See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/285639041

Experimental Determination of Statistical Parameters Associated with

Uniaxial Compression Behaviour of Brick Masonry

Article in Current Science · December 2015

DOI: 10.18520/v109/i11/2094-2102

CITATIONS READS

14 854

6 authors, including:

Balasubramanian S R Rupen Goswami

CSIR Structural Engineering Research Centre Indian Institute of Technology Madras
15 PUBLICATIONS 49 CITATIONS 249 PUBLICATIONS 434 CITATIONS

SEE PROFILE SEE PROFILE

P. Sivakumar
CSIR Structural Engineering Research Centre
24 PUBLICATIONS 148 CITATIONS

SEE PROFILE

All content following this page was uploaded by Rupen Goswami on 14 December 2015.

The user has requested enhancement of the downloaded file.

RESEARCH ARTICLES

Experimental determination of statistical

parameters associated with uniaxial
compression behaviour of brick masonry
S. R. Balasubramanian1,*, D. Maheswari2, A. Cynthia3, K. Balaji Rao1,
A. Meher Prasad4, R. Goswami4 and P. Sivakumar1
1
CSIR-Structural Engineering Research Centre, CSIR Campus, Taramani, Chennai 600 113, India
2
Department of Civil Engineering, Kumaraguru College of Technology, Coimbatore 641 006, India
3
Department of Civil Engineering, VelTech Multi Tech Dr Rangarajan Dr Sakunthala Engineering College, Chennai 600 025, India
4
Department of Civil Engineering, Indian Institute of Technology Madras, Chennai 600 036, India

The Bureau of Indian Standards (BIS) provides a table

In view of practical significance of the compression
behaviour of brick masonry, this article discusses the (table 8 in the respective code of practice) in which basic
evolvement of an experimental programme based on a compressive stress of masonry is prescribed for a limited
survey of the literature. Also, it is known that large set of brick and mortar combinations based on their com-
scatter is expected in the mechanical properties of pressive strengths1. Whereas in the international scenario,
masonry and studies characterizing these statistical a number of expressions have been proposed2 among
variations are scant in India. Using the evolved ex- which the following expression is being widely used
perimental programme and results of tests conducted,
the statistical parameters, namely mean and coeffi-
cient of variation (COV) associated with the uniaxial f m  kf b f j (1)
compression behaviour of typical brick masonry used
in South India have been determined in this article.
For the masonry considered in this study, the mean where f m is the compressive strength of masonry, fb the
values of peak compressive stress, strain correspond- normalized compressive strength of brick, f j is the com-
ing to peak stress and elastic modulus are 2.82 MPa, pressive strength of mortar and k,  and  are constants.
0.009 and 0.4 GPa respectively. The corresponding Equation (1) is not considered to be rational as it does
values of COV are 0.15, 0.2 and 0.12 respectively. In
not take into account the factors that directly affect the
addition, a trilinear curve has been suggested as an
idealized stress–strain relation for the brick masonry compressive strength of masonry and an expression based
used in South India. on two-stage homogenization has been proposed in the
literature3. However eq. (1) is known to give satisfactory
results for practical purposes3. Hence for characterizing
Keywords: Clay brick masonry, compressive strength, the compressive strength of masonry based on the com-
elastic modulus, uniaxial compression, statistical para- pressive strengths of brick and mortar, this type of
meters. expression may be useful.
In most of the western countries where this type of
BRICK masonry structures are one of the oldest forms of expression (i.e. eq. (1)) is being used, the bricks are rela-
construction. Even today, they are widely used especially tively stronger and stiffer than the mortar. In such condi-
for low rise and residential buildings in developing coun- tions, in accordance with the elastic sandwich model4,
tries like India. Despite the long history, brick masonry is when a stack bonded prism is subjected to uniaxial com-
least understood with respect to its behaviour. The design pression, due to Poisson effect the mortar in the bed
of steel or reinforced concrete structures is based on joints tends to laterally expand more than the bricks, but
mechanics-based principles, whereas the design of the mortar will be confined by the bricks because of the
masonry structures is empirical in general1. Hence, there bond between them. Thus, the bricks will be subjected to
is ample scope for understanding brick masonry and relative tension and in turn the mortar will be subjected to
establish mechanics-based inter-relationship between relative compression. Whereas, in India, often the mortar
brick, mortar, nature of interface and masonry used in used is relatively stronger and stiffer than the bricks. This
India. would result in relative compression of bricks and rela-
tive tension of the mortar in contrast to the scenario in
western countries. This issue has been pointed out in the
*For correspondence. (e-mail: [email protected]) literature5,6. Hence, adopting an expression such as

