BEHAVIOUR OF BRICK MASONRY COLUMNS WITH FERRO CEMENT OVERLAY .• -. -~ .... -._ ..• THESIS SUBMITTED TO THE DEPARTMENT OF CIVIL ENGINEERING IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY by TohurAhmed 1111111111111111111111111 111111 III #92763# CIVIL ENGINEERING DEPARTMENT BANGLADESH UNIVERSITY OF ENGINEERING AND TECHNOLOGY (BUET) DECEMBER, 1997
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
BEHAVIOUR OF BRICK MASONRY COLUMNS
WITH
FERRO CEMENT OVERLAY
.•-. -~....-._ ..•
THESIS SUBMITTED TO THE DEPARTMENT OF CIVIL ENGINEERING IN
PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
by
TohurAhmed
1111111111111111111111111 111111 III#92763#
CIVIL ENGINEERING DEPARTMENT
BANGLADESH UNIVERSITY OF ENGINEERING AND TECHNOLOGY
(BUET)
DECEMBER, 1997
BEHAVIOUR OF BRICK MASONRY COLUMNS WITH FERROCEMENT OVERLAY
Approved a&to &tyleand content by
Dr. Jamilur Reza ChoudhuryProfe&&OrCivil Engineering Dept.Banglade&h Univer&ity ofEngineering and Technology
Dr. Sk. Sekender AliProfe&&OrCivil Engineering Dept.Banglade&h Univer&ity ofEngineering and Technology
Dr. Md Hossain AliProfe&&orand HeadCivil Engineering Dept.Banglade&h Univer&ity ofEngineering and TechnQlogy
Dr. Sohrabuddin AhmadProfe&&OrCivil Engineering Dept.Banglade&h Univer&ity ofEngineering and Tec!m<?logy
Dr. Md Wahhaj UddinProfe&&or --Mechanical EngineeringI?ept.Banglade&h Univer&ity ofEngineering and TechnQlogy
Dr. Md A. MansurM&ociate Profes&OrCivil Engineering Dept.National Univer&ity of SingaporeSingapore
~At.'Member (Co-Supervi&Or)
!Y .it. ~.Member
~-!MMember
.~. ,
Member (Extep1al)
CERTIFICATE
I hereby certify that the work embodied in this thesis is the result of original research
and has not been submitted for a higher degree to any other University or Institute.
Tohur Ahmed
ABSTRACT
This thesis presents a comprehensive finite element study of the composite behaviour
of brick masonry column with ferrocement overlay. The ANSYS package available
at BUET Computer Centre has been used to perform this fmite element analysis. The
model considers the columns as a composite of three types of materials viz. bricks,
mortar matrix and ferrocement.
The nonlinear response of the column is produced by a combination of nonlinear
deformation characteristics and progressive failure of the constituent materials. The
material properties for the analytical model are determined from tests on samples of
ferrocement, bricks, mortar and small brick masonry specimens.
A total of 48 columns were tested in this study. The investigations were carried out
under only concentric and eccentric. short term static loading. The application of
ferrocement overlay on the bare masonry column increases the load carrying
capacity of the column quite significantly. With the increase of cross sectional area
of column by 46% due to addition of ferrocement overlay, the average increase in
strength is found to be 184% for axial loading and 158% for eccentric loading.
The results of the finite element analysis are verified by comparison with the
uniaxial behaviour of masonry column with ferrocement overlay. Several
parameters like type of ferrocement overlay, discontinuity in ferrocement overlay,
mortar proportion and number of layers of wire mesh were varied to produce
variations in behaviour and in failure modes of the specimens. In general, the
agreement between finite element analysis and experiment in case of failure load,
failure pattern and load-deformation characteristics is good.
vii
Sensitivity analyses of various parameters like bed joint thickness, tensile and
compressive strength of mortar, brick and ferrocement, number of layers of wire
mesh and modulus of elasticity and Poisson's ratio of brick and ferrocement, defining
the material model were carried out using the finite element model. The following
parameters were found to be the most significant in the range of values considered in
the study:
i) the compressive strength of ferrocement mortar (range: from 15MPa to 20 MPa)
ii) the tensile strength of brick (range: from 0.5 MPa to 2.25 MPa)
iii) the thickness ratio of bed joint and brick (range: from 0.045 to 0.30).
The finite element program available in ANSYS package has also been used to carry
out a comprehensive parametric study of the composite behaviour of ferrocement-
coated masonry column subjected to axial and eccentric short term static loading.
The parameters related to the strength and deformation characteristics of the
constituent materials of the composite (as listed in the preceding paragraph) were
considered in this study. From the parametric study, empirical equations are derived
to predict the ultimate loads of the ferrocement coated masonry column.
ACKNOWLEDGMENT
The author wishes to express his indebtedness to Dr. J.R. Choudhury, Professor of
Civil Engineering, Bangladesh University of Engineering and Technology, under
whose supervision the work was carried out. The author also expresses his
indebtedness to the Co-supervisor of this work Dr. Sk. Sekender Ali, Professor of
Civil Engineering, Bangladesh University of Engineering and Technology. Without
their constant guidance and invaluable suggestions at every stage, this work could
not have possibly materialized.
The author is grateful to the members of his Doctoral committee for their
suggestions and fruitful discussions at various phases of progress of this research.
The author acknowledges with gratitude the valuable suggestions of Dr.
Sohrabuddin Ahmad, Professor of Civil Engineering, BUET and a member of the
Doctoral Committee during the preparation of the thesis.
The support of the laboratory staff of the Department of Civil Engineering during
the course of experiments is gratefully acknowledged. The author also likes to
acknowledge the suggestions and services received from the Director and other
members of staff at the Computer Centre of BUET. The donation of the software
package ANSYS to BUET by Dr. S.M. Yunus and Dr. Ashraf Ali of ANSYS, USA
is very much appreciated as this software played a key role in the finite element
analyses carried out as part of this thesis.
The author gratefully acknowledges the financial support received during this work
from BUET and BIT, Rajshahi. The author also acknowledges the moral support
and suggestions from the faculty members of BIT, Rajshahi as well as from the
tm Thickness of MortarNDA Normal, double layer of wire mesh and axial loading
xv
LIST OF FIGURES
3.1 Three-Dimensional Finite Element Idealization(Quarter Column Cross-Section) 26
3.2 Transverse Stress Distribution Along Column Center Line(Top and Bottom Nodes Restrained Laterally) 29
3.3 Transverse Stress Distribution Along Column Center Line in Case ofUniform Displacement 30
3.4 Variation of Top Displacement 31
3.5 Transverse Stress Distribution Along Column Center Line in Case ofPlaster Overlay 33
3.6 Transverse Stress Distribution Along Column Center Line in Case ofFerrocement Overlay 34
3.7 Transverse Stress Distribution Along Column Center Line by VaryingType of Overlay 36
3.8 Transverse Stress Distribution Along Column Center Line forDifferent Thickness of Ferrocement Overlay 38
3.9 Transverse Stress Distribution Along Column Center Line for DifferentNumber of Mesh Layers 39
3.10 Transverse Stress Distribution Along Column Center Line for DifferentModulus of Elasticity of Ferro cement Overlay 41
3.11 Transverse Stress Distribution Along Column Center Line forDifferent Thickness of Bed Joint 43
3.12 Finite Element Mesh for Ferrocement Coated Column with Vertical Groove 45
3.13 Transverse Stress Distribution Along Column Center Line in Case ofDiscontinuous Ferrocement Overlay 46
3.14 Finite Element Mesh for Eccentric Load Case 47
3.15 Transverse Stress Distribution Along Column Center Line in Case ofEccentric Loading 48
4.1 Uniaxial Compression Test on Hollow Ferrocement Block 54
4.2 Average Stress-Strain curve of Ferro cement Loaded in Uniaxial Compression 55
4.3 Lateral vs. Longitudinal Strain for compression Test of Ferro cement 56
4.4 Uniaxial Compression Test on Stack Bonded Prism 59
4.5 Stress-Strain Curve for Brick, Brickwork and Mortar Obtained fromPrism Test 60
xvi
4.6 Lateral vs. Longitudinal Strain of Brick 61
4.7 Uniaxial Compression Test on Mortar Cylinder 64
4.8 Stress-Strain Curve of Mortar (1:5) 64
4.9 Stress-Strain Curve of Mortar (1:2) 65
4.10 Lateral vs. Longitudinal Strain for Mortar (1:5) 67
4.11 Lateral vs. Longitudinal Strain for Mortar (1:2) 67
5.1 Different Types of Column 74
5.2 Column During Construction 78
5.3 Mould for Ferrocement Hollow Column 78
5.4 Column (DDAlDSA) During Construction 79
5.5 Column (DDAlDSA) After Construction 79
5.6 Curing of Column in the Laboratory 81
5.7 The Specimen Before Test 81
5.8 Failure of Column Coated with Ferrocement Series NDA (Axial Loading) 87
5.9 Experimental Stress-Strain Curves of Different Columns 89
6.1 Predicted Failure Mode of Column Series NDA 98
6.2 Predicted Failure Mode of Column Series NSA 99
6.3 Stress-strain Curve of Different Columns with Axial Load Obtained fromFEM Analysis 104
6.4 Load-strain (Compressive Side)Curve for Different Columns withEccentric Load Obtained from FEM Analysis 105
7.1 Failure Mode of Column Series NSA 113
7.2 Stress-Strain Curve of Column Series NDA 114
7.3 Stress-Strain Curve of Column Series NSA 114
8.1 Mode of Failure of Column Series NDA 121
8.2 Mode of Failure of Column Series NSA 122
8.3 Mode of Failure of Column Series NZAW 123
9.1 Comparison of Proposed Design with Finite Element Analysis forAxial Load Case 144
9.2 Comparison of Proposed Design with Finite Element Analysis forEccentric Load Case 145
XVll
LIST OF TABLES
4.1 Summary of Ferro cement Properties 52
4.2 Summary of Brick Properties 61
4.3 Summary of Mortar (1:2) Properties 68
4.4 Summary of Mortar (1:5) Properties 68
5.1 Test Programme 72
5.2 Description of Specimens 76
5.3 Summary of Load Tests 82
5.4 Experimental Cracking and Failure Loads (Axial Loading) 84
5.5 Experimental Cracking and Failure Loads (Eccentric Loading) 85
5.6 Load Increase in Failure Due to Confinement Effect (Derived from
Experiments) 90
5.7 Material Cost of Different Columns 92
6.1 Analytical Cracking and Failure Loads (Axial Loading) 100
6.2 Analytical Cracking and Failure Loads (Eccentric Loading) 101
6.3 Load Increase in Failure Due to Confinement Effect (Derived from Analysis) 103
7.1 Analytical and Experimental Cracking and Failure Load (Axial Loading) 109
7.2 Analytical and Experimental Cracking and Failure Load (Eccentric Loading) 110
7.3 Load Increase in Failure Due to Confinement Effect 116
8.1 Influence of Linear Elastic Fracture Analysis 124
8.2 Parametric Study of Elastic Properties (EmlEnJ and Vfin 125
8.3 Parametric Study of Elastic Properties (EJEm) Vb 125
8.4 Influence of Bed Joint Thickness 127
8.5 Influence of Tensile Strength of Mortar 128
8.6 Influence of Compressive Strength of Mortar 129
8.7 Influence of Tensile Strength of Brick 130
8.8 Influence of Compressive Strength of Brick 131
8.9 Influence of Tensile Strength of Ferro cement 132
8.10 Influence of Compressive Strength of Ferro cement 133
xviii
8.11 Influence of Number of Mesh Layers 1348.12 Influence of Mesh Size in Finite Element Discretization 1358.13 Influence of Slenderness Ratio 1368.14 Influence of Boundary Conditions 1379.1 Failure Load Obtained from Different Methods 147
CHAPTER!
INTRODUCTION
l.1GENERAL
Structures built of stone or stone-like materials are known as masonry structure. In a
primitive form, they were among the earliest types of structures erected by man. The
enduring character of masonry structure, the relative simplicity of the processes
involved, the pleasing outlines usually obtained, together with the almost universal
availability of the materials and the consequent moderate cost, render masonry
construction one of the most important of the civil engineer's activities. Moreover,
the importance of masonry construction is likely to be enhanced in future by the
growing scarcity of other structural materials, notably steel and timber, and the fact
that the ingredients of masonry are almost unlimited in their raw state.
Now-a-days some buildings are used for purposes other than those in the original
design. Sometimes, intermediate floors are added and this involves higher loads on
slabs, beams, columns and foundations. These structures were usually constructed
from bricks or concrete blocks and in older cases stones. These units are tied together
by mortar (e.g. sand-cement mixture) and in some cases, steel or other
reinforcement.
There are several types of masonry structural elements within a building, e.g.
columns and walls. Columns are the primary vertical load carrying members of a
typical multi-storey building. The loads coming on the floors and beams are
transmitted to the foundation through these columns. These columns are also called
upon to resist lateral loads on the building due to wind and/or earthquake. Efforts
have been made in recent years to improve the performance of brick columns in
2
seismic areas by applying ferrocement overlay. The concept has been intuitively
applied for repair of distressed elements as well.
