strengths
obtained from the compression tests on these specimens are different. Consequently,
considering these differences in the compressive strengths, the need of use of a
correction factor reveals. The correction factor should be specified in a way to reflect
the effects of size, bond type and of damages to be occurred in the specimen during
test preparation. According to this study, correction factors are determined to
transform average compressive strengths of the cores, wallets, and walls to those of
the prisms.
271
For evaluating the compression tests quantitatively, the compressive strength and
corresponding strain, the compressive stress at proportional limit and corresponding
strain, the Young’s modulus and the ductility of each specimen are introduced for
the bricks, three-bricks, mortars, cores, wallets and prisms. The relationships of
Young’s modulus-compressive strength, compressive stress at proportional limitcompressive
strength and Young’s modulus-secant modulus at peak are expressed
with various linear functions. Modeling the relationships between compressive stress
and compressive strain, parabolic functions are proposed for the three-bricks,
mortars, cores, wallets and prisms based on the regression analysis conducted on the
corresponding test data. In the analysis and/or assessment of the existing masonry
structures, these functions can be utilized.
The effectiveness of code equations given for the prediction of the compressive
strength of historical masonry are evaluated considering the strengths of the tests. It
should be also noted that these equations estimate the masonry compressive strength
depending on the compressive strengths of the masonry components. The predictions
of the equations given by EN 1996-1-1 (2005) and TSDC (2007) are more close to
the test results with respect to the prediction of the equation given by BCRSMS
(2008).
To have knowledge on the shear behavior of the historical masonry, the in-situ shear
tests of the walls, and the laboratory shear tests of the cores are carried out. By
adopting Mohr-Coulomb criterion, the shear bond strengths and the coefficients of
friction for core and in-situ wall specimen groups are calculated from the tests. One
of the expected conclusions to be inferred from both shear tests is that there is a clear
inclination for the wall specimens under the high vertical stresses to display higher
values of the shear strength. It is understood from the average values that even in
case of similar vertical stresses for both test types, the shear strengths of the cores
are lower than those of the in-situ walls. This may be attributed to several factors
such as material variability and possible damages took place during extraction,
transportation, and/or preparation processes of the core specimens.
In addition to these factors, the differences between the test techniques and between
the specimen composition and sizes (while in-situ walls included two mortar bed,
and several head joints, the cores included one bed and no head joints) might lead to
be attained different values.
272
The triplets and the walls, which are reproduced masonry specimens, are built and
tested under shear loads taking account the proposals of TS EN 1052-3 (2004) and
ASTM E 519-02 (2003), ASTM E 72-05 (2005), respectively. The main purposes
were to figure out shear components and strengths and to compare the results with
the other shear test results. The prism and core shear tests may be conducted to
determine shear strength components of masonry since it is considered that the insitu
wall tests are destructive and that the results of the wall shear tests at laboratory
are more dependent on the boundary conditions. However, it is considered that the
results of the in-situ wall tests are more reliable, so the test results obtained from
prisms/cores should be corrected with respect to the in-situ wall tests.
Using compressive strength, shear strengths and corresponding compressive stresses,
the interaction curves of shear and compressive stresses are obtained for the cores,
the prisms, and the walls. To express these curves, parabolic functions are proposed.
It is possible to estimate the shear strength of a specimen under a specified
compressive stress utilizing these functions. A conclusion to be derived from these
curves is that the application of Mohr-Coulomb criterion should be limited. This
limit can be calculated as a percent of the compressive strength, that is, the criterion
can be utilized for the compressive stresses smaller than the limit. According to this
study, the limit may be taken as 30-40% of the compressive strength.
Several buildings built in the same period as the Akaretler Row Houses are also
included in this experimental study. By conducting tests on the cores of each
building, the compressive strength, the shear bond strength, and the friction
coefficient of the masonry walls of these buildings are obtained. It is seen that the
test results of these buildings are in good agreement with each other. The
relationships of shear bond strength-compressive strength and friction coefficientcompressive
strength are expressed with exponential functions for the core
specimens. Using these relationships, the shear components of a core specimen can
be predicted, if the core compressive strength is known.
Although large numbers of specimens are tested in different ways, as expected, due
to the variation of production techniques of masonry materials and/or workmanship,
as well as variations of exposures to the atmospheric conditions, the standard
deviations or the coefficients of variation of mechanical test results are high.
