American Journal of Engineering Research (AJER)

2017
American Journal of Engineering Research (AJER)
e-ISSN: 2320-0847 p-ISSN : 2320-0936
Volume-6, Issue-1, pp-328-336
www.ajer.org

Research Paper

Open Access

Code Comparison for the Assessment of Masonry Capacity

Mir Abdul Kuddus1*, Pere Roca Fabregat2
1

Assistant Professor, Department of Civil Engineering, Khulna University of Engineering &Technology,

Khulna-9203, Bangladesh
2
Professor, Department of Civil and Construction Engineering, Technical University of Catalonia, Barcelona08034, Spain

ABSTRACT: Masonry load bearing wall subjected to vertical concentric and eccentric loading may collapse
through instability. In this Paper the buckling behavior of masonry load bearing wall of different slenderness
ratio were investigated via testing a series of scale masonry wall subjected to concentric and eccentric vertical
loading. A total of thirty six masonry walls were tested in the Laboratory of Technical University of Catalonia
(UPC), which was the basis of the numerical study. A numerical finite element model was developed based on
the simplified micro model approach and calibrated by using those results found from experimental study. The
influence of nonlinear behavior of interface element, slenderness ratio and various end conditions have been
investigated together with the effect of different end eccentricity of vertical load. However, a series of analytical
studies were conducted in order to access the accuracy and performance of formulations provided by
EUROCODE 6 and ACI-530 for vertical capacity of masonry wall.
Keywords: Masonry load bearing wall, buckling failure, eccentric load, slenderness ratio, micro-modeling.

I.

INTRODUCTION

The buckling response of masonry walls depends on parameters such as the cross-sectional area, the
material properties, the slenderness and the effective eccentricities of loads at each end of the element. In
masonry design standards, these factors are treated in different ways. For example, the north-American ACI-530
code and the European EUROCODE-6 (EC-6) show significant differences in the analysis of brickwork walls
under eccentric axial load. Whereas ACI-530 code analyzes the buckling failure and the cross-section material
failure separately, EC-6 deals with both failure modes in a single analytical formulation.
The literature on the subject shows large number of studies carried out on axially loaded walls with
varying slenderness ratio. Among the first, Chapman and Slatford (1957) obtained closed form solutions for the
load deformation behaviour of brittle elastic wall by assuming that masonry material has no tensile strength and
that cracking occurs whenever a tensile stress would develop. After that Yokels (1971) results on the buckling
of walls made of no-tension material are well known. De Falcos proposal (2002) on the stability of columns
using an elastic-plastic material model stands among the most recently presented analytical approaches. More
recently, Mura (2008) has utilized a parabolic stress-strain relationship to describe the behaviour of the
brickwork under compression loads. Shalin (1978) reviewed the results of analysis carried out by a number of
authors and presented experimental evidence in support of the calculations.
Further work was carried out by Sawko and Towler (1982) who proposed a numerical procedure for
calculating the failure load of a no-tension material wall. An analytical solution has been carried out by Romano
et al. (1993), considering no tension bearing masonry with a monomial stressstrain relationship in compression.
Parland et al. (1982) proposed a method for determining buckling failure load of a slender wall, taking into
account the effect of tension stress field which exists between the cracked joints. However, the linear elastic
materials were used in this analysis.
The micro-modelling strategy has been adopted in the present research in order to carry out the needed
numerical simulations. The prediction of the ultimate capacity of walls obtained by means of micro-modelling
approach compared with experimental and analytical results. Moreover, results obtained in the parametric
studies by considering different end support condition and effect of support codition on failure loads of walls
were investigated. Conclusions are drawn on the relative importance of end support condition, non-linear
geometrical and material properties with different end eccentricity.

www.ajer.org

Page 328

American Journal of Engineering Research (AJER)

II.