Ferrocement is a type of thin wall reinforced concrete commonly constructed of
hydraulic cement mortar reinforced with closely spaced layers of continuous and
relatively small wire diameter mesh. The mesh may be made of metallic or other
suitable materials. Ferrocement is a versatile construction material and confidence in
the material is building up resulting in its wider applications especially in developing
countries in housing, sanitation, agriculture, fisheries, water resources, water
transportation in freshwater and marine environment, biogas structures, repair and
strengthening of older structures, and others.
Considered to be an extension of reinforced concrete, ferrocement has relatively
better mechanical properties and durability than ordinary reinforced concrete. Within
certain loading limit, it behaves as a homogeneous elastic material and this limit is
wider than those for normal concrete. The uniform distribution and high surface area
to volume ratio of its reinforcement result in better crack arrest mechanism, and
hence better use of its strength.
Although developed more than one hundred and fifty years back, use of ferrocement
composites is comparatively recent and is expanding. Ferrocement overlay is used
now-a-days to increase the ductility of masonry columns and walls. Ferrocement
composite column means a column having a ferrocement casing and a core of
brickwork or concrete (plain or reinforced). The casing may be circular, square or
rectangular depending on the shape of the column. Such columns may have
applications in three situations. Firstly, for prefabricated construction, where the
outer casing will be precast in a factory and concrete is poured into it at the site.
Secondly, for normal construction in which the ferrocement casing is cast in short
heights and filled with concrete. The third application may be for strengthening of
3
existing columns. In the first two cases the need for column formwork is eliminated
but there is a major structural difference between the two. In the first case, the core
shrinks after casting while the casing undergoes shrinkage before casting of the core.
Hence a weak interface may exist between the core and the shell. This is ail
important factor for the confinement effect of the central core.
In formulating design recommendations for axial and eccentric loads on columns,
simplifications are usually made because of the difficulty in obtaining sufficient
experimental data and realistic analysis of column behaviour. The prediction of
failure of such composite columns is difficult due to a large number of parameters
influencing the ultimate behaviour of column, the lack of a suitable material model,
and an efficient numerical technique.
To establish the critical parameters which influence the behaviour of bare masonry
columns and columns with ferrocement overlay, a linear elastic finite element
analysis has been performed using ANSYS package (26). A three-dimensional
analysis is carried out assuming brick masonry to be a homogeneous continuum as
well. as an assemblage of elastic bricks and joints, each with differing material
properties.
To realistically predict failure of such composite column a more sophisticated
material model is required. This model must reflect the inelastic nature of the
constituents as well as the progressive failure that occurs as the applied load is
increased. TIle material model required for such an analysis is microscopic rather
than macroscopic in nature with bricks and joints being modeled separately. The
properties needed to defme this material model available in ANSYS package is
obtained here from various simple tests on sample of bricks, mortar, small brick
masonry specimens and ferrocement specimens thus avoiding the need for more
4
complex testing apparatus. The model consists of elastic and inelastic deformation
characteristics, cracking and crushing of the constituent materials.
Both coated and uncoated brick masonry columns have been analysed in this study
by using finite element technique available in the ANSYS package. Sensitivity
analyses of the various parameters used in the fInite element analyses have also been
carried out. These analyses highlight the importance of the accurate evaluation of
deformation characteristics and strength characteristics of the masonry constituents.
The [mite element analysis is used here to carry out a comprehensive parametric
study of the behaviour of the composite columns. This study illustrates the potential
of the model both as a research tool and as a means of preparing design procedures
for practical use. From the results of the parametric study, design formulae for
predicting the ultimate loads for both axial and eccentric cases of the column have
been proposed.
1.2 OBJECTIVES OF THE RESEARCH
The principal objectives of the study are as follows:
1. The establishment of critical parameters influencing the behaviour of both
coated and uncoated brick columns using three-dimensional finite element
analysis.
2. VerifIcation of the accuracy of the material model used in the [mite element
analysis in predicting the failure loads by conducting experiments on columns
with ferrocement overlay subjected to axial as well as eccentric loading.
3. Investigation of the influence of various parameters involved in the composite
behaviour of brick column and ferrocement overlay.
5
4. The development of a design method for brick columns with ferrocement
overlay.
1.3 SCOPE OF WORK
The study is limited to masonry columns with burnt clay bricks and sand-cement
mortar, subjected to concentric and eccentric short term static loading.
A fInite element computer program (ANSYS) is used to examine the behaviour of
brick masonry column with ferrocement overlay. The nonlinear fInite element
computer program (ANSYS) used in this study will be broadly applied to investigate
its adequacy to:
(i) develop the complete load-deformation response of the columns; and
(ii) predict the ultimate load carrying capacity of the columns.
The effects of the following different parameters will be studied:
(i) Elastic properties of the constituents
(ii) Bed joint thickness
(iii) Tensile strength of mortar
(iv) Compressive strength of mortar
(v) Tensile strength of brick
(vi) Compressive strength of brick
(vii) Tensile strength offerrocement
(viii) Compressive strength offerrocement
(ix) Number oflayers of wire mesh
(x) Element size
6
(xi) Slendernessratio
Themajor parameters in the experimental investigationare
(i) Types of overlay
(ii) Number oflayers of wire mesh in the overlay
(iii) Discontinuities in the ferrocementoverlay
(iv) Types ofloading
CHAPTER 2
LITERATURE REVIEW
2.1 INTRODUCTION
Brick masonry columns are common in low-rise masonry buildings. They may be
reinforced or unreinforced, and commonly of square or rectangular and occasionally
circular cross-section. Although the performance of masonry column under axial
load may be satisfactory, they possess very limited capacity in bending. Encasement
by ferrocement can be a simple way.to increase the axial load carrying as well as
moment resisting capability of brick masonry columns.
This chapter reviews the literature on various aspects of this problem. Since brick
masonry and ferrocement have been used in this investigation, the material properties
of ferroCement and brick masonry are also reviewed. This is followed by a review of
failure mechanism of the constituent materials. Previous finite element models for
the analysis of ferrocement and masonry are then described. Since the number of
investigations made on masonry structural elements coated with ferrocement is very
limited, relevant works on the behaviour of concrete columns are also reviewed
here, because many of the parameters are very similar.
The fIrst large scale modem experimental research on reinforced brickwork (RB) has
been reported by Brebner in 1973 (17). This had established the applicability Of
working stress theory. However, subsequent research justify the use of ultimate
strength theory for the design of RB in flexure. The current British practice is to
design RB on the basis of ultimate strength design for both flexure as well as
compression (37). The Indian practice is similar, but makes use of the limit state
concept for flexure (23). Experiments at University of Roorkee (2, 24, 45, 57) have
shown that Whitney's ultimate stress block is somewhat approximate for RB in
"
8
flexure but, with some modifications, can be applied to it. If a RB column is
subjected to purely axial load, then its ultimate compressive load will be the sum of
the ultimate loads of masonry and reinforcement. However, as loads are commonly
eccentric, the effect of moment has to be considered also. The interaction diagram for
compression and flexure on the basis of ultimate strength has been provided by
Curtin (22).
Ferrocement encasement of a column can considerably increase its capacity to resist
compressive load and moment. This is a new concept of reinforcing a masonry
column and holds out a promise for economy. It can also be applied for repairing
distressed columns. However, the existing load should be removed before using
ferrocement for repairs (74).
The procedure consists of wrapping the layers of wire mesh around the columns
using some suitable arrangement like driving the V-nails through the mesh into the
column (82). A rich mortar having cement:sand ratio of 1:1.5 to 1:3 is then applied
over the mesh and made to penetrate the mesh and adhere to the column surface. On
setting, it forms a casing around the column.
It is well known that in a vertically loaded brick masonry column, cracks always
initiate from the vertical mortar joints and propagate through the bricks. TransverSe
stresses are mainly responsible for influencing the fracture process in this case. To
realistically predict the strength of such column, a numerical model capable of
predicting the ultimate failure of masonry subjected to any general loading is
required.
To develop a representative fmite element model for the analysis and design of
masonry structures with ferrocement overlay, a thorough knowledge of the
deformation and strength characteristics of the materials as weil as the masonry
9
assemblage is needed. Some models of this type have been developed (8), but they
are not necessarily suited to the analysis of brick masonry with ferrocement overlay.
Therefore a three-dimensional modeling is necessary to predict the composite
behaviour of brick masonry with ferrocement overlay.
The existing design rules for masonry columns are empirical and are not applicable
to masonry columns with ferrocement overlay. The design rules vary from country to
country, justifying the need for a comprehensive study in this area.
According to BNBC (97) the allowable axial compressive stress for unreinforced
masonry column is
F. = f~[1-(i-J 3]a 5 42t
where,
F. = allowable average axial compressive stress
f~= specified compressive strength of masonry
h' = effective height of column
t = effective thickness of column
The compressive strength of brickwork varies, roughly, as the square root of the
nominal brick crushing strength, and as the third or fourth root of the mortar cube
strength (37).
10
2.2 MATERIAL PROPERTIES
Brick masonry column with ferrocement overlay is a composite of three materials
and its properties are therefore dependent upon the properties of its constituents- the
ferrocement, the brick and the mortar joint. A brief review of the properties relevant
to the in-plane behaviour of masonry and the behaviour of ferrocement are carried
out in this section.
2.2.1 Properties of Ferrocement
In general, a composite material consists of a matrix and a reinforcement which act
together to form a new material with characteristics superior to either one of its
constituents alone (86). Ferrocement is a composite material which contains a high
percentage of ductile steel wire mesh with a high surface area to volume ratio in a
brittle cemcut-mortar matrix. This enables the matrix to assume the ductile
characteristic of the reinforcement. Usual range of wire diameter from 0.5 mm to 1.5
mm. Provision of reinforcement in excess of about 2 to 2.5% is uneconomical in
ferrocement as the proportional increase in strength is not achieved (86).
2.2.1.1 Strength Properties
The strength of ferrocement, as in ordinary concrete, is commonly considered as the
most valuable property, although in many practical cases other characteristics, such
as durability and permeability may in fact be more important. Nevertheless, strength
always gives an overall picture of the quality of ferrocement, as strength is directly
related with the properties of its hardened cement paste and reinforcement.
11
Tensile Strength
The tensile characteristics of ferrocement have not yet been fully defined and
standardized. In tension, the load-carrying capacity is essentially independent of
specimen thickness because the matrix cracks well before failure and does not
contribute directly to composite strength. The influence of types, sizes and volumes
of wire meshes on elastic cracking and ultimate behaviour offerrocement in uniaxial
tension have been studied by Naaman and Shah (58). They observed that the ultimate
tensile strength of ferrocement is the same as that of mesh alone while its modulus of
elasticity can be predicted from those of mortar and mesh (46, 58, 66, 90). The
specific surface of the reinforcement strongly influences the cracking behaviour of
ferrocement. Some technical information have been released (51, 59), but their
results seem to be specific to certain types of mesh reinforcement. In general, the
optimal choice of reinforcement for ferrocement strength in tension depends on
whether the loading is essentially uniaxial or significantly biaxial. Expanded metal in
its normal orientation is more suitable than other reinforcing meshes for uniaxial
loading because a higher proportion of the total steel is effective in the direction of
applied stress (46). For biaxial loading, square mesh is more effective because the
steel is equally distributed in the two perpendicular directions, although the weakness
in the 45 deg direction may govern in this case.
Compressive Strength
In this mode, unlike tension, the matrix contributes directly to ferrocement strength
in proportion to its cross-sectional area. Compressive strength of ferrocement
(regardless of the amount of mesh reinforcement) seems to be much the same as that
of mortar alone (70). The experimental results showed that under compression the
ultimate compressive strength is lower than that of equivalent pure mortar (66). The
compressive strength at ultimate condition is assumed to be 0.85fc where fc is the
ultimate compressive strength of the mortar. An investigation into the behaviour of
12
ferrocement specimen in direct compression has been discussed by Rao (70).
Conclusions were drawn with respect to the effect of percentage of reinforcement
and the size of reinforcement on the behaviour of ferrocement. Smaller diameter wire
mesh would be preferable to use as this gives higher elasticity and higher ultimate
compressive strengths for the same percentage of reinforcement, all the other factors
remaining essentially the same. When mesh reinforcement is arranged parallel to the
applied load in one plane only, no improvement in strength is observed (66). The
only forms of reinforcement likely to result in significant strength gains in
compression are square mesh reinforcements (86) fabricated in closed box or
cylindrical arrangements which restrain the matrix, thus forcing it to adopt a triaxial
stress condition with associated higher strength.
2.2.1.2 Deformation Characteristics
Following the consideration of ultimate and cracking strengths, it is appropriate t6
examine the overall load-deformation behaviour of ferrocement under various forms
of loading, in particular its modulus of elasticity.