273
This large deviation may cause several vital problems in the determination of the
material properties of historical and/or ordinary masonry structures and in the
decision regarding material selection for the restoration/strengthening works, if
sufficient amount of samples are not taken into consideration. If the increment in the
number of specimens does not decrease the high deviation, the mechanical parameter
to be used in structural assessment may be taken as the average value obtained from
the tests (such as strength) minus the corresponding standard deviation. The high
coefficient of variation should also be defined depending on several parameters such
as the knowledge level required and the assessment method of the in-place materials.
The obtained results are evaluated according to TSDC (2007) and are compared with
the corresponding statements given in the code. The points to be highlighted are as
follows: The specimens to be tested should be clearly described and the knowledge
on the masonry buildings built with lime mortar should be provided. The statements
concerning Young’s modulus of the masonry wall should be more clear. The design
procedure of masonry structures is generally referred as the assessment procedure of
the existing masonry structures by the code. This approach may be changed in a way
to include the common characteristics of the existing masonry structures.
In order to complete and develop the subjects studied in the thesis, further
investigations are proposed as follows:
 to understand the effects of test technique, specimen size and specimen
composition on the test results of the low strength masonry, fabricated
material for providing uniform material characteristics with low strength may
be tested,
 to study the out-of-plane behavior of the historical masonry,
 to investigate the role of the bond type in the behavior of the historical
masonry, considering the common bond types,
 to investigate historical stone and/or alternating brick-stone masonry
structures in terms of material characterization and structural behavior, and
 to research the efficiency of the strengthening methods by taking into account
the characteristics of the low strength masonry.
274
275
REFERENCES
Abaqus, 2009. Version 6.9-1. Hibbit, Karlsson & Sorensen Inc.
Ahunbay, M., Ahunbay, Z., Almesberger, D., Rizzo, A., Toncic, M. and
İzmirligil, Ü., 1997. Non-destructive testing and monitoring in Hagia
Sophia / İstanbul. International Congress on Studies in Ancient
Structures, Yıldız Technical University, Istanbul, Turkey, 14-18 July,
pp. 197-206.
Akman, M.S., 1978. Deney ve ölçme tekniğine giriş, İstanbul Teknik Üniversitesi,
İstanbul.
Akman, M.S., Güner, A. and Aksoy, İ.H., 1986. The history and properties of
khorasan mortar and concrete. The 2nd International Congress on the
History of Turkish and Islamic Science and Technology, Istanbul
Technical University Research Center of History of Science and
Technology, Istanbul, Turkey, 28-02 May, Vol.1, pp. 101-111.
Andersen, H.D., Zimmermann, H.D., Friis, H., and Schnell, U., 1999.
Experimental approach to a procedure for the investigation of historic
mortars. Proceedings of the International RILEM Workshop on
Historic Mortars: Characteristics and Tests, Advanced Concrete and
Masonry Centre University of Paisley, Scotland, 12-14 May, pp. 37-
42.
Aprile, A., Benedetti, A. and Grassucci, F., 2001. Assessment of cracking and
collapse for old brick masonry columns, Journal of Structural
Engineering, 127, 1427-1435.
Arce, P.L., Guinea, J.G., Gracia, M. and Obis, J., 2003. Bricks in historical
buildings of Toledo city: characterisation and restoration, Materials
Characterization, 50, 59-68.
ASTM C 1314-03b, 2003. Standard test method for compressive strength of
masonry prisms, American Society for Testing Materials, USA.
ASTM C 1531-03, 2003. Standard test methods for in situ measurement of masonry
mortar joint shear strength index, American Society for Testing
Materials, USA.
ASTM C 348-02, 2002. Standard test methods for flexural strength of hydrauliccement
mortars, American Society for Testing Materials, USA.
ASTM C 349-02, 2002. Standard test methods for compressive strength of
hydraulic-cement mortars (using portions of prisms broken in flexure),
American Society for Testing Materials, USA.
276
ASTM C 62-05, 2006. Standard specification for building brick (solid masonry units
made from clay or shale), American Society for Testing Materials,
USA.
