Structural Brickwork-Macmillan Education UK (1981)

STRUCTURAL BRICKWORK
Structural
Brickwork
Arnold W. Hendry
B.Se., Ph .D., D.Se., E LCE.,
E I.Struet.E., E R.S.E.,
Professor 0/ Civil Engineering,

University 0/ Edinburgh
M
©Arnold W. Hendry 1981
Softcover reprint ofthe hardcover 1st edition 1981 978-0-333-25748-7
All rights reserved. No part of this publication may be

reproduced or transmitted, in any form or by any means,
without permission.
First published 1981 by

THE MACMILLAN PRESS LTD
London and Basingstoke
Associated companies in Delhi Dublin
Hong Kong Johannesburg Lagos Melbourne
New York Singapore and Tokyo
Typeset in 10/11 PR by
Reproduction Drawings Ltd, Sutton, Surrey
ISBN 978-1-349-81441-1 ISBN 978-1-349-81439-8 (eBook)

DOI 10.1007/978-1-349-81439-8
CONTENTS
Preface viii
Structural design of brickwork buildings I

LI Introduction I
1.2 Wall layout 2
1.3 Plain and reinforced brickwork 4
1.4 Limit state design of brickwork 4
1.5 Derivation of partial safety factors 6
1.6 Analysis of brickwork structures 6
2 The strength of brickwork 12

2.1 Compressive strength : general 12
2.2 Factors affecting cornpressive strength 13
2.3 Failure in compression : indications from standard tests 13
2.4 Interaction of brick and bed materials 13
2.5 Failure theories based on elastic analysis 15
2.6 Failure theories based on the strength of brick and mortar
under multi-axial stress 17
2.7 Empirical studies of the compressive strength of brickwork 26
2.7.1 Brick characteristics 29
2.7.2 Effect of brickwork bond and wall type 31
2.7.3 Concentrated loads on brickwork 32
2.7.4 Chases in brickwork 33
2.8 The effect of workmanship factors on compressive strength 34
2.8.1 Incorrect proportioning and mixing of mortar 35
2.8.2 Incorrect adjustment of suction rate 36
2.8.3 Incorrect jointing procedures 40
2.8.4 Disturbance of bricks after laying 40
2.8 .5 Failure to build wall 'plumb and true to Une and level' 41
2.8.6 Failure to protect work from the weather 41
2.8 .7 Overall effects ofworkmanship on brickwork strength 43
2.9 The deformation properties of brickwork in compression 45
2.10 The strength and deformation ofbrickwork in shear
2.11 The strength of brickwork in tension 49
v
vi CONTENTS
3 The strength of brick masonry compression elements 62
3.1 Factors affecting the compressive strength of walls and piers 62
3.2 Empirical studies of the strength of walls and piers 62
3.3 Theoretical studies of the strength of compression elements 65
3.3 .1 Differential equation for brittle columns 67
3.3 .2 Solutions assuming deflection and stress-strain curves 72
3.4 Wall-floor slab interaction 72
3.4.1 Wall strength in terms of end rotation 74
3.4 .2 Wall-floor slab ioints 77
3.5 Moment magnifier method 81
3.6 Special wall types 83
3 .6.1 Cavity walls 83
3.6 .2 Stiffened walls 85
4 Design analysis of unreinforced brickwork structures 91

4.1 General 91
4.2 Vertical load analysis 91
4.2.1 Load distribution on walls 91
4 .2.2 Analytical models for vertical load analysis 93
4 .2.3 Frame analysis applied to brickwork structures 94
4.2.4 Experimental verification offrame action in brickwork
structures 102
4 .2.5 Approximate calculation of eccentricities 110
4.3 Lateral load analysis 114
4.3.1 Frame analysis for lateral loads 116
4.3 .2 Benjamin 's method for irregular wall arrays 116
5 Laterally loaded unreinforced walls 122

5.1 General p2
5.2 The strength of brickwork panels without precompression 122
5.2.1 Experimental studies 122
5.2.2 Calculation of strength of laterally loaded panels 123
5.3 Lateral strength of walls with precompression 126
5.3.1 Experimental studies 126
5.3 .2 Theoretical treatment 127
5.4 The lateral strength of infill panels 131
5.4.1 Arehing theories for strip walls 131
5.4.2 Walls supported on four sides 137
5.4.3 An approximate theory for infill panels 139
6 Reinforced and prestressed brickwork 144
6.1 The application of reinforced and prestressed brickwork 144
6.2 Reinforced brickwork flexural elements 144
6.2.1 Flexural strength ofreinforced brickwork 146
CONTENTS vii
6.2.2 Shear strength ofreinforced brickwork beams 149
6.2.3 Calculation ofdeflection ofreinforced brickwork beams 154
6.3 Reinforced brickwork compression elements 155
6.4 Reinforced brickwork shear walls 160
6.5 Prestressed brickwork 160
7 The resistance of brickwork structures to accidental damage 165

7.1 Abnormalloading incidents 165
7.2 Direct design for accidental damage 166
7.3 Indireet design for accidental damage 169
7.4 Experimental studies of accidental damage 170
8 Brick masonry walls in composite action 176

8.1 Composite wall-beam elements 176
8.1.1 Structural action ofwall-beams 176
8.1.2 Theoretical solutions 177
8.1.3 Experimental results and verification 0 f wall-beam theories 186
8.2 Infilled frames 189
8.2.1 Structural action ofinfill panels 189
8.2.2 Calculation of strength and stiffness of infilled frames 190
8.2.3 Infill panels with openings 197
Author Index 205
SubjectIndex 208
PREFACE
It is commonplace to observe that brick masonry is one of man's oldest building

materials , but the design of brickwork building structures in accordance with
modern structural engineering principles is quite new. It is true that isolated series
of tests on walls and piers were carried out in various countries in the late nine-
teenth and early twentieth centuries but it was not until around the middle of the
twentieth century that sufficient information was available to permit the prepara-
tion of codes of practice containing essential data on masonry strength and
reduction factors for slenderness and eccentricity. The early codes provided a
basis for the design of compression elements and were used on the structural
design of many high-rise buildings from the late 1950s onwards.
The application of engineering principles, however , necessarily remained in-
complete, as the theoretical basis for the design of masonry elements had not then
been evolved and analyti cal procedures were rudimentary as compared to those
applied to structural design in steel and concrete. The essential problem in masonry
design is that of the compression element of brittle materia l. The theory of brittle
columns having idealised end cond itions was therefore the first to receive att ention.
A solution was in fact produced as early as 1937 but seems to have been largely
overlooked and the main development of brittle column theory took place from
the mid-1950s with new and modified solutions still being produced.
The practical application of these theories is limited by the difficulty of relat-
ing the assumed end conditions to the actual conditions in a building where the
eccentricity of loading and the behaviour of the element are influenced by inter -
action between masonry walls and concrete floors . The codes of practice include
empirical rules which permit the designer to arrive at a practically satisfactory
result, but which have little established correlation with the real behaviour of
structures. Research work on wall-floor slab interaction was undertaken in
Scandinavia in the early 1960s and is reflected in the Swedish code which takes
into account the position of a wall relative to the top of the building in relation
to the capacity reduction factors for slenderness and eccentricity. This example
has not yet been followed in other national codes but further research work has
been carried out on the problem with a view to improving the rational foundation
for the design of brick masonry walls and columns.
The interaction between walls and floor slabs is also of importance in relation
to wind loading. This became of importance with the construction of high-rise
viü
PREFACE ix
buildings with slender walls for which accurate methods of lateral load analysis
were required . The earliest methods omitted consideration of the bending
moments in inter-connecting floors or beams and distributed the wind moment
among the wails in proportion to their stiffness. This method was extended in the
1950s to deal with torsional effects . Later , more refined methods, originally
derived for reinforced concrete shear walls, were applied to brickwork structures
and their validity was checked by large-scale tests on sections of brickwork
construction.
While these analytical methods were being developed, a more complete under-
standing of the behaviour of brickwork under compressive and shear stresses was
being promoted and, although still incomplete, has strengthened the purely
empirical basis of the design codes.
Experience has shown that brickwork structures are remarkably resistant to
accidental forces, such as gas explosion and vehicle impact, but during the late
1960s and early 1970s public concern developed as a result of a number of
building failures from such causes. A considerable amount of research work was
therefore carried out during this period to provide quantitative information on
the resistance of brickwork structures and elements to accidental damage. This
work has been carried out mainly in the United Kingdom, where the concern
seems to have been greatest.
The strength of brick panels unde r lateralloading has also attracted a great deal
of attention during the past decade . This has arisen from the requirement in a
number of countries to justify the stability of such walls against prescribed wind
loadings, which have been increased in recent years, and to arrive at wall thick-
nesses that would be economically acceptable and consistent with existing
experience . This has proved to be a difficult task and although partially solved
by extensive testing still awaits a cornprehensive theoretical solution. Laterally
loaded walls with precompression or with boundary restraint giving rise to in-plane
forces have also been investigated . These are capable of resisting very much higher
lateral pressures than the non-loadbearing panels referred to above, and in this
case theoretical treatment has proved easier to develop and is relevent to considera-
tion of resistance of brickwork structures to explosive forces.
Reinforced brickwork has been used on occasions for a great many years
especially in countries such as India and the United States , which have to contend
with earthquakes. Research work on reinforced masonry beams was carried out
early in the twentieth century, but only recently has interest in this form of
construction emerged in Western Europ e, giving rise to research programmes in
this field. It is probable that economic factors will favour this development
which is likely to extend to the application of prestressing techniques to brick-
work elements.
The whole field of research in brickwork construction has developed remark-
ably since the mid-1960s, and has been reflected by the holding of regular
conferences devoted to the subject since 1967. Many papers have also appeared
in technical journals and in the proceedings of national symposia during the
same period, so that a considerable literature exists on most of the problems out-
lined above. The object of this book is to review existing knowledge of the
x PREFACE
structural engineering aspects of brick masonry construction . It is hoped that
this will be of service to structural engineers by enabling them to extend their
knowledge of the subject beyond the limits of codes of practice, and that it will
help to place the structural design of brick masonry on a scientific and technical
basis comparable to that of steel and concrete.
I acknowledge my debt to the numerous authors whose work I have quoted .
I am particularly indebted to my colleagues and research students in Edinburgh
University who, over the years, have contributed greatly to our practical and
theoretical knowledge of brickwork. Finally, I express my thanks to the brick
industry in the United Kingdom for their support of my research work and to
my many friends in the masonry world for their comradeship and for their
encouragement in pursuing my studies of brickwork construction.
ARNOLD W. HENDRY
1 STRUCTURAL DESIGN OF
BRICKWORK BUILDINGS
1.1 Introduction
A large proportion of brickwork buildings for residential and other purposes are
satisfactorily designed and built in accordance with empirical rules.and practices
without the need for special structural consideration. However, the limits of this
approach cannot be extended much beyond the scale of two-storey houses of
very conventional construction without having to use very thick walls, which in
turn result in waste of materials and other disadvantages. Indeed for a considerable
time this led to the eclipse of masonry as a structural material for larger buildings,
and it is only since the 1950s that the application of structural engineering
principles to the design of brickwork has resulted in the re-adoption of this
material for certain classes of mult i-storey buildings , and to its use in situations
which would have been precluded by reliance on rule of thumb procedures.
The economic success of brickwork construction has been achieved not only
by the rationalisation of structural design, but also because it is possible for the
walls which comprise a brick building structure to perform several functions
which in a framed structure have to be provided for separatel y . Thus , brick walls
simultaneously provide structure, sub-division of space, thermal and acoustic
insulation, as weIl as fire and weather protection. The material is relatively cheap
and durable, can provide infinite flexibility in plan form and offer an attractive
external appearance. Furthermore , brickwork buildings can be constructed with-
out heavy capital expenditure on the part of the builder.
To make the best use of these inherent advantages it is necessary to apply
brickwork construction in cases where the accommodation gives rise to moderate
or small floor spans and where it is possible to continue the loadbearing walls
uninterrupted from foundations to roof. In some buildings where there has been
a need for large spans on the first and second floors (for example, in hotels), these
floors have been built in framed construction with loadbearing walls above this
level. It is likely, however, that with the development of reinforced brickwork this
limitation will be removed and that the entire structure will be of brick masonry .
Types of multi -storey building compatible with the adoption of brick masonry
structures include hotels, hostels , flats and other residential buildings, but
engineered brickwork is frequently advantageous in low-rise buildings where its
use can, for example, reduce wall and column sizes and thus increase the flexibility
of the design while retaining the advantages of the material.
2 STRUCTURAL BRICKWORK
1.2 Wall layout
The first consideration in the design of a brickwork building is to determine the
plan arrangement of the walls in accordance with the function of the building .
From the structural point of view the wall arrangement is important: firstly , as a
means of providing lateral strength and rigidity, and secondly, in order to ensure
that the building is generally robust in the sense that any local damage to the
structure does not result in catastrophic collapse.
Possible wall arrangements are almost unlimited but it may be helpful to
distinguish three basic categories
(1) Cellular
(2) Simple or double cross-wall systems
(3) Complex service core structures.
A cellular arrangement is one in which both internal and external walls are
loadbearing and in which these walls form a cellular pattern in plan as indicated
in figure 1.1.
Figure 1.1 Cellular wall layout
(a) Simple cross-wall structure

(b) Double cross-wall system
Figure 1.2 Cross-wall structure
The second category includes cross-wall structures of various kinds . These may
be used either for slab or point blocks: in the case of slabs, (figure 1.2a)
longitudinal stability is frequently achieved by means of internal corridor walls
or partition walls. Buildings have been designed in which longitudinal stability
depends on a stair-well or lift-shaft somewhere along the length of the structure,
STRUCTURAL DESIGN OF BRICKWORK BUILDlNG 3
but this is unsatisfactory in that failure of one wall could precipitate collapse of a
large part of the building ; this practice should therefore be avoided even if it is
not precluded by the requirements of the Building Regulations.
Point blocks require double cross-wall systems as suggested in figure 1.2b.
In both of the above systems the external walls may be in masonry , curtain
wall or indeed any other external walling at the choice of the architect , thus
giving considerable freedom of elevational treatment. It will be observed, however ,
that there is a limit to the depth of a cross-wall building if the rooms are to have
effective day-lighting.
Category (c) is typified in figure 1.3; lateral stability is here provided by a
complex service core incorporating lift-shafts, stair-wells, service rooms, toilets,
etc ., the walls of which, acting together, form a strong tower-like structure .
Surrounding walls or columns need only transmit verticalloading from the floors
while the external, perimeter walls can be non-loadbearing.
The arrangement chosen tends to evolve from the site plan and the required
sizes and disposition of rooms . It is not particularly critical from the structural
point of view, provided that a reasonable balance is allowed between walls
orientated in the principal directions of the building so as to permit the develop-
ment of adequate lateral strength and rigidity against forces applied in these
directions.
Very unsymmetrical wall arrangements are, however, to be avoided as these
will give rise to torsional effects, which are difficult to calculate and which may
produce undesirable stress conditions. Slender piers and cantilevered slabs can be
accommodated but will inevitably add to the cost and may give rise to structural
problems .
General robustness and stability are not usually difficult to ensure in any type
of brickwork structure, but the desirability of incorporating returns at the ends of
walls and other similar strengthening features will be apparent. Detailed considera-
tion of the means of checking and controlling the effects of local damage is
contained in chapter 7.
Figure 1.3 Core-wall structure

1.3 Plain and reinforced brickwork
The essential difference between plain and reinforced masonry is that the former
is incapable of resisting significant tensile stresses whereas the latter acts in a
manner similar to reinforced concrete, tensile stresses being taken by suitably
placed steel.
Most brickwork buildings are constructed in plain masonry without reinforce-
ment, but in seismic areas it is essential to use reinforced brickwork in order to
provide resistance to dynamic forces of considerable magnitude. In Europe and
Australia , therefore, brickwork structures are normally unreinforced whereas
in the United States and New Zealand , reinforced brickwork is generally required .
Apart from construction in seismic areas, however , there is a field of applica-
tion for reinforced brickwork elements in building construction in situations
where the nature of plain masonry imposes undue limitations on the design.
One such case has already been mentioned, namely the possible use of large span
reinforced brickwork wall beams in the lower floors of multi-storey buildings,
where it is not possible to continue the wall layout of the upper storeys to the
foundation level.
The limitation of plain masonry arising from lack of tensile strength can also
be overcome by prestressing , although this technique has not so far been widely
applied . The potentialities of reinforced and prestressed brickwork should there-
fore be kept in mind at the preliminary design stage of masonry structures which
are conceived primarily in terms of plain brickwork. The principles of design of
reinforced brickwork elements are discussed in chapter 6.
1.4 Limit state design of brickwork