Load Deformation Behaviour in Tension
For square mesh reinforcements, the load-elongation behaviour of ferrocement has
been characterised in three stages (58, 61, 66). In the initial stage, the matrix and
reinforcement act as a continuum having a composite elastic modulus approximately
equal to that predicted from the volumetric law of mixtures of the longitudinal
reinforcement and the matrix (55, 66). The second stage, associated with a fully
cracked matrix, is also linear. Its modulus is somewhat greater than the product of
the volume fraction and the modulus of the longitudinal reinforcement. The mortar
and the lateral reinforcement continue to play an active role after first cracking, either
13
individually or in combination (55,66). In the third stage, the matrix ceases to playa
role. Failure corresponds to the yielding of the reinforcement.
Load Deformation Behaviour in Compression
When the reinforcement is in one plane only, it has a minimal effect on the load-
deformation relationship, and the associated elastic modulus remains virtually the
same as that for the mortar matrix (66). When present in closed peripheral form, the
load-deformation relationship is curvilinear with the initial tangent modulus
increasing gradually with the amount of reinforcement (46). The initial elastic
modulus can be predicted quite accurately and conservatively on the basis of the
volumetric influence of the two material components acting together (71). Values of
the elastic modulus are slightly higher for specimens reinforced with welded mesh
than for their equivalents with expanded metal (46). The experimental results
obtained by various investigators (51, 70) show that the modulus of elasticity in
direct compression increases proportionately with the increase in steel content.
Studies on mechanical properties of ferrocement have been made since the early
1970s but studies on formulation of these properties based on fundamental material
properties has begun only recently. Some of its mechanical properties have not been
sufficiently investigated yet and not enough technical information is available to
suggest acceptable formulae for design.
2.2.2 Properties of Brick
2.2.2.1 Compressive Strength
Compressive strength of brick is one of the most important properties. Compressive
strength tests are easy to perform and give a good indication of the' general quality of
the brick and the compressive capacity of the resulting masonry. For these reasons,
14
the compressive strength test has been traditionally used for brick quality control and
specification.
The standard test for the determination of compressive strength can be influenced by
several factors such as loading rate (35), specimen size (9, 31, 69), perforation
pattern (9, 91, 93) and specimen end conditions (14, 29, 63). However, standard test
provides a basis for quality control and for making relative assessment of the strength
of brick masonry.
The standard test described in the Australian code for concrete masonry units (10)
requires the use of plywood packing on the top and bottom face of the specimen.
Such test results are influenced by the stiffness of the plywood and the frictional
restraint imposed by the solid platen, resulting in an artificial value of compressive
strength. Several investigators have attempted to minimize the effects of platen
restraint by using platens with variable stiffnesses (29, 62) and/or capping materials
(14, 79) on the specimens. Flexible steel brush platens have also been used
successfully for the testing of both concrete (49) and masonry (63, 64). An indication
of the magnitude of this strengthening effect has been given (65) from compression
tests on calcium silicate bricks. Steel brush and solid platens were used in tests on
bricks of varying size and shape. For standard size bricks, the unconfined
compressive strength (with brush platens) was found to be almost half the confined
compressive strength (with solid steel platens). This effect must therefore be
considered when assessing the compressive strength of a material.
2.2.2.2 Tensile Strength
Brick tensile strength has a significant influence on the in-plane behaviour of brick
masonry, as fmal failure usually occurs in some form of biaxial tension, split often
originating in the brick. When brick masonry is loaded in axial compression, for
15
example, lateral expansion of the brick and mortar takes place. Since the mortar
joints are typically more flexible than the bricks, the joint deformation is partially
restrained by the surrounding bricks due to the bond and friction at the brick-mortar
interface. This results in a triaxial compression stress state in the mortar and
compression iUldbilateral tension in the brick. Since the tensile strength of the brick
is low (much lower than its compression strength), failure of brick masonry is
initiated by tensile stresses.
Numerous attempts have been made to determine a convenient relationship between
the brick tensile strength obtained from a simple test and wall strength. Various
tension tests have been investigated, including modulus of rupture tests, splitting
tests (Double Punch or Brazil tests), and various form of shear tests including
indirect tension.
The effect of size, shape and disposition of the perforation on the tensile strength of
brick has been studied (9, 93). Significant reduction in the tensile strength of bricks
was reported in the case of bricks with perforation patterns which produce significant
stress concentration.
Despite the extensive research that has been carried out, no strong relationship
between brick tensile strength and brick masonry strength has emerged. As a result,
. brick compressive strength is still used as the prime indicator of the potential
compressive strength of the assemblage.
2.2.2.3 Other Properties of Brick
There are several other brick properties such as brick growth, pitting, efilorescence,
permeability, dimensional change, etc., which have a significant influence on the
satisfactory performance of masonry structure. However, most of these properties are
16
related to the physical and chemical characteristics of the brick and do not influence
the masonry strength. However the property that does significantly influence brick
masonry strength is the initial rate of absorption (IRA.) or brick suction. Brick
suction plays an important role in the achievement of bond and, as such, significantly
influences both the compressive and tensile strength of the masonry (8).
2.2.3 Properties of Mortar
Mortar in brick masonry has three main functions:
(i) To provide an even bed for the bricks
(ii) To bond the bricks together effectively
(iii) To seal the joint against weather
To perform these functions, mortar should possess suitable properties in both the
elastic and hardened states. These properties are briefly described in the following
section.
In its plastic state, the required properties are its good workability, good water
retentivity and sufficient early stiffening. Mortar workability depends upon the brick
and mortar properties, in particular, water retentivity of the mortar and the initial rate
of absorption of the brick. These two latter properties have also a marked effect on
the bond strength of the resulting brick masonry. Good water retention is required for
several reasons. It is needed to resist brick suction, to prevent bleeding of water from
the mortar, to prevent stiffening of the mortar bed before placement of the brick, and
to ensure retention of sufficient water in the mortar to allow hydration of the cement.
In its hardened state, the required properties are compressive strength, bond strength
and tensile strength. The compressive strength of the mortar is not an important
property since in masonry, the emphasis is on the achievement of adequate bond
17
between mortar and brick. However, it serves as an indicator of quality control. It
may be determined using either cube or prism tests. The compressive strength of
mortar as a function of shape, curing, age, air content and initial flow rate of mortar
(34).
2.3 BRICK MASONRY PROPERTIES
The two strength characteristics which ensure satisfactory performance of masonry
structures (specially for columns) are its compressive and tensile strengths. In
addition, the deformation characteristics of masonry are required for assessing the
stress distributions and relative movements under loads.
2.3.1 Masonry Compressive Strength
Since the majority of masonry structures are used to transmit loads in compression,
the compressive strength of the material is of prime importance.
The compressive strength of brick masonry is determined either from an approximate
relationship between brick strength, mortar type and brick masonry strength or from
compressive tests of prisms. When a more exact estimate of compressive strength is
required, a prism test is used. The strength is typically obtained from compression
tests on a series oftive high stack bonded prisms.
The influence of parameters such as brick and mortar properties, dimensional
variations, slenderness ratio, etc., on the strength of brick masonry have been
extensively reviewed by several investigators (34, 36,38, 39, 52, 56, 63, 75). The
effect of joint thickness on the compressive strength of brick masonry has been
reported (56) and a linear reduction in the compressive strength with the increase in
joint thickness has been suggested, whereas Francis et al (31) imd Chuxian (20)
suggested nonlinear relationS.
18
2.3.2 Masonry Tensile Strength
The tensile strength of brick masonry is inherently low and hence often restricts its
load carrying capacity (for both in-plane and out-of-plane loading). Under in-plane
loading, tensile failure may occur either normal or parallel to the bed joint depending
on the direction of loading. In-plane masonry tensile strength is critically dependent
on the loading direction with relation to the bed joints. This effect was first studied
by Johnson and Thompson (44) using diametrical splitting tests on circular
specimens sawn from a wall. Using this technique, by rotating the orientation of the
bed joint to the splitting force, tensile stress could be applied at varying angles to the
jointing directions.
2.3.3 Deformation Characteristics of Masonry
A knowledge of the deformation characteristics of masonry is needed to calculate the
deflection of masonry structures as well as to estimate differential movements in
buildings composed of masonry and other materials. Deformation characteristics are
also required for the numerical modeling of masonry behaviour in finite element
analysis.
Brick masonry typically exhibits nonlinear stress-strain relations. Most of the
nonlinear deformation occurs in the mortar joints with the bricks often exhibiting
linear stress-strain characteristics. Because of the influence of the bed joints and the
possible anisotropic properties of the bricks, the deformation characteristics of the
masonry are not necessarily isotropic and may vary markedly with loading direction.
Complete stress-strain relations for brick masonry have been experimentally
determined by Powel and Hodgkinson(68) and Scrivener and Williams (79) using
displacement controlled testing machines. Solid as well as perforated bricks were
19
used in their investigation. The shape of the stress-strain relations obtained was
parabolic.
2.4 EXPERIMENTAL INVESTIGATIONS OF COLUMNS
A small number of experimental investigations have been carried out to determine
the failure load of brick masonry column encased in ferrocement. Some tests have
been conducted by Nayak (60) and by Page (63), and it has been found that the
ferrocement helps in increasing the failure load. Tests were conducted by Singh et al
(82) for five different cases (3 specimens for each case) viz. plain, with plaster (1:6
and 1:2) and with two layers of square galvanished wire mesh in mortar with cement
to sand ratio of 1:6 and 1:2. Observations were made for failure loads, cracking load
and strains, and it was concluded that failure load is the lowest for unplastered
columns and highest for columns encased in ferrocement with 1:2 mortar. The failure
load of encased column was more than double the failure load of bare column.
The behaviour of ferrocement composite columns has some similarity with infilled
pipe columns. A number of investigations on R.C. pipe columns (hollow or filled)
and steel or plastic columns filled with concrete have been reported in literature. A
comparison of computer analysis of hollow R.C. spun pipe columns with
experimental results for axial and eccentric loads has been reported by Liu and Chen
(50). Furlong (30) carried out tests on steel pipe columns filled with concrete, for
axial and eccentric loading and found that there is little or no confmement effect of
cOncrete by the pipes. This might be because core concrete undergoes considerable
shrinkage, leaving a gap from the shell. An experimental investigation has been
conducted by Ghosh (33) on concrete filled steel tubes. For axial load, the strength of
concrete increases due to the effect of lateral confmement (33). This may occur as
value of Poisson's ratio of concrete near failure is considerably higher than that of
steel. Thus, under compressive load, both the core and pipe expand laterally due to
20
the Poisson's ratio effect. If the core tends to expand more than the pipe, then it
would be subjected to a lateral compression by the pipe, which in turn would be in,-
tension. An experimental investigation has been conducted by Bertero and Moustofa
(IS) on steel pipes filled with expansive cement concrete. The concrete would
expand instead of shrinking and would thus be subjected to an initial precompression
by the pipe. It is well known that compressive strength of concrete increases
considerably due to the confmement pressure. Tests on square columns with square
hoop steel (85), clearly indicate the strength increase due to lateral confinement.
Experiments on concrete filled plastic tube columns also show some effect of
confmement (48).
From the previous investigations it is clear that the confmement provided by
ferrocement should also contribute to a substantial increase in the axial load carrying
capacity of a masonry ferrocement composite column. For eccentric loads, there can
also be some strength increase. Spun ferrocement pipe columns filled with concrete
and subjected to axial and eccentric loads have been tested (77). The pipes had 0,3,5
and 7 layers of mesh. The most significant conclusion was that ultimate failure load
did not depend on number of mesh layers. In case of column without mesh, the
failure was of brittle type while with mesh it was ductile. The failure load of the
composite was not substantially different from the sum of the individual failure loads
of the core and casing, tested separately. All of these investigations (77) indicate that
the central core may not have confinement effect due to lack of proper bonding
between the core and the outside shell. This is possible due to uneven shrinkage of
these elements which are normally constructed at different times.
Bett et al conducted an experimental investigation and examined the effectiveness of
the different repairing and/Or strengthening techniques in enhancing the lateral load
response of identical reinforced concrete short columns (16). Both the strengthened
and the repaired columns performed better than the original column. Columns
21
strengthened and repaired by jacketing, with or without supplementary crossties,
were much stiffer and stronger laterally than the original unstrengthened columns.
Ersoy et al carried out two series oftests to study the behaviour of jacketed columns
(28). The main objective of the first series (uniaxially loaded) was to study the
effectiveness of repairing and strengthening of columns by jacketing with a new
layer of concrete that is reinforced with both longitudinal and transverse
reinforcement, both under sustained load and after removal of the load from the
column. Specimens repaired and strengthened by jacketing behaved well when
jacketing was introduced after the removal of the load. However, when jacketing was
made under load, the columns exhibited poor behaviour. In the second series,
jacketed columns were tested under combined axial load and bending. Two
monolithic specimens were also tested to serve as reference specimens. The
influence of load history on the behaviour of jacketed columns was also studied and
concluded that repair and strengthening jackets behaved well, both under monotonic
and reversed cyclic loadings.
An experimental investigation into the behaviour of short reinforced concrete
columns was described by Scott et al (78). Results presented include an assessment
of the effect of strain rate, amount and distribution of longitudinal steel, and amount
and distribution of transverse steel. Scott et al concluded that the longitudinal strain
rate influenced both the peak stress and the slope of the falling branch of the stress-
strain curve of the concrete core, an increase in the volume ratio of transverse
reinforcement increased the peak concrete core stress, and an increase in the number
oflongitudinal reinforcing bars resulted in better confmement of the core concrete.