ASTM C 67-05, 2005. Standard test methods for sampling and testing brick and
structural clay tile, American Society for Testing Materials, USA.
ASTM E 111-04, 2004. Standard test method for Young’s modulus, tangent
modulus, and chord modulus, American Society for Testing Materials,
USA.
ASTM E 519-02, 2003. Standard test method for diagonal tension (shear) in
masonry assemblages, American Society for Testing Materials, USA.
ASTM E 72-05, 2005. Standard test methods of conducting strength tests of panels
for building construction, American Society for Testing Materials,
USA.
Atkinson, R.H., Amadei, B.P., Saeb, S. and Sture, S., 1989. Response of masonry
bed joints in direct shear. Journal of Structural Engineering, 115,
2276-2296.
Ayyub, B.M. and McCuen, R.H., 1996. Numerical methods for engineers, Prentice
Hall, New Jersey, USA.
Baronio, G. and Binda, L., 1995. Experimental approach to a procedure for the
investigation of historic mortars. Proceedings of the Joint International
Workshop proposed by RILEM TC 127-MS Tests for Masonry
Materials and Structures and CIB W23 Wall Structures on Evaluation
and Strengthening of Existing Masonry Structures, University of
Padua, Italy, 28-29 June, pp. 107-115.
Baronio, G., Binda, L., and Saisi, A., 1999. Mechanical and physical behaviour of
lime mortars reproduced after the characterisation of historical mortar.
Proceedings of the International RILEM Workshop on Historic
mortars: characteristics and tests, Paisley, Scotland, 12-14 May, pp.
307-325.
Baronio, G., Binda, L., Tedeschi, C. and Tiraboschi, C., 2003. Characterisation of
the materials used in the construction of the Noto Cathedral,
Consruction and Building Materials, 17, 557-571.
Bayülke, N., 1980. Yığma yapılar (in Turkish), T.C. İmar ve İskan Bakanlığı,
Deprem Araştırma Enstitüsü Başkanlığı, Ankara.
Beall, C., 2000. Masonry and concrete. McGraw-Hill Book Company Inc., New
York, USA.
Berktay, İ., 1995. Betonarme I-Taşıma gücü ve kesit hesapları. Yapım Matbaası,
İstanbul.
Binda, L. and Saisi, A., 2001. Non destructive testing applied to historic buildings:
The case of some Sicilian Churches. 3rd International Seminar on
Structural Analysis of Historical Constructions, University of Minho,
Portugal, 7-9 November, pp. 29-46.
277
Binda, L., Baronio, G. and Tedeschi, C., 1999. Experimental study on the
mechanical role of thick mortar joints in reproduced Byzantine
masonry. Proceedings of the International RILEM Workshop on
Historic Mortars: Characteristics and Tests, University of Paisley,
Scotland, 12-14 May, pp. 227-247.
Binda, L., Saisi, A. and Tiraboschi, C., 2000. Investigation procedures for the
diagnosis of historic masonries, Construction and Building Materials,
14, 199-233.
Brencich, A. and Gambarotta, L., 2005. Mechanical response of solid clay
brickwork under eccentric loading. Part I: Unreinforced masonry,
Materials and Structures, 38, 257-266.
Brencich, A., and Sterpi, E., 2006. Compressive strength of solid clay brick
masonry: calibration of experimental tests and theoretical issues. Proc.
of International Conference on Structural Analysis of Historical
Constructions. New Delhi, India, 6-8 Nov, pp. 757-765.
Çakmak, T., 2001. İstanbul sıraevleri ve bir sosyal konut modeli olarak Surp Agop
sıraevleri, MSc Thesis, I.T.U. Institute of Science and Technology,
Istanbul.
Çoban, S. ve Günay, T., 2006. Jeolojik ve Jeoteknik inceleme raporu, Tarihi
Akaretler sıraevleri gelişim projesi, Geo Zemin Etüd Maden Arama
Şti., İstanbul.
Croci, G., 2000. The conservation and structural restoration of architectural heritage,
Computational Mechanics Publications, Southampton, UK and
Boston, USA.
Drysdale, R.G., Hamid, A.A. and Baker, L.R., 1994. Masonry structures
behaviour and design. Prentice Hall Inc., Englewood Cliffs, New
Jersey, USA.