The basic aim of structural design is to ensure that a structure should fulfil its
intended function throughout its lifetime without excessive deflection , cracking
or collapse , and this aim must of course be met with due regard to economy.
The designer is assisted in his task by the availability of a code of practice which
is based on accumulated experience and research. Up to the present time , such
codes have sought to ensure the safety and serviceability of masonry structures
by specifying permissible stresses for various types and combinations of materials.
Thus codes generally give basic compressive stresses for a range of brick-mortar
combinations; the basic stress in a particular case has then to be adjusted for the
slenderness ratio of the element and the eccentricity of the loading. The basic
stresses are derived from tests on walls or piers, the ultima te stresses having been
divided by an arbitrary factor of safety sufficiently large to avoid cracking at
working loads. Thus, to this extent, brickwork design has always been related to
ultimate strength and to a serviceability limit state .
In recent years a more rational procedure has been evolved for dealing with
structural safety and serviceability through consideration of the relevant 'limit
states'. A structure, or part of a structure, reaches a limit state when it becomes
incapable of fulfilling its function or when it no longer satisfies the conditions
for which it was designed . Two categories oflimit state normally have to be con-
sidered, namely, ultimate limit states corresponding to failure or collapse and
STRUCTURAL DESIGN OF BRICKWORK BUILDING 5
serviceability limit states at which deflections or cracking become excessive.
The general method of applying the limit states approach to the design of
structures is outlined in a publication of the International Organization for
Standardization I in which the criterion for a satisfactory design is expressed in
terms of design loading effects (S*) and design strengths (R*), as follows
R*~S* (U)
Design loading effects are determined from the characteristic actions from the
relationship
S* = effects of (1'rQk) (1.2)
where 1'r is a multiplier (or partial safet y factor) and Qk is a characteristic load
which , if defined in statistical terrns , is given by
Qk = Qm (1 +k<5)
where Qm is the value of the most unfavourable load with a 50 per cent probability
of its being exceeded once in the expected life of the structure
<5 is the standard deviation of the distribution of the maximum loading
k is a coefficient depending on a selected probability of maximum loadings being
greater than Qk
It is usual to take the characteristic load as that which will have a 5 per cent
prob ability of being exceeded during the lifetime of the structure. In many
situations, however , statistical data are not available and the characteristic loads
have to be based on nominal values given in codes of practice or other regulations.
The factor 1'r is a function of several partial coefficients
1'n which takes account of the possibility of unfavourable deviation of the
loads from the characteristic externailoads, thus allowing for abnormal
or unforeseen actions
1'r2 which takes account of the reduced probability that various loads acting
together will all be simultaneously at their characteristic values
1'r3 which is intended to allow for possible modification of the loading
effects due to incorrect design assumptions (for example, introduction
of simplified support conditions, hinges , neglect of thermal and other
effects which are difficult to assess) and constructional discrepancies
such as dimensions of cross-section, deviation of columns from the
vertical and accidental eccentricities
Similarly, design strengths of materials, R*. are defined by
R
R* =_k (1.3)
1'm
where R k = R m - ks is the characteristic strength of the material
R m is the arithmetic mean of test results
s is the standard deviation
k is a coefficient depending on the prob ability of obtaining results less
than R k
The characteristic strength of a material is usually taken as the 95 per cent
confidence limit of the material strength in a relevant test series. The reduction
coefficient 'Y m is a function of two coefficients
'Ym I which is intended to cover possible reductions in the strength of the
materials in the structure as a whole, as compared to the characteristic
value deduced from the control test specimen
'Ym2 which is intended to cover possible weakness of the structure arising
from any cause other than the reduction in the strength of the materials
allowed for by coefficient 'Yml' including manufacturing tolerances
Additionally, ISO 2394 allows for the introduction of a further coefficient 'Y c
which may be applied either to the design values of loadings or material strengths.
This coefficient is in turn a function of two partial coefficients
'Y ct which is intended to take account of the nature ofthe structure and its
behaviour, for example , structures or parts of structures in which
partial or complete collapse can occur without warning, where
redistribution of internal forces is not possible, or where failure of a
single element can lead to overall collapse
'Y c2 which is intended to take account of the seriousness of attaining a
limit state from other points of view, for example economic con-
sequences, danger to the community, etc.
Usually y, is incorporated into either 'Yf or 'Ym and therefore does not appear
explicitly in design calculations.
The advantage of the limit state approach is that it permits a more rat ional and
flexible assessment of structural safety and serviceability; the various relevant
factors are identified and up to a point can be expressed in numerical terms .
Ideally , loadings and strengths should be available in statistical terms but this is
seldom possible, so that characteristic values have to be determined on the basis
of available evidence. In the case of loads, the evidence generally results from
surveys of buildings in service. Characteristic strengths of materials , on the other
hand , are derived from laboratory tests, the results of which can sometimes
provide a statistical basis for characteristic strength. In the absence of such
statistical data, characteristic strengths have to be based on nominal values
proved by experience .
1.5 Derivation of partial safety factors

In principle , the 'Y-factors can be derived by prob ability methods, provided that
the necessary statistical information is available. Again, this information is at best
incomplete, but consideration of calculated safety factors is valuable in assessing
the relative importance of variables, and in reducing the purely arbitrary nature
of selecting suitable values for design codes. The following discussion, in simplified
terms , may be helpful in appreciating this approach.
Considering the ultimate limit state of a particular structure, we have the
condition that for failure to occur
R* -S*";;;;O (104)
where
R*= R k
rm
and
S* = [CrrQk)
This situation is illustrated graphically in figure 104. In statistical terrns, the
safety requirement is satisfied by ensuring that the probability of failure is very
small , that is
P [R* - S * ";;;;0] =p (1.5)
where Pis the probability of occurrence of the expression within the brackets and
pis the required, small value, of this probability. In practice, this will generally be
in the range 10- 5 to 10-6 .
Theproblem has been examined in terms of a global safety factor, r, mean
values Rand Sand their standard variations by Macchf and Beech ' . On the basis
of assumed normal distributions of strength and loading about their mean values,
Ultimate M ean
limit st reng th
Mean state
load
Frequ ency
01
occurren ce
Figure 1.4 Ultimate limit state condition
Macchi produced the set of curves shown in figure 1.5 showing the relationship
between the safety factor and the coefficient of variation of the strength of the
material, for coefficients of variation of the loads Cs = 0 .15 , of the level of
uncertainty of eccentricity and slenderness Cv; = 0, 0 .05,0.1 and 0.15, and for
probabilities of failure of 10-6 . From these curves it is clear that, other things
being equal , the safety factor required to ensure a given prob ability of failure
rises quite rapidly with the coefficient of variation of the strength of the material.
However, Beech has suggested that the assumption of normal distribution is
0 .15 0 .1 0 .05
6
C~
4
}'
0 .1 0 .15 0 .2
Co
Figure I.S Global safety factors for normal distribution (Macchi). Coefficient of
variation of area Ca = 0, of strength Ca , of eccentricity C w' of loading Cs = 0.15;
prob ability of failure 10-6
4
Y
3 I:.-_--::-:;-r'-
---:;; -
--
0.1 0 .15 0 .2
Figure 1.6 Global safety factors for lognormal (fullline) and truncated normal
(broken line) distributions (Beech) . Coefficients of variation and prob ability of
failure as in figure 1.5
unrealistic in practical terms unless the coefficients of variation are smalI. He
expresses the view that lognormal and truncated normal distributions have greater
validity and has shown (figure 1.6) that these result in a much less steep rise in
the factor of safety with an increase in the coefficient of variation of material
strength.
The investigations outlined above are in terms of a global safety factor whereas
in the application of limit state design two or more partial safety factors are used.
Assuming homogeneity of units , and introducing the partial safety factors in
equation 1.4, at the ultimate limit state
Rk
- ='YrQk (1 .6)
'Ym
or
Rk
- =Qk (1.7)
'Ym'Yr
It would appear from this that the product 'Ym 'Yr is equivalent to a global safety
factor but, as Beech has pointed out, the probabilityof failure associated with the
product of two partial safety factors is generally different from the product of
the partial probabilities. Thus if the probabilities of 'Y m and 'Yr are both 10-3 the
probability associated with their product is not necessarily 10-6 . It has been
suggested by Baker4 that the following approximate relationship holds between
the global and partial safety factors
'Y = 'Ym +'Yr- 1
provided that the probabilities involved are not much less than 10-5 and the
coefficients of variation are not greater than 0.4.
Table 1.1
Values of partial safety factors calculated by Beech
(a) Partial safety factor 'Ym from lognormal distribution
Unit strength coefficient n Coefficient of variation

of variation in masonry strength
0.15 0.22
0.055 1.0 2.25 3.08
0.8 2.22 3.04
0.6 2.19 3.01
0.08 1.0 2.37 3.19
0.8 2.29 3.11
0.6 2.23 3.05
Assumed probability of failure 10 - 6

(b) Partial safety factor 'Ym from truncated normal distribution
Unit strength coefficient n Coefficient of variation

of variation in masonry strength
0.12 0.16
0.05 1.0 2.22 2.77
0.8 2.15 2.68
0.6 2.08 2.59
0.07 1.0 2.40 2.98
0.8 2.28 2.84
0.6 2.18 2.71
Beech has derived equations for calculating the partial safety factor 'Ym on the
basis of lognormal and truncated normal distributions for any combination of
variability in material strength and other factors , and on the assumption that the
masonry strength ais related to the unit strength by a power law of the form
a = u" L , where u is the compressive strength of the units, n an exponent less than
unity and L areduction factor relating masonry strength to unit strength. The
results so obtained are given in table 1.1, from which it will be seen that there is
little practical difference between the results obtained using the alternative
distributions. The determination of safety factors is by no means an exact science,
but considerable progress has been made in establishing a rational framework for
their selection and the values calculated by Beech and quoted in table 1.1 are
very similar in magnitude to those recommended in the British limit state code
for masonry BS 5628 :1978 (table 1.2). Assuming that practical conditions are
reasonably reflected by Beech's theoretical model , there is no doubt that
satisfactory levels of safety will result from the use of the 'Ym values specified
in this code. Reassurance on this point comes in any case from a rough com-
parison of the overall safety factor implicit in the previous permissible stress
code (CP 111 :1964 , 1970) with the global safety factor resulting from the com-
bination of 'Y m and 'Yf in the new code .
Table 1.2
Values of 'Ym recommended in British masonry code BS 5628 : 1978
Category of construction control

Special Normal
Category of manufacturing: special 2.5 3.1

Control of structural units : normal 2.8 3.5
In theBritish limit state code for masonry BS 5628 :1978, the values of 'Yf are
taken to be the same as those specified for concrete construction. This is a
convenience for design engineers, but may not be strictiy valid in terms of the
ISO definitions of partial safety factors, since 'Yf is intended to provide for
inaccuracies in design calculations, deviation of columns from the vertical and
accidental eccentricities, etc ., which may or may not be the same for concrete as
for masonry.
1.6 Analysis of brickwork structures

Limit state design calls for the comparison of the effects of load actions with the
strength of the material. This in turn calls for the use of suitable methods of
structural analysis to estimate the effects of loads on the structure in terms of
forces, bending moments and deformations, and for the use of appropriate
methods of calculating the resistance of masonry elements and of establishing
deformation limits.
Unreinforced masonry is a brittle material and although its stress-strain
relationship is non-linear, it is customary to use elastic analysis to determine the
forces at particular seetions of a structure. Calculation of resistance is now
sometimes based on the assumption of a reetangular internal stress block,
neglecting ten sile strength. In the past, methods of analysis have been relatively
crude, but the construction of taller buildings in brick masonry and the general
need for increased econ omy in the use of materials have led to the development
of more refined methods. Similarly , the resistance of brickwork was calculated,
in the past , on an entirely empirical basis but is now increasingly supported by
analytical studies.
Reinforced brickwork has always followed the principles of reinforced concrete
but , in the absence of any extensive research results, has done so with consider-
able caution. This situation is, however , changing , particularly with the
transference of the basis of design to limit state methods.
Methods of analysis and estimation of the strength of brickwork elements will
be discussed in subsequent chapters of this book.
References
1. General Principles for the Verification of the Safety of Structures, ISO 2394
(1973).
2. G. Macchi, 'Safety Considerations for a Limit State Design ofMasonry',
Proceedings of the Second International Brick Masonry Conference (Stoke-on-
Trent) 1971 , ed. H. W. H. West and K . H. Speed (British Ceramic Research
Association , Stoke-on-Trent , 1971) pp. 229-32.
3 . D. G. Beech, 'Some Problems in the Statistical Calculation of Safety Factors',
Proceedings of the Fourth International Brick Masonry Conference (Brugge)
1976 , Paper 4 .b.8.
4. A. L. L. Baker , quoted by E. Rosenbleuth and L. Esteva in Reliability Basis for
some Mexican Codes (American Concrete Institute, Detroit, Mich., 1972)
Publication SP-31.
2 THE STRENGTH OF MATERIALS
2.1 Compressive strength: general

The strength of brickwork in compression, tension and shear has been the subject
of systematic investigation over a very considerable period. As brickwork structures
are primarily stressed in compression, there has naturally been a concentration of
interest in the resistance of brick masonry to this type of loading and many
investigations have been carried out with a view to establishing the relationship
between available brick types and materials, and a variety of mortar mixes. These
tests have formed the basis for the brickwork strengths used in structural design
codes, and in order to reduce the almost unlimited range of brick and mortar
combinations to manageable proportions, tables of basic compressive strength
have been evolved in which the principal variables are the brick compressive
strength and the mortar mix. The strengths of the component materials are
defined by standardised tests, which do not necessarily reproduce the state of
stress in the component material in brickwork, but which serve as index values in
the selection of design stresses.
Although this purely empirical approach has provided an acceptable basis for
the structural design of brickwork, it is not altogether satisfactory as it gives little
insight into the behaviour of the brick-mortar composite and therefore little
guidance either as to the most appropriate kinds of materials test, or as to the
ways in which the strength of brickwork could be improved. Before examining
the results of some of these investigations it may therefore be useful to identify
the principal variables which affect the strength of brickwork in compression, and
to discuss the mechanism of failure of the material in compression. This discussion
will be concerned in the first instance with the crushing strength of short
specimens under uniform stress , but consideration will be given to the effects of
slenderness and eccentricity of loading on the strength of walls and columns in
chapter 3.
2.2 Factors affecting compressive strength

Research work has shown that the following factors are of importance in
determining the compressive strength of brick masonry
12
THE STRENGTH OF MATERIALS 13
Strength of unit
Geometry of unit
Strength of mortar
Deformation characteristics of unit and mortar
Joint thickness
Suction of units
Water retentivity of mortar
Brickwork bonding.
Some of these factors, such as the unit characteristics, are determined in the
manufacturing process, while others such as mortar properties, are susceptible
to variations in constituent materials, proportioning, mixing and accuracy of
construction. Not all are of equal significance, and in assessing their likely effect
on the overall strength of brickwork we may consider the nature of the failure
of brickwork in compression.
2.3 Failure in compression: indications from standard tests

A number of irnportant points have been derived from compression tests on brick-
work and from associated standard materials tests . These include, firstly, the
observation that brickwork loaded in uniform compression usually fails by the
development of tension cracks parallel to the axis of loading , that is, as a result of
tensile stresses at right angles to the primary compression. This fact has been weIl
known since the early years of brickwork testing. Secondly, it is evident that the
strength of brickwork in compression is much smaller than the nominal cornpres-
sive strength of the bricks from which it is built, as given by a standard
compression test-on the other hand , brickwork strength may greatly exceed
the cube crushing strength of the mortar used in it. Finally, it has been shown
that 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.
From these observations it has been inferred : (1) that the secondary tensile
stresses which cause splitting failure of the brickwork result from the restrained
deformation of the mortar in the bed joints of the brickwork; (2) that the apparent
compressive strength of bricks in a standard crushing test is not a direct measure
of the strength of the unit in brickwork, since the mode of failure is different in
the two situations ; and (3) that mortar is able to withstand higher compressive
stresses in a brickwork bed joint because of the multi-axial nature of the stressing
in this situation.
2.4 Interaction of brick and bed materials

It is possible to detect a paradox in the evidence from the normal brickwork
testing summarised above, namely that while the cube crushing strength of mortar
is only weakly related to brickwork strength by a third or fourth root relationship,
the properties of the bed material do exert a controlling influence on the brick-
work strength achieved ; this has been demonstrated by a number of investigations.
Francis et al. 1 showed that brickwork prisms consisting of loose bricks, the bed-
ding planes of which had been ground flat , achieved compressive strengths
approximately twice as high as those obtained from prisms with normal mortar
joints. Astbury and West? reported a similar effect in relation to brickwork cubes .
Aseries of experiments conducted by the Structural Clay Products Research
Foundation in the United States," examined the effect on the compressive strength
of brick couplet specimens in which the jointing materials ranged from 0 .8 mm
aluminium sheet placed between ground surfaces to anormal cement :lime:sand
mortar , and included a dry sand joint contained by adhesive tape . Some of the
results of these experiments are summarised in table 2.1.
Table 2.1
Effect of different joint materials on the compressive strength
of brick couplets (Monk 3 )
Joint material Compressive strength Ratio to brick strength

(N/mm 2 )
Aluminium sheet 106 0 .96

1 1
Mortar (1 :2" :42" ) 44 0.40
Sand 65 0.59
Ground surfaces 98 0 .89
The effect of bed material on brick prism strength was also investigated by
Morsy." In his experiments the bed material in aseries of model brick prisms was
varied, from rubber at one end to steel at the other, and the results of these
experiments, summarised in table 2.2, show that there is an eight-fold change in
the prism strength with the substitution of steel for rubber in the bed joints. In
the case of rubbe r jointing material the bricks failed in tension as a result of
Table 2.2
Effect of different joint materials on the compressive strength of three
brick stack prisms; one-sixth scale model bricks , faces ground flat ;
six specimens of each type tested (Morsy" )
Joint material Compressive strength Ratio to brick strength

(N/mm 2 )
Steel 56.48 1.4

Plywood 46.39 1.15
Hardboard 43.89 1.09
Polythene 16.99 0.42
Rubber with fibres 11.71 0.29
Soft rubber 6.99 0.17
No joint material 37 .20 0.93
Mortar (l :t :3) 14.0 0.35
tensile stresses induced by the deformation of the rubbe r. Steel in the bed joints,
on the other hand , had the effect of restraining lateral deformation of the bricks
and this induced astate of triaxial compressive stress in them. Failure in th is case
was by crushing as in a typical compression test on a brittle material.
The value of these investigations lies in the light which they throw on the basic
behaviour of the brick-rnortar composite, in particular on the interpretation of
the results of compressive tests on bricks ; they are also useful in drawing attention
to the potential for developing high-strength masonry using improved mortars.
2.5 Failure theories based on elastic analysis

