You are on page 1of 11

Downloaded from ascelibrary.org by "Indian Institute of Technology (Indian School of Mines), Dhanbad" on 06/05/22. Copyright ASCE.

For personal use only; all rights reserved.

STRESS-STRAIN CURVES FOR B R I C K MASONRY IN


BIAXIAL COMPRESSION
By Krishna Naraine 1 and Sachchidanand Sinha2

ABSTRACT: A generalized approach is proposed to determine the envelope, com-


mon-point, and stability-point curves for brick masonry in cyclic, biaxial compres-
sion. The approach uses a family of interaction curves to predict the peak stresses
of the envelope, common-point, and stability-point curves at different principal
stress ratios. The corresponding strains at the peak stresses are determined using
an empirical relationship involving the principal stress ratio and the strain at the
peak stress of the envelope curve. A general stress-strain equation is then used to
determine the envelope, common-point, and stability-point curves for each prin-
cipal stress ratio. The method is developed using the test results of 45 and 18 half-
scale brick masonry models tested under biaxial and uniaxial cyclic compressive
loading, respectively. The test specimens are 360 mm x 360 mm x 115 mm
constructed with half-scale clay bricks 115 mm x 55 mm x 35 mm and with 6-
mm thick bed joint mortar. The proposed generalized approach is simple and
accurate.

INTRODUCTION

Many masonry elements are subjected to biaxial states of stress produced


by in-plane loading. Most of the in-plane loading patterns are cyclic, and
this results in cyclic, biaxial stress states. Only recently there has been
attempts to study the material properties of brick masonry subject to biaxial
stress states. These studies have concentrated more on obtaining a failure
criterion rather than on studying the deformation characteristics of the ma-
terial. Furthermore, these studies have been restricted to tests under mon-
otonic loading conditions.
Bernardini et al. (1982) and Ganz and Thurlimann (1983) developed
failure surfaces for hollow brick masonry. Samarasinghe (1980) experimen-
tally established a three-dimensional failure surface for one-sixth-scale brick
masonry subjected to biaxial tension-compression. Page (1980) proposed a
failure surface for brick masonry subjected to biaxial tension-tension using
a nonlinear finite element model accounting for joint failure only. Page
(1981, 1982, 1983) also reported experimentally derived failure surfaces for
half-scale brick masonry subjected to compression-compression and tension-
compression stress states. Dhanasekar et al. (1985a, 1986b) proposed a
failure surface in terms of the stresses in a reference system aligned with
the bed joint directions. This formulation yielded a closed surface that could
be idealized as three intersecting elliptic cones. Hamid and Drysdale (1982)
proposed a failure criterion to account for the anisotropic nature of brick
masonry under in-plane biaxial compression loading.
Recently, an investigation of the behavior of brick masonry under cyclic,
uniaxial and biaxial compressive loading has been presented (Naraine and
Sinha 1988,1989,1991). It has been established that the uniaxial and biaxial
cyclic stress-strain history possesses a locus of common points where the

2
•Res. Scholar, Civ. Engrg. Dept., Indian Inst, of Tech., New Delhi 110 016, India.
Prof., Civ. Engrg. Dept., Indian Inst, of Tech., New Delhi 110 016.
Note. Discussion open until November 1, 1992. To extend the closing date one
month, a written request must be filed with the ASCE Manager of Journals. The
manuscript for this paper was submitted for review and possible publication on March
15, 1991. This paper is part of the Journal of Structural Engineering, Vol. 118, No.
6, June, 1992. ©ASCE, ISSN 0733-9445/92/0006-1451/$1.00 + $.15 per page. Paper
No. 1584.
1451

J. Struct. Eng., 1992, 118(6): 1451-1461


Downloaded from ascelibrary.org by "Indian Institute of Technology (Indian School of Mines), Dhanbad" on 06/05/22. Copyright ASCE. For personal use only; all rights reserved.

reloading part of any cycle crosses the unloading part of the previous cycle
(Fig. 1). Stresses above the common point produce additional strain, while
stresses below this point will result in the stress-strain path looping where
the intersection point of the reloading curve and unloading curve descends
and stabilizes at a lower bound, called the stability point (Fig. 2). The locus
of stability points provides a rational design tool in defining the permissible
stress levels for brick masonry under cyclic loading.
In this paper, a generalized approach is proposed to determine the en-
velope, common-point, and stability-point curves on the absolute stress-
strain coordinate system. This generalized approach uses a family of inter-
action curves to predict the peak stresses of the envelope, common-point,
and stability-point curves at different principal stress ratios. The corre-