EN 1996-1-1, 2005. Eurocode 6: Design of masonry structures - Part 1-1: General
rules for reinforced and unreinforced masonry structures, European
Committee for Standardization, Brussels, Belgium.
EN 1998, 2005. Eurocode 8: Design of structures for earthquake resistance - Part 3:
Assessment and retrofitting of buildings, European Committee for
Standardization, Brussels, Belgium.
EN 1998-1, 2004. Eurocode 8: Design of structures for earthquake resistance - Part
1: General rules, seismic action and rules for buildings, European
Committee for Standardization, Brussels, Belgium.
Erenoğlu, G., 1998. Kentsel kimlikte sürekliliğin rolü ve Akaretler evleri-Teşvikiye
bölgesi örnek alanında tarihi çevrede modern yapı incelemesi, MSc
Thesis, I.T.U. Institute of Science and Technology, Istanbul.
Felice, G., 2006. Experimental investigation on historic brickwork subjected to
eccentric axial loads, Proceedings of the 5th International Conference
on Structural Analysis of Historical Constructions, New Delhi, India,
6-8 November, pp. 809-816.
278
FEMA 356, 2000. Prestandard and commentary for the seismic rehabilitation of
buildings, Federal Emergency Management Agency, USA.
Güleç, A. and Tulun, T., 1997. Physico-chemical and petrographical studies of old
mortars and plasters of Anatolia, Cement and Concrete Research, 27
(2), 227-234.
Güleç, A., Acun, S. and Ersen, A., 2005. A characterization method for the fifthcentury
traditional mortars in the land walls of Constantinople,
Yedikule, Studies in Conservation, 50, 295-306.
Gumaste, K.S., Rao, K.S.N., Reddy, B.V.V. and Jagadish, K.S., 2007. Strength
and elasticity of brick masonry prisms and wallettes under
compression, Materials and Structures, 40, 241-253.
Güner, A., 1984. Compressive tests on khorasan-burnt brick masonry specimens (in
Turkish), Structural Material Lab. in Istanbul Technical University
Report, 186, Istanbul, Turkey.
Hendry, A.W., 1990. Structural masonry. Macmillan Education, London, England.
Hendry, A.W., Khalaf, F.M., 2003. Masonry wall construction, Spon Press,
London, England.
Ilki, A., Ispir, M., Demir, C., Karamuk, E., As, F., Cini, N., Aydin, M.,
Toponder, G., Tulun, T., Kumbasar, N. and Akman, S., 2007.
Seismic safety of a historical row house complex built during ottoman
period, the Ninth Canadian Conference on Earthquake Engineering,
Ottawa, Ontario, Canada, 26-29 June (CD-ROM).
Ispir, M., and Ilki, A., 2010. Compressive behavior of historical masonry subjected
to monotonic and cyclic loads, 3rd International Workshop on
Conservation of Heritage Structures Using FRM and SHM, CSHM-3,
Ottawa-Gatineau, Canada, 11-13 August. (accepted to be published
and to be presented).
Ispir, M., Demir, C., Ilki, A. and Kumbasar, N., 2008. Evaluation of test methods
for seismic safety assessment of masonry buildings, 8th International
Seminar on Structural Masonry (ISSM 08), Istanbul Technical
University, Istanbul, Turkey, 05-07 Nov., pp. 387-394.
Karavezirogluo, M., Barboutis, C. and Kranas, V., 1997. Behaviour of masonry
with full bricks and lime mortars. International Congress on Studies in
Ancient Structures, Yıldız Technical University, Istanbul, Turkey, 14-
18 July, pp. 225-233.
Karavezirogluo-Weber, M., Barboutis, C. and Kranas, B. and Zombou, A.,
1998. Experimental research on the compressive strength of masonry
repairs in Archeiropoietus Church, Thessaloniki, The Structural
Engineer, 76, 353-356.
Kaushik, H.B., Rai, D.C. and Jain, S.K., 2007. Stress-strain characteristics of clay
brick masonry under uniaxial compression, Journal of Materials in
Civil Engineering, 19(9), 728-739.
279
Kırşan, Ç., 1996. 19. Yüzyıl İstanbul dizi konutlarının morfolojik analizine dayalı
bilgi-tabanlı tasarım modeli, MSc Thesis, I.T.U. Institute of Science
and Technology, Istanbul.