Consideration of qualitative evidence such as that discussed in the preceding
paragraphs has prompted a number of investigators to derive formulae for brick-
work strength in compression, based on the elastic analysis of the brick-mortar
complex. The earliest attempt would appear to be due to Haller,5 published in
1959 . Haller's forrnula , however , results in values of brickwork strength greater
than the uniaxial strength of brick , and is thus not valid in a quantitative sense.
Formulae based on elastic behaviour were also proposed by Lenczner" and by
Francis et al. 1
The formulae proposed by the latter were derived by considering a brick-
mortar prism subjected to an axial compressive stress Gy, as shown in figure 2.1a.
The lateral stresses induced in a central brick and in the adjacent mortar beds are
indicated in figure 2.1 b, and the extensional strains in the brick in the x and z
directions are
(2.1)
(2 .2)
(al (b )
Figure 2.1 Stresses in brick-mortar composite

and similarly in the mortar joints
1
e xrn = - [- axrn + v rn (a y - azrn)] (2.3)
Ern
1
e zrn = - [- azrn + vrn (a y - axrn)] (2.4)
Ern
where E b and Ern are the elastic moduli of the brick and mortar respectively, and
!Jb and vrn the corresponding Poisson's ratios. For equilibrium, the total lateral
tensile force in the brick is equal to the total lateral compressive force in the
mortar; hence
axrn = a axb (2.5)
and
azrn = a azb (2.6)
where o is the ratio of the height of the brick to the thickness of the mortar bed .
As the lateral strains in the bricks and mortar are the same , equating equations 2 .1
and 2.3, 2.2 and 2.4 , and using equations 2.5 and 2.6 gives
ay (ßv rn - Vb) (2 .7)
axb = azb = - - < - - - - - -
1 + aß - vb - aßv rn
\
, ° ult
° ult
Lateral Lateral
compressive stress A tensile stress
o at '
orazt>
Figure 2.2 Failure envelope for brick material in biaxial compression-tension

assumed by Francis et al.
Assuming a linear relationship between ultimate longitudinal compressivc stress
and lateral tensile stress, as shown in figure 2.2, gives the relationship
I ,
axb = ; (ault -- ault) (2.8)
whcre c/J = a ult' ja,' . Substituting für axb in equation 2.7 and ncglccting (I -- Vb)
gives
ault
(2.9)
a'ult I +!...(ßVm =-~ 2
aß (I - - Vm )
C'omparison with experimental results using joint th ickncss as a variable (figurc
2.3), shows that this formula gives a fair reprcsentation 01' the actual behaviour 01'
a set 01' specimens tested by its originators. Discrcpaucics between theory and
experiment were thought to arise from approximations in cstimating the truc
12
°ul t
• Solid bricks
O u~t
D Perforated br icks
2 c
o ~_"",,:-_"""'-_"""_"""' _
Q2 Q4 Q6 QS 1.0
Average joint thickness t m (in .l
Figure 2.3 Effect of joint thickness on brickwork compressive strength
values of the parameter c/J and Poisson's ratio. This in fact suggests one 01' the
limitations of this approach, together with the fact that the materials, particularly
the mortar, are not elastic up to the point of failure . Furthermore, the theory is
derived for stack prisms and may not be applicable to bonded brickwork.
2.6 Failure theories based on the strength of brick and mortar under
multi-axial stress
An alternative approach to the definition of brickwork strength was proposed by
Hilsdorf,7 based on an assumed linear relationship between lateral biaxial tensile
strength and local compressive stress equal to the mean external compressive stress
(J . =0 r
lateral
tension
/
I
/
8./ Minimum lateral tension
/ A::=-_---~,-in brick (Iine Cl
/
/
/
local compression
Figure 2.4 Hilsdorf's failurc theory
multiplied by a 'non-uniformity' factor U. Referring to figure 2.4, linc Ais the

failure criterion envelope, and when external compression is applied to the brick-
work, the internal tensile stresses induced follow some line such as B• . When this
line intersects the failure criterion envelope, a local crack is developed in the
brickwork . Further local cracks will appear on subsequent increase of load, but
general failure will not take place until the brick can no longer provide the biaxial
restraint necessary to prevent failure of the mortar, or alternatively, when the
state of stress developed in the brick exceeds its resistance to the combination of
stresses developed . This will occur when the line defining the triaxial strength of
the mortar, line ein figure 2.4 , intersects the failure line for the brick.
Hilsdorf assumed that the triaxial strength of mortar eould be represented by
the equation (obtained originally for eonerete)
I.' = I~ + 4.1 02 (2 .10)
where 11 r is the eompressive strength of a laterally eonfined eylinder
I c' is the uniaxial eompressive strength of a eylinder
02 is the lateral eonfinement stress of a eylinder
This leads to the minimum lateral eonfinement of the mortar joint of

1 ,
Oxj= 4j(Oy-ij) (2.11 )
in whieh Oxj is the lateral eompressive stress in the mortar joint

Oy is the loeal stress in the y direetion
ij' is the uniaxial eompressive strength of mortar
Taking into aeeount the equilibrium of lateral forees in brieks and mortar, the
equation of line ein figure 2.4 is then
TIlE STRENGTH or MATERIALS 19
0x = .t -(0 - f';') (2.12)

4.1b Y J
where b is the height of the brick, and j the mortar bed thickness. Une A is
expressed by
°x = Oz = fbt (1 - ;f) (2 .13)
in which fbt' is the strength of the brick under biaxial tension and fb' is the uniaxial
compressive strength of brick .
The magnitude of the local stress at failure, that is th e intersection of lines A
and C, is therefore given by
(2.14)
where a = j /4.1b .
The average masonry stress at failure is then
o
o = ~ (2.15)
ym (J,
u
where Uu is a coefficient of non-uniformity, which Hilsdorf established experi-

mentally for various brick-mortar combinations. This varies according to the
brickwork strength, but for cement mortar it has been shown to have a value of
around 1.3 in the medium strength range.
This approach was developed by Khoo and Hendry" who investigated the
behaviour of brick mat erial und er astate of biaxial compression-tension, and of
mortar under astate of triaxial compression ; these characteristics had to be
assumed by Hilsdorf in the absence of direct experimental data. They established
that the biaxial compression-tension strength envelope for brick can be
represented by the relationship
__: _ ) = 1 _ (~)0 .546 (2 .16)

(
Co to
This curve is shown in figure 2.5, and was based on the results of tests on a large
number of specimens of bricks ranging in crushing strength from 31 .63 Njrnm?
to 92 .66 N/mm 2 • It will be noted that comparing the concave shape of this
curve to the linear relationship assumed by Hilsdorf, the compression strength of
brick is severely reduced by the presence of an orthogonal tensile stress.
The effect on the compressive strength of mortar of a confining pressure was
investigated by Khoo and Hendry for l :i :3 and 1:1:6 mortars using a triaxial
test cell,? The increase in strength so found was less than that found for con-
crete l O, lI , 1 2 and is shown in figure 2.6. The principal stress relationship is
non-linear and may be defined by the expression
1.0 i:"'oi.
,,'~-
" -, ~....... ./Griffith·s
0 .8 " "
"'"
<, .. . .......
"
"" '-,',
'-" ' ' . / Modified Coulornb's
"',
, ......
'.
-, ' .~.
"
0
~ 0 .6
c
0
'iii " " Original " .......
l
Vl
"~
Ql
Ci Coulornb's •\
u
E 0.4
0
"" .,
""
0 .2
Model bricks ""
""
0 .....- -........-----'.......- -.......- -.......- - - - "
0 .2 0.4 0,6 0 ,8 1.0
Tension t/t o
Figure 2.5 Biaxial compression-tension failure envelopes
=1+2 .91
a )0,805
(2.17)
(o
oZ
where al is the major principal stress

az is the minor principal stress
00 is the uniaxial compressive strength
On the basis of these studies a failure theory for brickwork has been
devcloped .8 , 1 3 Thus in figure 2.7, which shows an assumed failure envelope for
brick in biaxial compression-tension in a brickwork prism, any state of stress to
the right of the envelope curve denotes failure, As thc vertical compression acting
on the brickwork prism increases, the state of stress in the brick element proceeds
along the dashed line OA. Failure occurs within the brick element when the line
OA intersects the failure envelope at A, and hence the compressive strength of the
brickwork prism is given by the ordinate of this point. The stress path taken by
the line OA depends on the properties of the mortar joint under tri axial com-
pression. For a weaker mortar, whose lateral strain is greater under load, the
stress path would travel along the lower line OB, and in this case the point B would
define the compressive strength of the brickwork prism .
It is not, however, necessary to determine the stress paths OA and OB: the
failure point for a given brick-mortar combination can be located graphically by
superimposing on the brick failure envelope a curve derived .frorn the triaxial
compressive strength relationship for the mortar, which defines the tensile stress
'b
x
.-
CIi
~
e'"'"
;;;
iö
a.
'ü
.s 4
Ci.
Ci
'iö
~
O~ _ _""" _ _-L._ _"""" ..I..-_ _"""_ _- I ._ _--I_ _ ~
0 .4 0 .8 1.2 1.6
M inor principal stress 03 (p.s.i. x l 0 3 )
Figure 2.6 Principal stress relationship for mortars in triaxial cornpression
induced in the brick . Such curves have been established by Khoo taking the
lateral tensile strain in brick material as approximately 225 x 10- 6 • Relating this
value to the stress-strain curves for mortar obtained from triaxial tests, as shown
in figure 2.8, a relationship between axial and lateral compressive stresses in the
mortar is obtained. Introducing the ratio, Q , between the mortar joint thickness
and the depth of the brick then gives the lateral tensile stress in the brick material ,
assuming that these stresses are uniformly distributed on vertical sections through
the mortar and brick. The resulting axial compression-lateral tensile stress curves
can now be plotted on the same axes as the biaxial failure envelope for the brick
material, as indicated in figure 2.9, the intersection of the two curves defining
failure of th e particular brick-mortar combination.
An analytical solution is readily obtained by replacing equations 2.16 and 2.17
for the brick failure envelope and the mortar triaxial strength curve, respectively
by the polynomials
Co
Compression
Tension
Figure 2.7 Failure envelope for brick in biaxial compression-tension
CJ
2
!-. = 0.9968 - 2.0264(-~)+ 1.2781 -0 .2487 (2.18)

to Co
)
2 3
~= -0.1620 + 0.1126 C~ + 0.0529 (:~) -- 0.0018 ( : ;) (2.19)
°0
Introducing the condition
t = a03 (2.20)
and combining equations 2.18, 2.19 and 2.20 gives
(2.21)
60
I
I
I 2
50 II . =10.4 N/mm
2
~t - 8 .30 N/mm
2
N~ ~40I ~ 6 .25 N/mm ..,::I:
tT'l
_Ll ()I'; N /",,,,22
30 . 4.06 N/mm CI>
Z~ I '~~ ~
..,
:;l:l
Q) tT'l
- Q)
2
VI
- /"f'"I - ~ ~ ~1 .94 N/mm Z
C)
Q)
..,::I:
I~ 2
'S"",--O N/mm (later al compressionl o'Tl
;>
..,s::
tT'l
:;l:l
;;
r-
CI>
6 3 o 3 6 9 12 15 18 21
Compression Tension
Total lateral strain I xl 0 2,u1
Figure 2.8 Axial stress-lateral strain relationship für I :~ :3 mortar
IV
W
80
Failur e en velopes for

var io u s br ic k t ype s
60
E
E
......
z 1 " "3
m ~~ tar
s:
Öl ]
c
e
üi
Q)
>
"iij
<Il
Q)
Ci
E 1 :1 :6
o
u rn o r t ar
Cl 0.4 0 ]
o 2 4 6
2
T en sil e st re ng t h IN / m m )
Figure 2.9 Graphical solution for brickwork prism compressive strength
where x = c = al represents the compressive strength of the prism .

Comparison of brickwork prism strengths calculated by the above theory shows
reasonable agreement with experimental results . Such a comparison with the
results obtained in aseries of tests conducted by the Structural Clay Products
Research Poundation!" in the United States is shown in figure 2.10. The
influence of joint thickness on compr.essive strength has been examined experi-
mentally , and as may be seen from figure 2.11 the theory gives a reasonable
repre sentation of this effect.
60
~
S--..; -v. . . . .
50 ~ ~
'I" ::,,~
,Il' ~
.: n,,' •
E
E
.. ...... CO'~
.'-:
<,
6
s: . ..
0, 40
c
~
..
;;;
.,
>
.
'iii
.,'" J.
C- A
E 30
0
o
E
'"
.~
"'Ci"
3:
"''<:o" 20
CD
10
o 50 100 150
2
Brick com pres sive strength (N/mm )
Figure 2.10 Brickwork prism compressive strength (S.C.P.R.F. National Testing

Program)
At this stage, failure theory can predict the strength of brickwork prisms with
reasonable certainty , but prism strength is not necessarily equal to the compressive
strength of bonded brickwork, which is usually somewhat lower. Research work
in Australia,' 5 where a prism test is used as a basis for determining brickwork
design strength, has indicated that the ratio of wall strength to prism strength is
on average 0.9 . It should also be noted that the failure theory is only valid where
the brick strength exceeds the mortar strength, otherwise the deformation
properties of the materials will be such that the brick element will no longer be
1.0
Bri ckwork
pri sm/ 4 Kh o o
bri ck
compressive o S CPRF
stre ng t h 0 .8
rat io
Sol id
Perf o rat ed )(
-J Fran ci s
0 .6
•
0 .4 111
121
0 .2 141
13)
o 0 .1 0 .2 0 .3 0.4 0 .5 0 .6 0 .7
M ort ar/bri ck th ick ness rat io , Cl
Figure 2.11 Effect of mortar{brick thickness ratio on brickwork compressive

strength: comparison with experimental data
subjected to bilateral tension as envisaged in the theory. The 'cut-off" implied by

this limitation is indicated in figure 2.12, which shows the theoretical relationship
between brick and brickwork prism strength for a range of mortar strengths.
Finite element analysis of the brick-mortar composite subject to axial com-
pression, carried out by Khoo 13 revealed the existence of stress concentrations at
the edge of the brick-mortar interface, causing high stress gradients in both
materials in this region. The end face of an axially loaded prism, where the
width/depth ratio of brick is lower, contains higher values of lateral stresses in
the region of the stress concentration than the front face section , and
consequently suggests that initial tensile cracking in the brickwork prism will
occur along a plane at the edge of the end face in the direction of the brick length .
This gives rise to what is usually referred to as 'spalling' failure , which is frequently
observed in tests.
2.7 Empirical studies of the compressive strength of brickwork

Numerous experimental investigations have been carried out to examine the
variables affecting brickwork strength. The literature pertaining to these studies is
extensive; references 17 to 47 relate to work reported during the 1960s and 1970s.
The primary variables are brick strength and mortar strength and early work on
which codes of practice were based concentrated on these. A well-known example
of such work is that reported by Thornas.P the results of which are summar ised
in figure 2.13. The general similarity ofthese results to the theoretical curves of
figure 2.1 will be immediately apparent.
Brickwork prism 60
co m pressive
strength
(N/mm 2 )
50
,
I
I
40
I
I
I
"S,I
30 t: ,
$,
~I
E'
.c: ,
20 0, ,
&I
:1
;tl
,,
I
10
I
I
o 50 100
Brick com pressive str eng th (N/mm 2 )
Figure 2.12 Theoretical brickwork prism compressive strength (0: = 1/7)
Basic studies of brickwork compressive strength have been carried out in many
other countries including Switzerland,!" the United States ,18-20 Germany ,"'
Belgium.P Australia 23,24 and Yugoslavia.P Together with earlier tests, a large
body of data has been accumulated on the compressive strength of brick masonry
and efforts have been made to relate the variables concerned by mathematical
formulae. Earlier formulae, summarised by Sahlin,26 were in terms of brick
strength and mortar strength, although some included the height and density of
the brick . A formula for the compressive strength of brick masonry prisms, based
Sri ckwork
strengt~
(N imm )
30
20
10
o 50
Sri ck strength (N /mm 2 )
Figure 2.13 Crushing strength of brickwork walls and piers (after Thomas)
on an analysis of a large number of tests in the United States, has been

presented by Grimm,"? as follows
Im' = 1.42~11/b' 10-8 (f; +9.45 x 106)(1 +Er 1 (2.22)
where Im' is the compressive strength of a brick masonry prism in Ibf/in?
Ib' is the average compressive strength of brick (tested according to ASTM
C67) , not exceeding 12000 lbf/in."
t; is the compressive strength of mortar cubes (tested according to ASTM
C270)
~ is the brick masonry prism slenderness factor, given by
(2.23)
in which 6 > hs/ts > 2
11 is the materials size factor, given by
0.0048 [273 - (ha/tj - 14)2] (2.24)
in which 10 > ha/tj > 2.5
E is the workmanship factor , which is given by
8 x 10-5 (12000 - Ib ') for uninspected work and 0 for inspected work
In the above, ~ = Lif hs/t s = 5 and 11 = 1 when h a = 2.25 in. and tj = 0.38 in.
The various strength theories described in sections 2.5 and 2.6 , and empirical
formulae such as those quoted above, are useful in obtaining a quantitative
understanding of the main factors affecting brickwork strength in compression.
However, it is necessary to take into account a variety of secondary variables
including : the effect of brick characteristics other than crushing strengtli;" -34
brickwork bond,3S particular wall types 36- 42 and various practical details. 4 3 ,44
2.7.1 Brick characteristics

The effect of the ratio of brick height to joint thickness has already been discussed,
and it has been shown that for a brick of given height, brickwork strength is
reduced as the joint thickness is increased . Houston and Grimm/" have carried out
tests on bricks of the same material, but with heights varying between 41 mm and
127 mm. The crushing strength of the units decreased as the height increased, as
would be expected , as the effect of platen restraint became of less importance
(figure 2.14). The effect ofunit height on brickwork prism strength in this
investigation is shown in figure 2.15 which is based on a brick height of 57 mm
and a joint thickness of 9.5 mm. The figure shows that if the unit height was
increased to , say, 100 mm, giving a ratio of brick height to mortar joint thickness
of 10.5, the compressive strength of the brickwork wou1d be increased by about
25 per cent, or a1ternatively, the same compressive strength could be obtained
using a unit strength about 25 per cent lower than the original. This effect is
allowed for in the British Code of Practice BS 5628:Part 1:1978, where strength
tables are given for a number ofunit height to thickness ratios. This code, however,
does not refer to the thickness of the bed joint, which is a highly relevant factor.
Height
of unit 125 Q
I +Sample A
(mm) QSample B
100
75
50
25
oL-_ _.L.-_ _.........._ _..........- - - " " ' - -

60 70 80 90 100
2
Brick strength (qro ss areal (N/mm )
Figure 2.14 Compressive strength of brickwork as a function of unit height

(after Houston and Grimm)
Brick height (mm) when
laid in 10 mm mortar joint
40 6070 100 125
% Change % Change in
in ultimate - 3 0 required compressive
40
compressive strength of brick
~
strength / w ith no change
30 / in ultimate
of masonry
/ i' - 2 0 compressive strength
with
/
no change 20 • .-- of masonry
in I ;'
",.
required I lC ,1 - 10
10
/
brick I /
strength
0 I / o
I /
/ 10
- 10
~
/ 20
- 20 I
i'
30
-30 I 40
Allen , 1965.
I 50
Sampie A •
- 40 Sampie B Je
2 4 6 8 10 12 14
Ratio of brick height to
rnortar joint thickness
Figure 2.15 Change in compressive strength of masonry and requ ired compressive
strength of brick as a function of brick height to joint thickness (after Houston
and Grimm)
A number of investigators have studied the effect of different types of bricks
on compressive strength-an extensive series of tests was undertaken by West et
al. 30 at the British Ceramic Research Association to examine the compressive
strength of brickwork built with a variety of wire-cut bricks having different per-
foration patterns. The results of these tests showed that if the brickwork strength
was calculated on the basis of a standard crushing test on the unit, the perforation
pattern made little practical difference . In these tests, the perforations were either
circular holes or slots with round ends, but in some tests reported by Monk 3 the
units had reetangular slots, and these tests showed reduced compressive strength
in prisms . It would seem probable that such slots would introduce stress
concentrations, not only in service but also in manufacture, which would be a
source of weakness . Apart from this, however, there seems to be little clear
evidence of a connection between perforation geometry and compressive strength.
Variability in the properties and dimensions of bricks is an obvious source of
variability in masonry strength. The effect of low-strength bricks on high-strength
brickwork has been systematically investigated by Fisher'" who showed that use
of the normal strength classification, based on statistical quality control, will be
sufficient to allow for random variations in the strength of bricks in walls. Variable
height of bricks is likely to introduce variability in strength as the joint thickness
in the masonry will vary correspondingly, but this again can be adequately con-
trolled by the manufacturer.
2.7.2 Effect of brickwork bond and wall type