2094 CURRENT SCIENCE, VOL. 109, NO. 11, 10 DECEMBER 2015

 RESEARCH ARTICLES

eq. (1) needs thorough investigation; also the values of Survey of Standards and the literature on the
the constants k,  and  need to be evaluated in the test procedure
Indian scenario by carrying out tests on a wide variety of
brick and mortar combinations. Salient steps involved in the test and the corresponding
In order to characterize the compression behaviour recommendations as given in the Standards1,11,12 and
of brick masonry, other parameters, viz. elastic modulus, adopted in the literature5,6,9 concerned with brick masonry
and deformability are also important. Hence, the experi- used in India, are shown in Box 1.
mental results on compression stress–strain relation
of brick masonry are of practical significance. Attempts
Experimental programme on compressive
towards testing of brick masonry under compression have
stress–strain behaviour of brick masonry
been made in the recent past5–10.
Testing of brick masonry under uniaxial and biaxial
In this section, we provide details of the experimental
cyclic loading has been carried out by Naraine and Sin-
programme, evolved based on a survey of the literature.
ha7, and AlShebani and Sinha8. The scope of the present
A critical discussion on the literature is also included at
article is the characterization of stress–strain relation of
relevant places. More details of experimental studies car-
brick masonry subjected to monotonic, uniaxial compres-
ried out are given below. Some of the highlights of the
sion loading. Hence the studies on cyclic loading are not
present experimental programme are:
included in further discussions.
Stress–strain relation for masonry has been obtained (1) It includes a procedure to control the moisture con-
by Sarangapani et al.5 through compression test on stack tent in the bricks and mortar during fabrication of
bonded prisms and the initial tangent modulus, secant prism.
modulus at 25% of ultimate stress and the strain corre- (2) It includes the usage of compressometer that is cus-
sponding to ultimate stress also have been obtained. tomized for measuring strain in the brick masonry
Gumaste et al.6 have tested four types of specimen (viz. prisms.
stack bonded prism, English bonded prism, stretcher (3) It proposes a rate of displacement to be adopted in
bonded wallettes and English bonded wallettes). In order to obtain the post-peak behaviour properly.
addition to the values of initial tangent modulus, secant
modulus at 25% of ultimate stress and strain correspond- In the present experimental programme, stack bonded
ing to ultimate stress, the constants k,  and  as given in prisms used have been fabricated out of five bricks as
eq. (1) have been evaluated by them. Kaushik et al.9 have shown in Figure 1. It is to be noted that the prisms meas-
tested stack bonded prisms and have presented the chord uring 420–440 mm height, 219–226 mm width and 100–
modulus between 5% and 33% of ultimate stress and 105 mm thickness satisfy the requirement pertaining to
strains corresponding to eight different levels of normal- the aspect ratio as given in IS1 and ASTM11. Since the
ized stress. Also, the expressions for compressive main concern of this article is to obtain statistical para-
strength of masonry, strain corresponding to peak stress meters associated with the stress–strain relation, it is
and simplified models for stress–strain relation of considered the sample size of 66 and of 79 may not be suf-
masonry have been presented. Though there are com- ficient. If a sample size of 7 is considered, the length of
monalities in the aspects considered in these studies, confidence interval for mean value at a 5% level of sig-
there are differences in the procedures adopted. It is nificance for a typical brick masonry with mean value of
also noted that there is no standard test procedure in the 4 MPa and standard deviation 0.5 MPa is 0.71, whereas
BIS for obtaining the stress–strain relation of masonry. the length of confidence interval comes down to 0.45 (a
Keeping these in view, a survey of the literature and reduction of about 57%) if a sample size of 15 is consid-
standards has been carried out and based on this a ered. Hence, in the present programme, a sample size of
detailed experimental programme has been evolved for 15 specimens has been considered.
carrying out test on compression behaviour of brick We decided to use a poor quality brick and cement
masonry. Actual test on a typical brick masonry fabri- mortar (1 : 5) by weight that is commonly used in the
cated using soft and weak bricks from South India has construction of small residential buildings. The bricks
been performed and the results have been compared with have been subjected to monotonically increasing com-
those obtained from the expressions proposed in the litera- pression of 14 N/mm 2/min; Table 1 presents the mean
ture. Results of these comparisons have been presented. and COV of compressive strength computed based on test
It is known that the variation in the mechanical proper- results. In this programme, it has been decided to pre-wet
ties of bricks, mortars and hence the masonry is large. the bricks to a level of approximately 75% of saturation.
Therefore, reliable estimates of mean and coefficient of This is to achieve a good brick–mortar interface by
variation (COV) of important parameters associated with avoiding the floating of bricks and limiting the loss of
uniaxial compression behaviour are necessary and are of water from the mortar as recommended by Groot and
practical significance. Larbi 13. In Table 1, the water absorption corresponding to
CURRENT SCIENCE, VOL. 109, NO. 11, 10 DECEMBER 2015 2095
RESEARCH ARTICLES