2017

LABORATORY INVESTIGATION

This chapter deals with the description of the testing program carried out in Structural Technology
Laboratory of the Technical University of Catalonia which was the basis of numerical simulation of this
research. The following is the characterization of materials used for the manufacture of the walls and a
description of the manufacturing process, with emphasis on those details because it is a 1:4 scale study. Thirty
six walls were tested with slenderness ratio (calculated as h / t) 6, 12, 18 and 25 and values of eccentricity of
load was e = 0, e =t/6 and e = t/3. A summary of the number and type of the test specimens and test program is
shown in Table 1.
Table 1: Summary of number and type of test specimens.
Wall series
W-0-6
W-0-12
W-0-18
W-0-25
W-1/6-6
W-1/6-12
W-1/6-18
W-1/6-25
W-1/3-6
W-1/3-12
W-1/3-18
W-1/3-25

Slenderness ratio
6
12
18
25
6
12
18
25
6
12
18
25

Height (cm)
21
42
63
87.5
21
42
63
87.5
21
42
63
87.5

Eccentricity of load
0
0
0
0
t/6
t/6
t/6
t/6
t/3
t/3
t/3
t/3

Support condition
Hinge-hinge

Observation
Slenderness ratio
calculated as H/t

Hinge-hinge

Slenderness ratio
calculated as H/t

Hinge-hinge

Slenderness ratio
calculated as H/t

The walls were constructed as single leaf with scale of 1:4. The width was of 297 mm and thickness of
35 mm. In order to introduce different slenderness ratio, walls were built with heights of 210 mm, 420 mm, 630
mm and 875 mm. The thickness of the vertical and horizontal joints was approximately 2.5 mm. Figure 1 shows
the layout of the tested wall. One fourth scale bricks were used for the construction of walls. The dimensions of
the bricks (Length x width x thickness) are 72.5 x 35 x 12.5 mm. The average density of the brick was found,
1717.17 kg/m3, after the drying process. The compressive strength of the brick is f b = 32.45 MPa. The mortar
used for the construction of the masonry walls was an M-8 prepared mortar. In order to adjust the fineness of the
mortar for 1:4 scale sieving is made of it, removing all the percentage of material retained by the sieve1 mm
aperture, since the presence of larger sizes difficult to maintain size of the joint.

Figure 1: Layout of tested wall.

The average value of flexural strength is 3.05 MPa and the average compressive strength of the mortar
is 7.29 MPa which was taken as the value for the further calculation. To determine the uniaxial compression and
Youngs modulus of elasticity five specimens of 147.5 mm x 147.5 mm (height of 10 rows and width of 2
pieces) considered. The average value of the compressive strength and Youngs modulus obtained from the test
are 14.2 MPa and 3458 MPa respectively.
www.ajer.org

Page 329

American Journal of Engineering Research (AJER)

2017

III. ADOPTED MODELING STRATEGY

The numerical simulation presented is performed with the well-known micro-model proposed by
Lourenco & Rots (1997) requires more specific software oriented to masonry analysis. For all cases, micromodels assume 2D plain-stress and a hinged-hinged configuration. The hinges are modeled by means of stiff
triangular objects placed at the bottom and at top of the wall, whose end vertex is allowed to freely rotate. In
addition, a minimum eccentricity of 1mm is always applied in order to account for possible irregularities of the
wall geometry of the load positioning. Basically, the model assigns an elastic behavior to the units whereas
masonry inelastic behavior is transferred to the joints. This analysis was performed with DIANA software. The
integration schemes used are 2x2 points Gauss integration for the continuum elements and 3 points Lobato
integration for the interface elements. An interface allows discontinuities in the displacement field and its
behavior is described in terms of a relation between the traction t and relative displacement u across the
interface. In the multisurface interface model for the masonry proposed by Lourenco and Rots, the quantity of
traction and displacement is denoted as generalized stress and generalized strain, . In this case the elastic
constitutive relation between stresses and strain is given by:
D
(1)
For 2D configuration D diag k n , k s , ( , ) and (u n , u s )
Where, n and s is the normal and shear components respectively. The terms in the elastic stiffness
matrix can be obtained from the properties of both masonry components and thickness of the joint as:

Eu Em
Gu Gm
;k s
t m ( Eu Em )
t m (Gu Gm )

(2)

Where, Eu and Em = Youngs moduli; Gu and Gm = Shear moduli and tm = thickness of the joint.