In general the effect of various brickwork bonds on basic compressive strength has
been shown to be relatively small/" but the results of a few tests 42 on walls built
in what is sometimes described as stretcher bond , that is, two leaves or wythes of
brickwork without headers, suggested that this type of masonry is of lower
strength than normally bonded brickwork of the same materials. The comparisons
leading to this conclusion were, however, rather limited, but some further work
reported by Beard41 indicated that the strength of a 219 mm wall, consisting of
two 105 mm leaves, was reduced by about 16 per cent when butterfly wire ties
giving 40 mm 2/m 2 of wall were usedand by about 10 per cent when using special
reetangular ties giving 340 mm? 1m 2 of wall. On the other hand , comparative tests
by James 23•24 on 219 mm walls with headers every fourth course, and 219 mm
stretcher bond walls with various kinds of ties, showed no significant difference
between the two types. Also some model -scale tests,46 in which standard twisted
steel ties were simulated, showed no significant reduction in strength, while
stretcher bonded walls without ties showed about an 11 per cent reduction.
Although the number of tests is still rather limited, it would appear that the
difference between the strength of stretcher bond walls and normally bonded
walls is unlikely to be greater than 10 per cent, with normal spacing of twisted
steel ties, and therefore unlikely to be important considering the safety factors
currently used in brickwork construction.
Closely related to the stretcher bond wall is the open cavity wall built of two
stretcher bond leaves of equal thickness. Tests on cavity walls equally loaded on
both leaves"? indicated that the strength of the wall was about 70 per cent of the
strength of two single-leaf walls of the same materials. The probable reasons for
this were suggested to be : (1) that it is difficult to ensure that loading is equally
applied to the two leaves, (2) it is difficult to construct both leaves of a cavity wall
equally weIl in terms of ensuring that the joints are completely filled , and (3) the
strength of the cavity wall will be limited by the strength of the weaker leaf,
failure of which will precipitate failure of the whole wall. An investigation con -
ducted by Fisher 43 showed about a 14 per cent reduction in cavity wall strength
for walls built of perforated bricks, and no difference for single frog, semi-dry
pressed bricks . These tests also indicated that the strength of cavity walls is not
greatly affected by tie spacing, even when the cavity wall was loaded only on one
leaf. Cavity walls tested by James 23 under axial load were practically equal in
strength to two single-leaf walls. The number of tests reported is again rather
small and is insufficient to resolve the apparent discrepancy between the various
results . It would, however, be reasonable to assurne that the brickwork in a
270 mm cavity wall was equal in strength to that of a 229 mm bonded wall of
the same height.
It has been found by many investigators 2 3 ,2 4 , 3 2 , 3 6 , 3 7 , 3 9 that , in terms of
ultimate compressive stress, walls whose thickness is equal to the width of the
bricks used are stronger than bonded walls. Thus a 105 mm thick masonry wall
is significantly stronger than a 219 mm bonded masonry wall built with the same
units . This point may be illustrated by reference to the results of comparable tests
reported by James.P Pisher " and Bradshaw and Hendry .:" In each case 105 Olm
and 219 Olm walls built of the same materials were tested, and the load factor for
each wall was calculated on the basis of the code of practice in use at the time and
place of the tests, with the following results
105mm 2l9mm
James 10.6 7.2
Fisher 7.4 4.3
14.9 5.6
Bradshaw and Hendry 7.5 3 .6
An extensive series of tests on 105 Olm walls built of wire-cut bricks carried
out by West et al. 32 and the earlier results of Bradshaw and Hendry" ? provide
further evidence of the consistently high load factors found experimentally for
this type of brickwork.
2.7.3 Concentrated loads on brickwork
In considering the compressive strength of bri ckwork it has been assumed , and in
experimental work arranged , that the overall stressing of the material has been
uniform . It frequently happens in practice that there are areas of high local com -
1 .0
o • Full scale
Bri ck w ork
strength : + 1sca le
br ick strength a ~ scale
0 .8
06
0 .4
~:----_ ...a
.,. +
+
0 .2 "'---- ........ ...1..- ........ ....&.. ......1._
jt
Width 01 be aring plate (t is
thickness 01 w all)
Figure 2.16 Concentrated loading.on brickwork : effect of width of bearing plate

pressive stress under beam bearings and the like. In these locations it is to be
expected that higher stresses could be accepted on account of the restraint
provided by surrounding lightly stressed material, and codes of practice make
corresponding provision. Published information on the problem at present seems
to be limited to three papers 4 0 , 4 8 ,4 9 one of which describes the results of a
photo-elastic investigation of the stresses in plates loaded at varying distances
from a corner. The results obtained by Rutherford on full-scale and model brick-
work specimens are summarised in figure 2.16, which shows the effect of bearing
plate length for central and end bearings in terms of the ratio of brickwork
strength to brick strength. It is clear from these results that as the plate width is
decreased the bearing pressure at failure increases. This effect is considerably
greater where the load is applied away from the end of the wall, so that allowance
for concentrated loading in design calculations should take into account both the
spread of the load , and the distance of the bearing from the end of a wall.
This investigation also showed that the 'bulbs' of compressive strain that exist
under a concentrated load are contained within a 45° fan drawn from the ends of
the bearing plate but are not uniformly distributed on horizontal planes within
these limits. Failure may take place by the development of vertical cracks at some
distance below the loaded area, by horizontal 'tearing' at the surface , or by
spalling of the brickwork under the load.
The effect of bed joint reinforcement was examined but the particular type
and amounts of steel introduced showed little or no advantage . It is possible that
relatively fine wire reinforcement rather than a small number of, say, 3 mm rods
is necessary in this situatiorr'" to achieve significant increase in brickwork corn-
pressive strength.
Although useful, and to a large extent reassuring in relation to present practice,
available results are somewhat limited, in that they have been obtained for only a
few brick types and for rigid bearing plates to which the loading was applied
cent rally . Further information covering a range of brick types and practical bear -
ings with allowance for eccentric loading arising from beam deflection would be
desirable.
2.7.4 Chases in brickwork

In practical building construction the need to accommodate electrical wiring con-
duit, switch boxes and other fittings frequently leads to the cutting of chases and
holes in brickwork; if the walls are slender this may lead to an appreciable
reduction in wall strength. Tests by Prasan et al. 36 indicated that vertical chases
19 mm wide by 12 mm deep in a 105 mm wall had little effect on strength. In a
more extensive investigation, Fisher'" tested walls of 102.5 mm, 178 mm and
215 mm thickness with 38 mm wide by 25 mm deep chases cut in the patterns
shown in figure 2.17. The decrease in strength as com pared to corresponding
walls without chases may be seen in table 2.3.
While no very clear pattern of behaviour emerges from these tests , it would
appear that chases may reduce the strength of a wall by 20 per cent or more. In
the case of the walls tested by Fisher , the safety factor was still quite adequate ,
but it would seem prudent to limit the cutting of chases and the like in highly
34 STRUCTURALßR~KWORK
1
1 .,
I I(
'1
11
1
0
1 '"
"" .,"
"I
~ .
"I
:
I'-
5 6 0 111 111
I.
178 1/8 560 1 1 4 58
22 9 229
~~
~
co
'"
<t
TYP E A TYPE 8 TYPE C
",I
,I
'" '"
o
'"
,I
11
o
ec
M
'"
N
~
cn
'"
N
~
m
"
'I
I
458
'- .-
ii-'-'"
W r"'tI -f~ (")
~I'- I
I
- 5 60 178
56 0 178
15 3 LJ ~ ~ ~
W
153
610 I
'"
cn
TY PE D TY PE E TY PE F
Figure 2.17 Effect of chases on cornpressive str ength ; patt erns of chases tests by
Fisher
stressed walls. The Model Specification published by th e British Ceramic Research
Association" 1 suggests that, as far as possible, chases and holes should be provided
durin g the erection of brickwork using, where appropriate , purpose made chased
unit s. It is also recommended that if chases are cut in completed walls this should
be done with a special tool, and that horizontal or diagonal chases should be
avoided .
2.8 The effect of workmanship factors on compressive strength

In common with other construction mat erials and techniques, th e strength of
brickwork is affected by site workmanship. To some degree this has been
exaggerated , and has led to the adoptio n of very large safety factors in design
Table 2.3
Decrease in strength with chasing
Wall type % Decrease in strength
178 mm walls 102 .5 mm walls 215 mm walls

A 11.2* 16.9 23.8*
B 6 .7 1.1 *
C 1.2 14.7
D 20.1 4.0 6.4
E 13.1 * 13.3*
F 9.5 7.7* 28 .5*
*based on one result only : in other tests mean of three walls
codes that do not take into account the factors involved, and thus the possibility
of achieving laboratory strengths by appropriate control measures. It is there-
fore essential to identify the workmanship factors that are significant in relation
to strength and serviceability , and from there to devise the necessary specification
clauses and site control measures. It should be made quite clear that we are not
concerned he re with gross errors or omtssions.-such as the use of the wrong bricks
or mortar materials, or with defective materials-but with the identification of
various defects in site work and an assessment of their effect on the performance
of masonry. The most obvious workmansip factors are as follows
Incorrect proportioning and mixing of mortar
Incorrect adjustment of suction rate of bricks
Incorrect jointing procedures
Disturbance of units after laying
Failure to build walls 'plumb and true to !ine and level'
Unfavourable curing conditions.
These factors will be discussed in relation to the available information in the
following sections.
2.8 .1 Incorrect proportioning and mixing of mortar

The effect of mortar strength on the compressive strength of brickwork has been
referred to in previous sections of this chapter. Generally , the mortar strength, as
defined by cube crushing strength, is not a very critical factor in masonry strength;
for example, with bricks of crushing strength 35 Njrnm" , a halving of the mortar
cube strength from 14 N/mm 2 to 7 Njmm? may be expected to reduce the com-
pressive strength of the brickwork from about 16 N/mm 2 to 14 N/mm 2 • This
corresponds roughly to a change in mortar mix from 1:3 cement to sand to
1:41- or, say , 30 per cent too !ittle cement in the mix. A similar reduction in
mortar strength could, of course, be brought about by an excess of water-moving
from a water .cement ratio of about 0 .6 to one of about 0 .8 in a typical case.
Mclntosh'" has stated that 'the cement content of a 1: 1:6 mix composed of
cement gauged with ready-mixed !ime :sand for mortar , could vary from about
131- to 19% of the weight of dry !ime plus dry sand; the corresponding range in
water :cement ratio required to produce mortar of standard consistence was from
about 1.6 to 1.1 resulting in a change of 7 day strength from 2.9 to 6.5 Nrmm" .
Greater differences might be expected if all the materials are batched separately
on site.' Thus it is clear that while a 2 :1 variation in the strength of site-produced
mortar to the same nominal specification is possible, the effect on masonry com-
pressive strength is proportionately much less, at any rate when the units are of
low to medium strength. In the case of high-strength units, however, the effect of
variation of mortar strength is likely to be more significant.
2.8.2 Incorrect adjustment of suction rate

In order to achieve optimum brickwork strength it has long been realised that the
suction rate of bricks should be controlled to prevent excessive removal of water
from the mortar. It seems probable that the water absorbed by the bricks leaves
cavities in the mortar, which fill with air and result in a weakened material on
setting. On the other hand, brickwork built with saturated bricks develops poor
adhesion between bricks and mortar, and is of course susceptible to frost damage
and other troubles. Some specifications recommend a limit ing suction rate, or
alternatively the use of a high retentivity mortar to control the extraction of water
from the mortar. In so far as water extraction affects the final strength of mortar ,
one would not expect it to result in a serious weakening of brickwork in compres -
sion . However, Haller l ? has demonstrated that , in certain circumstances, suction
rate has a considerable effect on brickwork strength because de-watered mortar
tends to form a rounded joint during building owing to a loss of 'elasti city' . It
would appear , for example, that with eccentric loading an increase in the suction
rate from 2 kg/m? Imin to 4 kg/m 2 Imin could halve the compressive strength of
the brickwork.
It is clear, therefore , that suction rate is a factor to be taken very seriously ,
especially in the case of slender walls built in relatively low-strength bricks . If the
bricks being used have a high initial rate of absorption, it is clearly essential to
adjust this by wetting them before laying.
2.8.3 Incorrect jointing procedures

A variety of defects can arise from incomplete filling of joints and some evidence
is available on the structural effects of these defects . The effect of incomplete
filling of perpend joints has been investigated by the British Ceramic Research
Assoctation.P and by the Building Development Laboratories of Australia.t" A
total of thirty walls were tested at B.C.R.A. with unfilled vertical joints, using
two types of brick and mortar with the results shown in table 2.4. Statistical
examination of the results showed that there was no significant difference
between corresponding sets of walls with joints filled and unfilled . The Australian
tests (table 2.5) also showed that unfilled vertical joints had no significant effect
on the strength of walls. There are theoretical reasons for expecting that this
would be the case, but careless filling of vertical joints may be indicative of poor
workmanship in other respects , aud would certainly reduce non-structural per-
Ta ble 2.4
Effect of unfilled perpend joints on strength of brick walls
(tests carried out by the British Cerarnic Research Association 53)
Mean brick Mortar Mean rnortar Wall thickness Cross- Wall strength
strength and water cube strength and bond joints (MN/rn 2 )
absorption (MN/rn 2 ) (rnrn) filled or
(MN/rn 2 ) unfilled
92 I :t :3 19.3 0 102.5 (4tin .) 18.40

7 .9% 15.30 Stretcher Filled 21.16
19.65 20 .89
18.13 20.15
16. 89 Unfilled 20.75
16.27 13.99
20 .54 23.3 7
17.92 19 .37
92 l :t :3 15.65 215 (9 in .) 23.64
22.27 Englis h Filled 16.75
7 .9% 19 .03 24. 06
18 .96 2 1.44
20 .54 Unfilled 17.99
19.30 19.37
18.34 16.4 1
19.37 17.92
92 l :t :3 15.72 215 (9 in .) 18.27
17.03 Flernish Filled 23 .85
7.9 % 15 .10 16.89
15.92 19 .65
92 I :t:3 16 .68 275 CI I in.) 21.37
13.85 Cavity Filled 21 .09
7 .9% 13 .37 19.92
14.61 20.75
46 I :t :3 13 .79 102 .5 (4t in.) 15 .65
16 .27 Stretcher Filled 16 .68
14 .5% 17 .5 I 13 .72
15.85 15.30
15.44 Unfilled 13.79
18.0 5 12.7 5
10.41 10.75
14 .6 1 12.41
46 I : 1:6 5.94 102 .5 (4t in.) 10.48
4 .27 Stretcher Filled 9.72
14.5% 4.37 9.44
4 .86 9 .85
Table 2.4 cont
Mean brick Mortar Mean mortar Wall thickness Cross- Wall strength
strength and water cube strength and bond joints (MN/m 2 )
2
absorption (MN/m ) (mm) filled or
(MN/m 2 ) unfilled
5.37 Unfilled 8.27

4.88 10.20
4 .56 7.30
4.96 8 .61
Table 2.5
Summary of wall and specimen strengths and their relationship to the
'no faults' wall and specimen strengths. (From investigation of the
effect of workmanship and curing conditions on the strength of
brickwork , Building Development Laboratories Pty Ud, Melboumev")
Wall type Wall strength Im prisms Bond piers Strength relationship

(MN/m 2 ) (MN/m 2 ) (MN/m 2 ) to no faults specimens
Walls Im Bond
prisms piers
No faults 21.2 18.1 0.524

21.5 18.9 0.613
Average 21.4 18.5 0.565 1.00 1.00 1.00
Outside curing 19.0 15.0 0.310
20.8 18.6 0.351
Average 19.9 16 .8 0.330 0.93 0.91 0.59
Furrowed bed 16.1 15.0 0 .841
16.1 14.3 0 .792
Average 16.1 14.6 0 .813 0 .76 0.79 1.44
No perpends 21.9 19.2 0 .717
21.7 21.9 0.579
Average 21.8 20 .5 0 .648 1.02 1.11 1.15
10 mm bed 16.6 14.1 0.448
joints 15.2 14.8 0.482
Average 15.9 14.4 0.468 0.75 0.78 0.83
12 mm bow 19.8 19.0 0.620
17.5 18.8 0 .565
Average 18.6 18.9 0 .592 0.87 1.02 1.05
All faults 8.27 6 .75 0 .158
8.20 8.13 0.186
Average 8.27 7.44 0.172 0.39 0.40 0.30
formance in terms of sound insulation and resistance to rain penetration to a

serious extent.
Incomplete filling of bed joints is, from the structural point of view, much
more serious and has been investigated by the Structural Clay Products Institute
in the United States.i" and by the Building Research Laboratories in Australia.l"
In the S.C.P.I. tests , the results ofwhich are summarised in table 2.6, the 'uninspe-
ted' workmanship included unfilled vertical joints, as well as deeply furrowed bed
joints, and resulted in a reduction of strength of about 33 per cent. As it is known
that unfilled vertical joints do not affect strength significantly, it may be assumed
that most of this reduction arose from the furrowed bed joints. The Australian
tests (table 2.5) show a reduction of similar magnitude from this cause.
Table 2.6
Effect of workmanship on the compressive strength of non-reinforced
brick walls* (from Gross et al. 55)
Wall thickness and Mortar hit Workmanshipj Average Relative effect

type type ultimate stress of workman-
MN/m 2 ship
90 mm (3 .6 in.) M 22.7 I 18.4 1.00

single wythe ] U 12.2 0.66
N 22.7 I 10.8 1.00
U 6.84 0.63
200 mm (8 .0 in.) M 20.5 I 16.0 1.00
multi wythe U 11.0 0.68
(metal-tied) t
S 20.5 I 12.2 1.00
U 7.92 0.65
N 20.5 I 8.82 1.00
U 6.02 0.68
*Data from unpublished S.C.P.I. tests. Walls were tested with hinged ends; eccentricity at top
=t/6 and eccent ricity at bottom =o. Walls tested at age of 14 days. Metal-tied walls contained
one 4 .8 mm (3/16 in.) steel tie for each 2.7 sq.ft. of wall area.
tI =Inspected : U =Uninspected.
t Brick compressive strength = 81 MN/rn' , 11 760 p.s.i.
The third factor in brickwork jointing is that of thickness , which has been
already discussed in sections 2.5 and 2.6. This has been investigated by the
Building Research Laboratories in Australia and at the Universities of Edinburgh"?
and Melbourne ,' and elsewhere.! It is difficult to compare the results ofall these
investigations , but it has been shown beyond doubt that excessively thick bed
joints, say 16-19 mm, may be expected to reduce the strength of brickwork by
something of the order of 30 per cent, as compared to normal 10 mm thick joints.
This is of the same order of magnitude as the reduction caused by deep furrowing,
but excessively thick joints are at least easily seen.
Another laying defect arises from the practice of spreading too long a bed of
mortar-only sufficient mortar should be spread as will permit bricks to be set in
plastic mortar. There is, however , no quantitative data on the effect of this defect
on brickwork strength.
2.8.4 Disturbance of bricks after laying
Any disturbance ofbricks after they have been placed wilrr'esult in the bond
between bricks and mortar being broken, with possible adverse effects on strength
and resistance to moisture penetration. This commonly happens at corners, when
the bricklayer attempts to correct plumbing errors by hammering bricks into a
true plumb position, but there is no quantitative data available on the effect of
disturbance on the strength of brickwork. It is, however, related to the effect noted
by Haller,s whereby bricks with high initial rate of absorption tend to result in
lowered strength of slender walls.
2.8.5 Failure to build wall 'plumb and true to line and level'
This type of defect can give rise to eccentric loading in a wall under compression
and thus to reduce strength ; information on this is available from tests carried out
at the University of Edinburgh,"? and at the Building Development laboratories
in Australia.!" A summary of the Edinburgh results is shown in table 2.7. In these
Table 2.7
Results of 105 mrn wall tests 3 7
Wall No. Brick Mortar Brickwork Loading Remarks

strength strength strength
(N/mm 2 ) (N/mm 2 ) (N/mm 2 )
2-7 42.9 12.3 14.9 Axial Av. of 6 walls