0.001 0.002 0.003 0.004 0.005 0.006 0.007

FIG. 1. Locus of Common Points

0.001 0.002 0.003 0.004 0.005 0.006 0.007

FIG. 2. Locus of Stability Points


1452

J. Struct. Eng., 1992, 118(6): 1451-1461


Downloaded from ascelibrary.org by "Indian Institute of Technology (Indian School of Mines), Dhanbad" on 06/05/22. Copyright ASCE. For personal use only; all rights reserved.

sponding strains at the peak stresses are determined using an empirical


relationship involving the principal stress ratio and the strain at the peak
stress of the envelope curve. With these stress-strain prameters known, a
general stress-strain equation can be used to compute the envelope, com-
mon-point, and stability-point curves for each principal stress ratio.

EXPERIMENTAL STUDY

The test results used in this paper are based on an experimental study
on the behavior of brick masonry under cyclic, uniaxial and biaxial com-
pressive loading (Naraine and Sinha 1988, 1991). The test specimens were
360 mm x 360 mm x 115 mm constructed with half-scale clay bricks 115
mm x 55 mm x 35 mm and with 6-mm thick bed joint mortar. Fig. 3 shows
the loading arrangement for testing the brick masonry models under cyclic,
biaxial loading. The mean strength of the bricks was 13.6 N/mm2, with a
standard deviation of 1.6 N/mm2 based on a sample of 54 half-scale bricks.
A 1:5 mortar (cement:sand-by-volume) mix was used, and the mean com-
pressive strength of 70 mm mortar cubes was 6.6 N/mm2, with a standard
deviation of 1.1 N/mm2 based on a sample of 54 mortar cubes.
Biaxial compression tests on 45 brick masonry specimens were conducted
for five values of the principal stress ratio, fjfp, equal to 0.2, 0.6,1.0, 1.67,
and 5.0. The angles between the principal stresses and bed joint were re-
stricted to 90° and 0°. The quantities/,, and/, represent the principal stresses
applied normal and parallel to the bed joint, respectively. Three tests were
conducted for each ratio of principal stress. In the first, the load was in-
creased steadily to failure and a critical envelope stress-strain curve was
established. For any given principal stress ratio, either the/,, versus e„ curve

FIG. 3. Loading Arrangement

1453

J. Struct. Eng., 1992, 118(6): 1451-1461


Downloaded from ascelibrary.org by "Indian Institute of Technology (Indian School of Mines), Dhanbad" on 06/05/22. Copyright ASCE. For personal use only; all rights reserved.

or the fp versus ep curve defined the critical envelope curve, depending on


which one of these stress-strain curves governed the failure of the brick
masonry specimen. The parameters en and e represent the strains normal
and parallel to the bed joint, respectively. Tne second and third tests were
conducted to establish the loci of common points and stability points. The
test results have been reported (Naraine and Sinha 1991). The test results
for the uniaxial tests (f„/fp = 0.0 and °°) on half-scale models have been
reported earlier (Naraine and Sinha 1988).
The failure of the brick masonry assemblage is quantified by the maximum
compressive strain obtained from the measured, principal orthogonal strains.
These critical principal strains and corresponding stresses are used to obtain
the critical stress-strain envelope curves. For principal stress ratios of 1.0,
1.67, and 5.0, the critical principal strain e„ and corresponding stress /„
govern the critical stress-strain envelope curve. For principal stress ratios
of 0.2 and 0.6, the critical principal strain ep and corresponding stress fp
govern it.

FAMILY OF INTERACTION CURVES

The general form given in (1) can be used to predict the peak stresses of
the envelope, common-point, and stability-point curves at each principal
stress ratio. This general form is given as follows:

c
(f-?) ! + <1-c)(? + ?) + c ( ^ ) ^ (1)
where

<*« = T- •-..(2a)

«P = f- (26)
J mo
where the values /moc and fmo = the average uniaxial compressive strengths
normal and parallel to the bed joint, respectively; and the parameter C =
a constant equal to 1.6. The variation of the factor (3 in (1) accounts for the
change in the stress at the maxima of the envelope, common-point, and
stability-point curves for a given principal stress ratio. A statistical method
for determining (3 for these curves based on the experimental data has been
reported (Naraine and Sinha 1991). Based on these statistically computed
values of (3 for the curves from the uniaxial and biaxial tests (Table 1), the
average values of (3 equal to 1.0, 0.85, and 0.67 are used to define the
interaction curves of the peak stresses of the envelope, common-point, and
stability-point curves, respectively, for brick masonry in the compression-
compression stress state region. This is a reasonable approximation, because
the maximum variations of the computed values of |3 for the common-point
and stability-point curves are approximately 6% and 12%, respectively. The
values of (3 remains 1.0 for the envelope curves.
Fig. 4 shows the family of interaction curves presented on the u„ versus
(Tp coordinate system. The average experimental failure (peak envelope)
stresses and peak stresses of the empirically determined common-point and
stability-point curves (Naraine and Sinha 1991) are also shown on the a„
versus crp coordinate system for each principal stress ratio. It is observed
1454

J. Struct. Eng., 1992, 118(6): 1451-1461


Downloaded from ascelibrary.org by "Indian Institute of Technology (Indian School of Mines), Dhanbad" on 06/05/22. Copyright ASCE. For personal use only; all rights reserved.