Figure 2: Proposed modeling strategy.

The interface model includes a compressive cap where the complete inelastic behavior of masonry in
compression is lumped. This is a phenomenological representation of masonry crushing because the failure
process in compression explained by the microstructure of units and mortar and the interaction between them. In
the model the failure mechanism represented in such way that the global stress strain diagram is captured. The
model was justified by Lobato et al. and found that the model is efficiently able to reproduce the experimental
results. For this reason, the proposed micro-model was selected by the author to simulate the wall for buckling
failure.

IV. MODEL DESCRIPTION

In the numerical simulation, the units were modelled by using plain-stress continuum 8-node elements
and for the mortar joints adopted 6-node zero-thickness line interface elements. In addition, hinges are modelled
by means of stiff triangular objects. Each unit was modelled with 12 x 3 elements. The geometry and meshing of
the wall for slenderness ratio 6 and eccentricity 0, t/6 and t/3 are shown in the Figure 3. For all cases, micromodels of wall considered hinged-hinged configuration. The vertical load was applied concentrically and
eccentrically as unit deformation. The boundary condition and loading configuration is also shown in the Figure
3.
www.ajer.org

Page 330

American Journal of Engineering Research (AJER)

2017

Figure 3: Geometry, Meshing and Load of wall for slenderness ratio 6.

4.1 Material Properties
The material parameters used for the numerical simulation are shown in the Table 2. Some parameters
such as G I f and C S have been taken directly from the previous research. The fracture energy for mode I,

G I f have been taken from the test carried out by Van der Pluijm (1992) and for the parameter of shape of
elliptical cap C S a value of 9 has been adopted from Lourenco (1996). The interface elastic stiffness values
were calculated from thickness of the joint h j , the Youngs moduli of unit and joint E u and E j , respectively,
and the shear moduli of unit and joint Gu and G j , respectively as CUR (1994):

Eu E j
h j ( Eu E j )

;kt

Gu G j
h j (Gu G j )

(3)

The different strength values f f , c and f m have been obtained from the experimental study carried
out in UPC (2010). The compressive fracture energy G fc and equivalent relative displacement k p calculated
according to Model Code 90 and Eurocode 6, respectively by using followings formula (Lourenco, 1996):

15 0.43 f m 0.0036 f m ; k p {0.002 f m (

2

fc

1
1

)}
Eu k n (hu h j )

(4)

Table 2: Material parameters adopted for numerical analysis.

Components
Brick

Joint

Parameter
Elastic modulus
Poison ratio
Tensile strength
Normal stiffness
Shear stiffness
Bond tensile strength
Mode I fracture energy
Cohesion
Mode II fracture energy
Angle of internal friction
Angle of dilatancy
Compressive strength of masonry
Compressive fracture energy

Symbol
Eb

ftb
kn
kt
ft
GIf
c
GIIf
tan
tan
fm
Gfc

Units
N/mm2
N/mm2
N/mm2
N/mm2
N/mm2
Nmm/mm2
Nmm/mm2
N/mm2
Nmm/mm2

Values
4800
0.15
3.95
2800
1900
0.554
0.02
0.45
0.175
0.812
0.009
14.20
20.38

4.2 Validation of Model

The micro-models were validated next by a comparison with experimental results obtained from UPC
(2010). The main concern of this work was, to demonstrate the ability of the model to capture the behaviour
observed in the experiments and close quantitative reproduction of the experimental results.

www.ajer.org

Page 331

American Journal of Engineering Research (AJER)

2017

Figure 4: Comparison of compressive stress for different load eccentricity.