10-11 42.9 5.96 12.7 20mm Av. of 2 walls
eccentric
12-13 42.9 5.03 12.6 20mm Av. of 2 walls
off-plurnb
tests 105 mm brickwork walls were tested in compression between reinforced

concrete slabs to give realistic end conditions. Two storey-height walls were tested
with the applied load 19 mm eccentric with respect to the axis of the wall, and
three walls were built 19 mm off-plumb . Comparing the strength of the walls with
eccentrically applied loads to corresponding axially loaded walls indicates a
reduction in strength of the order of 15 per cent ; the reduction for those built
off-plumb is similar. In the Australian tests (table 2.5) similar walls were built with
a 12 mm bow resulting in a 13 per cent strength reduction as compared to a truly
plumb wall.
A survey of ten buildings in England carried out by the Building Research
Establishment'" showed that the following levels of accuracy were attained in
brickwork construction
Wall plumb over a storey height ± 13 mm

Vertical alignment between to.p and bottom of walls of successive storeys ±
20mm
These figures are similar in magnitude to those used in the tests referred to in the
previous paragraph , which therefore give a reasonable indication of the maximum
probable reduction in strength arising from lack of plumb and vertical alignment.
2.8 .6 Failure to protect work from the weather

Newly completed brickwork can be adversely affected by exposure to unfavourable
weather conditions such as curing under very hot conditions, frost and rain
damage, and some information is available on the effect of the first two of these
conditions. The Building Development Laboratories in Australia reported'" a
series of tests on walls that were built in temperatures between 78°F and 100°F
and cured in the sun for five to six days. These walls showed about a 10 per cent
reduction in strength as compared to walls cured in the shade under polythene.
At the other end of the climatic scale, tests have been carried out in Norway
and Finland to examine the effect on brickwork properties of laying and curing
at low temperatures.!? Masor.ry piers of I m height were built in various mortars
at room temperature and in cold rooms at temperatures down to -15°C; curing
of the piers built at low temperatures was carried out at -15 °C and the results
showed , perhaps surprisingly, no deterioration in strength as between the walls
built at room temperature and those in the cold conditions (table 2.8). On the
other hand the liability of masonry built under freezing conditions to develop
undesirable deformations is pointed out , and one would suspect that this could
give rise to indirect reduction of strength as a result of bowing, or lack of plumb .
No information is available about the effects of damage by rain.
2.8.7 Overall effects of workmanship on brickwork strength

In the foregoing sections the separate effects of a number of workmanship factors
have been discussed. In any particular case, these defects will be present in varying
degrees and the overall strength of the brickwork will reflect their combined effect.
Various efforts have been made to assess the overall effect of workmanship on the
strength of brickwork, the most systematic being the programme carried out by
the Building Development Laboratories in Australia already referred to ,54 in
which controlled defects were introduced separately and in combination. As may
be seen in table 2.5, the combined effect of outside curing, deep bed furrowing,
unfilled perpends 16 mm bed joints and 12 mm bow was to reduce the wall
strength from 21.4 N/mm 2 to 8.3 Njmm" , that is, a 61 per cent reduction. This
is generally consistent with experiments at the U.S. National Bureau of Standards
where unsupervised site brickwork was from 55 to 62 per cent of the strength of
supervised brickwork. (A further study of the relative strength of supervised and
unsupervised brickwork will be found in reference 58.)
The Australian report makes the following assessment of the relative importance
of the various defects in terms of the probable reduction in strength of a wall
built under laboratory conditions
Outside curing (warm conditions) 10 per cent

Furrowed bed 25 per cent
Table 2.8
Compression tests on piers 5 7
Mortar I, KC 50/50/610
Pier no. Laid in Brick Breaking strength Cracking load
mortar type (MN/m 2 ) breaking load
1 S 9.88 0.57
2 0.53
3 S v 14.2 0.79
4 0.73
5 K2 10.3 0.59
6 0.56
7 K2 v 13.5 0.65
8 0.72
9 K1 11.1 0.64
10 0 .63
11 K1 v 13.6 0.70
12 0.73
Mortar 2, KC 35/65/520
13 S 11.1 0.55
14 0.62
15 S v 19.1 0.87
16 0.91
17 K2 t 12.3 0.63
18 0.60
19 K2 v 18.0 0.77
20 0.87
21 K1 11.7 0.57
22 0.63
23 K1 v 15.5 0.79
24 0.70
25 K1 vv 15.4 0.81
26 0.69
Mortar 3, KC 20/80/440
27 S 12.9 0.72
28 0 .70
29 S v 21.3 0 .89
30 0 .90
31 K2 13.9 0.78
32 0 .62
33 K2 v 21.9 0.87
34 0.72
35 K1 13.8 0.64
36 0.85
37 K1 v 21.6 0.82
38 0.90
S = laboratory ; K, = room at - 15° C; K1 =roorn at + 6-7 °c during laying then reduced

slowly to - 15°C; vv =warm, wet brick ; v =wet brick; t =dry brick .
16 mm thick bed joints 25 per cent
Perpend joints unfilled Nil
12 mm bow 15 per cent
It was concluded that these effects are not interactive and that the separate
factors are additive .
2.9 The deformation properties of brlckwork in compression

Knowledge of the stress-strain relationship for brickwork in compression is
frequently required in structural design, and numerous measurements have been
made on small specimens and on walls and piers to establish the nature of the
stress-strain curve and the value of Young's modulus.
The stress-strain relationship for brickwork loaded in compression to failure
has been determined for four brick types by Powell and Hodgkinson.t" The
results of these tests are summarised in table 2.9 and figure 2.18a.
Table 2.9
Stress-strain relationship for brickwork i"
(mortar-I :i :3, mean compressive strength 15.24 N/mm 2 )
Brick type Brick Brickwork E1astic modulus

compressive compressive
strength strength tangent* secantj
(N/mm 2 ) (N/mm 2 ) (N/mm 2 ) (N/mm 2 )
A-16 hole 69.6 19.93 18230 11900

perforated
B-Class A, 71.7 27.65 17370 12930
blue engineering
C-Fletton 25.5 9.33 4960 3740
D-Double frogged, 45.3 20.10 16830 11 610
stiff plastic
*Initial tangent modulus

t Secant modulus at 2/3 of maximum stress
Powell and Hodgkinson were able , by using a suitable load control technique,
to determine the stress-strain relationship past the maximum compressive stress
to failure. There was some variation of results between specimens of the same
materials , but reasonable consistency was obtained, and the curves in figure 2.18a
are average values for three tests in each case. By plotting these stress-strain curves
on a dimensionless basis (figure 2.18b); it is found that the curves for the four
types of brick are, for practical purposes, of the same form, which in turn is in
good agreement with that reported by Turnsek and CacoViC. 2 5 The relationship
is closely represented by the parabola
(2.25)
t
Stress
(N/mm 2 )
o
G m aK
20
B 1.0
:i'
~ vO
(,'1> 0<:-
O.B
'*~ ~<:-o;
~q, b($
10 0 .6 Vl
".;§<:- ~o ""l
:;<:l
0.4 q;.~~
c ~
~O~ ""l
c::
:;<:l
>
r-
o w:: '" C I o I:C
0 .002 0 .004 0.006 0 .2 0.4 0 .6 0 .8 1.0 1.2 1.4 1.6 :;<:l
(=i
St rain ch m ..
(a) :;<:l
(b)
~
~
Figure 2.18 Str ess-strain curves for brickwork in compression . (a) Four types of bri ck in 1:* :3 mortar (Powell
and Hodgkinson) ; (b) dim ensionl ess str ess-strain curv es
where o' and e' are respectively, the stress and the strain at the maximum point of
the curve. The initial tangent modulus is given by
,
o
E=2 - , (2.26)
e
and the secant modulus at 0.75a' is three-quarters of this value.
A number of authors 60,61 have related the modulus of e1asticity of brick
masonry to its compressive strength on an empirical basis. This has resulted in
values of E between 400 and 1000 times the masonry crushing strength. This kind
of relationship would of course give only an approximate estimate of the elastic
modulus since, for example, quite different combinations ofbrick strength and
mortar strength might have the same crushing strength but not necessarily the
same elastic modulus .
The difference between the initial tangent modulus and the secant modulus
at two-thirds to three-quarters of the maximum compressive strength is indicative
of the non-linearity of the stress-strain curve. A number of investigators have
noted an apparent increase in the tangent modulus for brickwork with an increase
in stress at low stress levels. This effect is mentioned by SaWin61 in relation to
lime and other relatively weak mortars, but it has also been demonstrated in high-
strength brickwork by Beard.62 The reason for an initial increase in the elastic
modulus is not altogether clear, but it is almost certainly connected with deforma-
tions in the mortar bed, possibly resulting from uneven bedding. Strain measure-
ments in walls are, in fact, usually found to be very variable both in terms of the
point or line of measurement and the general stress level.
The Evalues discussed in the previous paragraphs are the result of short-term
measurements. For long-term loading, creep strains are significant and have been
investigated by Lenczner ,63-67 whose results have shown that the ratio of long-
term to instantaneous strain is between 2 and 4 . Higher values were found for
walls than for piers, and the presence of a damp-proof course led to the
development of measurably higher creep strain, resulting no doubt from reduced
lateral restraint. Bricks with a high suction rate laid saturated showed higher creep
strain than iflaid dry. Creep was found to continue in brickwork for about one
year after loading. These observations give some guide as to the allowance which
should be made for creep in brickwork subjected to long-term loading, although
knowledge of the actual mechanism of creep deformation is as yet incomplete.
2.10 The strength and deformation of brickwork in shear

Brick masonry walls are frequently subjected to racking shear in addition to com-
pressive loads. Consequently, investigations of the shear strength of brickwork
have been undertaken in a nu mber of countries, both on small and large-scale
specimens.
The results of aseries of tests on storey-height shear walls, reported by Hendry
and Sinha,68,69 are summarised in figure 2.19 . These tests were carried out on
full-scale and on model structures built of wire-cutbricks in l :t :3 lime mortar.
~
0\
2 .0
J(M od el stru ct ures: sing le she ar panel

Sh ear • Mod el stru ctures: double shear panel
stress r • Full -scale struct ures: double shear pan els
(N/mm 2 )
y Full- scal e structures : without open ing s
+ Full -s cal e struc t ures: complete tests
CI:l
....,
::0
.R
...,
1.0 c
~
>
r-
tl:l
::0
n
;;0::
s
::0
;;0::
JC JC
o 1.0 2 .0
Precompress ion a; (N/mm 2 )
Figure 2.19 Shear strength against precompression : results of full-scale, model and small
specim en tests
The shear strength of this type of brickwork was found to be
T = 0 .3 + O.5oe N/mm 2 (2.27)
where Oe is the precompression; this relat ionshi p holds up to values of Oe = 2
N/mm 2 •
Similar tests have been carried out by Chinwah,"? by Piepe r and Trautsch." !
and by Schn eider 72 which have also led to expressions of the form T = To +[oe'
but with constants depending on the pro perties of the materials used, and
possibly on the form of the test walls and other factors. The wide range of results
obtained is indieated in table 2.10.
Table 2. 10
Variation of shear strength and precompression
T + [oe N/mm 2
= TO
Source Type of brick Mortar TO [
Hendry and Sinha Wire-cut clay 1+3 0.3 0 .5

Chinwah Wire-cut clay I :~ :3 0.25 0 .34
Pieper and Trautsch Solid sand-lime 1:2:8 0.2 0 .84
Solid sand -lime 1:0 :4 0.7 1.04
Schneider Calcium silicate I : I :6 0.14 0.3
Pieper and Trautsch showed that the shear strength decreased somcwhat as the
length of wall tested was increased , whereas it increased with thickness. It has
also been shown 76 that the degree of saturation of solid clay bricks at the time of
laying has an effect on the initial shear bond strength; the strength is very small and
variable in the case of saturated or completely dry brieks.
Failurc of shear panels is generally in diagonal tension . At low precompressions,
failure tends to develop stepwise along the brick -mortar joints on an approxi-
mately 45° angle, although the crack may run along a bed joint for some distance
at some places. As the precompression increases, the principal tensile stress causing
the crack becomes inclined at a greater angle to the horizontal and passes verti-
cally through the brieks as well as along the bed joints. A number of investigators
have suggested that the criterion of failure is the attainment of a critical value of
the principal tensile stress. Turnsek and Cacovic2 5 have prod uced evidence to show
that the appearance of the first crack in brick masonry is consistent with this
criterion, but it would appear that the value of the principal tensile stress at
failure increases with precompression . This effect, has been noted by Chinwah 7 0
and by Schneider,72 and presumably reflects the anisotropie nature of briekwork.
Diagonal splitting tests on briekwork discs, tested on several diarneters.?" clearly
indicate the increase in diagonal tensile strength as the inclination of the splitting
plane moves from 45° towards the normal relative to the bed joint.
3.0
.j:>.
00
2.0
:;. ~~\~~~
0, 0
v.J
>-l
T,
:>:'
.. Hendry & Sinha (full-scale tests)
°t + Hendry & Sinha (model tests)

:>:'
Oe x Cacov ic & Turnsek
~
:>
Q
r-
p Pieper & Trautsch t:D
:>:'
1.0 4 Yorulmaz & Atan (=)
~
~
:>:'
~
~
~.0~"<9
'1:-.-.;\0,
o 1.0 2.0 3.0 4 .0
a c/at
Figure 2.20 Failure criterion in shear with precompression

Examination of collected data from shear tests (figure 2.20) shows that
experimental results are consistent with the relationship
Tf = ..J (1 + a e ) (2.28)
at at
where Tf is the shear strength of the masonry
a e is the precompression normal to the bed joints
at is the principal tensile stress at failure
The value of the principal tensile stress at failure is not constant, as its inclina-
tion relative to the horizontal increases with increase in a e , and because of the
directional strength of properties of brickwork. Failure in combined shear and
compression must therefore be defined by a surface 8 6 •8 7 which is a function of a e ,
at , and 8 the inclination of the compressive principal stress to the horizontal , as
indicated quantitatively in figure 2.21. The problem has been studied by PageS 8
using finite element methods .
As an approximation, the following relationship, reported both by Chinwah 70
and Schneider 72 might be used
at = atO + 0.05ae (2 .29)
where atO is the value of the principal tensile stress at failure, when a e = O. As atO
is equal to the shear stress in a pure shear loading test, it would be possible to find
its value from a small-scale test, and to use equations 2 .28"and 2.29 to calculate
the strength of the brickwork in any state of combined shear and compression.
Figure 2.22 shows a suitable type of specimen that has been used by a number of
investigators, and which is easily tested without special equipment. Although the
failure produced in this specimen is by shearing of the bed joints, this has been
found to correlate satisfactorily with the diagonal tensile failure observed in panels
at low precompression .
The shear modulus of brickwork G was calculated from deflection measure-
ments made on full-scale, single-storey shear panel structures tested by Hendry
and Sinha.P? For the type of brickwork tested (31 N/mm 2 crushing strength,
wire-cut bricks set in I :4": 3 mortar) the value of G was in the region of 1500 to
2000 N/mm 2 • The value increased appreciably with precompression and, reflect-
ing the non-linear characteristics of the material, decreased with an increase of
shear stress. For approximate calculation, G might be taken as
E
- ---
2 (1 + v)
where v is Poisson's ratio, found by Hilsdorf?" to have an initial value of 0.2,
increasing to 0.35 near failure .
2.11 The strength of brickwork in tension

The tensile strength of brickwork built with normal cement :lime .sand mortar is
relatively small and variable; consequently, codes of practice discourage reliance
on this property. Nevertheless, there are occasions when tensile resistance has to
be used, particularly in the design of laterally loaded panels.
Ge VI
o
G,
(N/mm 2 ) . 8'-.... G,
0 .5
CJ:l
...,
;;tl
c
n
...,
c;;tl
>
t'"'
'\
;;tl
=
0.2 (')
'\ -,o = 2 2 t o
, ~
'\ ~;;tl
-,
~
0 .1
,,
'\
'\
,
-,
o 2 3 4 5 6 7 8 9 10
Ge (N(m m 2 )
Figure 2.21 Biaxial strength envelopes for brick panels for different bedjoin t
orientations (Samarasinghe)
, /.
.-- I
r-
I
~
. ./ I" 1/
Figure 2.22 Shear test on brick trip let specimen

Tensile stresses in brickwork may arise as a result of either in-plane or trans -
verse bending. In the case of in-plane stressing, tension is developed across the bed
joints of the brickwork and the strength is dependent on the adhesion between
bricks and mortar. There appear to be few systematic investigations of the factors
influencing the tensile bond strength of brickwork, but experiments by Grandet 7S
on the interaction between brick material and cement paste have shown that this
is critically affected by the formation of a micro-layer of ettringite (3CaS04 '
Ah 0 3 , 3CaO. 31 H2 0) at the clay-cernent interface, and by the respective mean
diameters of the pores of the brick and of the micro-crystals of the ettringite. It
is necessary for the pore size of the brick material to be greater than 0.05 mm for
a mechanical bond to be formed, and also for the cement to be properly hydra ted
behind the ettringite layer, despite the withdrawal of some of the water by
Bond tension
(N/mm 2 )
0 .5
0.4
0.3
0.2
0.1
o 2 4 6 B 10 12
Moisture content(%)
Figure 2.23 Relationship between moisture content of brick and bond tension
of brick masonry couplets (Sinha)
suction from the brick. If the brick is dry, and has a high suction rate, there will
be a partially hydrated zone in the cement paste , to a depth of several millimetres,
and possessing poor mechanical strength. The movement of water between the
brick and the mortar, and the resulting effects on the development of the
mechanical bond between these components, was shown to be considerably
affected by their specific surfaces and capillary dimensions . Grandet concluded
that it is possible on the basis of these observations to obtain some indication of
the likely behaviour of cement mortars, which would, however , also be influenced
by such other parameters as the compaction of the mortar, its cement content
and water retentivity.
The results of various phenomenological investigations 76, 77 have confirmed that
the moisture conten t 0 f the bricks at the timeoflaying is of importance in deterrnin-
(J
•
,., 1
1:
3.0 1 :5/ •
•
~
l- G
o •
I e
1 :2
Q
/
I /
2.0
1 • /
/
.~ "
G~
/
I
°parallfll
(N/mm 2 )
p /
/
G) B.C.R.A. 1 : *: 3
• B.C.R.A. 1 : 1 :6
1.0 + Sinha & Hendry
4 Satt i 1 : * : 3
ä Satti 1 :1 :6
~ Lawrence
J;1 Lawrence & Morgan