Krstevska, L., Tashkov, L., Arun, G. and Aköz, F., 2007. Evaluation of seismic
behavior of historical monuments. International Sympoosium on
Studies on Historical Heritage, Yıldız Technical University, Antalya,
Turkey, 17-21 September, pp. 411-418.
Kuban, D., 2002. Mimarlık kavramları: Tarihsel perspektif içinde mimarlığın
kuramsal sözlüğüne giriş, Yapı-Endüstri Merkezi Yayınları, İstanbul.
Lourenço, P.B and Pina-Henriques, J., 2006. Validation of analytical and
continuum numerical methods for estimating the compressive strength
of masonry, Computers and Structures, 84, 1977-1989.
Lourenço, P.B., 1994. Analysis of masonry structures with interface elements-
Theory and applications, TU-DELFT Report No. 03-21-0-01,TNOBOUW
Report No. 94-NM-R0762, TNO Building and Construction
Research, Computational Mechanics, Portugal.
Matovic, V. and Milovanovic, D., 2001. Durability and decay type of sandstone
from the façade of the St. Marco church in Belgrade (Serbia). 2nd
International Congress on Studies in Ancient Structures, Yıldız
Technical University, Istanbul, Turkey, 9-13 July, pp. 599-606.
Mulligan, J.A., 1942. Handbook of brick masonry construction, McGraw-Hill Book
Company Inc., New York, USA and London, England.
Naraine, K. and Sinha, S., 1991. Cyclic behavior of brick masonry under biaxial
compression, Journal of Structural Engineering, 117 (5), 1336-1355.
Oliveira, D.V.C., 2003. Experimental and numerical analysis of blocky masonry
structures under cyclic loading, PhD Thesis, University of Minho,
Portugal.
Omurtag, M.H., 2007. Mukavemet, Cilt 2 (in Turkish), Birsen Yayınevi Ltd. Şti.,
Istanbul.
Orton, A., 1992. Structural design of masonry, Longman Publishers, London,
England and New York, USA.
Palieraki, V., Vintzileou, E. and Fezans-Miltiadou, A., 2007. The use of radar
technique and boroscopy in investigating historic masonry:
Application of the techniques in Byzantine monuments in Greece.
International Symposium on Studies on Historical Heritage, Yıldız
Technical University, Antalya, Turkey, 17-21 September, pp. 403-
410.
Pande, G.N., Kralj, B. and Middleton, J., 1994. Analysis of the compressive
strength of masonry given by the equation    
k b m f  K f f , The
Structural Engineer, 71, 7-12.
280
Papayianni, I. and Hatzitrifonos, E., 1997. A survey of the Pazar Hamam in
Thessaloniki. Proceedings of International Symposium on Studies in
Ancient Heritage, Yıldız Technical University, Istanbul, Turkey, 14-
18 July, pp. 283-293.
Papayianni, I. and Stefanidou, M., 2001. Porosity and structure of old mortars.
Proceedings of 2nd International Congress on Studies in Ancient
Structures, Yıldız Technical University, Istanbul, Turkey, 9-13 July,
pp. 509-517.
Papayianni, I. and Stefanidou, M., 2007. The influence of mixture design
parameters on the long term strength of lime-based mortars.
International Symposium on Studies in Historical Heritage, Yıldız
Technical University, Antalya, Turkey, 17-21 September, pp. 355-
362.
Papayianni, I., 1997. Technology of mortars and bricks used in Ottoman
monuments of Thessaloniki. Proceedings of International Congress on
Studies in Ancient Structures, Yıldız Technical University, Istanbul,
Turkey, 14-18 July, pp. 245-253.
Paulay, T. and Priestly, M.J.N., 1992. Seismic design of reinforced concrete and
masonry buildings, Wiley-Interscience, New York, USA.
Penelis, G.G., 2002. Structural restoration of historical buildings in seismic areas,
Prog. Struct. Engng Mater., 4, 64-73.
Pohle, F. and Jager, W., 2003. Structural restoration of historical buildings in
seismic areas, Construction and Building Materials, 17, 651-667.
Postacıoğlu, B., 1981. Tests on khorasan jointed burnt brick masonry specimens
from vaults (in Turkish), Structural Material Lab. in Istanbul
Technical University Report, 177, Istanbul, Turkey.