An experimental study was performed by Sandowicz and Grabowski (77) on
columns both under eccentric and axial load. They presented a comparison of the
tests results of columns made of ferrocement pipes and mortar pipes, with or without
22
concrete infill, as well as concrete core alone. Ultimate load carrying capacity of
columns made of ferrocement pipes filled with concrete is higher than that of
reinforced concrete columns having similar diameters and volumetric percentage of
reinforcement. In the authors opinion, these columns are superior to spirally
reinforced concrete columns due to their ability to sustain tensile stresses.
2.5 REMARKS
It is apparcnt from the available literature that a considerable number of researchers
studied thc composite behaviour of masonry columns and reinforced columns
encased in fcrrocement overlay under uniaxial loading. From these studies they made
some conclusions about the overall performance of the composite column but no
empirical rclations of the performance emerged from their investigations. In most of
the caseS the development of these relations are handicapped due to the noninclusion
of all the relevant parameters in their observations. No theoretical investigation is
performed in these areas. A series of studies have, therefore, been initiated in the
Department of Civil Engineering, BUET, to study the behaviour of masonry column
encased in ferracement overlay. The studies include extensive laboratory
investigations and analytical studies using 3-D nonlinear finite element technique.
CHAPTER 3
ELASTIC FINITE ELEMENT ANALYSIS
3.1 INTRODUCTION
The preliminary elastic finite element study outlined in this chapter is aimed at
establishing the important parameters which influence the behaviour of brick
masonry columns with ferrocement overlay subjected to axial loads. In the analysis,
a uniform displacement of all the nodes on the top face of the column is applied to
simulate the effect of axial load. A three-dimensional analysis has been performed in
this study. Since the analysis is based on linear elastic material response, it gives
information about the nature of stress distribution and cannot be used to predict
failure.
Two types of three-dimensional fmite element analyses have been performed. One
aSsumes masonry to be a homogeneous continuum, the other considers masonry to
be an assemblage of elastic bricks and mortar joints, each with differing material
properties. In both the cases, ferrocement overlay has been modeled separately with
two component materials, viz. the mortar and the steel wire mesh as a smeared
equivalent steel thin sheet. The influence of the following parameters on the stress
distribution of ferracement coated masonry colunm is studied using the three-
dimensional fmite element analysis:
(i) Types of overlay
(ii) Thickness of ferrocement overlay
(iii) Number of mesh layers in the overlay
(iv) Modular ratio offerrocement overlay and brick masonry
(v) 111ickness of bed joint
(vi) Confinement effect offerrocement overlay
24
(vii) Eccentricity ofloading
In the comparison, emphasis has been given on the variation of transverse tensile
stresses, as these influence critically the fracture initiation of the composite column.
3.2 FINITEELEMENT METHOD
The present analytical procedure is based on the fmite element method. The
application of this method has become very common and many texts have been
written on the subject (21, 43, 96). It would thus be inappropriate to repeat the
elaborate description of the method here. The finite element method can be thought
as a general method of structural analysis by means of which the solution of a
problem in continuum mechanics may be approximated by analysing a structure
consisting of an assemblage of properly selected finite elements interconnected at a
fmite number of joints or nodal points. Recognized as one of the most versatile
methods of structural analysis, it is capable of analysing plates or solid bodies with
any irregularity in shape and physical properties. Particular advantages of this
method relating to the brickwork with ferrocement overlay are that the size, modulus
of elasticity and Poisson's ratio can be varied from element to element throughout
the composite structural element thus allowing the mortar joints and ferrocement
element to be clearly distinguished from those of bricks.
For the purpose of structural analysis, the continuum is divided by imaginary lines or
surfaces into a number of finite elements. These are assumed to be interconnected at
discrete number of nodes situated on their boundaries. The objective of the analysis,
with specified joint loading, geometry of the structure (location of joints) and
stiffness properties of the structural elements, is to fmd the resulting joint
displacements and the internal stresses in the structural elements. The size of the
elements is one of the major factors influencing the accuracy of the solution. As a
general rule, the size of the elements should be as small as possible.
25
3.3 THREE-DIMENSIONAL LINEAR ELASTIC FINITE ELEMENT MODEL
The analysis of brick columns coated with ferrocement requires three-dimensional
effects to be considered. The particular finite element computer program used for this
analysis is taken from ANSYS package (26). Eight noded three-dimensional solid
elements with three degrees of freedom at each interconnected node and linear
displacement function along their edges have been used with 2x2x2 Gaussian
integration. The details of element formulation are given in Appendix III. The finite
element idealization of the masonry column with overlay is shown in Fig. 3.1. The
computer output includes the nodal displacement, the direct stresses and shear stress
at the Gauss points as well as at the centroid of each element.
3.4 CASES ANALYSED
The Structures Laboratory at BUET is equipped with a 200 ton capacity Universal
Testing Machine. The maximum distance between platens of this machine is 1220
mm. In view of this limitation it was decided to test columns with a height of 1220
mm. The brick columns investigated were of 244 mm x 244 mm cross-section and
were made of244 mm x 116 mm x 70 mm bricks. A 25 mm thickness was used both
for ferrocement overlay and plain mortar plaster applied on the bare columns. lfthe
height of column is considered 3050 mm instead of 1220 mm no change of stress
distribution occurs, since the platen effect is confined within 100 mm to 120 mm
from the ends. This has been verified in Chapter 8 Art 8.3.11. Considering
symmetry, one fourth of the column is considered in the analysis. The typical finite
element mesh with appropriate boundary conditions used in the analysis is shown in
Fig. 3.1. Two types of three-dimensional analyses have been performed in this study.
One considers brickwork as a homogeneous continuum, the other treats bricks and
mortar-joints separately. In each case the overlay (ferro cement or plaster) was
considered separately from the brick column.
26
"..~34.9
6.1$
y
.,,~7.'~. "-
57.9
Bmm Z
cation 1
1220mm
x
FIG. 3.1 TYPICAL THREE DIMENSIONAL FINITE ELEMENT MESH(QUARTER COLUMN CROSS-SECTION)
27
Complete restraint was applied to the nodes at the base of the column while those at
the top were restrained in the horizontal direction only so as to simulate the effect of
applying the load to the column through a beam or slab. All nodes lying on vertical
lines passing through A, B, C and D excepting the top and bottom nodes were
restrained in V-direction and the nodes lying on vertical lines passing through F, G,
H and I excepting the top and bottom nodes were restrained in X-direction (Fig.
3.1). TIle nodes lying on vertical lines passing through E excepting the top and
bottom nodes were restrained both in X and Y direction (Fig. 3.1). In all the analyses,
the load was applied as a uniform displacement of the loading plate, simulating a
rigid loading platen. It would have been possible also to include the distributed
weight of the column. However, the effects of self weight of the small sized columns
were neglected.
3.5 METHOD OF LOAD APPLICATION
The method of load application influences the stress distribution in the column.
Depending on the stiffuess of the loading device, the transverse stress within the
column immediately beneath the load is markedly non-uniform if the loading plate is
flexible, and approximately uniform if the loading plate is stiff (the latter case
corresponds to a prescribed displacement of the loading plate). Furthermore,
depending on the frictional characteristics at the interface between the loading plate
and top of the column, the loaded surface will be either laterally restrained or
unrestrained with relation to the loading plate.
The influence of these parameters was investigated on the column described in
section 3.4. At first the column was loaded by subjecting the plate to a uniform
displacement (rigid loading plate) and then to a uniform load (flexible loading plate).
For each of the loading cases, two different analyses were performed - one with the
28
top and bottom nodes restrained laterally; the other with no restraint against lateral
expansion.
The distribution of transverse stresses, considering top and bottom of the column
restrained against lateral expansion, is shown in Fig. 3.2. It is seen from the figure
that with the application of uniform load instead of uniform displacement (leading to
nodal forces corresponding to same intensity of loading) there is a significant
variation of transverse stress down column centre line at location-2 (Fig. 3.2b) and
also through the centre line of ferrocement overlay (location-I), but only near the top
of the column (Fig. 3.2a).
The transverse stress distributions at centre lines of column (location-l and location-
2) both with and without lateral restraint at top and bottom of the column are shown
in Fig. 3.3. There is a significant variation of transverse stress near top and bottom of
the column at both locations (Figs. 3.3a and 3.3b).
To illustrate the difference between the two load cases (prescribed load and
prescribed displacement), variation of vertical displacement is also presented in Fig.
3.4. Obviously, the vertical displacement at every nodal point of top surface of the
column is same in case of applied uniform displacement. But in case of applied
uniform load (same magnitude as for uniform displacement case) there is a
significant difference in nodal displacements at the top of the column.
It is seen from Figs. 3.2 and 3.3 that there is a significant difference of stress.
distributions at the top and bottom of the column due to different methods of load
application. It is also seen from Fig. 3.4 that significant difference of top
displacement due to different method of load application. However, these differences
are conJined to regions near the ends, having a height approximately equal to the
width of the column.
C. L
Location-2 Ic.L.
tv\!l
----"- "-
\ '\
I
R
- II
II
II~
II
Jj
- II
TI
/I
- II
I'"77
1000
200
800
1200
400
EE~ 600:E.9''"J:
Location-I
QUARTER SECTIONOF COLUMN
~niform Load(10MPA)------1JniformDisplacement
PositivevaluesindicatesTensilestresseso ! ! i , I I ! ! I 0
NDA ,..Normal. double layer a/wire mesh and axial loadingNSA. =- Normal, single layer a/wire mesh and axial/aDdingNZA =- Norma!, zero layer a/wire mesh and axial loadingNZ4. rv "'"Normal, zero layer a/wire mesh, axial loading and weak mortarDDA ,. Discontinuous, double layer a/wire mesh and axial/aDdingDSA :0 Discontinuous, single layer a/wire mesh and axial loadingHDA = Hollow, double layer a/wire mesh and axial/aodingHSA ""Hollow, single layer a/wire mesh and axial/aDdingGDA ~ Top and bottom gop. double layer a/wire mesh and axial loadingGSA = Top and hallam gap. single layer a/wire mesh and axial loadingBM = Bare, zero layer a/wire mesh and axial loadingNDE = Normal, double layer a/wire mesh and eccentric loadingNSE = Normal, single layer a/wire mesh and eccentric loadingNZE =Normal, zerO layer a/wire mesh and eccentric loadingNZEW =Normal, zero layer a/wire mesh, eccentric loading and weak mortarBZE = Bare, zero layer a/wire mesh and eccentric loading
77
mesh respectively, was applied around the masonry column with a 25 mm gap at the
top and bottom of the masonry column. For NZA, NZE, NZA-W and NZE-W series,
plain mortar plaster was applied. For construction of hollow ferrocement column
(series HDA and HSA), a mould was fabricated using 1/16 inch thick mild steel
plates (Fig. 5.3). The mould was covered by polythene paper and the ferrocement
shell was constructed by similar procedure as of the ferrocement overlay for bare
masonry column. On the second day, the mould was taken out and the hollow
ferrocement column was available for curing.
For construction of ferrocement overlay with a longitudinal groove along the column
axis (series DDA and DSA), a steel plate of 2 mm thick, 25 mm wide and 1300 mm
long was placed vertically at the mid point of each of the four faces of the specimen.
An 8 mm thick layer of mortar was applied around the column on which one layer of
discontinuous wire mesh was wrapped on the column as shown in Fig. 5.4 and tied
it. Then another 9 mm thick layer of mortar was applied. On this mortar layer,
another layer of discontinuous wire mesh was wrapped as before, on which
additional 8 mm mortar was applied to make a 25 mm thick ferrocement overlay for
series DDA. All steel plates were removed just after the construction of ferrocement
overlay. The same procedure was adopted for single layer of wire mesh for series
DSA. In this case the wire mesh was wrapped on the column after the application of
12.5 mm thick layer of mortar as shown in Fig. 5.4. Then another 12.5 mm thick
layer of mortar was applied to make a 25 mm thick ferrocement overlay. The
specimens of series DDAIDSA are shown in Fig. 5.5.
After construction, the columns were moist cured for 14 days by wrapping all the
faces with gunny bags and air cured in the laboratory for 14 days before testing (Fig.
5.6). Demec targets were attached at mid height to monitor longitudinal strains
before testing.
78
FIG. 5.3 MOULD FOR FERROCEMENT FIG. 5.2 COLUMN DURING CONSTRUCTIONHOLLOW COLUMN
79
FIG. 5.4 COLUMN (DDAlDSA) FIG. 5.5 COLUMN (DDAfDSA)DURING CONSTRUCTION AFTER CONSTRUCTION
80
5.2.4 Testing of Columns,
All the specimens were tested with a monotonically increasing vertical load upto
failure, A typical test arrangement is shown in Fig. 5.7. A summary of the tests
performed is schematically presented in Table 5.3.
The specimens were tested using universal testing machine of 1,800 kN capacity.