GI Baker
4' James
o 0 .2 0.4 0 .6 0 .8 1.0 1.2

2
a no,mal (N/mm )
Figure 2.24 Moduli of rupture of brickwork in orthogonal directions

THE STRENGTH OF MATERIALS S3
ing the tensile bond strength. The results of experiments by Sinha are shown in
figure 2.23. The extreme variability of tensile bond strength is immediately
apparent in this diagram, and although there is no clear relationship between
moisture content and tensile bond strength, it will be noted that only very low
values were found as the bricks approached their saturation moisture content.
Flexural tensile strength, as defined by the modulus of rupture, is of greater
practical importance than direct tensile strength and has been investigated in some
detail in relation to the resistance of wall panels to wind loads.78 - 8 s The flexural
8 +
o normal
a parallel
7
(!) B.C.RA 1 : t: 3
• B.C.R.A. 1 : 1 : 6
+ Sinh a 1 : 1 : 6
6 " Satti
" Lawrence 1 : 1 : 6
J( Lawren ce & Mo rgan
EI Baker
5
• ~James
..
4
41 et
~ es
• "
.
11
3 CI
GI
e CI •
, •
•
• et
Q •
2 e
o 0 .5 1.0 1.5
(J nonn.' (N/mm 2
)
Figure 2.25 Ratio of moduli of rupture in orthogonal directions

54 STRU CTURAL BRICKWORK
strength of brickwork is, of course , different for bending in a plane at right angles
to the bed joints, and for bending in a plane parallel to this direction , being
several tirnes greater in the latter case. The ratio is not constant but varies with
the strength achieved. Thus in figure 2.24, values from a numb er of sources of the
two moduli of rupture are plotted against each other, and in figure 2.25 their ratio
is plotted against that for bending across the bed joints. These plots reveal the
great variability of flexural tensile strength, but a definite trend can be distin-
guished in the value of the orthogonal strength ratio , which decreases markedly
with increase in the flexural tensile strength across the bed joints.
Many investigators have attempted to establish a relationship between material
properties and flexural tensile strength. Thus West8 3 has shown correlations of
flexural strength with suction rate and with water absorption , for thirty-three
different bricks and two mortar mixes. None of the correlations are very elose
but, in statistical terms , it was possible to find a relationship between flexural
tensile strength and water absorption, and this is illustrated in figure 2.26 . West's
results indicated that for flexure parallel to the bed joints there was not a great
difference in strength between 1:* :3 and 1: 1:6 mortars, although a difference
was noticeable in the orthogonal direction. James'" has reported a similar result,
as may be seen from table 2.11.
These results also indicate the high variability of flexural strength, particularly
in the weaker direction, and suggest that the form of test specimen is important.
This latter effect may be seen from the difference in results for flexure parallel to
the bed joint obtained from specimens respectively three course and four courses
in width , the latter giving consistently lower values.
Flexural
stren gt h
(N/ m m 2 ) + Undocked
• Oocked
...,
1.2
+
... +
0 .8 J-l
... ,
..... ••
0.4
' ,
.... ... ...
.........
- - 95% confidence
limit
o 10 20 30
Water absorpt ion (%)
Figure 2.26 Flexural tensile strength of brickwork related to water absorption.

Specimens tested in flexure parallel to bed joints ; mortar mix 1:* :3 (West)
Table 2.11
Flexural tensile strengths of small specimens
Mortar Statistic Flexural strengths ( N rrnm")

Normal to bed joi nt Parallel to bed joint
Sta ck prism 3-course 4-course
specimen specimen
1:2 :9 Mean 0.39 2.08 1.78

C. of Y. % 23.3 20 .6 26.1
1: 1:6 Mean 0 .594 2.40 2.03
C. ofY. % 22.9 15.5 18.5
1:* :3 Mean 0.984 2.7 4 2.29
C. ofY. % 25.4 18.0 16.5
Fixed fram e
___- - S crew c1amp
I.~-+-Jo int be ing tested
IL-::H--" F
Screw c1amp
Figure 2.27 Arrangement for multiple flexural tensile tests on brickwork

specimen (Baker)
Baker 8 S has praposed a test on a sta ck prism in which each joint is tested in
turn using the type of rig shown in figure 2.27. This permits the testing of a fairly
large number of joints with reasonable economy , and thus to the definition of a
statistically significant characteristic strength for use in structural design .
References
I. A. J. Francis, C. B. Horman and L. E. Jererns, 'The Effect of Joint Thickness
and other factors on the Compressive Strength of Brickwork', Proceedings of
the Second International Brick Masonry Conference (Stoke-on-Trent) 1971,
ed . H. W. H. West and K. H. Speed (British Ceramic Research Association ,
Stoke-on-Trent, 1971) pp. 31-7.
2. N. F . Astbury and H. W. H. West, 'Tests on Storey-height Brickwork Panels
and Development of Site Contral Test for Brickwork' , in Designing, Engineer-
ing and Constructing with Masonry Products , ed. F. B. Johnson (Gulf,
Houston, Tex., 1969) pp . 216-20.
3. C. B. Monk, 'A Historicai Survey and Analysis of the Compressive Strength
of Brick Masonry', Research Report No. 12 (Structural Clay Products
Research Foundation, Geneva, Ill., 1967).
4. E. H. Morsy, 'An Investigation of Mortar Properties Influencing Brickwork
Strength', Ph.D. Thesis, University of Edinburgh , 1968.
5. P. Haller, 'The Physics of the Fired Brick : Part One Strength Properties',
Libr. Commun. Bldg. Res. Stn, 929 (1960) (trans. G. L. Cairns).
6. D. Lenczner , Elements of Load Bearing Brickwork (Pergamon, Oxford,
1972).
7. H. K. Hilsdorf, 'An Investigation into the Failure Mechanism of Brick
Masonry Loaded in Axial Compression', in Designing, Engineering and
Constructing with Masonry Products, ed . F. B. Johnson (Gulf, Houston ,
Tex. , 1969) pp. 34-41.
8. C. L. Khoo and A. W. Hendry , 'A Failure Criterion for Brickwork in Axial
Compression', Proceedings of the Third International Brick Masonry
Conference (Essen) 1973, ed. L. Foertig and K . Gobel (Bundesverband der
Deutschen Ziegelindustrie , Bonn , 1975) pp. 139-45 .
9. C. L. Khoo and A. W. Hendry , 'Strength Tests on Brick and Mortar under
Complex Stresses for the Development of a Failure Criterion for Brickwork
in Compression ' ,Proc. Br. ceram. Soc., 21 (1973) 57-66.
10. G. G. Balmer, 'Shearing Strength of Concrete und er High Triaxial Stress,
Computation of Mohr's Stress Envelope as a Curve' , Report No. 5P23
(Bureau of Reclamation , Denver , Colo., 1949).
11. T. N. W. Akro yd, 'Concrete under Triaxial Stress', Mag. Concr. Res. , 13
(1961) 111. .
12. N. J. Gardner , 'Triaxial Behavior of Concrete ',J. A m. Concr. Inst., 66
(1969) 136.
13. C. L. Khoo, 'A Failure Criterion for Brickwork in Axial Compression', Ph.D .
Thesis, University of Edinburgh, 1972.
14. - 'Small Scale Specimen Testing' , National Testing Program (Structural
Clay Products Research Foundation, Geneva, Ill., 1964).
15. G. W. Anderson , 'Sta ck Bonded Small Specimens as Design and Construc-
tion Criteria', Proceedings of th e Second International Brick Masonry
Conference (Stoke-on-Trent) 1971, ed. H. W. H. West and K. H. Speed
(British Ceramic Research Association , Stoke-on-Trent , 1971) pp. 38-43.
16. F. G. Thomas, 'The Strength of Brickwork' ,Struct. Engr., 31 (1953) 35-46.
17. P. Haller, 'Load Capacity of Brick Masonry', in Designing, Engineering and
Constructing with Masonry Products, ed. F. B. Johnson (Gulf, Houston ,
Tex ., 1969) pp . 129-49.
18. - 'Compressive and Transverse and Racking Strength Tests in Four-inch Brick
Walls', Research Report No. 9 (Structural Clay Products Research Founda-
tion, Geneva, Ill., 1965).
19. - 'Compressive and Transverse Strength Tests of Eight-inch Brick Walls',
Research Report No. 10 (Structural Clay Products Research Foundation,
Geneva, BI., 1966).
20 . - 'Compressive and Transverse Tests of Five-inch Brick Walls', Research
Report No . 8 (Structural Gay Products Research Foundation, Geneva, Ill.,
1965).
21 . W. Albrecht and H. Schneider, 'Der Einfluss der Saugfähigkeit der
Mauerziegel auf die Tragfähigkeit von Mauerwerk', Heft 46 Berichte aus
der Bauforschung (Wilhelm Ernst , Berlin, 1966).
22 . H. Motteu, ' Research on Load-bearing Masonry in Belgium since 1963', in
Designing, Engineering and Constructing with Masonry Products , ed . F. B.
Johnson (Gulf, Houston, Tex ., 1969) pp. 171-84.
23. J . A. James, 'Investigation of the Behaviour of Single Leaf, 9" and 11"
Cavity Storey Height Walls under Axial Load', Report W/3/A (Building
Development Laboratories, Morley , W. Australia, 1972).
24. J . A. James, ' Investigation of the Behaviour of Storey Height Single Leaf
Walls, 9" Walls and 11" Cavity Walls under Eccentric Compressive Load ',
Report W/4/A (Building Development Laboratories, Morley, W. Aust ralia,
1973). v
25 . V. Turnsek and F. Caöovic, 'Some Experimental Results on the Strength of

Brick Masonry Walls', Proceedings 0/ the Second International Brick
Masonry Conference (Stoke-on-Trent) 1971, ed . H. W. H. West and K. H.
Speed (British Ceramic Research Association , London, 1971) pp . 149-56.
26. S. SaWin, Structural Masonry (Prentice-Hall, Englewood Cliffs, N.J ., 1971)
pp. 36-8.
27 . T. Grimm, 'Strength and Related Properties of Brick Masonry' ,J. struct.
Div. Am. Soc . civ. Engrs., 107 (1975) 217 -32.
28. L. G. Sims, 'The Strength of Walls Built in the Laboratory with some Types
of Clay Bricks and Blocks',Proc. Br. ceram Soc ., 4 (1965) 81-92.
29. J. Y. Houston and C. T. Grimm , 'Effect of Brick Height on Masonry Com-
pressive Strength',J. Mater. ASTM, 7 (1972) 388-92 .
30 . H. W. H. West, M. R. Hodgkinson, D. G. Beech and S. T. E. Davenport, 'The
Comparative Strength of Walls Built of Standard and Modular Bricks' Pro-
ceedings 0/ the Second International Brick Masonry Conference (Stoke-on-
Trent) 1971 , ed . H. W. H. West and K. H. Speed (British Ceramic Research
Association , Stoke-on-Trent , 1971) pp. 172-6.
31. P. Szabo, 'The Effect of Brick Size on the Load Bearing Capacity and
Shear Strength of Masonry Construction , Proceedings 0/ the Third
International Brick Masonry Conference (Essen) 1973, ed. L. Foertig and
K. Gobel (Bundesverband der Deutschen Ziegelindustrie, Bonn , 1975)
pp . 124-6.
32. H. W. H. West, H. R. Hodgkinson, D. G. Beech and S. T. E. Davenport,
'The Performance of Walls Built of Wirecut Bricks with and without
Perforations', Parts i and H, Proc. Br. ceram. Soc., 17 (1970) 1-39.
33. A. Ilantzis , 'Mechanical Strength of Walls of Hollow Bricks and Hollow
Blocks of light and Heavy Aggregate under Axial and Eccentric Load',
Proceedings 0/ the Third International Brick Masonry Conference (Essen)
1973, ed . L. Foertig and K. Gobel (Bundesverband der Deutschen
Ziegelindustrie, Bonn, 1975) pp . 115-19.
34 . A. Huizer and M. Ward, 'The Effect of Brick Type on the Compressive
Strength of Masonry', Proceedings of the Fourth International Brick
Masonry Conference (Brugge) 1976, Paper 4.a.2.
35 . W. Albrecht and H. Schneider, 'Der Einfluss des Mauerverbandes von 30 cm
dicken Hochlochziegel werden auf deren Tragfahigkeit', Heft 46 Berichte aus
der Bauforschung (Wilhelm Ernst, Berlin, 1966).
36 . S. Prasan, A. W. Hendry and R. E. Bradshaw , 'Crushing Tests on Storey
Height Walls 4t" Thick',Proc. Br. ceram. Soc., 4 (1965) 67-81.
37 . R. E. Bradshaw and A. W. Hendry, 'Further Crushing Tests on Storey
Height Walls4t" Thick ', Proc. Br. ceram. Soc., 11 (1968) 25-54.
38 . R. E. Bradshaw and A. W. Hendry , 'Preliminary Crushing Tests on Storey
Height Cavity Walls', in Designing, Engineering and Constructing with
Masonry Products, ed. F. B. Johnson (Gulf, Houston, Tex ., 1969) pp . 101-9 .
39. K. Fisher , 'The Effect of Wall Ties on the Compressive Strength of Cavity
Walls', Proceedings of the Second International Brick Masonry Conference
(Stoke-on-Trent) 1971, ed. H. W. H. West and K. H. Speed (British Ceramie
Research Association, Stoke-on-Trent, 1971) pp . 177-85.
40. A. W. Hendry, R. E. Bradshaw and D. J. Rutherford, 'Tests on Cavity Walls
and the Effect of Concentrated Loads and Joint Thickness on the Strength
of Brickwork ', Res. Note Clay Prod. tech. Bur. (Lond.), 1 (1968).
41. R. Beard, 'The Compressive Strength of some Grouted Cavity Walls', Proc.
Br. ceram. Soc ., 21 (1973) 113-40.
42 . - ' Loading Tests on Brick Walls Built in Stretcher Bond, Res. Note Clay
Prod. tech. Bur. (Lond.), 1 (1968).
43. K. Fisher, 'The Effect of Chasing on the Compressive Strength of Brickwork',
Proceedings of the Third International Brick Masonry Conference (Essen)
1973, ed . L. Foertig and K. Gobel (Bundesverband der Deutschen Ziegel-
industrie, Bonn, 1975) pp . 106-14.
44. K. Fisher, 'The Effect of Low-Strength Bricks in High-strength Brickwork',
Proc. Br. ceram. Soc., 21 (1973) 79-98 .
45 . i
A. W. Hendry and C. K. Murthy, 'Comparative Tests on ~ and Scale Model
Brickwork Piers and Walls', Proc. Br. ceram. Soc., 4 (1965) 45-66 .
46. B. P. Sinha and A. W. Hendry , 'The Effect of Brickwork Bond on the Load
Bearing Capacity ofModel Brick Walls', Proc. Br. ceram. Soc., 11 (1968)
55-68 .
47 . B. P. Sinha and A. W. Hendry, 'Further Tests on Model Brick Walls and
Piers',Proc. Br. ceram. Soc., 17 (1970) 83-96 .
48. - 'Bearing Stress Investigation : Report on Bearing Pressures on Brick Walls',
Struct. Engr., XVI (1938) 242-68.
49. R. Hiltscher and G. Florin, 'Photoelastic Investigation of Stress Concentra-
tion, Splitting and Tear Tensile Stresses in Reetangular Plates Loaded at
Varying Distances from the Plate Corner', Technical Translation 1137
(National Research Council of Canada, Ottawa, 1964).
50 . A. C. Armstrong and A. W. Hendry , 'The Compressive Strength of Brick-
work with Reinforced Bed Jo ints', Technical Note 209 (British Ceramic
Research Association, Stoke-on-Trent , 1973).
51 . - 'Model Specification for Load Bearing Clay Brickwork', Special
Publication 56, Revised Edition (British Ceramic Research Association,
Stoke-on-Trent, 1975).
52. J. D. McIntosh, 'Specifying the Quality of Bedding Mortars', Proc. Br.
ceram. Soc., 17 (1970) 65-82.
53. - 'Investigation of the Effect on Brickwork of not Filling Vertical Mortar
Joints', Internal Report (British Ceramic Research Association, Stoke-on-
Trent, 1972).
54. J . A. James, 'Investigation of the Effect of Workmanship and Curing
Conditions on the Strength of Brickwork',Proceedings ofthe Third
International Brick Masonry Conference (Essen) 1973, ed. L. Foertig and
K. Gobel (Bundesverband der Deutschen Ziegelindustrie, Bonn, 1975)
pp . 192-201.
55. J. G. Gross, R. D. Dikkers and J. C. Grogan, Recommended Practice for
Engineered Brick Masonry (Structural Clay Products Institute, McLean, Va.,
1969).
56. R. M. Milner and R. P. Thorogood, 'Accuracy of Load Bearing Brick Con-
struction and its Structural Implications', Proc. Br. ceram. Soc. , 21 (1973)
231-42.
57. S. D. Svendsen and A. Waldum, 'Some Remarks on Winter Masonry',
Technical Translation 1456 (National Research Council of Canada, Ottawa,
1971 ).
58 . I. C. McDowall, T. N. McNeilly and W. G. Ryan, 'The Strength of Brick
Walls and Wallettes', Special Report No. 1 (Building Development Research
Institute, Melbourne, 1966).
59 . B. Powell and H. R. Hodgkinson , 'The Determination of Stress/Strain
Relationship of Brickwork', Proceedings of the Fourth International Brick
Masonry Conference (Brugge) 1976, Paper 2.a.5.
60 . J. M. Plowman, 'The Modulus of Elasticity of Brickwork ', Proc. Br. ceram.
Soc., 4 (1965) 37-44.
61 . S. Sahlin, Structural Masonry (Prentice-Hall, Englewood Cliffs, N.J., 1971)
pp. 59-61 .
62. R. Beard, 'Compressive Strains in Bricks and Brickwork of a 4t in Test Wall
under Axial Loading',Proc. Br. ceram. Soc., 17 (1970) 137-54.
63. D. Lenczner, 'Creep in Brickwork', Proceedings ofthe Second International
Brick Masonry Conference (Stoke-on-Trent) 1971 , ed . H. W. H. West and
K. H. Speed (British Ceramic Research Association, Stoke-on-Trent, 1971)
pp . 44-9.
64. D. Lenczner, 'Creep in Brickwork with and without Damp Proof Course',
Proc. Br. ceram. Soc., 21 (1973) 39-49 .
65 . D. Lenczner, J. Salahuddin and K. Wyatt, 'Effect of Stress on Creep in
Brickwork Piers',Proc. Br. ceram. Soc., 24 (1975) 1-10.
66 . K. Wyatt, D. Lenczner and J. Salahuddin, 'The Analysis of Creep Data in
Brickwork', Proc. Br. ceram. Soc., 24 (1975) 11-21.
67. D. Lenczner and J. Salahuddin, 'Creep and Moisture Movements in Masonry
Piers',Proceedings ofthe First Canadian Masonry Symposium (Calgary)