Box 1. Summary of survey of standards and literature

Recommendation in standards

Description Adopted in the literature in India

11
Specimen type ASTM : Stack bonded prisms of aspect ratio (height to thickness ratio) between
1.3 and 5.
12
BS : Wallettes of 1.2–1.8 m length having a minimum cross-sectional area of
2
0.125 m and 2.4–2.7 m in height.
1
IS : Stack bonded prisms of aspect ratio (height to thickness ratio) at least 2 but
not more than 5.
5
Sarangapani et al. : Stack bonded prism, 435 mm  225 mm  105 mm.
6
Gumaste et al. : Stack bonded prism: 460 mm  230 mm  105 mm; English
bonded prism: 460 mm  230 mm  230 mm; stretcher bonded wallettes:
600 mm  520 mm  105 mm; English bonded wallettes: 665 mm 
520 mm  230 mm.
9
Kaushik et al. : Stack bonded prisms, 400–410 mm height, 110 mm thickness.
11
No. of specimens ASTM : To be decided in accordance with ASTM E122-09 (ref. 23).
12
BS : Minimum of 2 nos.
1
IS : Minimum of 5 nos.
5
Sarangapani et al. : Not mentioned.
6
Gumaste et al. : Six specimens for each sample. Twenty-one samples out of four
types of specimen, two types of bricks and six types of mortar have been
considered.
9
Kaushik et al. : Seven specimens for each sample. Twelve samples out of four
types of bricks and three types of mortar have been considered.
11
Conditioning of units ASTM : Surface of the units shall be free of moisture at the time of laying.
12 2
BS : Bricks with IRA more than 1.5 kg/m /min may be docked or be water con-
tent in the mortar may be adjusted.
1
IS : Moisture content of the units at the time of laying shall be same as that used
in the structure.
5
Sarangapani et al. : Not mentioned.
6
Gumaste et al. : Table moulded bricks are soaked under water for 15 min and
wire cut bricks for a duration of 45 min. This is to achieve 75% of saturation.
9
Kaushik et al. : Not mentioned.
11
Preparation of mortar ASTM : No specific recommendation.
12 24
BS (BS ): 35–140 litre/50 kg of cement depending on the proportion of the mix.
1
IS : Consistency of the mortar and thickness of joint shall be same as that used
in the structure.
25
IS : The working consistency of a mortar or plastering mix as judged by the worker
from its behaviour during application. Generally, only as much quantity of ce-
ment mortar as would be sufficient for 30 min of work shall be mixed at a time.
5
Sarangapani et al. : Not mentioned.
6
Gumaste et al. : Not mentioned.
9
Kaushik et al. : Not mentioned.
11
Fabrication of the specimen ASTM : Orient the units in the prism as in the corresponding conduction. In the
case of hollow blocks with multiple cells, only one cell shall be trimmed flush
and used. In the case of grouted masonry, the grouting shall be carried out
between 4 and 48 h after fabrication of the specimen.
12
BS : Construction and bond of the wallettes should correspond to those to be
used in practice.
1
IS : Workmanship and quality of construction shall be same as that used in the
structure.
5
Sarangapani et al. : Not mentioned.
6
Gumaste et al. : Not mentioned.
9
Kaushik et al. : Not mentioned.
(Contd)