The vertical load was applied through a rigid steel bearing plate located on the top of
the specimen. Plywood sheets (2x1l4" thick) were placed at the top and bottom of
the column to absorb any local irregularities at the contact surfaces. After placing the
specimen in the testing machine, vertical alignment was adjusted to eliminate any
eccentricity.
At the start of each test, a small load (20 kN) was applied and then released. The zero
readings of the Demec Gauges were then recorded. The load was then applied in
increments. Different load increments were used for different types of columns.
Loads were recorded using machine dial gauge. Deformations were measured on a
central 200 rom gauge length on opposite faces of the columns using a Demec gauge.
The readings were averaged to eliminate the bending effects. The load was applied
incrementally until the final failure occurred. The total duration of loading was about
20 minutes. The cracking load, failure load, failure pattern and stress-strain
characteristics were recorded for every specimen during testing. A summary of the
observations is contained in Tables 5.4, and 5.5 and Fig. 5.8 and Appendix V. In case
of columns subjected to eccentric loading, a steel plate (12 rom x 12 rom x 300 rom)
was placed at the top of the column. The plate was placed at a distance 'e' (e = 77
rom for coated masonry columns and e = 64 rom for bare masonry columns) from the
centroidal axis of the columns as shown in the Table 5.3. The load was then applied
on the plate.
•
81
FIG. 5.6 CURING OF COLUMN IN FIG. 5.7 THE SPECIMEN BEFORE TESTTHE LABORATORY
ConcentricLoad
(NDA,NSA, 21 1 0.0NZA,NZAW,
BZA) ,~
Concentric 6 0.75 0.0Load
(GDA,GSA)
ConcentricLoad
(HDA,HSA)
EccentricLoad
(NDE,NSE,NZE,NZEW)
EccentricLoad
(BZE)
6
12
3
1
S2
0.0
0.26
0.26
AI=Loaded area = Shaded areaAt = Total areah = total width of column
NDA =Normal, double layer of wire mesh and tuia/loadingNSA ""Normal, single layer of wire mesh and ax/a/loadingNZA ""Normal, zero layer a/wire mesh and axial loadingNZA.W '" Normal, zero layer of wire mesh, axial loading and week mortarDDA ""Discontinuous, double layer a/wire mesh and ax/alloadingDSA ""Discontinuous, single layer a/wire mesh and axial loadingHDA ""Hol/ow, double layer a/wire mesh and axial loadingHSA = Hollow, single layer a/wire mesh and axfa/loadingGDA ""Top and bottom gap. double layer a/wire mesh and axial loadingGSA ""Top and bottom gap, single fayer a/wire mesh and axiaJloadingBZ4. == Bare. zero layer of wire mesh and ax/a/loadingNDE ""Normal. double layer of wire mesh and eccentric loadingNSE '" Nonnal, single layer a/wire mesh and eccentric loadingNZE ""Normal, zero layer of •••.lrt mesh and eccentric loadingNZEW ""Normal. zero layer o/wlre mesh, eccentric loading and week morlarBZE ""Bare. zero layer 0/ wire mesh and eccentric loading
83
5.2.5 Failure Load
The failure loads of different columns are shown in Tables 5.4 and 5.5. As expected,
failure load was minimum for bare column and maximum for column with
ferrocement overlay. It may be seen from Table 5.4 that in case of axial loading the
nominal stress at ultimate load of column with ferrocement overlay is 1.94 times of
that of the bare column and 1.89 times of that of the column with plaster. This
indicates that ferrocement has a good potential to be used as an overlay for
strengthening brittle structural elements like brick columns. It may be seen from
Table 5.5 that for eccentric loads the nominal stress at failure of masonry column
with ferroccment overlay is 2.54 times of that of the masonry column with plaster. It
is also interesting to note that the effect of number of layers of wire mesh is not
significant both for axial and eccentric loading. From Table 5.4 it is seen that the
load carrying capacity of axially loaded brick columns coated with rich mortar
(cement : sand = 1:2) is 1.07 times that of the columns coated with weak mortar
(cement: sand = 1:5).
5.2.6 Modes Of Failure
The failure pattern for axially loaded ferrocement coated columns (NDA series) has
been shown in Fig. 5.8 and the failure pattern for other series of axially loaded and
eccentrically loaded columns are given in Appendix V. It may be seen from the
figures that failure modes are different even for the same type of specimen and
loading, although they are usually of the same general form. This variation is due to
the variability in the properties of the joints, the bricks, the ferrocement and the
plaster.
In case of bare masonry columns, cracks were initiated in the vertical joints in the
region near mid-height of the columns as shown in Fig. A.V.1. In.most of the cases
local failure occurred near the ends of the specimens due to platen effect. The failure
84
Table 5.4 Experimental Cracking and Failure Loads(Axial Loading)
h = total width of columnNDE Normal, double layer of wire mesh and eccentric loadingNSE Normal, single layer of wire mesh and eccentric loadingNZE = Normal, zero layer of wire mesh and eccentric loadingNZEW = Normal, zero layer of wire mesh, eccentric loading and weak mortarBZE = Bare. zero layer of wire mesh and eccentric loading
86
of columns with ferrocement overlay occurred slowly, whereas failure of the bare
column or columns coated with plaster took place suddenly. The ultimate failure of
ferrocement coated masonry columns occurred mainly due to the formation and
propagation of a few dominant vertical cracks at the centre of the column, (Figs. 5.8
andA.V.2).
In case of columns with plaster the crack propagation was very rapid. In most of the
cases the cracks widened very quickly after their formation and spalling occurred
very rapidly near the ultimate load (Figs. A.V,3 and A.VA). In case of ferrocement
hollow columns, the fracture process was mainly confined at the ends of the
specimen. This can be attributed to the restraint provided by the platens. The spalling
failure normally occurred in this case due to high out-of-plane tensile stress, (Figs.
A.V.5 and A.v.6). The failure pattern for columns with discontinuous ferrocement
overlay (series DDA and DSA), which were tested mainly to see the confinement
effect of ferrocement overlay, is shown in Figs. A.V.7 and A.v.8. In most of the
cases the gap at the discontinuity widened gradually and the ferrocement overlay
separated totally from the masonry column near the ultimate load. No cracks were
found in the four separate pieces of the ferrocement overlay. The failure pattern of
columns from series GDA and GSA are the same as that of columns from series,
NDA and NSA (Figs. A.V.9 and A.V.lO).
The failure patterns of each column subjected to eccentric loading are shown in Fig.
A.V.l1 to Fig. A.V.15. The type of failure for the compression face of all the
specimens is the same as that for the concentric load case. No major crack appeared
at the tension face of the specimens.
87
,Y-"
I!\ 3
3 £ ,- .; . ,
I'~ I3 . ',~( , n• ~~~ u,. ~~ n
\
u.~. u.
8~ f
/i
,0° I i'rt: \ ' ',
,
.: ';.•'! I I " . , ,. !~:<'I. ' ..
FIG, 5.8 FAlLURE OF COLUMN COATED WITH FERRO CEMENT SERJES NDA(AXIAL LOADING)
88
5.2.7 Stress-strain Characteristics
The stress-strain curves for all the columns are shown in Fig. 5.9. Due to the nature
of the testing machine the falling branch (unloading portion) of the stress-strain
curves in these cases could not be obtained. From Fig. 5.9 it can be seen that all the
columns, with or without overlay, show a distinct non-linear stress-strain response
over almost the entire loading range, while the hollow ferrocement columns
maintains almost a linear behaviour upto the ultimate load. From the figure it can be
seen that the stiffuess of the columns with ferrocement overlay is higher than the
bare column. The columns with different wire mesh layers showed similar stress7
strain behaviour as can be seen from the figure. It may also be seen that the stiffuess
of masonry column coated with rich mortar (cement: sand = 1:2) is higher than the
masonry columns coated with normal mortar (cement: sand = 1:5).
5.3 CONFINEMENT EFFECT OF FERRO CEMENT OVERLAY ON MASONRY CORE
The behaviour of masonry in direct compression is improved if it is subjected to
compressive stresses in the transverse direction, resulting from confmement. When a
column is subjected to an axial load it is compressed in the vertical direction and
tends to expand in the lateral direction due to Poisson's effect. However, the
ferrocement overlay opposes the lateral expansion and imposes compressive stress
on the core. The structural behaviour of the column changes due to this confIDing
effect.
To obtain the strength increase of masonry column due to confinement effect of
ferrocement overlay, two types of investigations have been performed. In the first
case a vertical groove in the ferrocement overlay at each face of the column was
provided to discontinue the ferrocement overlay (Fig. 3.12) and hence to eliminate
the confmement effect of overlay. The difference in load carrying capacity of the
ferrocement encased masonry column and the column with discontinuous, ~- '~
FIG. 5.9 EXPERIMENTAL STRESS-STRAIN CURVES FOR DIFFERENT COLUMNS
90
ferrocement overlay was considered to be the increase of load due to confinement
effect in this case. In the second case the constituent members of the composite
system (ferrocement hollow column and bare masonry column) were tested
individually to failure. The difference in load carrying capacity of the ferrocement
encased masonry column and the total capacity of the individual members was
considered to be the increase ofload due to confinement effect in this case.
Table 5.6 Load Increase in Failure Due to Confinement Effect(Derived from Experiment)
NDA 769
DDA 653NSA 769
DSA 635NDA 769
BZA+HDA 727NSA 769
BZA+HSA 706
116
134
42
63
18
21
6
9
NDA = Normal, double layer of wire mesh and axial loadingNSA =Normal, single layer of wire mesh and axial loadingDDA =Discontinuous, double layer of wire mesh and axial loadingDSA = Discontinuous, single layer of wire mesh and axial loadingHDA = Hollow, double layer of wire mesh and axial loadingHSA =Hollow, single layer of wire mesh and axial loadingBZA = Bare, zero layer of wire mesh and axial loading
From the test results (Table 5.6) it can be seen that the failure load of ferrocement
encased column (specimen series NDA) is 18% higher than that of column encased
91
with discontinuous ferrocement overlay (specimen series DDA). Failure load of
ferrocement encased column (specimen series NSA) is 21% higher than that of
column encased in discontinuous ferrocement overlay (specimen series DSA). It can
also be seen from Table 5.6 that the failure load of column encased in ferracement
(specimen series NDA) is 6% higher than that of the sum of failure load of bare
column (specimen series BZA) and hollow ferrocement shell (specimen series
HDA). Failure load of column encased in ferrocement (specimen series NSA) is 9%
higher than that of the sum offailure load of bare column (specimen series BZA) and
hollow ferrocement shell (specimen series HSA). From this study it can be
conCluded that there is a confmement effect on the masonry column due to the
provision offerrocement overlay.
It is found that introduction of vertical grooves separating the overlay along the
centre line of the four faces of the columns represents a case where the overlay
consists of four separate angle sections with no transfer of stresses from one angle to
the other; this does not represent a hollow ferrocement section where there would be
sOmedirect stresses in the lateral direction. Although the failure pattern of the hollow
sections are different from those in the composite column, it appears that the second
case gives a better indication of the confinement effect.
5.4 INFLUENCE OF OVERLAY ON COST OF COLUMNS
The material cost of columns with different types of overlay has been shown in Table
5.7. It can be seen from the table that with the increase of 46% cross-sectional area of
bare masonry column (BZA/BZE) due to the application of different types of
overlay, the material cost increases by 233% for column with ferrocement overlay
containing double layer of wire mesh (NDA/NDE), 164% for column with
ferrocement overlay containing single layer of wire mesh (NSAlNSE), 48% for
column with plaster (I :5) (NZAWINZEW), and 95% for column.with plaster (1:2)
(NZAlNZE). The results of Table 5.7 show that, the material cost increases 124%
92
due to ferrocement overlay containing double layer of wire mesh (NDAlNDE)
The procurement costs have been calculated on the basis of prices of bricks, sand, cementand wire mesh used in the test specimen.
NDA = Normal, double layer a/wire mesh and axial loadingNSA = Normal, single layer a/wire mesh and axial loadingNZA = Normal, zero layer a/wire mesh and axial loading, i.e. piane morfarNZAW = Normal, zero layer a/wire mesh, axial loading and weak morfarBZA = Bare, zero layer a/wire mesh and axial loadingNDE = Normal, double layer 0/wire mesh and eccentric loadingNSE = Normal, single layer 0/wire mesh and eccentric loadingNZE = Normal, zero layer a/wire mesh and eccentric loadingNZEW = Normal, zero layer 0/wire mesh, eccentric loading and weak morfarBZE = Bare, zero layer 0/wire mesh and eccentric loading
93
It is also seen from Table 5.7 that, the material cost increases 78% due to
ferrocement overlay containing single layer of wire mesh (NSAlNSE) instead of
plaster (1:5) (NZAWINZEW). From Table 5.7 it is seen that, the material cost of
column series NDAlNDE, series NSAlNSE, series NZAWINZEW and series
(NZAlNZE) increases in Tk. 745, Tk. 525, Tk. 155 and Tk. 305 over bare masonry
column (BZNBZE). It is also seen from table that, the material cost of column series
NDAlNDE, series NSAlNSE, and series (NZAlNZE) increases in Tk. 590, Tk. 370,
and Tk. 150 over masonry column with plaster (1:5) (series NZAWINZEW).