TABLE 1. values of p and a

Stress ratio Envelope Curve Common-Point Curve Stability-Point Curve


fJfP P a- P a P a
(D (2) (3) (4) (5) (6) (7)
0 1.0 1.0 0.87 0.90 0.68 0.75
0.2 1.0 1.0 0.83 0.90 0.65 0.79
0.6 1.0 1.0 0.85 0.87 0.69 0.73
1.0 1.0 1.0 0.84 0.89 0.64 0.76
1.67 1.0 1.0 0.83 0.87 0.66 0.73
5.0 1.0 1.0 0.83 0.89 0.68 0.79
GO 1.0 1.0 0.88 0.89 0.71 0.74

that the computed interaction curves compare well with the points from the
test results. However, at low levels of restraining stress (fjfp = 5.0, 0.2)
the interaction curves generally underestimate the values of the peak stresses.
Given the uniaxial strengths of brick masonry perpendicular and parallel
to the bed joint, the interaction curves presented in Fig. 1 can be used to
compute the absolute stresses at the peaks of the envelope, common-point,
or stability-point curves for any principal stress ratio in the compression-
compression stress state region.

DETERMINATION OF STRAINS AT PEAK STRESSES

The strain at the peak of the envelope curve on the/„ versus e„ coordinate
system is represented by e„,„. Similarly, the strain at the peak of the envelope
curve on the fp versus ep coordinate system is represented by emp. It has
been observed that there is a general tendency for the e„„, values to decrease
as the principal stress ratio f„/fp decreases from °° to 1.0. Similarly, the value
of e decreases asf„/fp increases from 0 to 0.60 (Naraine and Sinha 1991).
Fig. 5(a) shows a plot of the stress ratio fp/f„ versus the average values of
e„m. Observe that the values of em„ decrease approximately linearly as/ p //„
increases from 0 to 1.0. Fig. 5(b) shows a plot of the stress ratio f„/fp versus
the average values of emp, from which it can be observed that the values of
emp decreases approximately linearly as/„// p increases from 0 to 0.6. Hence,
linear expressions can be used to relate/„//„ to emn and fjfp to emp. For the
critical stress-strain curve normal to the bed joint

f
f = - 4 3 5 ( 0 + 2.87 (3a)
In
For the critical stress-strain curve parallel to the bed joint

f
f = - 6 2 5 ( 0 + 3 - 62 (3fo)
Ip
On the /„ versus e„ coordinate system, the strain at the peak of the
common-point curve and the strain at the peak of the stability-point curve
can be expressed as aemn, where the variation of a accounts for the change
of the strain at the peak of the common-point and stability-point curves.
Similarly, on the fp versus ep coordinate system, the strains at the peaks of
the common-point and stability-point curves can be expressed as aemp. A
statistical method for determining a for the envelope, common-point, and
1455

J. Struct. Eng., 1992, 118(6): 1451-1461


Downloaded from ascelibrary.org by "Indian Institute of Technology (Indian School of Mines), Dhanbad" on 06/05/22. Copyright ASCE. For personal use only; all rights reserved.

Peak slresses o1
x Envelope
A Common poinl
_ © Slabilily poinl

-y-v \

J
0.4
°2/ /iA /
1—/ I /
0.8
I
1.2
L J L
2.0

FIG. 4. Family of Interaction Curves for Peak Stresses of Envelope, Common-


Point, and Stability-Point Curves

1.0 r

IS. °' 6
fn In.
0.4

0.2

_1 I I I I l_i_l I
0.002 0.004 0.006 0.008 .006 0.0OS

FIG. 5. Plot of Stress Ratio versus Strain at Peak Stress: (a) fplfn versus emn; (b)
f„lfp versus emp

stability-point curves based on the experimental data has been reported


(Naraine and Sinha 1991). Based on the computed values of a for the
common-point curves from the uniaxial and biaxial tests (Table 1), an av-
erage value of a is used to define the strain at the peak of the common-
point curve for any given principal stress ratio. This approximation is rea-
sonable, because the maximum variation in the computed values of a for
the common-point curves is approximately 4%. This average value of a is
1456

J. Struct. Eng., 1992, 118(6): 1451-1461


Downloaded from ascelibrary.org by "Indian Institute of Technology (Indian School of Mines), Dhanbad" on 06/05/22. Copyright ASCE. For personal use only; all rights reserved.