2096 CURRENT SCIENCE, VOL. 109, NO. 11, 10 DECEMBER 2015

 RESEARCH ARTICLES

Box 1. (Contd)
Recommendation in standards

Description Adopted in the literature in India

11
Curing ASTM : Up to 48 h covered with moisture-tight bag and left undisturbed/do not
transport. After 48 h cure at 24  8C and humidity less than 80%.
12
BS : Up to 3 days: covered with polyethylene sheets and then left uncovered
until tested. Recommended age of testing: 28 days; when required, this may
be extended to 35 days.
1
IS : No specific recommendation.
5
Sarangapani et al. : Method of curing has not been mentioned. But the duration
of curing is 28 days and the specimen is soaked under water for 48 h before
testing.
6
Gumaste et al. : Method of curing has not been mentioned, but the duration of
curing is 28 days and the tests have been carried out under saturated condi-
tion.
9
Kaushik et al. : Not mentioned.
11 26
Capping ASTM (ASTM ): Either gypsum cement materials or sulphur capping materials.
Age of the caps shall at least be 2 h and possess a compressive strength of
at least 24.1 MPa at an age of 2 h.
12
BS : No specific recommendation.
1
IS : No specific recommendation.
5
Sarangapani et al. : Not mentioned.
6
Gumaste et al. : Specimens were capped with rich cement sand mortar (1 : 1)
before curing and allowed to gain strength during curing itself.
9
Kaushik et al. : Not mentioned.
11
Requirements for testing machine ASTM : Accuracy of 1% of the anticipated load range. The upper platen shall
be spherically seated and attached to the centre of the head of the machine.
The centre of the sphere shall be free to move in any direction. The diameter
of the upper platen shall be at least 150 mm.
12
BS : No specific recommendation.
1
IS : The upper platen shall be spherically seated.
5
Sarangapani et al. : Not mentioned.
6
Gumaste et al. : Not mentioned.
9
Kaushik et al. : Not mentioned.
11 27
Requirements for strain gauge ASTM (ASTM ): Class B1 or better quality.
12
BS : No specific recommendation.
1
IS : No specific recommendation.
5
Sarangapani et al. : Demec gauge of 200 mm gauge length.
6
Gumaste et al. : Demec gauge of 200 mm gauge length.
9
Kaushik et al. : Epsilon extensometers are positioned such that they covers
three mortar joints.
11 27
Rate of loading ASTM (ASTM ): The speed of testing shall be low enough that thermal effects
of adiabatic expansion or contraction are negligible and that accurate deter-
mination of load and extension is possible yet the speed shall be high enough
that creep will be negligible. In loading with dead weights, avoid temporary
overloading due to inertia of the weights. The strain rate should be reported.
12
BS : No specific recommendation for obtaining compression behaviour, but for
2
obtaining compressive strength alone, a rate of 1 N/mm /min is prescribed.
1
IS : No specific recommendation for obtaining compression behaviour, but for
obtaining compressive strength alone, a rate of 350–700 kN/min is pre-
scribed.
5
Sarangapani et al. : Constant displacement rate of 1.25 mm/min.
6
Gumaste et al. : Not mentioned.
9
Kaushik et al. : Not mentioned.

CURRENT SCIENCE, VOL. 109, NO. 11, 10 DECEMBER 2015 2097

RESEARCH ARTICLES
Table 1. Summary of results of tests on brick, mortar and masonry prism

Material

Parameter* Mean COV

Brick
Dimension
Length [6 measurements  50 bricks#=300#] 219.45 mm 0.007
Breadth [6 measurements  50 bricks#=300#] 102.41 mm 0.012
Height [6 measurements  50 bricks#=300#] 75.5 mm 0.026
Water absorption
@24 h immersion [15#] 12.21% 0.037
@15 min immersion [5#] 9.59% –
Compressive strength [15#] 5.31 N/mm2 0.19