5.5 SUMMARY
The experimental investigation of columns subjected to axial and eccentric loading
has been described in this chapter. The types of ferrocement overlay were changed to
observe the variations in the behaviour and failure modes. The following conclusions
can be drawn from this experimental study.
1. The application of fetrocement overlay on bare brick masonry column
enhances the load carrying capacity both for axial and eccentric loadings.,With the increase of 46% croSs sectional area of column due to application of
ferrocement overlay, the average increase in failure load is found to be 184%
for axial loading and 158% for eccentric loading. The average increase in
failure nominal stress is found to be 95% for axial loading
2. With the increase of 46% cross-sectional area of bare masonry column due
the application of ferrocement overlay, the material cost increases by 233%
for column with ferrocement overlay containing double layers of wire mesh
and 164% for column with ferrocement overlay containing single layer of
wire mesh.h.
94
3. The application of plaster with a weak mortar (1:5) over bare masonry
column increases the failure load by 50% and increases the nominal stress at
failure only by 2.8%.
4. The use of plaster with rich mortar (1:2) instead of weak mortar (1:5) for
brick columns does not increase the load carrying capacity significantly. The
increase in strength (compared to a column encased in weak mortar) was
approximately 7%.
5. The cracking resistance of bare columns is improved due to the provision of
ferrocement overlay. The average increase in cracking nominal stress is found
to be about 200% for axial loading
6. The failure of bare column and column coated with plaster is sudden and
crack widths increase very rapidly after their formation, leading to brittle
failure for the system, whereas the failure of columns with ferrocement
overlay Occurred slowly.
7. There is a confmement effect on bare column due to the provision of
ferrocement overlay and the effect of confinement varies from 6% to 9%.
CHAPTER 6
NONLINEAR FINITE ELEMENT ANALYSIS
6.1 INTRODUCTION
With the advent of sophisticated numerical tools for analysis like the finite element
method (FEM), it has become possible to model the complex behaviour of brick
masonry column encased in ferrocement and plaster with appropriate constitutive
relationships. In this study all the columns were analysed using the finite element
method (FEM) available in ANSYS package. The objective of this chapter is to carry
out the fInite element analysis of the columns in order to obtain a better understanding of
the stress redistribution and progressive cracking of the ferrocement and the masonry.
6.2 ANALYSIS OF THE COLUMNS
In order to simulate the actual behaviour of columns, it is essential that the three-
dimensional nature of stress distribution is recognized. A three dimensional fInite
element analysis has, therefore, been performed for the columns mentioned in the
previous chapter. As mentioned earlier the finite element computer program used for
this analysis is taken from ANSYS package.
The constituent materials were assumed to be isotropic and a perfect bond is assumed at
the interface between the brick and the mortar joint, the ferrocement and the brick. In
case of ferrocement, the wire meshes were assumed to be smeared throughout the
element. The brickwork has been considered as nonhomogeneous continuum Le. bricks
and mortar joints were treated separately.
In the analyses the prescribed displacement has been applied at the nodes of the interface
between the machine platen and the specimen to simulate the load applied from bearing
96
plate of the machine. To simulate the friction between column and the rigid plate the
nodes were restrained along the horizontal directions (X and Y). The nodes at the base
of the column were restrained along X, Y and Z directions and the nodes at the plane of
symmetry were assigned appropriate restraint as shown in Figs. 3.1 and 3.12 for the
axial loading case. In case of eccentric loading the nodes at the base of the column were
restrained in X, Y and Z directions whereas the nodes at the plane of symmetry were
restrained in X direction only as shown in Fig. 3.14.
6.2.1 Material Deformation Characteristics and Failure Criteria
The nonlinear behaviour of brick masonry is caused mainly due to plasticity, cracking
and crushing type of fracturing. For modeling plasticity the Besseling model, also called
the sublayer or overlay model has been adopted in this analysis. Fig. A.III.3 and Table
A.III.l illustrates typical stress-strain curve and data input is demonstrated by an
example. The material model can predict elastoplastic behaviour through to fracture of
the constituent materia!. To predict cracking or crushing type of failure the stress-strain
matrix is adjusted as discussed in Appendix III. The failure criterion of William and
Warnke (97) has been used in this analysis and is discussed in Appendix III.
6.3 METHOD OF LOAD APPLICATION
The load was applied in the form ofincrementai prescribed displacements of the loading
plate for axial load case only. For the eccentric loading case, distributed load was
applied. The corresponding displacement/load increment varied from test to test. The
displacement/load increment was maintained constant until the cracks initiated after
which a smaller increment of displacement/load was considered. This procedure of
applying small increment of displacement/load after the initiation of fracture, allows
faster convergence Ofthe solution and more cracks to propagate in a narrow band.
97
6.4 FAILURE MODES
It is well known that in a vertically loaded masonry column tensile cracking type of
fracture dominates the whole fracture process. Transverse stresses are mainly
responsible for influencing the ultimate failure load of the column. The crack in this case
will start from the vertical mortar joints and then propagate through the bed joints, other
header joints and the bricks. The failure pattern for each case of axial loading has been
shown in figures 6.1, 6.2 and Appendix VI. It should be pointed out that in case of
ferrocement coated columns (series NDA and NSA) crack initiated from the vertical
joints but the propagation of cracks was delayed to some extent. All the columns
continued to sustain further load until the crack or cracks propagated through a
substantial portion of the column. The failure pattern of columns NDA and NSA are
shown in Figs. 6.1 and 6.2. In case of columns with plaster (series NZA and NZA W) the
crack propagation was very rapid. In these cases the cracks propagated very quickly
after their formation and failure occurred shortly afterwards. The failure patterns are
shown in Figs. A.VI.I and A.VI.2 respectively. In case of masonry columns coated with
discontinuous ferrocement overlay (series DDA and DSA) cracks initiated from the
vertical joints and then propagated through the bed joints and the bricks. In this case no
crack appeared in the ferrocement overlay as shown in Figs. A.VI.3 and A.VIA. In case
of ferrocement hollow column (series HDA and HSA) the fracture process is mainly
confmed at the ends of the column. This is due to the friction developed at the interface
between the specimen and the machine platen. The failure patterns at ultimate load for
column series HDA and HSA are shown in Figs. A.VI.5 and A.VI.6 respectively. The
failure patterns for columns series GDA and GSA (Figs. A.VI.7 and A.VI.8) are very
similar to the failure patterns of column series NDA and NSA. In case of bare masonry
columns (series BZA) the cracks initiated in the vertical joints near the top of the column
and then propagated through the bed joints and the bricks. The failure pattern for this
case is shown in Fig. A.VI.9. The failure patterns for each case of eccentric loading have
been shown in Figs. A.VI.!O, A.VI.!I, A.VI.12, A.VI.13 and A.VI.14. For eccentrically
98 ,
•
.,
Outer Ferrocement Shell Centra! Masonry Core
• Cracked Blement
Mortar Joint
FIG. 6.1 PREDICTED FAll.URE MODE OF COLUMN SERIES NDA
99
Outer Ferrocement Shell Central Masonry Core
• Cracked Element
••••
Mortar Joint
FIG. 6.2 PREDICTED FAILURE MODE OF COLUMN SERIES NSA
100
loaded columns failure occurs near the top of the column as shown in figures from
A.VI.I0 to A.VI.14.
6.5 FAILURE LOAD
In this analysis complete failure of the columns is assumed when the solution failed to
converge. The failure load of columns (both axial and eccentric loading) with different
types of overlay has been shown in Table 6.1 and 6.2.
o~~SJ[.~'~.@-;~~~~~..- ..~[_.=~Table 6.1 Analytical Cracking and Failure LoadAxi
Load(kN) Nominal Load(kN) Nominalstress a stress a
NDA 354 4.07 788 9.06
NSA 346 3.98 725 8.34
NZA 343 3.94 525 6.04
NZAW 314 3.61 468 5.38
DDA 318 3.65 639 7.35
DSA 314 3.61 617 7.01
HDA 301 10.99 396 14.46
HSA 279 10.19 365 13.33
GDA 256 2.94 550 6.32
GSA 255 2.93 528 6.07
BZA 250 4.19 324 5.44
NDA ~ Normal, double layer a/wire mesh and axiallaadingNSA = Normal, single layer a/wire mesh and axial loadingNZA =Normal, zero layer a/wire mesh and axial loadingNZA W =: Normal, zero layer of wire mesh, axial loading and week mortarDDA = Discontinuous, double layer a/wire mesh and axial loadingDSA = Discontinuous, single layer a/wire mesh and axial/oatlingHDA = Hollow, double layer a/wire mesh and axiallaadingHSA = Hollow, single layer a/wire mesh and axial loadingGDA = Top and hollom gap, double layero/wire mesh and axiallaadingGSA ~ Top and bollom gap, single layer o/wire mesh and axial loadingBZ4. = Bare, zero layer a/wire mesh and axial loading
101
As expected the failure load was minimum for unplastered masonry column and
maximum for column coated with ferrocement overlay. From Table 6.1 it can be seen
that in case of axial loading the ultimate load carrying capacity of masonry column with
ferrocement overlay (series NDA) is about 1.6 times than the corresponding column
coated with plaster (series NZAW). This indicates that the ferrocement overlay can be
used to strengthen the brittle structural element like brick columns. In case of eccentric
loading it can be seen from Table 6.2 that the ultimate load carrying capacity of masonry
column with ferrocement overlay (series NZE) is about 1.9 times the masonry column
coated with plaster. From Tables 6.1 and 6.2 it can be also seen that there is no
appreciable increase of ultimate load due to the provision of double layers of mesh
instead of single layer of wire mesh in ferrocement overlay for both axial and eccentric
load cases. This has also been observed experimentally as mentioned in Chapter 5.
Table 6.2 Analytical Cracking and Failure Loads(Eccentric Loading)
NDE 335 2.15 9.85 481 3.09 14.14 77 0.26
NSE 335 2.15 9.85 454 2.92 13.35 77 0.26
NZE 252 1.62 7.41 434 2.78 12.76 77 0.26
NZEW 125 .80 3.67 250 1.60 7.35 77 0.26
BZE 71 .66 3.05 185 1.72 7.95 64 0.26
h = total width of columnNDE =Normal, double layer of wire mesh and eccentric loadingNSE =Normal, single layer of wire mesh and eccentric loadingNZE =Normal, zero layer of wire mesh and eccentric loadingNZEW =Normal, zero layer of wire mesh, eccentric loading and weak mortarBZE =Bare, zero layer of wire mesh and eccentric loading
102
6.6 CONFINEMENT EFFECT OF FERROCEMENT OVERLAY
A nonlinear fInite element analysis has been performed to see the confInement effect of
ferrocement overlay on masonry column. In this case the lateral continuity of the overlay
has been disrupted by providing longitudinal groove of 6.35 mm wide through the
height of the column. It is assumed that the difference in load carrying capacity of this
composite column and the laterally confIned composite column is mainly due to the
confInement effect of ferrocement overlay. From the results shown in Table 6.3, the
failure load of masonry columns with ferrocement overlay (series NDA and NSA) is
higher than that of masonry columns with discontinuous ferrocement overlay (series
DDA and DSA). It can also be seen from Table 6.3 that the sum of the failure load of
bare masonry column (series BZA) and hollow ferrocement shells (series HDA and
HSA) is lower than that of ferrocement encased masonry columns (series NDA and
NSA).
From the results (Table 6.3) it can be seen that the failure load of ferrocement encased
column (specimen series NDA) is 23% higher than that of column encased with
discontinuous ferrocement overlay (series DDA). Failure load of ferrocement encased
column (series NSA) is 17% higher than that of column encased in discontinuous
ferrocement overlay (series DSA). It can also be seen from Table 6.3 that the failure
load of column encased in ferrocement overlay (series NDA) is 9% higher than that of
the sum of failure load of bare column (series BZA) and hollow ferrocement shell (series
HDA). Failure load of column encased in ferrocement overlay (series NSA) is 5%
higher than that of the sum of failure load of bare column (series BZA) and hollow
ferrocement shell (series HSA). From this study it can be concluded that there is a
confInement effect on the masonry column due to the provision of ferrocement overlay;
however, this is not very signifIcant and varies from 5% to 9%.
103
Table 6.3 Load Increase in Failure Due to Confinement Effecterived from Anal sis
NDA 788
DDA 639
NSA 725
DSA 617
NDA 788
BZA+HDA 720
NSA 725
BZA+HSA 689
149
108
68
36
23
17
9
5
NDA =Normal, double layer a/wire mesh cind axial loadingNSA =Normal, single layer a/wire mesh and axial loadingDDA = Discontinuous, double layer o/wire mesh and axial loadingDSA =Discontinuous, single layer a/wire mesh and axial loadingHDA =Hollow, double layer a/wire mesh and axial loadingHSA = Hollow, single layer o/wire mesh and axial loadingBZA = Bare, zero layer o/wire mesh and axial loading
6.7 STRESS-STRAIN CHARACTERISTICS
The stress-strain curves for axial and eccentric loading cases have been provided in Figs.