Prakash, S.S. and Alagusundaramoorthy, P., 2008. Load resistance of masonry
wallets and shear triplets retrofitted with GFRP composites, Cement &
Concrete Composites, 30, 745-761.
Radivojevic, A., Dervissis, D., 2001. Production and testing of bricks for repair
work, 2nd International Congress on Studies in Ancient Structures,
Yıldız Technical University, Istanbul, Turkey, 9-13 July, pp. 581-587.
Romano, A., 2005. Modelling, analysis and testing of masonry structures, PhD
Thesis, Università degli Studi di Napoli Federico II. Naples, Italy.
Sağdıç, Z., 1999. The investigation of the idea of row houses and the Akaretler row
houses in Ottoman Architecture (in Turkish), MSc Thesis, I.T.U.
Institute of Science and Technology, Istanbul.
Sahlin, S., 1971. Structural masonry, Prentice-Hall Inc, Englewood Cliffs, New
Jersey, USA.
Sarangapani, G., Reddy, B.V.V. and Jagadish, K.S., 2005. Brick-mortar bond and
masonry compressive strength, Journal of Materials in Civil
Engineering, 17, 229-237.
281
Thomas, K., 1971. Structural brickwork-materials and performance, The Structural
Engineer, 49, 441-450.
TMS 402-08/ACI 530-08/ASCE 5-08 and TMS 602-08/ACI 530.1/ASCE 6-08,
2008. Building Code Requirements and Specification for Masonry
Structures, American Concrete Institute, USA.
Tomazevic, M., 2006. Earthquake-resistant design of masonry buildings, Vol.1.
Imperial College Press, London, England.
TS 500, 2000. Betonarme yapıların tasarım ve hesap kuralları, Türk Standardları
Enstitüsü, Ankara.
TS EN 1015-11, 2000. Kagir Harcı - Deney metotları - Bölüm 11: Sertleşmiş harcın
basınç ve eğilme dayanımının tayini, Türk Standardları Enstitüsü,
Ankara.
TS EN 1052-1, 2000. Kagir - Deney metotları - Bölüm 1: Basınç dayanımı tayini,
Türk Standardları Enstitüsü, Ankara.
TS EN 1052-3, 2004. Kagir - Deney metotları - Bölüm 3: Başlangıç kayma
dayanımının tayini, Türk Standardları Enstitüsü, Ankara.
TS EN 771-1, 2005. Kagir birimler - Özellikler - Bölüm 1: Kil kagir birimler
(tuğlalar), Türk Standardları Enstitüsü, Ankara.
TS EN 772-1, 2002. Kagir birimler - Deney metotları - Bölüm 1: Basınç
dayanımının tayini, Türk Standardları Enstitüsü, Ankara.
TS EN 772-16, 2002. Kagir birimler - Deney metotları - Bölüm 16: Boyutların
tayini, Türk Standardları Enstitüsü, Ankara.
Turkish Seismic Design Code, 2007. Deprem bölgelerinde yapılacak binalar
hakkında yönetmelik, Türk Standardları Enstitüsü, Ankara.
UNDP/UNIDO PROJECT RER/79/015, 1984. The regional project: Building
construction under seismic conditions in the Balkan region, Design
and construction of stone and brick-masonry buildings, Vol.3, United
Nations Development Programme.
Url-1 <http://www.ecologreen.com/testresearch.html>, retrieved at 24.02.2006.
Valluzzi, M.R., Tinazzi, D. and Modena, C., 2002. Shear behaviour of masonry
panels strengthened by FRP laminates, Construction and Building
Materials, 16, 409-416.
Yorulmaz, M. and Atan, Y.T., 1971. Çeşitli forme yapı taşlarıyla yapılmış duvar
numunelerinin iki istikametli yükleme altında davranışları, İstanbul
Teknik Üniversitesi Matbaası, Gümüşsuyu, İstanbul.
282
283
APPENDICES
APPENDIX A : The Normalized Compressive Strength of the Brick Specimens
APPENDIX B: The Compressive Stress-Strain Relationships of the Wallets
APPENDIX C: The Compressive Stress Contours and Deformed Shapes of the
Prisms and Walls
284
285
APPENDIX A : The Normalized Compressive Strength of the Brick Specimens
Table A.1 : The conversion factors for the brick compressive strength.