Mortar
Consistency: Slump test (in accordance with RILEM 14 )
@ w/c ratio 1.0: slump = 203 mm
@ w/c ratio 0.9: slump = 79 mm
@ w/c ratio 0.85: slump = 68 mm (chosen in this experimental programme)
@ w/c ratio 0.70: slump = 0
Peak compressive stress [10#] 19.14 N/mm2 0.09
Strain corresponding to peak stress [10#] 2.664  10–3 0.12
Modulus of elasticity [10#] (chord modulus between 5% and 33% of peak compressive stress) 16695.29 N/mm2 0.07

Masonry
Peak compressive stress [15#] 2.82 N/mm2 0.15
Strain corresponding to 5% of the peak stress i [15#] 0.2213  10–3 0.263
Strain corresponding to 33% of the peak stress e [15#] 2.222  10–3 0.216
Strain corresponding to 50% of the peak stress [15#] 3.568  10–3 0.190
Strain corresponding to 70% of the peak stress [15#] 5.173  10–3 0.171
Strain corresponding to 90% of the peak stress [15#] 7.081  10–3 0.153
Strain corresponding to peak stress p [15#] 9.063  10–3 0.115
Strain corresponding to 90% of the peak stress in recession limb [11#] 10.499  10–3 0.179
Strain corresponding to 60% of the peak stress in recession limb u [8#] 13.599  10–3 0.264
Modulus of elasticity [15#] (chord modulus between 5% and 33% of peak compressive stress, in 401.72 N/mm2 0.127
accordance with the recommendations9,11,22 )
Ratio u /p [15#] (a parameter similar to ductility) 1.526 0.187

*Values presented in the square brackets indicate sample size.

corresponding to 24 h immersion. Hence, the bricks have

been pre-wetted by soaking under water for 15 min
before fabrication.
RILEM14 recommends a slump of 170 mm which is too
fluidic and generally not used in India. Hence, the slump
requirement for the mortar used in India has been as-
sessed, based on the slump test on mortar (results of the
slump test have been presented in Table 1). Mortar with
w/c ratio of 0.85 which gives rise to a slump of 68 mm,
and which has been adjudged by three different expert
masons as ‘the convenient workable mix and normally
used in the field’, has been used. Importantly, the water
content has been controlled by mixing water with small
batches of dry mix and the resulting mix has been used for
fabrication within 30 min. In order to determine the pa-
Figure 1. Fabrication of bricks masonry stack bonded prism. rameters required to describe the compression behaviour of
the mortar, cylinders of 150 mm diameter and 300 mm
24 h immersion represents the saturated condition. The length have been subjected to monotonically increasing
percentage of moisture corresponding to 15 min immer- displacement at the rate of 0.15 mm/min. Table 1 presents
sion is 9.59%, it is around 78% of the value (12.21%) the results, mean and COV of peak compressive stress.

2098 CURRENT SCIENCE, VOL. 109, NO. 11, 10 DECEMBER 2015

 RESEARCH ARTICLES

Figure 2. a, Curing of fabricated specimen under water up to 7 days. b, Curing of specimen under wet burlap during 8–28 days.