6.3 and 6.4 respectively. From these figures, it can be seen that all the columns (with or
without overlay) show a distinct nonlinear stress-strain curve over almost the entire
loading range. The nominal E value of hollow ferrocement columns is higher than all
other columns. The stiffness of the columns with ferrocement overlay is higher than that
of the bare columns and columns coated with plaster. From Figs ..6.3 and 6.4 it can be
NDA = Normal, double layer o/wire mesh and axialloodingNSA '" Normal, single layer o/wire mesh and axialloodingNZA = Normal, zero layer o/wire mesh and axialloodingNZAW = Normal, zero layer o/wire mesh, axiallooding and weak mortarDDA = Discontinuous, double layer o/wire mesh and axialloodingDSA = Discontinuous, single layer o/wire mesh and axialloodingHDA = Hollow, double layer 0/wire mesh and axialloodingHSA = Hollow, single layer o/wire mesh and axialloodingGDA = Top and bottom gap, double layer o/wire mesh and axialloodingGSA = Top and bottom gap, single layer o/wire mesh and axiallooding1JZA = Bare, zero layer o/wire mesh and axiallooding
110
load is from -40% to 94% in comparison with the analytical cracking load in case of
axial loading and the variation of experimental cracking load is from -23% to 42% in
comparison with the analytical cracking load in case of eccentric loading. It appears
from the above discussion that the analytical results do not agree with the
experimental results. This is because the analytical cracking loads are directly related
to the strength of the vertical mortar joints, since the fIrst crack always formed in
these elements.
Table 7.2 Analytical and Experimental Cracking and Failure Loads(Eccentric Loading)
NDE 477 335 1.42 522 481 1.08
NSE 427 335 1.27 517 454 1.13
NZE 300 252 1.19 417 434 0.96
NZEW 97 125 0.77 205 250 0.82
BZE 122 71 1.71 218 185 1.17
NDE = Normal, double layer o/wire mesh and eccentric loodingNSE =Normal, single layer o/wire mesh and eccentric IOodingNZE =Normal, zerO layer o/wire mesh and eccentric loodingNZEW =Normal, zero layer o/wire mesh, eccentric looding and weak mortarBZE ""Bare, zero layer o/wire mesh and eccentric loading
The fIrst cracking in the analytical investigation always occurs in the vertical mortar
joints inside the columns. Obviously these cracks are not visible from the outside.
The experimental cracking load represents the load at which the fIrst crack is visible
on the exterior surface of the column. The exact experimental iIiitial cracking load
111
corresponding to the failure of the vertical joints could not be observed due to the
presence of plaster or ferrocement overlay. In the analytical investigation, cracks
occurred in the vertical joints long before the cracks were detected experimentally in
the plaster or ferrocement overlay by naked eyes. In the experimental investigation
local cracks appeared in the surface of the columns (series NZA, NZAW, BZA and
NZEW) long before the cracks were occurred analytically in the vertical joints.
Analytical model assumes uniform properties of the constituent materials throughout
the whole column. However, the material properties may vary within the actual
column. The fIrst initial crack may be in a location which has the weakest mortar.
For the columns encased in discontinuous ferrocement overlay (series DDA and
DSA) and for the hollow ferrocement column (series HDA and HSA), the initial
cracking load could not be measured experimentally because of sudden failure.
7.3 FAILURE LOAD
The analytical failure load of the columns was obtained from fInite element analysis.
Tables 7.1 and 7.2 show the comparisons between the analytical results and the
experimental results for axial and eccentric loading respectively.
It can be seen from Table 7.1 that the experimental failure loads were 0.97 and 1.06
times the analytical failure loads in case of column coated with ferrocement overlay
containing double and single layer Ofwire mesh respectively. From Table 7.1 it can
be seen that the experimental failure loads of all other columns coated with
ferrocement overlay (specimen series DDA, DSA, GDA and GSA) were 1.02, 1.03,
1.13 and 1.13 times the analytical failure loads. It is noted from these comparisons
that analytical results compare reasonably well with the experiment. It can also be
seen from Table 7.1 that the experimental failure loads in case of bare columns and
columns coated with plaster were 0.83, 0.83 and 0.87 times the analytical failure
loads (specimen series BZA, NZA and NZAW). These variations occur due to
inadequacy of the bond failure criterion of the material model available in ANSYS
112
package. In case of eccentrically loaded ferrocement encased masonry column it can
be seen from Table 7.2 that the experimental failure loads were 1.08 and 1.13 times
the analytical failure loads (specimen series NDE and NSE). It can also be seen from
Table 7.2 that the experimental failure loads in case of columns coated with plaster
were 0.96 and 0.82 times the analytical failure loads (specimen series NZE and
NZEW). From Table 7.2 it can be seen that the experimental load in case of bare
masonry column was 1.17 times the analytical failure load (specimen series BZE).
This variation may occur due to non-inclusion of bond failure criterion in the model
used in FEA. However, further investigation is required incorporating the appropriate
bond failure criterion in the ANSYS package. Overall the variation of failure loads in
case of eccentric load is almost similar to that of axial load. The variation of
experimental failure load is from -17% to 19% in comparison with the analytical
failure load in case of axial loading and the variation of experimental failure load is
from -18% to 17% in comparison with the analytical failure load in case of eccentric
loading. From the comparison of experimental loads with finite element prediction,
it is concluded here that the agreement between the finite element analysis and
experiment is reasonable.
7.4 FAILuRE PATTERN
Final cracking patterns rather than the sequence of cracking are compared in this
study, since it was not possible to record the experimental cracking sequence in
many cases. Comparisons between experimental and analytical failure patterns are
presented for different cases (Fig. 7.1 and Appendix Vll). Only failure pattern of
peripheral elements are presented. From these figures it can be seen that there is
reasonably a good agreement between the experimental failure patterns and the
failure patterns predicted by the finite element analysis.
Strain (10")FIG. 7.2 STRESS-STRAIN CURVE OF COLUMN SERIES NDA
Strain (1(y')
FIG. 7.3 STRESS-StRAIN CURVE OF COLUMN SERIES NSA
115
7.5 STRESS-STRAIN CURVES
Comparisons are made between the stress-strain curves of different columns obtained
from experiment and those obtained from finite element analysis. The comparison
for NDA and NSA series is shown in Figs. 7.2 and 7.3 and the comparison for other
cases is shown in the Appendix vn. It is seen from Figs. 7.2, 7.3 and Appendix vnthat the value of "modulus of elasticity" for the composite column derived from
analytical curves are higher than those obtained from the experimental curves for
almost all the columns. This is due to the fact that the stiffuess of the elements in the
finite element model is higher than the actual stiffuess of the structure. In case of
columns NZAW and BZA (Fig. A.Vll.17 and Fig. A.Vll.24) there is a considerable
difference between the two curves. This variation occurs due to the non-
incorporation of bond failure criterion in the finite element analyses. However, from
the comparison of experimental and analytical stress-strain curves, it can be
concluded that in general, the agreement between fmite element analyses and
experiment is reasonably good.
7.6 CONFINEMENT EFFECT
To examine the confmement effect of ferrocement overlay, both experimental and
analytical investigations have been made in this study. Two different cases were
considered to eliminate the confinement effect of ferrocement overlay on masonry
column as mentioned in Chapter 5. As outlined in Art. 5.3, in the fIrst case a vertical
groove in the ferrocement overlay at each face of the column was provided to
discontinue the ferrocement overlay (Fig. 3.12) and hence to totally eliminate the
capability of overlay to resist direct stresses in the lateral direction. The difference in
load carrying capacity of the ferrocement encased masonry column and the column
with discontinuous ferrocement overlay was considered to be the increase of load
due to confmement effect. In the second case the constituent members of the
116
composite system (ferrocement hollow column and bare masonry column) were
tested individually to failure. The difference in load carrying capacity of the
ferrocement encased masonry column and the total capacity of the individual
members was considered to be the increase of load due to confinement effect in this
case. The results for the cases are presented in Table 7.3.
Table 7.3 Load Increase in Failure Due to Confinement Effect
NDA 769 788116 149
DDA 653 (17%) 639 (23%)NSA 769 725
134 108DSA 635 (21%) 617 (17%)NDA 769 788
42 68BZA+HDA 727 (6%) 720 (9%)NSA 769 725
63 36BZA+HSA 706 (9%) 689 (5%)NDA =Normal, double layer o/wire mesh and axial loadingNSA =Normal, single layer o/wire mesh and axial loadingDDA =Discontinuous, double layer o/wire mesh and axial loadingDSA =Discontinuous, single layer o/wire mesh and axial loadingHDA =Hollow, double layer o/wire mesh and axial loadingHSA =Hollow, single layer o/wire mesh and axial loadingBZA = Bare, zero layer o/wire mesh and axial loading
From the table it is seen that the strength increase of masonry column encased in
ferrocement overlay due to the confinement effect is between 17% to 23% for the
first case (composite column with grooves in ferrocement overlay). It is also seen
117
from Table 7.3 that the strength increase of masonry column encased in ferrocement
overlay due to the effect of confinement is between 5% to 9% for the second case
(masonry column and ferrocement hollow column). As discussed in Art. 5.3, the
confinement effect determined from bare masonry column and hollow ferrocement
column has been considered to be more representative in this study.
7.7 SUMMARY
A comparison between the experiment and the finite element analysis of composite
behaviour of masonry column coated with ferrocement overlay has been presented in
this chapter. The cracking load determined from tests is usually higher since it is not
possible to observe cracks (which are usually initiated inside the columns) during
tests; cracking loads determined from tests represent the values when cracks appear
on the surface of the column. However, in case of failure loads, the agreement
between finite element analysis and experiment is good. The finite element model
available in the ANSYS package will, therefore, be used onwards for predicting
failure loads. The following conclusions can be drawn from this study.
1. The variation of experimental cracking load IS from -40% to 94% in
comparison with the analytical cracking load.
2. The variation of experimental failure load is from -17% to 19% in comparison
with the analytical failure load.
3. The experimental failure patterns and the failure patterns predicted by the
finite element analysis are almost similar.
4. The value of modulus of elasticity for the composite column derived from
analytical curves are higher than those obtained from. the experimental
curves for almost all the columns
118
5. The strength increase of masonry column encased in ferrocement overlay due
to the confmement effect is between 5% to 9% analytically and the strength
increase of masonry column encased in ferrocement overlay due to the effect
of confmement is between 6% to 9% experimentally.
CHAPTER 8
SENSITIVITY ANALYSIS OF CRITICAL PARAMETERS
8.1 INTRODUCTION
The material properties determined from experiment on individual components and
small brick masonry specimens have been incorporated in the three-dimensional
finite element analysis. The appropriateness of the material model adopted in the
analysis have already been verified by carrying out experiments on masonry columns
coated with ferrocement. Although, the results of the experiments agreed well with
the prediction of the fmite element analyses, it is also important to determine which
parameters of the material model and the fmite element analysis are particularly
significant. In this chapter a study of the parameters affecting the fracture behaviour
of column is performed by analysing the behaviour of a ferrocement coated column
subjected to axial load, distributed uniformly over the cross-section. Individual
parameters are changed in turn and the influence of each change is investigated. In
the analysis a colunm encased in ferrocement, containing double layers of wire mesh,
is investigated (column series NDA).
Two groups of parameters are considered for this sensitivity analysis, one which
affects the material model and the other which relates directly to the finite element
analysis. The parameters which directly affect the material model are,
(i) Elastic properties (viz. modulus of elasticity and Poisson's ratio) of the
constituents
(ii) Bed joint thickness
(iii) Tensile strength of the constituent materials
The proposed formulae involve seven important parameters, namely, the tensile
strength of brick, compressive strength offerrocement mortar, thickness ratio of bed
joint and brick, cross-sectional area of brick masonry, cross-sectional area of
ferrocement, eccentricity and depth of column. The agreement between the proposed
formulae and the finite element analysis for different column sizes is shown in Figs.
9.1, 9.2 and Appendix IX.
9.3 EXAMPLE PROBLEMS
To check the adequacy of the proposed design formulae, four example problems are
considered in this section. The problems have already been investigated both
analytically and experimentally before (Chapter 6 and Chapter 5). In this section only
the results are compared with those from the proposed design formulae.