The comparison between the experimental collapse load and collapse load obtained from the micromodel is presented in the Figure 4. The figures shows that the experimental behavior is satisfactorily reproduced
and the collapse load estimated within a 15 % range of the experimental values. The micro-modeling approach
is being able to provide a very satisfactory estimation of the experimental capacity of the walls particularly, for
the case with e = 0. The average errors are 7.85%, 12.6% and 11.93% for the eccentricity of 0, t/6 and t/3
respectively. For all cases, one tendency is clear that with the increasing of slenderness ratio and application of
load eccentricity the capacity of the wall decreased.

V.

ANALYTICAL APPROACH

5.1

Eurocode 6 (EC 6)
According to Eurocode-6, the resistance of a masonry wall subjected to vertical load depends on the
geometry of the wall, the eccentricities of the load and constituent material properties. This development
allowed the following assumptions:
After each cross-section deformation remains plane and normal to the deformed axis (Bernoulli-Navier
hypothesis);
The resistance of the wall in tension perpendicular to the bed joints is zero.
5.1.1 Determination of Vertical Load Resistance
The vertical load resistance of a single leaf wall per unit length, N RD , can be calculated as:

N RD

i ,m tf k

(5)

Where, i,m is the capacity reduction factor i (top or bottom of wall) or m (in the middle one
fifth of the height of wall), allowing for the effects of slenderness and eccentricity of loading, f k is the
characteristic compressive strength of masonry according to paragraph 3.6.2 of EC 6, if the cross-sectional area
www.ajer.org

Page 332

American Journal of Engineering Research (AJER)

2017

A is less than 1 m2, this property is multiplied by the factor (0.7 + 3A), M is the partial safety factor for the
material, under paragraph 2.3.3.2 of EC 6 and t is the thickness of the wall, taking into account the depth of
recesses in joints greater than 5 mm.
5.1.2 Determination of Reduction Factor for Slenderness ratio and Eccentricity

At the top or bottom of the wall.

i 1

2ei
t

(6)

Where, ei is the eccentricity at the top or the bottom of the wall;

ei

Mi
ehi ea 0.05t
Ni

M i is the design bending moment at the top or the bottom of the wall resulting from the eccentricity of the
floor load at the support, according to 4.4.7 (Figure 4.1) of EC 6;
N i is the design vertical load at the top or bottom of the wall;

e hi is the eccentricity at the top or bottom of the wall, if any, resulting from horizontal loads (for example,
wind);

ea

hef
450

; is the accidental eccentricity and

t is the thickness of the wall.

In the middle one fifth of the wall height.

m A1e

u2
)
2

(7)

emk
t
hef

2
t

u is the Numerical factor, u ef

emk
23 37
t
hef is the effective height, obtained from 4.4.4 of EC 6 for the appropriate restraint or stiffening condition;
t is the thickness of the wall;
t ef effective thickness of the wall in accordance with paragraph 4.4.5 of the EC 6;
e is the base of natural logarithms, approximately, e = 2.71828;
emk is the eccentricity within the middle one fifth of the wall height;

Where, A1

with

Numerical factor, A1 1 2

emk em ek 0.05t
M
em m ehm ea
Nm

em is the eccentricity due to loads;

M m is the greatest moment within the middle one fifth of the height of the wall resulting from the moments at
the top and bottom of the wall, see Figure 6 (a);
N m is the design vertical load within the middle one fifth of the height of the wall;

ehm is the eccentricity at mid-height resulting from horizontal loads (for example, wind);

ek is the eccentricity due to creep;

em 0.002

www.ajer.org

hef
t ef

tem

Page 333

American Journal of Engineering Research (AJER)

2017

is the final creep coefficient from Table 3.8 of EC 6.