in this programme it has been decided that for the first 7

days the specimens shall be soaked under water (Figure
2 a) and from 8 to 28 days, the specimens shall be cured
under moist gunny bags (Figure 2 b). This will help mi-
nimize the variation of moisture content and its effect on
the structural behaviour of brick masonry. Just before
transporting the specimen from the curing yard for test-
ing, it is thoroughly watered so that the tests are carried
out on wet specimens.
For measuring strain in the masonry prism, a custom-
Figure 3. Set-up for carrying out test on compression behaviour of ized compressometer has been designed and fabricated.
brick masonry specimen. The compressometer is designed in such a way that the
gauge length can be adjusted, two LVDTs can be fitted
and 12 foot screws are used to fix the compressometer
with the prism specimen. In this test programme, 170 mm
gauge length has been used in such a way that two mortar
joints and approximately two bricks are covered (Figure
3). LVDTs having least count of 0.001 mm have been
used. For measuring compressive force, an external load
cell with a least count of 0.01 kN has been used.
The specimens have been placed between the platens
of the servo-controlled compression testing machine with
a 3 mm plywood at top and bottom. A rate of displace-
ment of 1.25 mm/min has been adopted by Sarangapani et
al.5. However, trial tests indicated that 1.25 mm/min was
too high to capture the post-peak behaviour properly,
Figure 4. Compressive stress–strain curves of masonry obtained from
while a value of 0.25 mm/min was found to be reason-
tests on 15 different specimens. able. Hence, the compression load has been applied at a
constant displacement rate of 0.25 mm/min throughout the
strain ranges reported here using the servo-controlled
In order to apply a uniform compression force on any compression testing machine. Table 1 and Figure 4 pre-
cross-section of the masonry prisms, a thin cap has been sent the results.
provided at the top and bottom of the specimen using a
rich mortar (1 : 2). The capping has been done during fa-
brication itself, so that the cap also attains sufficient Discussion of results
strength during the curing of the specimen. As far as cur-
ing of brick masonry prisms is concerned, it is common The compressive strength of the bricks used in this pro-
to cure the specimens under moist gunny bags; however, gramme is quite low and hence, as expected, the peak

CURRENT SCIENCE, VOL. 109, NO. 11, 10 DECEMBER 2015 2099

RESEARCH ARTICLES
Table 2. Deviation of the mean value of peak stress obtained from tests with estimations according to the literature

Reference Estimated compressive strength (N/mm2 ) Deviation of the result of tests* (%)

Dayaratnam18 : fm = 275 f b0.5 f j0.5 2.772 +1.73

Eurocode19 : fmk = 0.6(1  0.89  fb)0.65 ( fj )0.25 3.444 –18.11
(characteristic compressive strength)
Bennett et al.16 : fm = (3/10)fb 1.593 +77.02
MSJC 17 : fm = 1.0(400 + 0.25fb) (in psi) 4.085 –30.97
Gumaste et al.6 : fm = 0.317 f b0.866 f j0.134 1.998 +41.14
Kaushik et al.9 : fm = 0.63 f b0.49 f j0.32 3.671 –23.18
Freeda Christy et al.10 : fm = 0.35 f b0.65 f j0.25 2.167 +30.13

*Mean value of peak compressive stress as obtained from the tests: 2.82 N/mm2 .

value of strain corresponding to peak stress as obtained is

comparable with that presented by Gumaste et al.6
(prisms fabricated using table-moulded bricks). The mean
value of modulus of elasticity is approximately 150 times
that of the mean peak stress of brick masonry, which is
much less than the range presented by Kaushik et al.9 (i.e.
250–1100 times the compressive strength of masonry).
These observations indicate that the masonry fabricated
using soft and weak brick possesses larger deformability
than those fabricated using stiff and strong bricks.
The measured mean peak stress of brick masonry
showed a COV of 0.15. Measured mean peak stress of
brick masonry is compared with that estimated using the
expressions given in the literature (Table 2). Bennet et
al.16 do not take into account the compressive strength of
the mortar. Similarly MSJC17 is also less sensitive to the
compressive strength of the mortar. Hence these may not
be advisable for countries like India, where wide varieties
of mortar are used. From the results presented in Table 2,
it is noted that the predictions made by Dayaratnam18 are
close to the experimental value. He gives equal weightage
for compressive strengths of mortar and brick18 . Euro-
Figure 5. Mean normalized stress–strain curve of brick masonry code19 , Gumaste et al.6 and Kaushik et al.9 give higher
obtained from the tests. weightage to compressive strength of brick. While for the
expression of Gumaste et al.6 , the summation of the con-
stants  and  is equal to ‘1’ (see eq. 1)), which makes it
compressive stress of the masonry is also quite low. It is dimensionally balanced, the same is not true for the ex-
noted from the literature6,9,15 that the peak compressive pressions of Eurocode19 and Kaushik et al.9. Hence, it is
stress of the masonry is lower than that of the bricks proposed that the constants  and  may have to be ar-
used. Also it is worth noting that though strain was not rived based on the thickness ratios of brick and mortar
recorded when the bricks were subjected to compression, joint. It will not only fulfil the dimensional balancing, but
visual observations indicated that the bricks undergo also to some extent the mechanics of the behaviour. Never-
large deformations. The relative movement of the heads theless, this needs further experimental studies on differ-
of the compression testing machine around the peak ent brick and mortar combinations, including relative
stress was approximately 9–14 mm. These observations strength and stiffness between brick and mortar.
indicate that the bricks used are ‘soft’ and also ‘weak’. The stress–strain curves for brick masonry shown in
The brick masonry prisms used have undergone high Figure 4 were normalized with respect to peak stress and
levels of strain at peak stress (mean value of 0.0093) strain corresponding to peak stress. The curve corre-
compared to those presented by Kaushik et al.9 (0.0036 sponding to mean and the variation of peak stress and
for brick masonry prisms fabricated using cement mortar strain are presented in Figure 5. There is variation of
1 : 6 and 0.0029 for brick masonry prisms fabricated us- COV in the normalized strain at different levels of the
ing cement mortar 1 : 3). On the other hand, the mean normalized stress; however, for engineering purpose, an
2100 CURRENT SCIENCE, VOL. 109, NO. 11, 10 DECEMBER 2015
 RESEARCH ARTICLES