144
3500
3000
381 mm x 381 mm column
25.4 mm overlay
2500
Z 20006"C
'"0..J
1500
•
508 mm x 508 mm column38.1 mm overlay
Proposed Design Fonnula
Proposed Design Formula
1000
Finite Element Analysis ---_..-500 •. =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=~~~~~-:~-=:=:=~-~p~ro~p:os:e~d~o:eS:ig~n~F:ormUla
~ 244 mm x 244 mm column19.05 mm overlay
15 16 17 18 19Compressive Strength of Ferrocement Mortar (MPa)
20
FIG.9.1 COMPARISON OF PROPOSED DESIGN FORMULA WITnFINITE I~LEMENT ANALYSIS FOR AXIAL LOAD CASE
1.0
0.9
0.8
0.7
0.6
0.5
145
-Prop05ed Design Formula-e- 381mm x 381mm column, overlay =25,<lmmn
t",tt,,=O.273-A-244m x 244 mm column, overlay:38.1mm
t,A=O.273
-O-244mm x 244mm column, overlay=19.05mm'm,"15MPa
-.- 244mm x 244mm column, overlay:19.OSmmt,A"'O.091
0.00 0.05 0.10 0.15
e/h
0.20 0.25 0.30 0.35
FIG. 9.2 COMPARISON OF PROPOSED DESIGN FORMULA WITHFOOTE ELEMENT ANALYSIS FOR ECCENTRIC LOAD CASE
146
Problem 1
A 295 nun X 295 nun column cross section is chosen as an example problem for this
case. The overlay thickness is 25.4 nun. Two layers of woven square wire mesh of
1.2 nun diameter wire and opening size 11.3 nun X 11.3 nun are used, Le. volume
fraction Vf becomes 0.014 in this case. Thickness of bricks and bed joints is 70 mm
and 6.35 mm respectively. Tensile strength of brick, compressive strength of
ferrocement mortar and yield strength of wire are 2.21 MPa, 18.5 MPa and 285 MPa
respectively. A uniformly distributed load at the top of the column is applied in this
case.
Problem 2
A 295 mm X 295 nun column with 25.4 nun thick ferrocement overlay is analysed.
One layer of woven square wire mesh of 1.2 nun diameter wire and opening size
11.3 mm X 11.3 nun is used, Le. volume fraction Vf becomes 0.007 in this case.
Thickness of bricks and bed joints was kept constant as problem NO.1. There were
no changes in the values of the parameters like tensile strength of brick, compressive
strength of ferrocement and yield strength of wire. A uniformly distributed load at
the top of the column is applied.
Problem 3
In this case the column of problem No. 1 has been considered. All the parameters
related with the strength characteristics and the deformation characteristics of the
component materials were kept constant. Only the loading type has been varied. In
this case an eccentric load with an eccentricity of 77 nun at the top of the column is
applied.
Problem 4
In this case the column of problem NO.2 has been considered. The values of all the
parameters were kept constant. Only the loading type has been varied. In this case an
eccentric load with an eccentricity of77 nun at the top of the column is applied.
'~/'.
\ \
147
From Table 9.1 it is seen that the results obtained from the proposed formulae are
. very close to the experimental and the analytical values. The failure load obtained
from experiments are on an average 7% and 20% higher than the failure loads
obtained from proposed design formulae for axial and eccentric loading
respectively. The failure load obtained from fmite element analyses are on an
average 1% and 10% higher than the failure loads obtained from proposed design
formulae for both axial and eccentric loading respectively. This indicates the
suitability of the proposed design formulae for the uniaxial compressive strength of
masonry colunm coated with ferrocement.
454
1.18
1.04
481
435 435
1.20
1.10
725
1.01
1.07
788
714 714
1.10
1.07
Table 9.1 Failure Load Obtained from Different Methods
2 3 4
59536 59536 59536 59536
27489 27489 27489 27489
1.4 0.7 1.4 0.7
0.091 0.091 0.091 0.091
2.21 2.21 2.21 2.21
18.5 18.5 18.5 18.5
0.0 0.0 77 77
769 769 522 517
148
9.4 SUMMARY
The formulation of a design procedure for ferrocement encased masonry column is
difficult due to the interaction of a large number of variables which influence the
composite behaviour of these columns. However, a review of the parametric study
discussed in Chapter 8 shows that not all the parameters have significant influence on
the failure load. The important variables are tensile strength of brick, compressive
strength of ferrocement mortar and thickness of bed joint. Considering these
important parameters two formulae have been proposed for two different loading
cases. The following conclusions can be drawn from this study.
1. For axial loading the failure load obtained from experiment is on an average
7% higher than the failure load obtained from proposed design formula.
2. For eccentric loading the failure load obtained from experiment is on an
average 20% higher than the failure load obtained from proposed design
formula.
3. For axial loading the failure load obtained from finite element analysis is 1%
higher than the failure load obtained from proposed design formula.
4. For eccentric loading the failure load obtained from finite element analysis is
10% higher than the failure load obtained from proposed design formula.
CHAPTER 10
SUMMARY AND CONCLUSIONS
10.1 SUMMARY
Linear elastic finite element analysis has been performed to establish the important
parameters which influence the behaviour of ferrocement encased columns,.subjected to vertical loads. A three-dimensional analysis was used for various types
ofloading. Two types of finite element analyses were used - one assumes masonry to
be a homogeneous continuum and the other considers masonry to be an assemblage
of bricks and joints. Two loading types were considered in this investigation - one of
uniformly distributed load at the top and the other of uniform displacement at the top.
In order to defme the material model of the fmite element analysis an extensive
investigation into the properties of brick, mortar, brick masonry and ferrocement has
been carried out. The model is microscopic in nature and considers the bricks and the
mortar joints separately. The material parameters were established from various
types of tests performed on representative samples of bricks, mortar, mortar joints,
brick masonry and ferrocement. These involve compression tests on bricks, mortar
cylinders, stack bonded prisms, hollow ferrocement block, splitting tests on brick and
tensile test on ferrocement plate.
A total of 48 columns were tested in the investigation. These included six bare
masonry columns (244 mm x 244 mm x 1220 mm), 24 bare masonry columns coated
with 25 mm thick ferrocement overlay and 12 bare masonry columns coated with
25 mm thick plaster. Six hollow columns, 1220 mm in height with ferrocement shell
thickness 25 mm and 244 mm x 244 mm inside dimension were also investigated.
The number of layers of wire mesh in ferrocement overlay, type of loading and type
of specimen were varied to produce variations in strength characteristics and'
150
deformation characteristics of the composite column. With the increase of 46% cross
sectional area of column due to addition of ferrocement overlay, the increase in
strenith is found to be 184% for axial loading and 158% for eccentric loading (with
e/h = 0.26). With the increase of 46% cross-sectional area of bare masonry column
due to the application of ferrocement overlay, the material cost increases by 233%
for column with ferrocement overlay containing double layer of wire mesh and 164%
for column with ferrocement overlay containing single layer of wire mesh. The
increase of number of layers of wire mesh in the ferrocement overlay does not
produce significant change in the failure load for axial loading. The influence Of
number of mesh layers in ferrocement overlay on the load carrying capacity of
eccentrically loaded column has also been found to be negligible.
In order to determine the increase in strength of masonry column due to confinement
effect of ferrocement overlay two series of tests were performed. In the first series a
vertical groove in the ferrocement overlay at each face of the column was provided
to discontinue the ferrocement overlay (Fig. 3.12) and hence to eliminate the
confinement effect of overlay. The difference in load carrying capacity of the
ferrocement encased masonry column and the column with discontinuous
ferrocement overlay was considered to be the increase of load due to confinement
effect in this case. In the second Series the constituent members of the composite
system (ferrocement hollow column and bare masonry column) were tested
individually to failure. The difference in load carrying capacity of the ferrocement
encased masonry column and the total capacity of the individual members was
considered to be the increase of load due to confinement effect in this case. A more
realistic assessment of the confmement effect can be made by testing the bare
masonry column and hollow ferrocement shell. The strength increase of bare
masonry column due to the confinement of ferrocement overlay is fOund to be
around 5% to 9%.
151
A comparative study of the experimental and analytical results has been carried out.
In general, the agreement between finite element analysis and experiment is good,
thus indicating that the material model used in the finite element package (viz.
ANSYS as outlined in Appendix A Ill) is suitable for the analysis of composite
behaviour of masonry column coated with ferrocement overlay. The finite element
analysis is capable of predicting the failure load and the failure pattern with
reasonable accuracy and thus considered appropriate for a comprehensive parametric
study for the design recommendation.
A sensitivity analysis of the parameters which influence the nonlinear behaviour of
the column has been carried out. This study revealed that the tensile strength of
brick, the ratio of joint thickness and brick (tm/tb) and compressive strength of
ferrocement mortar are of prime importance. The model is found to be less sensitive
to modulus of elasticity, Poisson's ratio, compressive strength of brick, the tensile
strength of mortar and number of layers of wire mesh.
From linear elastic fracture analyses of ferrocement coated columns subjected to
static axial and eccentric loading the following two formulae have been developed.
For axial load, the ultimate column load, Pult> is taken as,
Pult = 0.80 (PI + P2) ••••••••••••••••• (l0.1)
where:
PI = Axiillioad carrying capacity of the masonry core in Newton (N),
/BATCHIFILNAME,file/PREP7mTLE, BEHAVIOUR OF BRICK MASONRY COLUMNS WITHFERROCEMENT OVERLAYET,1,65KEYOPT, 1,6.3MP,EX,I,17187 *BRICKMP,NUXY,I,0.16 *BRICKMP,EX,2,2900 *MORTARMP,NUXY,2,O.2 *MORTARMP,EX,3,19000 *FERROCEMENTMP,NUXY,3,0.17 *FERROCEMENTMP,EX,4,206896 *WIREMP,NUXY,4,0.3 *WIRE/COM*************************************************************VR=0.007 *DOUBLE LAYERS OF WIRE MESH/COM VR=0.007/2 *SINGLE LAYER OF WIRE MESH/COM VR=O.O *ZERO LAYER OF WIRE MESHR,2,4,VR,90,O,4,VRRMORE,0,90R,3,4,VR,90,0,4,VRRMORE,0,90,4,VR,0,0R,4,4,VR,0,0,4,VRRMORE,O,90/COM*************************************************************TB,CONCR,lTBDATA,1,.5,.5,2.21,20.5TB,CONCR,2TBDATA,1,.5,.5,.6,4.95TB,CONCR,3TBDATA,1,.5,.5,3.75,18.5 *2.5,18.5/COM*************************************************************TB,BKIN,4TBDATA,1,285,O/COM*************************************************************TB,MKIN,lTBTEMP" STRAIN
*NUMBER OF BRICK LAYER*NUMBER OF LAYER PER BRICK*NU11BER OF LAYER PER 110RTAR LAYER*NU11BER OF ELE11ENT PER ROW*FERROCE11ENT THICKNESS*VERTICAL 110RTAR THICKNESS*BED JOINT THICKNESS
A.40
TBDATA,1,1.164E-4,2.38E-4,2.99E-4,3.61E-4,5.8E-4TBTEMP,TBDATA,1,2,4,5,6,8/C011*************************************************************TB,l\1KIN,2TBTEMP "STRAINTBDATA,l ,3045E-4,21.85E-4,55.2E-4, 121.9E-4,266E-4TBTE11P,TBDATA,1,1,2.5,4,6,8/C011*************************************************************TB,l\1KIN,3TBTE11P"STRAINTBDATA,I, 1.579E-4,6.028E-4, 1o406E-3,2.612E-3,3.84E-3/C011 TBDATA, I, 1.613E-4,3.795E-4,7.1535E-4, 13.55E-4, 19.578E-4TBTE11P,TBDATA,I,3,7,11,15,18/C011 TBDATA, 1,1,2,3,4,4.5/C011*************************************************************NL=16BL=211L=1NE=4FT=250411TV=6.3511T=6.35T11L=11TI11LLB=9.625*2504/2 *LENGTH OF BRICK/2BT=3*2504.11T *THICKNESS OF BRICKNE1 =NE+ 1 *NUMBER OF NODE PER ROWNN=NEI *NE 1 *NUMBER OF NODE PER LAYERTNN=NN*«ML+BL)*NL+ I) *TOTAL NUMBER OF NODENEL=NE*NE *NUMBER OF ELE11ENT PER LAYERTNE=NL *(BL+11L)*NEL *TOTAL NUMBER OF ELE11ENTN,lN,2"FTN,NE"FT+LB-11TVFILL,2,NEN,NE1"FT+LBNGEN,2,NEI,1 ,NEI, 1,FTNGEN,NE-I ,NE1,NEI +1,2*NEI ,1,(LB-11TV)/(NE-2)NGEN ,2,NE 1,NE 1*NE.NE,NE 1*NE, 1,11TV
(t,,/!,,)X-Sectional area of 381 381 381 381 381 381 508 508 508 508 508 508 508 508 508brick masonry (A"",) x x x x x x x x x x x x x x x(mm x mm) 381 381 381 381 381 381 508 508 508 508 508 508 508 508 508
X-Sectional area of 244 244 244 244 244 381 381 381 381 381 508 508 508 508 508brick masorny (A"",) x x x x x x x x x x x x x x x(mrnxmm) 244 244 244 244 244 381 381 381 381 381 508 508 508 508 508
X-Sectional area of 244 244 244 244 244 244 244 244 244 381 381 381 381 381 381 381 381 381 508brick masonry (Am.) x x x x x x x x x x x x x x x x x x x(mmxmnl) 244 244 244 244 244 244 244 244 244 381 381 381 381 381 381 381 381 381 508
X-Sectionalareaof 508 508 508 508 508 508 508 508 244 244 244 244 244 244 381 381 381 244 244 244brickmasomy(AtmJ x x x x x x x x x x x x x x x x x x x x_x~ _ _ _ _ _ _ _ _ ~ ~ ~ ~ ~ ~ ~I ~I ill ~ ~ ~