0.89. Similarly, an average value of a is used to define the strain at the peak
of the stability-point curve. This approximation is also reasonable because
the maximum variation in the computed values of a for the stability-point
curves is approximately 10%. This average value of a is 0.75.

GENERALIZED STRESS-STRAIN CURVES

To compute the envelope, common-point, or stability-point curves on the


absolute stress-strain coordinate system, the following general stress-strain
equation can be used:

(4)
'-«"«£)•*('-«£))
where / and e = the absolute values of stress and strain, respectively; fm
= the failure (peak) stress; and e„, = the strain when the peak envelope
stress is attained. When failure is governed by strains normal to the bed
joint, stresses and strains in (4) are normal to the bed joint. Similarly, when
failure is governed by strains parallel to the bed joint, stresses and strains
in (4) are parallel to the bed joint. The values (}/m and aem in (4) represent
the peak stress and the strain at the peak stress of the envelope, common-
point, or stability-point curve.
Given the uniaxial strengths of brick masonry normal and parallel to the
bed joint, the failure interaction curve can be used to determine/ m for any
given principal stress ratio. The value of em can be evaluated using (3). In
the computation of the envelope curves, the values of a and (3 are both
unity. Thus the envelope curves can be computed on the absolute stress-
strain coordinate system. The computed envelope curve for/„// p = 1.0 is
shown in Fig. 6, together with the empirical envelope curve developed by
Naraine and Sinha (1991) using the test data.
The peak of the common-point and stability-point curves is given by the
9.0p

6.0-

7.0 -

6.0 -

tn 5.0-

^ Stability
4.0 - point curve

3.0 -

B Empirical curve
/ Generalized approach
10
/ I e .
: i i i—\te-i—i—i
0 0.001 0.002 0.003 0.004 0.005 0.006 0.007

en

FIG. 6. Comparison of Empirical and Generalized Stress-Strain Curves f„/fp = 1.0

1457

J. Struct. Eng., 1992, 118(6): 1451-1461


Downloaded from ascelibrary.org by "Indian Institute of Technology (Indian School of Mines), Dhanbad" on 06/05/22. Copyright ASCE. For personal use only; all rights reserved.

J I I I I I I I I I
10.0
enxlO3
la) Envelope curves

10.0

8.0 10.0
en»10J
Ic) Stability point curves

FIG. 7. Stress-Strain Curves on /„ versus e„ Coordinate System: (a) Envelope


Curves; (b) Common-Point Curves; (c) Stability-Point Curves

1458

J. Struct. Eng., 1992, 118(6): 1451-1461


Downloaded from ascelibrary.org by "Indian Institute of Technology (Indian School of Mines), Dhanbad" on 06/05/22. Copyright ASCE. For personal use only; all rights reserved.

ep»io
( b | Common point curves

- js-*~" """"-"---^

•"J

y
• /

2.0
~/ o.o
/* 0.2
J 0.6
r . i i i i i i i i i
2.0 4.0 _ 6.0 8.0 10,0
e„»10"r3
I c ) S t a b i l i t y point curves

FIG. 8. Stress-Strain Curves on /,, versus ep Coordinate System: (a) Envelope


Curves; (b) Common-Point Curves; (c) Stability-Point Curves

1459

J. Struct. Eng., 1992, 118(6): 1451-1461


Downloaded from ascelibrary.org by "Indian Institute of Technology (Indian School of Mines), Dhanbad" on 06/05/22. Copyright ASCE. For personal use only; all rights reserved.

coordinates aem and (3/m. The coordinates of the peak of the common-point
curve can be determined by using (3 = 0.85 and a = 0.89. The coordinates
of the peak of the stability-point curve can be determined by using p =
0.67 and a = 0.75. Then, using (4), the common-point and stability-point
curves on the absolute stress-strain coordinate system are determine. Fig.
6 also shows these computed curves for fjfp = 1.0, together with the
empirical common-point and stability-point curves developed by Naraine
and Sinha (1991). A reasonable comparison is observed between the en-
velope, common-point, and stability-point curves from the generalized ap-
proach and the empirical curves. The maximum variations between the
presently computed envelope, common-point, and stability-point curves and
the corresponding empirical curves using the test data (Naraine and Sinha
1991) are 10%, 5%, and 5%, respectively, for the values of principal stress
ratios considered.
Based on the generalized approach, the envelope, common-point, and
stability-point curves on the absolute stress-strain coordinate system are
shown in Fig. 7 for principal stress ratios f„/fp = 1.0, 1.67, 5.0, and oo,
where failure is governed by the /„ versus en curve. Similarly, Fig. 8 shows
these curves for principal stress ratios f„/fp = 0.0, 0.2, and 0.6, where failure
is governed by the fp versus ep coordinate system.