Figure 6. Comparison of normalized stress–strain curves. Desayi and Krishnan20 :

RMSD = 0.6240; Priestly and Elder 22 : RMSD = 0.1932; Kaushik et al.9 – parabolic:
RMSD = 0.0918; Kaushik et al.9 – trilinear: RMSD = 0.0499; Proposed: RMSD = 0.02632.

average value of 0.2 may be used throughout. The pro- mechanics-based inter-relationship between compressive
posed value of 0.2 for the COV of normalized strain at strengths of brick, mortar, nature of interface and ma-
different levels of the normalized stress could be suitable sonry. In order to achieve this, there is a need to carry out
while soft and weak bricks are used. While using strong similar experimental programmes using different brick
and stiff bricks, the COV of normalized strain at different and mortar combinations. While considering different
levels of the normalized stress needs to be assessed from combinations of brick and mortar, their relative strength
similar experimental studies. and stiffness need to be taken into account. Towards
By comparing the root-mean-square deviation (RMSD) these aspects, studies are being continued at CSIR-SERC.
values of the mean normalized stress–strain curve as ob-
tained from the tests with those estimated using equations
available in the literature9,20,21, it is found that the curve 1. IS:1905–1987 (reaffirmed 2002), Code of practice for structural
use of unreinforced brick masonry – guidelines. Bureau of Indian
obtained from the test is comparatively closer to that pro- Standards (BIS), New Delhi, 2002.
posed by Kaushik et al.9 (Figure 6). It is to be noted that 2. Dymiotis, C. and Gutlederer, B. M., Allowing for uncertainties in
the normalized form of the stress–strain relation for con- the modelling of masonry compressive strength. Constr. Build.
crete given by Desayi and Krishnan 20, is also considered Mater., 2002, 16(8), 443–452.
because of its simplicity. Between the two equations pro- 3. Pande, G. N., Kralj, B. and Middleton, J., Analysis of the com-
pressive strength of masonry given by the equation fk =
posed by Kaushik et al.9, the trilinear equation fits the K(fb)  ( fm) . Struct. Eng., 1994, 1, 7–12.
experimental curve better than the parabolic curve. How- 4. Haller, P., Die technische Eigenschaften von Backstein, Schweize-
ever, it is observed that for the weak and soft bricks the rische Bauzeitung, 1958.
initial linear part has to be more slanting and hence, a 5. Sarangapani, G., Venkatarama Reddy, B. V. and Jagadish, K. S.,
normalized strain level of 0.6 corresponding to 75% of Structural characteristics of bricks, mortar and masonry. J. Struct.
Eng. (CSIR-SERC), 2002, 29, 101–107.
peak stress as shown in Figure 6 has been proposed for 6. Gumaste, K. S., Nanjunda Rao, K. S., Venkatarama Reddy, B. V.
better agreement with the obtained experimental data. and Jagadish, K. S., Strength and elasticity of brick masonry
Also, this needs concurrence from similar experimental prisms and wallettes under compression. Mater. Struct., 2007,
studies. 40(2), 241–253.
7. Naraine, K. and Sinha, S. N., Loading and unloading stress–strain
curves for brick masonry. J. Struct. Eng. (ASCE), 1989, 115(10),
Conclusion 7631–7644.
8. AlShebani, M. M. and Sinha, S. N., Stress–strain characteristics of
brick masonry under uniaxial cyclic loading. J. Struct. Eng.
This article presents procedure for obtaining the com- (ASCE), 1999, 125(6), 600–604.
pressive stress–strain relationship of masonry. A trilinear 9. Kaushik, H. B., Rai, D. C. and Jain, S. K., Uniaxial compressive
stress–strain relation has been proposed for typical brick stress–strain model for clay brick masonry. Curr. Sci., 2007,
masonry used in South India. It is important to establish 92(4), 497–501.