Height (mm)
Minimum (width, length) (mm)
50 100 150 200 ≥250
40 0.80 0.70
50 0.85 0.75 0.70
65 0.95 0.85 0.75 0.70 0.65
100 1.15 1.00 0.90 0.80 0.75
150 1.30 1.20 1.10 1.00 0.95
200 1.45 1.35 1.25 1.15 1.10
≥250 1.55 1.45 1.35 1.25 1.15
286
Table A.2 : The normalized compressive strengths of the bricks.
Specimen min( , ) b b l b (mm) b h (mm)  bc   f (MPa)
BC-1 100 50 0.75 7.38
BC-2 81 55 0.82 4.67
BC-3 105 45 0.73 5.74
BC-4 105 50 0.75 4.14
BC-5 107 55 0.77 1.57
BC-6 104 50 0.75 1.51
BC-7 99 52 0.76 3.42
BC-8 85 49 0.77 5.42
BC-9 76 48 0.79 1.49
BC-10 107 60 0.91 3.59
BC-11 126 79 0.86 10.31
BC-12 111 72 0.86 3.33
BC-13 119 71 0.84 4.28
BC-14 123 75 0.84 3.91
BC-15 123 72 0.83 7.84
BC-16 111 76 0.88 8.18
BC-17 112 76 0.88 2.85
BC-18 112 79 0.89 4.57
BC-19 108 70 0.85 4.47
BC-20 124 83 0.88 10.4
BC-21 109 78 0.89 4.01
BC-22 109 72 0.86 1.78
BC-23 111 72 0.86 1.53
BC-24 111 75 0.87 2.27
BC-25 118 72 0.84 4.48
Table A.3 : The normalized compressive strengths of the bricks in parallel to the
bed joint.
Specimen min( , ) b b l b (mm) b h (mm)  bc p f ,   (MPa)
BCp-21 66 131 1.20 2.82
BCp-22 63 135 1.23 2.28
BCp-23 61 125 1.20 1.02
BCp-24 64 132 1.22 1.71
BCp-25 62 141 1.24 2.73
287
APPENDIX B: The Compressive Stress-Strain Relationships of the Wallets
Figure B.1 : The relationships of compressive stress-vertical strain.
0
1
2
3
0.00 0.01 0.02 0.03 0.04 0.05 0.06
Vertical strain
Compressive stress (MPa)
WtC-1
WtC-2
WtC-3
WtC-4
WtC-5
WtC-6
WtC-7
WtC-8
WtC-9
WtC-10
288
Figure B.2 : The relationships of compressive stress-vertical strain and their
envelope curves under cyclic loads.
0
1
2
3
0.00 0.01 0.02Vertic0a.0l 3strain 0.04 0.05 0.06
Compressive stress (MPa)
WtCC-1
WtCC-1-Envelope
0
1
2
3
0.00 0.01 0.02Verti0c.a0l3 strain0 .04 0.05 0.06
Compressive stress (MPa)
WtCC-2
WtCC-2-Envelope
0
1
2
3
0.00 0.01 0.02 Vert0i.c0a3l strain0 .04 0.05 0.06
Compressive stress (MPa)
WtCC-3
WtCC-3-Envelope
0
1
2
3
0.00 0.01 0.02Verti0c.a0l 3strain0 .04 0.05 0.06
Compressive stress (MPa)
WtCC-4
WtCC-4-Envelope
0
1
2
3
0.00 0.01 0.02Verti0c.a0l3 strain0 .04 0.05 0.06
Compressive stress (MPa)
WtCC-5
WtCC-5-Envelope
0
1
2
3
0.00 0.01 0.02Verti0c.a0l3 strain0 .04 0.05 0.06
Compressive stress (MPa)
WtCC-1-Envelope
WtCC-2-Envelope
WtCC-3-Envelope
WtCC-4-Envelope
WtCC-5-Envelope
289
APPENDIX C: The Compressive Stress Contours and Deformed Shapes of the
Prisms and Walls
Figure C.1 : The compressive stress contours and deformed shapes of PCF-2.
z
x
z
y
z
x
A1
A2 z
y
A3
A4
Max stress = 4.393E-1 MPa
Min stress = -4.060 MPa
290
Figure C.2 : The compressive stress contours and deformed shapes of PCF-3.