1976 ,pp.72-86.
68. B. P. Sinha and A. W. Hendry , 'Racking Tests on Storey-Height Shear Wall
Structures with Openings Subjectedto Pre-compression' , in Designing,
Engineering and Constructing with Masonry Products , ed. F. B. Johnson
(Gulf, Houston , Tex. , 1969) pp. 192-9.
69. A. W. Hendry and B. P. Sinha, 'Shear Tests on Full Scale Single Storey
Brickwork Structures Subjected to Precompressing', Civ. Engng publ. Wks
Rev. , 66 (1971) 1339-44.
70. J. C. G. Chinwah, 'Shear Resistance of Brick Walls', Ph.D. Thesis,
University of London , 1972.
71 . K. Pieper and W. Trautsch , 'Shear Tests on Walls, Proceedings of the Second
International Brick Masonry Conference (Stoke-on-Trent) 1971 , ed. H. W. H.
West and K. H. Speed (British Ceramic Research Association, Stoke-on-
Trent, 1971) pp. 140-3 .
72. H. Schneider, 'Tests on Shear Resistance of Masonry', Proceedings of the
Fourth International Brick Masonry Conference (Brugge) 1976, Paper 4.b.12.
73. F. B. Johnson, 'Development of Diametral Testing Procedures to Provide a
Measure of Strength Characteristics of Masonry Assemblages', in Designing,
Engineering and Constructing with Masonry Products, ed. F. B. Johnson
(Gulf, Houston , Tex., 1969) pp. 51-7 .
74. H. K. Hilsdorf, 'Untersuchungen über die Grundlagen der Mauerwerkfestig-
keit' , Bericht Nr. 40 Materialsprüfungsamt für das Bauwesen der Technischen
Hochschule (Technischen H/schule , München, 1965) .
75. B. Grandet , 'Physico-Chemical Mechanisms of the Bond between Clay and
Cement', Proceedings of the Third International Brick Masonry Conference
(Essen) 1973 , ed. L. Foertig and K. Gobel (Bundesverband der Deutschen
Ziegelindustrie, Bonn, 1975) pp. 217-21.
76. B. P. Sinha, 'Model Studies Related to Load Bearing Brickwork', Ph.D.
77 . S. V. Polyakov, 'Masonry in Framed Buildings', Gosudalst-Vennoe Izdatel'
stvo Literature po Straitel'stvu i Arkitecture, Moskva, 1956, trans. G. L.
Cairns (Building Research Station, Watford , Herts., 1963).
78. S. J . Lawrence and T. W. Morgan, 'Strength and Stiffness of Brickwork in
Lateral Bending',Proc. Br. ceram. Soc., 24 (1975) -79-90.
79. S. J. Lawrence, 'Flexural Strength of Brickwork Normal to and Parallel to
the Bed Joints',]. Aust. ceram. Soc., 11 (1975) 5-6.
80 . B. P. Sinha and A. W. Hendry, 'Tensile Strength of Brickwork Specimens',
Proc. Br. ceram. Soc., 24 (1975) 91-100 .
81. A. W. Hendry, 'The Lateral Strength of Unreinforced Brickwork', Struct.
Engr, 57 (1973) 43-50.
82 . L. R. Baker, 'Brickwork Panels Subjected to Face Wind Loads, M.Eng.Sci.
Thesis, University of Melbourne , 1972.
83. H. W. H. West, 'The Flexural Strength of Clay Masonry determined from
Wallette Specimens', Proceedings of the Fourth International Brick Masonry
Conference (Brugge) 1976 , Paper 4.a.6.
84. J. A. James, 'An Investigation of the Lateral Load Resistance of Walls of
Unreinforced Brickwork withoui Precompression Built in Clay Bricks',
Report W/Lat/1 (Building Development Laboratories, Morley, W. Australia) .
85 . L. R. Baker, 'Variability Aspects ofthe Flexural Strength of Brickwork' ,
Proceedings of the Fourth International Brick Masonry Conference (Brugge)
1976, Paper 2.bA .
86 . W. Samarasinghe, 'In Plane Strength of Brickwork', Ph.D. Thesis, University
of Edinburgh, 1980 .
87 . A. W. Page, W. Samarasinghe and A. W. -Iendry, 'On the Failure of
Masonry Shear Walls',Int. J. Masonry Constr. , 1 (1980).
88 . A. W. Page, 'A Model for the In-plane De formation and Failure of Brick-
work', Engng Bull., CE.8 (University of f'ewcastle , N.S.W. , Faculty of
Engineering) 1978 .
3 THE STRENGTH OF BRICK MASONRY
COMPRESSION ELEMENTS
3.1 Factors affecting the compressive strength of walls and piers

The discussion on the compressive strength of brick masonry in chapter 2 related
to short piers or columns , axially loaded. The strength of a wall of a given type of
brickwork will be influenced by the eccentricity of loading and the slenderness
ratio , which in turn depend on the geometry of the building, the relative stiffnesses
of the walls and floors, the nature of the joints between them and the distribution
of the loads. The calculation of the strength of brickwork compression elements is
further complicated by the low tensile strength of the material, which may crack
unde r certain loading conditions, leading to variations in effective sectional
properties.
Structural design is currently based on the results of tests on walls and piers of
various slenderness ratios and eccentricities, and with idealised end conditions.
The designer has then to make allowance for the actual end conditions by estimat-
ing the effective height of the wall or column and the eccentricity of loading on it,
generally on the basis of conventional rules or judgement. Covered by large safety
factors, these rather crude methods have given satisfactory results, but more
rational design methods are clearly desirable and will be discussed in this and the
following chapter .
3.2 Empirical studies of the strength of walls and piers

With the object of providing design data for structural codes of practice , generally
in the form of reduction coefficients on the basic masonry compressive strength to
allow for slenderness and eccentricity, tests have been carried out in many countries
over a long period of time. As mentioned in section 3.1 , these tests have achieved
their purpose and have resulted in a fairly reliable knowledge of the effect of the
primary factors affecting the strength of simple walls and piers.'
The literature on the subject shows that a large number of tests have been
carried out on axially loaded walls of varying slenderness ratio . Some of the
results, from references 2 to 8, are plotted in figure 3.1; these show a considerable
scatter which appears to increase with slenderness ratio . The reason for the scatter
is, presumably, that in testing walls it is difficult to avoid small unintentional
• Eclilllllll \ll1 U uivnr s itv
o S.C.P.R.F.. U .S.Ä .
• S.C.P.R.F.. U. S.Ä.
D B.R.S.. U.K .
R t'dU i"11\ ' fl

a
f.h"h' l
a
a
. • o a
- ••...ll...-. __." • a ""l
-.• ~. o :I:
t'r:I
• o CI>
""l
• ;:tI
_0_ -. .
l' ~ a. • a •• eid = 0 t'r:I
• Z
-- --- -
a Cl
-.- •• • -- ""l
:I:
l' l~
• o'Tl
I;l:l
;:tI
- -- ---.-
II ..
n
;><:
E::
eid = 1/6 ;I-
CI>
-- -- -- -.-
--.--- o
o .; Z
-- - -- --.- ;:tI
-<
= eid 1/3
---- - --.-
o 6
-- -- -- - 12 18 24
Si enderness rat io
Figure 3.1 Reduction factor agairrst slenderne ss rat io (Hasan)

0-
W
eccentricities, the effect of which increases with slenderness. There are fewer
available results for walls tested with eccentric loading, but some have been
plotted in figure 3.1. In the case of these results the scatter is rather less, no doubt
because small experimental deviations from a finite eccentricity are of less import-
ance than they are in the case of nominally axial loading,
It is found from wall tests that up to slenderness ratios (that is, ratio of height
to thickness) approaching 30 , which is a practical limit, failure under axial load is
usually limited by the strength of the material rather than by buckling . The walls
fail in this case by the development ofvertical cracks, as shown in figure 3.2.
Figure 3.2 Typical splitting failure of wall at failure in cornpression

THE STRENGTH OF BRICK MASONRY 65
Figure 3.3 Incipient buckling failure of asiender wall

Under eccentric load asiender wall may show considerable lateral deflection
(figure 3.3) before failure, which takes place by catastrophic collapse (figure 3.4).
These tests have been carried out either with hinged end conditions or 'flat
ends', the latter meaning that the walls have been tested between the rigid platens
of the testing machine . There have inevitably been differences in the test con-
ditions and in the interpretation of the results, as reflected in the reduction
factors used in the various national codes shown in figure 3.s .
3.3 Theoretical studies of the strength of compression elements

Although the strength of hinged-end compression elements is satisfactorily
described by test results, this is not really a sufficient basis for the design of walls
in an actual structure. Such walls are never hinged, and for a given load condition
their strength is critically influenced by the stiffness of contiguous walls and floor
slabs. In conventional design calculations, the real condition in the structure is
related to the hinged-end situation by 'guesstimation' of an equivalent height of
the member and of the eccentricity of the loading. To progress beyond this
empirical method of design, it is necessary to examine theoretically the
mechanics of brittle material columns.
Figure 3.4 Collapse of wall shown in figure 3.2 at ultimate load
- - U.SA S.C.P.R.F.
_ . _ .- W . Germany
- --- U.K. CP 111

Reduction
factor
<,
e 1 ......
d 3
o 6 12 18 24
Sienderness ratio
Figure 3.5 Comparison of reduction factors for slenderness and eccentricity in

various national codes
3.3 .1 Differential equation for brittle columns
Theories for the elastic buckling of brittle columns were first developed by
Royen? and by Angervo;'? the latter's theory was based on the solution ofthe
differential equation for a column without tensile strength. A similar solution by
Chapman and Slatford 11 considered pinned and fixed-end columns, with an
initial deformation that increased linearly from the ends to mid-height, and an
initially straight pinned-end column with eccentric loading . Restricting attention
to the last case, it will be observed that if an eccentric load is applied within the
middle third of the section, tension in the material will be produced only when
lateral deflection of the column causes the line of thrust to move out of the middle
third. As indicated in figure 3.6a, a cracked zone will extend over part of the
height of the column and the effective section will be reduced . (If the eccentric
load is applied outside the middle third of the section, the cracked zone will
extend over the whole height of the column.) The stress distributions at various
seetions of the column will be as shown, and are assumed to be linear. Failure of
p p
e
h
y ~---+--+--
p
P
(a) (b)
Figure 3.6 Eccentrically loaded pinned-end column of brittle material

the column will take place when the cracked zone reaches the line of thrust; at
the moment of collapse, a hinge forms at the mid-height of the column, and the
line of thrust passes through it as shown in figure 3.6b . Following the solution
proposed by Chapman and Slatford it is necessary to consider two differential
equations : one for the uncracked parts of the column, and the other for the
cracked length. Thus, for the uncracked part
d2 y
EI - -2 +P(e p +y)=O (3.1)
dx
where the symbols have the meaning shown in figure 3.6a, E is Young's modulus
and I the second moment of area . The reduced depth of a section within the
cracked part is
d' = 3 (t - y - ep )
and the corresponding moment of inertia is
I (:J
The eccentricity of the line of thrust with respect to the cracked section is
, d'
e =-
6
Therefore, the differential equation at any cracked section is
or
t
EI -
d' )
d
3 d2y
-
dx2
+Pe =0
,
2
EI d y + - - - - - - - =0 (3.2)
dx2 54 [(d/2) - Y - epl 2
Substituting z = [(d/2) - Y - epl , d 2 z/dx 2 =- d 2 y/dx2 and equations 3.1 and 3.2
become respectively
(3.3)
and
d2 Z Pd3
EI - - - = O (3.4)
dx 2 54z 2
The general solutions of equations 3.3 and 3.4 are, respectively
d
z = A cos IJX +B sin IJX + -
2
THE STRENGTH OF BRICK MASONR Y 69
and
f (c-Fz)~ = (Fc d)f

3" J.IX + G
in which Jl = V(P/EI) and A, B, C, Fand Gare constants determined by the end
conditions and the conditions at the common sections between the cracked and
uncracked parts of the column. Evaluation of these constants leads to the
following equations
a-_ -piz -_ (d)-f

- f (d)
- +sin- (4d 1 6e-
- - 3 )-~ -sin- - 1
p(4d
--3 ~ ~ (3.5)
2 3z c 3z c 3z c d 3z c
The maximum value of P occurs when da/dz c =0, and then zc/d and ep [d satisfy
the equation
Using the above eq:::alions, Chapman and Slatford were able to obtain, in
dimensionless terms , the load-deflection curves for eccentrically loaded brittle
columns reproduced in figure 3.7. In this diagram the load on the column is
expressed as a fraction of the Euler load. The broken lines indicate the behaviour
Cracking line
1.0 ......- _ _1,--,'"--
P
0 .8
0 .6
0.4
0 .2
o 0 .1 0 .2 0 .3 0 .5
Yc + ep
d
Figure 3.7 Load-deflection curves for column with eccentric load: broken lines
refer to wholly elastic columns (Chapman and Slatford)
of an elastic column with tensile resistance; compared to such a member the load-
deflection curve for a brittle column reaches a maximum. The theoretical collapse
occurs when the sum of the central deflection and the initial eccentricity become
equal to half the depth of the section. The vertical chain dotted line indicates the
onset of cracking , which of course , is exceeded as soon as any load is applied with
an eccentricity greater than df6.
Pm .. 1.0
0.4
o 0 .1 0 .2 0.3 0.4 0.5

e
d
Figure 3.8 Maximum load for column with load eccentricity (Chapman and
Slatford)
A second diagram from the Chapman and Slatford analysis, figure 3.8, shows
the buckling load of the column as a fraction of the Euler load , plotted against the
eccentricity ratio, emphasising the critical importance of eccentricity as a factor
influencing the strength of brickwork elements in compression. From this
diagram and on the assumption, based on experimental evidence, that axially
loaded walls having a slenderness ratio (heightfthickness) exceeding about 25 fall
by buckling, it .is possible to derive a set of reduction factors for slenderness and
eccentricity. This has been done, with the results shown in figure 3.9 . The curves
to the right of the diagram indicate buckling fallures as derived from figure 3.8,
taking the Euler criticalload at a slenderness ratio of 25 as unity and equating it
to the compression failure stress for a short column. The reduction factors for
short columns with various degrees of eccentricity have been calculated by con-
sidering the combined bending and axial stresses, and finding the reduced load
necessary to produce unit maximum stress in the material. Although very approxi-
mate, these curves serve to illustrate the behaviour of masonry elements in
comprcssion and generally correspond with experimental results .
Angervo's solution to the differential equation for a pinned-end brittle column
was based on linear and non-linear stress-strain curves, and Sahlin 12 extended the
analysis to inc1ude for the possibility of the eccentricity at each end of the
column being different.
e/d =O
1.0 I •
Stress
reduction
factor
0 .8
o-i
::r:
tT'l
0 .1 CIl
o-i
0 .6 :;l:l
tT'l
Z
0 .2 ~
::r:
004 o'Tl
0 .3 tI:l
:;l:l
()
0.2 ~
s::
>
CIl
o
Z
:;l:l
o 10 20
-<
30 40 50
h/d
Figure 3.9 Stress reduction factors for pinned-end brickwork compression elements with various eccentrici-
ties of loading
-....I
3.3.2 Solutions assuming dejZection and stress-strain curves
Haller's theory" is based on the assumption of a .sinusoidal lateral deflection curve
and a non-linear stress-strain relationship derived from tests on small masonry
prisms. The solution is in less formal mathematical terms , and results in expressions
for calculating the maximum load on the element which are rather less cumber-
some than those resulting from the general solution of the differential equation .
Haller's solution showed good agreement with experimental results, provided that
an additional initial eccentricity of 1/1000 of the height of the test column was
added to the eccentricity of the load . Although the experimental stress-strain
curve used by Haller was very close to parabolic, the substitution of the latter
relationship in Haller's calculation, rather surprisingly, appears to result in a con-
siderable difference in the estimated mean failing stress as compared to that
derived on the basis of the experimental curve.
Haller's theory related to specific masonry specimens and was not generalised;
other investigators however, including Monk'? and Turkstra'" have derived general
theories based on non-linear stress-strain relationships. The degree of non-linearity
is defined by a factor k = E oo'le' where E o is the initial tangent modulus, o' the
maximum stress and e' the corresponding strain . Thus k is the ratio of the elastic
strain to the ultimate strain at failure stress. These theories demonstrate (as did
Sahlin's) that the wall strengthis a function ofa non-dimensional parameter
(hld y( a'IEo) . Evaluation of wall strength at a given eccentricity ofloading thus
requires knowledge of three parameters of the stress.strain curve, namely o', Eo
and k. The curves shown in figure 3.10, calculated by Turkstra, show the effect
ofvariation in the parameter k with slenderness ratio, assuming a particular value
of a'IEo. Turkstra obtained reasonable agreement between test and theoretical
results using values of y(E 0 Io') of 15 and 19.4 , and of k in the range 1.0 to 1.5.
As shown in chapter 2 (p. 44) , accurately deterrnined stress-strain curves indicate
a second degree parabolic form, that is, k =2, in which case, Eola' =2Ie'. The
maximum strain e' is on average about 0 .003, suggesting a value of the order of
24 for Y(Eola'). Monk in fact suggests values of2 and 25 .8 for these parameters
and used these in deriving a set of design curves. Uncertainties concerning these
parameters may result in significant inaccuracies in the derived reduction factors ,
so that tests on specific materials may be required in applying these theories.
3.4 Wall-floor slab interaction