CURRENT SCIENCE, VOL. 109, NO. 11, 10 DECEMBER 2015 2101

RESEARCH ARTICLES
10. Freeda Christy, C., Tensing, D. and Mercy Shanthi, R., Experi- 21. Priestley, M. J. N. and Elder, D. M., Stress–strain curves for un-
mental study on axial compressive strength and elastic modulus of confined and confined concrete masonry. Am. Concr. Inst. J.,
the clay and fly ash brick masonry. J. Civ. Eng. Constr. Technol., 1983, 80(3), 192–201.
2013, 4(4), 134–141. 22. Drysdale, R., Hamid, A. and Baker, L., Masonry Structures:
11. ASTM C1314-12, Standard test method for compressive strength Behaviour and Design, The Masonry Society, USA, 1999, 3rd
of masonry prisms. American Society for Testing and Materials, edn.
West Conshohocken, USA, 2012. 23. ASTM E 122-09, Standard practice for calculating sample size to
12. BS 5628-1:2005, Code of practice for the use of masonry – estimate, with specified precision, the average for a characteristic
Structural use of unreinforced masonry. British Standards Insti- of a lot or process. West Conshohocken, USA, 2009.
tute, London, UK, 2005. 24. BS 5628-3:2005, Code of practice for the use of masonry – Mate-
13. Groot, C. and Larbi, J., The influence of water flow (reversal) on rials and components, design and workmanship. British Standards
bond strength development in young masonry. Heron, 1999, 44(2), Institute, London, UK, 2005.
63–78. 25. IS:2250-1981 (reaffirmed 2002), Code of practice for preparation
14. RILEM TC 127-MS B.4, Measurement of the shear strength index and use of masonry mortars. BIS, New Delhi, 2002.
for unit-mortar junction. Mater. Struct., 1996, 29(8), 459–475. 26. ASTM C 1552-14, Practice for capping concrete masonry units,
15. Hendry, A. W., Structural Masonry, Macmillan Education Ltd, related units and masonry prisms for compression testing. West
London, UK, 1990. Conshohocken, USA, 2014.
16. Bennett, R. M., Boyd, K. A. and Flanagan, R. D., Compressive 27. ASTM E 111-04, Practice for capping concrete masonry units, re-
properties of structural clay tile prisms. J. Struct. Eng. (ASCE), lated units and masonry prisms for compression testing. West
1997, 123(7), 920–926. Conshohocken, USA, 2004.
17. MSJC 2002, ACI 530-02/ASCE 5-02/TMS 402-022002 – Building
code requirements for masonry structures. Masonry Standards
Joint Committee, USA, 2002.
18. Dayaratnam, P., Brick and Reinforced Brick Structures, Oxford ACKNOWLEDGEMENT. We thank the Director, CSIR-SERC,
and IBH, India, 1987. Chennai for permission to publish this manuscript.
19. Eurocode 6 (ENV 1996-1-1), Design of masonry structures. Part
1-1: General rules for buildings – reinforced and unreinforced ma-
sonry. European Committee of Standardization (CEN), Brussels, Received 30 January 2015; revised accepted 30 August 2015
1996.
20. Desayi, P. and Krishnan, S., Equation for the stress–strain curve of
concrete. J. Am. Concr. Inst., 1964, 61(3), 345–350. doi: 10.18520/v109/i11/2094-2102

2102 CURRENT SCIENCE, VOL. 109, NO. 11, 10 DECEMBER 2015

View publication stats