(a)
(b)
Figure 6: (a) Moment according to EC 6 and (b) values of m against slenderness ratio for different
eccentricity.
This development has taken a modulus of the elasticity of masonry as a thousand times the compressive
resistance property of masonry ( E 1000 f k ).
5.2 ACI-530
ACI-530-05 code was reported by Masonry Standards Joint Committee (MSJC), the design of
unreinforced masonry has included a limit on the allowable axial compression force that may be applied. The
limit is that the maximum allowable compressive force P is not exceed one fourth of the buckling load P e as
defined in the code. The maximum compressive force is limited to:

Pe
4

(8)

Where,
3

Em I
e
(9)
Pe
1 0.577
2
r
h
In which, E m = modulus of elasticity; I = uncracked moment of inertia of the section; e = eccentricity
of the compressive force P; r = radius of gyration of the uncracked unit section; h = unbraced height of the
2

member under load. As the member deflects and bends under the action of eccentrically applied force, flexural
tension cracking occur wherever the bending stress due to moment exceed the axial compression stress. The
buckling equations for members subjected to compressive force are shown below:

Pe 2

Em I
e
1 2
2
t
h

(10)

For a solid rectangular cross-section, the radius of gyration is approximately equal to 0.289t . For
members having an slenderness ratio less than 99 and greater than 99, the allowable compression stress under
axial load Fa is given from the following equation 11 and 12, respectively:

1 h
Fa f m

4 140rh

(11)

1 70r
(12)
Fa f m

4 h
Where, f m = specified compressive strength of masonry, t = thickness of the wall.

www.ajer.org

Page 334

American Journal of Engineering Research (AJER)

2017

VI. COMPARISON OF COLLAPSE LOADS

A comparison between the experimental results obtained, the results calculated with standards ACI-530
and EUROCODE 6 (EC 6) and the proposed numerical micro-models is presented in Figure 7. As can be
observed, the method proposed in EC 6 underestimates substantially the bearing capacity of the walls. The other
standard considered, ACI-530, also underestimates the strength of walls in all cases.
ACI-530 code produce average error of 76.86%, 87.62% and 96.26% compared to experimental results
for the cases of load eccentricity 0, t/6 and t/3 respectively. For overall cases this code shows average error
86.92%. EC 6 underestimates the strength of the walls in all cases, although it is able to reproduce the general
tendency. This code provides more satisfactory estimations of collapse loads for the lower eccentricity and
lower slenderness ratio specifically, for the eccentricity t/6 which provides average error of 23.94%. Moreover,
EC 6 is the most conservative method to predict the collapse load of the higher eccentrically loaded walls
(e=t/3) with an average error of 53.6% while it produce overall average error of 34% when compared with
experimental results.

Figure 7: Comparison of compressive stress for different slenderness ratio and eccentricity.
The best fit occurs in the EC 6 (7.29%) and ACI-530 (38.47%) for the eccentricity t/6 and 0
respectively, with slenderness ratio 6 for both, while the bothe standards produce maximum error of 98.70% and
98.77% for eccentricity t/3 and slenderness ratio 25. On the other hand, both standards are underestimates the
collapse load of wall when compared to numerical micro-models. ACI-530 provides most conservative results
with average error of 87.19%, while EC 6 estimates collapse load by 31.30% of average error.
www.ajer.org

Page 335

American Journal of Engineering Research (AJER)

2017

VII. CONCLUSIONS
The diverse combinations of slenderness ratio and load eccentricity used in the experimental program
which provided the means for a comprehensive numerical analysis of the masonry wall. The micro-models
afford a satisfactory prediction of the ultimate load of walls taking into account the buckling behavior.
Simulations carried out by the micro-model provide the best fits for all load eccentricity, (with an average error
of 10.79%). It must be noted that some difference with respect to the experimental results is unavoidable
because of the influence of possible non-reported accidental eccentricities. In the case of fixed support, the load
capacity increased 2 to 6 times higher than hinge support depending on slenderness ratio and eccentricity. The
capacity of wall for hinge-fixed support lies between the both end hinge and both end fixed support. In the case
of hinge-hinge support with high eccentricity.
The comparison between experimental and the standards results shows significant errors of 86.92%
and 34.07% for ACI-530 and EC 6, respectively. In particular, this comparison suggests that the both method
proposed by EC 6 and ACI-530 tends to conservatively underestimate the strength of walls. The micromodeling approach has shown its ability to assess the bearing capacity of masonry walls subjected to concentric
or eccentric vertical loading. It has been observed that an accurate description of tensile cracking and opening of
mortar joints, by means of an appropriate interface element, is essential to obtain reliable results on the bucking
failure of walls.