CONCLUSION

A generalized approach proposed to determine the envelope, common-


point, and stability-point curves on the absolute stress-strain coordinate
system is simple and accurate. It uses a family of interaction curves to predict
the peak stresses of the envelope, common-point, and stability-point curves
at different principal stress ratios. The corresponding strains at the peak
stresses are determined using an empirical relationship involving the prin-
cipal stress ratio and strain at the peak stress of the envelope curve. With
these stress-strain parameters known, a general stress-strain equation is used
to compute the envelope, common-point, and stability-point curves for each
principal stress ratio. The computed curves using this generalized approach
compare well with the empirical curves obtained using the test data.

APPENDIX I. REFERENCES

Bernardini, A., Modena, C., and Vescovi, U. (1982). "An anisotropic biaxial failure
criterion for hollow clay brick masonry." Int. J, Masonry Constr., 2(4), 165-171.
Dhanasekar, M., Page, A. W., and Kleeman, P. W. (1985a). "The behavior of brick
masonry under biaxial stresses with particular reference to infill frames." Seventh
Int. Brick Masonry Conf., Melbourne, 2, 815-824.
Dhanasekar, M., Kleeman, P. W., and Page, A. W. (1985b). "The failure of brick
masonry under biaxial stresses." Proc, Inst, of Civ. Engrs., Pt. 2, London, U.K.,
79, 295-313.
Ganz, H. R., and Thurlimann, B. (1983). "Strength of brick walls under normal
force and shear." Proc. 8th Int. Symp. on Load Bearing Brickwork, London, U.K.
Hamid, A. A., and Drysdale, R. G. (1982). "Proposed failure criteria for brick
masonry under combined stresses." Proc, 2nd North Am. Masonry Conf., College
Park, Md., 9.2-9.11
Naraine, K., and Sinha, S. N. (1988). "Test of brick masonry models under cyclic
compressive loading." Proc. 8th Int. Brick Masonry Conf, Dublin, Ireland, 1,
395-405.
Naraine, K., and Sinha, S. N. (1989). "Behavior of brick masonry under cyclic
compressive loading." /. Struct. Engrg., ASCE, 115(6), 1432-1445.
1460

J. Struct. Eng., 1992, 118(6): 1451-1461


Downloaded from ascelibrary.org by "Indian Institute of Technology (Indian School of Mines), Dhanbad" on 06/05/22. Copyright ASCE. For personal use only; all rights reserved.

Naraine, K., and Sinha, S. N. (1991). "Cyclic biaxial behavior of brick masonry."
/. Struct. Engrg., ASCE, 15(5), 1336-1355.
Page, A. W. (1980). "A biaxial failure criterion for brick masonry in the tension-
tension range." Int. J. Masonry Constr., 1(1), 26-29.
Page, A. W. (1981). "The biaxial compressive strength of brick masonry." Proc,
Inst, of Civ. Engrs., Pt. 2, London, U.K., 71, 893-906.
Page, A. W. (1982). "An experimental investigation of the biaxial strength of brick
masonry." Proc. of the Sixth Int. Brick Masonry Conf, Rome, Italy, 3-15.
Page, A. W. (1983). "The strength of brick masonry under biaxial compression-
tension." Int. J. Masonry Constr., 3(1), 26-31.
Samarasinghe, W. (1980). "The in-plane failure of brickwork." PhD thesis, Univ.
of Edinburgh, Edinburgh, U.K.

APPENDIX II. NOTATION

The following symbols are used in this paper:

em„ =envelope strain normal to bed joint when peak stress is attained;
emp =envelope strain parallel to bed joint when peak stress is attained;
en, ep =strain normal and parallel to bed joint, respectively;
fm-^ fmo -average uniaxial compressive strength normal and parallel to
bed joint, respectively;
/„, fp = stress normal and parallel to bed joint, respectively;
a = strain ratio at peak of envelope curve, common-point curve, or
stability-point curve; and
(3 = stress ratio at peak of envelope curve, common-point curve, or
stability-point curve.

1461

J. Struct. Eng., 1992, 118(6): 1451-1461

You might also like