Figure C.3 : The compressive stress contours and deformed shapes of PCF-4.
A1
A2 z
y
A3
A4
z
y
z
x
Max stress = 6.444E-1 MPa
Min stress = -5.916 MPa
z
x
z
x
Max stress = 1.126E-1 MPa
Min stress = -2.002 MPa
A1
z A2
x
A4
A3
z
y
z
y
291
Figure C.4 : The compressive stress contours and deformed shapes of WCF-2.
Figure C.5 : The compressive stress contours and deformed shapes of WCF-3.
yz
x
A1
A2
A3
A4
Max stress = 2.945 MPa
Min stress = -6.709 MPa
y
x
y
x
A1
A2
A3
A4
Max stress 1.012 MPa
Min stress = -4.813 MPa
y
x
292
Figure C.6 : The compressive stress contours and deformed shapes of WCF-4.
y
x
A1
A2
A3
A4
Max stress = 2.945 MPa
Min stress = -6.709
M
P
a
y
x
293
CURRICULUM VITA
Candidate’s full name: Medine İSPİR
Place and date of birth: Elbistan-20.03.1974
Permanent Address: Kocasinan Merkez mah. İnönü cad. No:11/5
Yenibosna-Bahçelievler-İstanbul
Universities and Colleges attended: Istanbul Technical University
Publications:
 Ispir, M., Demir, C., Ilki, A. and Kumbasar, N., 2009. Material characterization of
the historical unreinforced masonry in İstanbul, ASCE, Journal of Materials in Civil
Engineering, 22, pp. 712-713.
 Ispir, M., Dalgic, K.D., Sengul, C., Kuran, F., Ilki, A. and Tasdemir, M.A., 2010.
Modulus of Elasticity of Low Strength Concrete, 9th International Congress on
Advances in Civil Engineering (ACE), Karadeniz Technical University, Trabzon,
Turkey, September 27-30.
 Ispir, M., and Ilki, A., 2010. Compressive Behavior of Historical masonry
subjected to monotonic and cyclic loads, 3rd International Workshop on
Conservation of Heritage Structures Using FRM and SHM (CSHM-3), Ottawa-
Gatineau, Canada, 11-13 August 2010.
 Ispir, M., Ozkaynak, H., Yuksel, E., and Ilki, A., 2010. Diagonal Tension Tests of
Hollow Brick Walls Retrofitted with CFRP, Proceedings of the Turkish-Saudi
Workshop on Structural & Earthquake Engineering, Istanbul Technical University,
Istanbul, Turkey, June 24.
 Ispir, M., Demir, C., Ilki, A. and Kumbasar, N., 2008. Evaluation of Test Methods
for Seismic Safety Assessment of Masonry Buildings, 8th International Seminar on
Structural Masonry ( ISSM08), Istanbul, November 05-07.
 Ilki, A., Ispir, M., As, F., Demir, C. and Kumbasar, N., 2008. FRP Retrofit of
Walls Constructed with Historical Bricks, International Conference on Challenges
for Civil Construction (CCC2008), Porto, April 16-18.
 Demir, C., Acar, D., Terzi, S.O., Ispir, M., Demirtas, B., Ilki, A. and Kumbasar,
N., 2007. Preliminary Structural Assessment of Kariye Monument-Northern Annex,
Int. Symp. Studies on Historical Heritage, Antalya, Turkey, September 17-21.
 Ilki, A., Ispir, M., Demir, C., Karamuk, E., As, F., Cini, N., Aydin, M., Toponder,
G., Tulun, T., Kumbasar, N. and Akman, S. 2007. Seismic Safety of a Historical
Row House Complex Built during Ottoman Period, 9th Canadian Conference on
Earthquake Engineering, Ottowa, Canada, June 26-29.
294
 Ilki, A., Ispir, M., Demir, C., Kumbasar, N., Akman, S., 2006. An Outline of the
Seismic Behavior of Historical Structures in North Western Anatolia, International
Conference on Structural Analysis of Historical Constructions (SAHC2006), New
Delhi, India, November 6-8.