ACKNOWLEDGEMENTS
In completing this research, I would like to express my grateful appreciation and special thanks to
Professor Paulo Lourenco, Professor Enrico Garbin and PhD student Cristian Sandoval. A financial support
from European Commission through Erasmus Mundus Scholarship is gratefully acknowledged.

REFERENCES
[1].
[2].
[3].
[4].
[5].
[6].
[7].
[8].
[9].
[10].
[11].
[12].
[13].
[14].
[15].
[16].
[17].
[18].

ACI-530 (2008). ACI Manual of Concrete Practice. ACI 530.1-05/ASCE 6-05/TMS 602-05. American Concrete Institute.
CEN (1995). Eurocode 6: Design of Masonry Structures ENV 1996-1-1:1995, CEN, Brussels, Belgium.
Chapman, J. C. and Slatford, J. (1957). The Elastic Buckling of Brittle Columns. Inst. Civ. Eng., Vol. 107, No. 6.
De Falco, A. and Lucchesi, M. (2007). No Tension Beam-Columns with Bounded Compressive Strength and Deformability
Undergoing Eccentric Vertical Loads. International Journal of Mechanical Sciences 49, p. 54-74.
Hasan, S. S. and Hendry, A.W. (1976). Effect of Slenderness and Eccentricity on the Compressive Strength of Walls. Proc. 4th
International Brick Masonry Conference (Brugge). Paper 4.d3.
Lourenco, P. B. (1996). Computational Strategies for Masonry Structures. PhD thesis, Delft University of Technology, Delft, The
Netherlands.
Lourenco, P. B. and Rots, J.G (1997). Multisurface Interface Model for Analysis of Masonry Structures. Journal of engineering
mechanics, 123(7), p. 660-668.
Lourenco, P. B. (1996). A User / Programmer Guide for the Micro-modeling of Masonry Structures. TU- DELFT, Report no. 0321-1-31-35, Delft University of Technology, Delft, The Netherlands.
Mura, I. (2008). Stability of Nonlinear Masonry Members Under Combined Load. Computers and Structures 86, p. 1579-1593.
Page, A.W. (1978). Finite Element Model for Masonry. J. Struc. Div., ASCE, 104(8), p. 1267-1285.
Parland, H. (1982). Basic Principles of the Structural Mechanics of Masonry: A Historical Review. Int. J. Mas. Constr., 2(2).
Payne, D. C., Brooks, D. S. and Dved, G. (1990). The Analysis and Design of Slender Brick Walls. Masonry International
Journal, 4(2), p. 55-65.
Pluijm, R. V (1992). Material Properties of Masonry and its Components Under Tension and Shear. Proc. 6th Canadian Masonry
Symposium, Eds. V.V. Neis, Saskatoon, Saskatchewan, Canada, p. 675-686.
Sahlin, S. (1978). Structural Masonry. Prentice-Hall Inc., New jersey, U. S. A.
Sawko, F. and Towler, F. (1982). Numerical Analysis of Walls and Piers with Different End Conditions. 6IBMAC. Ed.
Laterconsult s.v.l., Rome, Italy.
TNO DIANA (2010). Finite Element Analysis. The Netherlands.
Valladares, I. N. (2010). Estudi Experimental de la Fallada a Vinclament de Parets de Crrega de Fbrica de Ma. MSc. thesis,
Technical University of Catalonia, Barcelona, Spain.
Yokel, F. Y. (1971). Stability and Load Capacity of Members with no Tensile Strength. J. Struct. Div. ASCE, 97(7), p. 19131926.

www.ajer.org

Page 336