In the theories discussed above, it is assumed that thc compression element under
consideration has known end conditions and eccentricity of loading, whereas in
an actu al building these factors and the strength of a wall or column will be
influenced by interaction between the members of the structure. Thus, as
illustrated in figure 3.11, the rotation of the top of an eccentrically loaded wall
willdepend on the relative stiffnesses of the two elements, the loading on the slab
and the characteristics of the joint between them . In order to take wall-floor slab
interaction into account a solution is required for the strength of the walls, in
1.0
0 .8
--l
::x:
tr1
0 .6 tI.l
--l
:;>:l
p tr1
Z
a'd C)
--l
::x:
o.."
0.4 t%l
:;>:l
()
~
Ei:
>
tI.l
o
0 .2 Z
:;>:l
e/d =1 /120 -<
=>r
e/d =1 /6
o 0 .5 1.0 1.5 2 .0 2.5

.!!... I?:...
dY Eo -..I
W
Eigure 3.10 Effects of stress-strain relationships
~ ~ '\
~Ms
Figure 3.11 Restraining effect of a floor slab
which the rotation at the wall end is a variable in addition to the variables of load,
eccentricity and member dimensions.The problem may be considered in two
parts : (I) the strength of the walls in terms of end rotation and eccentricity, and
(2) the determination of the deflected form and eccentricity. The first of these
will be considered in this chapter and the second in chapter 4.
3.4.1 Wallstrength in tenns 0/ end rotation

A number of solutions for this problem have been published . Sahlin 1 S , working
from Angervo's solution of the differential equation for a brittle column, derived
the set of parametrie curves shown in figure 3.12, which connect the end rotation
of a wall, the applied load and the eccentricity . These particular curves are for the
case of zero eccentricity at the bottom of the wall. Superimposed on them in
dimensionless terms are curves of constant edge stress so that it is possible, for a
given load, to determine the maximum stress in the wall for known conditions of
end rotation and eccentricity or the buckling load.
Risager'" has produced a solution for walls compressed between floor slabs
with equal angles of rotation at the ends, based on consideration of an equivalent
column having a parabolic deflected form, as indicated in figure 3.13 . Using the
notation shown in this diagram, the deflection, Y, of the column of length H;
between points of inflection is
Y = 4yc -~ (1 + ~-) (3.7)

n, u,
where Y c is the maximum deflection of the equivalent column and x is the distance
along the load axis, measured from the point of inflection. The eccentricity and
angle of rotation at the level of the lower surface of the floor slab are then,
respectively
10
Full lines ind icate

~ co nstant valu es of
8 \ (Ah)2
- - bdo edge
\ EJv
(Ah)2
-p
EJv
o 0.2 0.4 0 .6 0 .8 1.0

Ah
d~ v
Figure 3.12 Interrelationship between load in a brittle column and angle of

rotation at its upper end . Load is eccentric at top and axial at base (after Sahlin)
Figure 3.13 Equivalent column in Risager's analysis

e=4yc ~ (1 - ;J (3.8)
w=4 ~: (1 - ~) (3.9)
Introducing the following parameters

e
e=- n= -H w
H
</>= -
d/2 d d
it can be shown from equations 3.8 and 3.9 that
(1 )
..' '\1 _ <' 4yc
dl2
1
(I +2~)'
(3.1 0)
and
S
e=2<1> - (3.11 )
H
Consideration of the stresses in a column of reetangular section without tensile
stress shows that the edge stress is
2 P
°max = - (3.12)
3 b{d/2 - e)
and the curvature
2 P
K = (3 .13)
9 Eb {d/2 - ei
Introducing the further parameters
E
W = 7.2 ( ; ) 2 (3.14)
P
v= (3.15)
bdObr
Risager established the formulae shown in table 3.1, from which the bearing
capacity , crack condition, mode of failure, eccentricity and end moments can be
determined for a given column, provided that the angle of rotation at its end is
known.
A similar analysis has been developed by Colville,17 ,18 who extended it to
include walls in single and double curvature and with several combinations of end
eccentricity. This solution resulted in parametric curves of [(P) - [(<I» , similar to
those derived by Sahlin. The results of Colville's analysis have been found to com-
pare reasonably weil with those obtained experimentally. A development of these
theories for the calculation of eccentricities is discussed in chapter 4 .
Table 3.1
Formulae for bearing capacity and eccentricity (Risager)
Zone Bearing capacity Eccentricity
II
III e=O
IV e=O
Mode of failure : Zones land 11: Stress failure. Zones III and IV: Buckling.
Wall seetion : Zones land IIl : Uncracked Zones 11 and IV: Cracked .
3.4 .2 Wa/l-floor slab joints

It is necessary in considering wall-floor slab interaction to take into account the
characteristics of the joints in terms of the forces and moments acting and the
resulting angles of rotation of the various elements. If the joints are rigid, the wall
end rotation will obviously be the same as the slab end rotation. In general these
rotations will be different , and if we consider the rotation of the slab and the wall
at a particular joint as the total load on the structure is increased, these rotations
cjJh and cjJy respectively , will increase as shown in figure 3.14 . The joint rotation at
e,
any particular load will then be represented by the difference, between the
abcissae of these curves. This behaviour would greatly complicate the analysis of
brickwork structures, but Sahlin 15 has found that the behaviour of joints of the
type shown in figure 3.15a in terms of M and (J can be approximated by a 'rigid-
Figure 3.14 Wall and floor slab rotations as a function of applied load on structure
p
M 4
(t rn/rnl
3
I
>
I
I
o 234 5 /
8( x 10 - 3 )
Ial Ib]
Figure 3.15 Wall-floor slab joints tested by Sahlin . (a) Type of joint ; (b) M-8
relationship
80
Percentage
fix ity
60
40
20
o 0 .1 0 .2 0 .3 0.4 0 .5 0.6 0 .7 0.8

Precompression stress (N/mm 2 )
Figure 3.16 Percentage fixity of wall-floor slab joints with increase in

precornpression
plastic' relationship (figure 3.I5b). The quasi-plastic angle of rotation at failure,
8 p , depends on the axial load on the wall above the joint-candmore specifically
on the ratio of this load to the carrying capacity of the wall under axial load .
Quite elearly, if a wall is loaded elose to its axial bearing capacity it will be able
to accommodate on1y a very small angle of rotation from a slab joint.
Some further information on joint rigidity has been obtained by Colville,' 7 ,19
who carried out tests on a structure incorporating joints between 225 mm bonded
brickwork walls and 152 mm reinforced concrete slabs. The results, shown in
figure 3.16, indicated that approximate1y 80 per cent fixity was obtained with a
precompression on the joint of 0.3 Njmm". Similar tests on a three-storey struc-
ture reported by Sinha and Hendry''? indicated that the wall-floor slab joints
behaved as if rigid, even at very low precompression. Figure 3.17 shows a corn-
parison between the deflections measured in one of these tests and the results of
a rigid frame analysis.
It would appear reasonable, in the light of the available evidence, to treat wall-
floor slab connections as rigid, that is, 8 = 0, provided that the precompression is
greater than about 0.3 N/mm 2 , and also that the joint does not fail.
Maurenbrecher"! has shown that, setting aside failure of the slab, joint strength
may be limited either by 'equilibrium' failure or by failure of the brickwork in
compression below the slab. These types of failure are illustrated in figures 3.18a
and b. In figure 3 .I8a, the precompression is relatively low and the joint has failed
".
..
\..;:tO~Nci
,0
deflection in mm
expertrnent e or .....
theory - -
vertical deflection
scale twice horizontal
0.100.1
'-'--' ......
00.1 0.1 00.1
L.......--....I
Figure 3.17 Comparison between measured and calculated deflections of

three-storey test structure
Figure 3.18 Joint failures. (a) Equilibrium failure ;(b) local compression failure. In both cases the test specimens are inverted in
relation to the situation in a building; thus in (b) damage to the brickwork occurs at the top of the wall, below the floor slab
by the two wall seetions being prised apart by the slab. In figure 3.18b, at a higher
preeompression, local failure of the brickwork has taken plaee below the slab.
Equilibrium failure ean be eheeked by staties, negleeting the tensile strength of
the brick-mortar joint. Loeal stress failure below the slab will be investigated in
the normal course of wall design.
3.5 Moment magnifier method

An alternative approaeh 2 2 - 2 6 to the ealculation ofthe eompressive strength of
masonry elements is the 'moment magnifier' method, which has been developed
für reinforeed eonerete and steel eolumns . In this method an interaetion eurve
defining the strength of a short eompression member subjeet to axial load and
bending moment is first derived. This eurve has the form indicated in figure 3.19,
in which the vertical axis is in terms of the applied load P divided by the ultimate
Po~ 1.0
0 .8
~~
0.6
r
P
Po 0.4
0.2
o 0.6 0 .8 1.0 1.2

MIM, L
M,V
Figure 3.19 Axial load-moment interaction curve
0 .8
P 0 .6
0.4
0 .2
o 0 .2 0.4 0.6 0 .8 1.0 1.2 1.4 1.6

MIM,
Figure 3.20 Effeet of stress-strain relationship on axial load-moment interaetion

(Turkstra)
load , PO , which the section can resist in the absence of bending , that is, the
maximum axial load for failure . On the horizontal axis the applied moment M is
divided by the moment capacity ,Mk , when a load is applied at the end of the
middle third , that is, at an eccentricity of d/6 . The stress conditions associated
with the various portions of the interaction curve are also shown . The curve in
figure 3.19 is based on the assumption that the ultimate tensile strength is equal
to 0.1 of the compressive strength ; if, as is usual, the tensile strength is neglected,
the interaction curve will pass through the origin. Turkstra and Ojinaga/? have
demonstrated the effect of different idealised stress-strain curves, assuming the
same ultimate strain in each case, with the results shown in figure 3.20 . Formulae
for the calculation of the interaction curve for asolid reetangular section are
given in table 3.2 .
End conditions and slenderness are allowed for by introducing a 'moment
magnifier'
Cm
L=--- (3.16)
1 -PIPe
in which Cm is a correction factor depending on the moment distribution in the
element , and is intended to give the equivalent uniform moment in the column
that would lead to the same long column strength as the actual moment diagram.
Pe is the Euler criticalload taking intoaccount th e effective height of the com-
pression member . The application of the moment magnifier will be understood
by reference to figure 3.21. Figure 3.21 a shows a pinned column which carries a
load P at an initial eccentricity e at each end. If the maximum eccentricity is ß,
the corresponding bending moment is
1
P(e +~) =Pe (3.17)
1 ~PIPe
In this case Cm = 1 and Pe = 1f2 Ellh 2 •
Ia) (b)
Figure 3.21 Moment magnifier method. (a) Equal end eccentricities; (b) general
case-veccentricities different at column ends
If the moments at the lower and upper ends of the column are MI and M 2 ,
respectively, the correction factor Cm is given by
(3.18)
The rat io of the moments is posit ive if th e column is bent in single curvature, and
in this case
where kh is the effective length .

Calculation of EI presents certain problems, as the stress-strain relationship is
non-linear, and the value of I may be reduced by cracking of the section. The
following empirical relationship has been suggested'" for calculating the flexural
;J
rigidity of a masonry element
EI=Eo/n (0.2 + <"0.7Eo/n (3.19)
where E 0 is the initial tangent modulus, and In the moment of inertia of the
uncracked section.
Comparison with experimental results shows that this method gives a reasonably
good estimate of the carrying capacity of masonry walls subjected to eccentric
verticalloading or to combined vertical and lateralloading . It is found, however ,
that the apparent ultimate compressive stress under eccentric loading calculated
on the basis of a linear stress distribution may be higher than the prism strength
under axial load . Fattal and Cattaneo" give the following values for the ratio a of
these two stresses for a particular type of brick masonr y
Eccentricity , ejd 1/12 1/6 1/4 1/3

a 1.40 1.37 1.36 1.18
The above results were obtained for bricks of 90 Njrnm? crushing strength and
1:+:4+ mortar of crush ing str ength 10.4 Njrnrrr" ; the compressive strength of
stack-bonded prisms was 31 .1 Njrnm" . Other investigator s have reported different
values, and the interaction curve for any specific masonry type would have to be
based on the results of aseries of tests on small prism specimens, tested under
various degrees of eccentricity, in conjunction with the equationsf" shown in
table 3.2.
This method has considerable advantages as a basis for the design of masonry
compression elements : it is easily comprehended, it can be applied to both plain
and reinforced masonry and in principle is applicable to the design of steel and
concrete columns.
3.6 Special wall types

3.6.1 Cavity walls
Cavity wall construction is frequently used for the outer walls of buildings in
order to achieve a higher degree of weather protection and thermal insulation than
would be obtained from a solid wall of the same material thickness . Experiments"?
Table 3.2
Formulae for interaction curves for solid reetangular brickwork section
For a cracked section (zero tensile strength)
Me = Pd (I
2 \
__ g !:-)
aPo
g =2 (1-~)
\: Ad 2
aPo
2
21
Ad
For an uncracked section

2
M = - - (Apo - P)
e Ad
The cracking line is given by

M k =Pkek
where Me is the maximum moment capacity
P cornpressive force on the section
A area of cross-section
1 moment of inertia of the section based on the uncracked section
a flexural compressive strength coefficient
d section thickness
ubr compressive strength of the brickwork.
have shown that , for practical purposes , the inner leaf of a multi -storey cavity wall
transmits by far the larger proportion of the loading from floors above, even when
the floor slabs are carried through the thickness of the wall. In considering the
strength and stability of a cavity wall therefore, it will be appropriate in most
cases to assurne that the structural role of the outer leaf is to stabilise the inner
leaf.
The effect of the ties is to constrain the two leaves to deflect laterally to the
same extent, thus doubling the effective stiffness of the loaded leaf, if both are of
the same thickness . The theoretical buckling load for the leaf of a cavity wall will
thus be twice that of a single-leaf wall of the same thickness , but under eccentric
2
oL_--':===========--
0 .05 0 .1 0 .15
e
d
Figure 3.22 Cracking loads of single-leaf and cavity walls related to the Euler load
for a single leaf (Sahlin)
loading a crack will develop in the unloaded leaf at a relatively low load. Referring
to figure 3.22 the maximum tensile stress in this leaf is given by
3a (eid)
aol = ----'-'----'-- (3.20)
(l - a/2a e )
where o =Plbd and a e is the Euler buckling stress. The edge stresses in the inner
leaf are given by
3a (eid)
ajJ=-a ± - - - -
(l - a/2a e ) (3.21)
The relative cracking loads for solid and cavity walls have been calculated by
SaWin2 8 and are shown in figure 3.22.
3.6.2 Stiffened walls

Theoretical and experimental studies of wall strengths have generally been limited
to consideration of reetangular cross-sections . However, many walls in practical
situations are stiffened by piers, or returns, and recently interest has developed in
the use of walls with fins or of cellular constructlon.i? as indicated in figure 3.23.
Codes of practice include arbitrary rules for dealing with some of these con-
structions, but the problem so far lacks systematic investigation.
86
\ u t ,
J r
Ia) Wall stiffened by piers
U
(d} Cellular or diaphragm wall
Figure 3.23 Cross-seetions of stiffened walls
In principle, it would be rational to consider the compressive strength of such

elements by calculating their slenderness ratios on the basis of their overall radii
of gyration.l? The strength and stability of the individual seetions of wall, and the
integrity of the bond between them may however be more critical, so that informa-
tion about the behaviour of brickwork panels with stiffened edges is of importance.
No theoretical solution for this problem appears to exist, but there are some
experimental results'" from tests on axially loaded walls with returns on both
vertical edges. The results of these tests, some of which are summarised in table
3.3 , led to the following conc1usions.
(1) Wallsstiffened along their vertical edges by returns did not 'Show
increased strength as compared to strip walls, up to a slenderness ratio of
32 . It would appear therefore, that up to this limit, no increase in the bear-
ing capacity ofaxially loaded walls should be made on account of their
edges being stiffened by bonded returns. This conc1usion holds whether or
not the returns are loaded to the same extent as the main wall.
(2) Before cracking and separation of returns, the central deflections of
walls with returns are smaller than those of corresponding strip walls, which
indicates effective stiffening up to this point. This stiffening effect decreases
with increasing aspect ratio.
(3) As returns provide effective stiffening at low axial loads, they may be
effective in increasing the strength of very slender walls that may be expected
to show buckling rather than strength failures.
(4) Where only the main wall was loaded, about 6 per cent of the total
applied load was transferred to each of the returns. The average ultimate
vertical shear stress which destroyed the bond between the main wall and
returns varied between 0.35 and 0 .68 N/mm 2 , calculated on an area equal
to the height times the thickness of the main wall.
Fattal and Cattaneo" have produced formulae for calculating moment-axial
load interaction curves for T-section piers, as shown in table 3.4. These could be
Table 3.3
Test results of axially lo aded strip walls and walls st iffened
along their vertical edges by returns
Test Height/ Sien derness a b St reng th of walls with return s

No. Length ratio Failure Failu re
stress for stress for Strength of st rip walls
strip wall walls with (b /a)
(Njrnm") ret urns
(N /mm 2 )
I 1.3 24 10.8 9 .9 1.1

9 .2
2 1.3 24 7.65 10.9 1.2
3 1.0 24 8.1 3 8.86 1.08
4 1.0 24 8.86 1.0
5 1.4 8 10.84 .24 0. 9 1
6 1.4 8 11.55 11. 19109 .8 0 .87
7 0 .8 8 11.0 0 .98
11.2
8 0 .8 8 10.56 0.94
9 2.8 16 11.15 11 15 10 .7 0.96
10 2.8 16 11.15 . 10.7 0.96
11 1.6 16 9.30 0.83
11.15
12 1.6 16 9 .72 0 .87
13 5.6 32 9.07 9 .89 1.05
9.35
14 5.6 32 9 .62 8 .50 0 .91
15 3 .12 32 7.9 0 .85
9.35
16 3.12 32 9 .2 0 .98
17 2.06 32 7.24 0 .77
9.35
18 2.06 32 9.08 0 .97
19 1.3 24 12.0 10.6 0 .73
14.5
20 1.3 24 17.0 9 .4 0.65
21 1.0 24 10.0 0.60
22 1.0 24 20.3 166 15.0 0.90
23 1.0 24 12.94 . 11.8 0.71
24 1.0 24 17 .7 1.10
25 1.0 24 15.0 0 .90
1 to 4: wallsbuilt with ~ scale brick } Return 5 to 16: wallsbuilt with +scale briCk}Return
+
17 to 18: walls built with scale brick walls 24 to 25: full-scale wall walls
19 to 23: full-scale walls not loaded
loaded
Table 3.4
Interaction equations for a T-section 2 6
Sidel
Side2 -
t
e,
:'.~I'"
J
For a cracked section For a uncracked section
aPo - P
M el =PCI
~- gl ~)
aPo
M el =Pklekl
aPo - P k l
aPo - P
~ -g2~)
M e2 = PC2 M e2 =Pk2 ek2
aPo aPo - P k2
The cracking line which separates the uncracked
~ -~)
aPo
gl - - - and cracked regions is defined by the equations
Pk l Cl
Mk l = M k2 = Pk l ek l =Pk2 ek2
0
aPo _ e k2}
g2 - - - where
P k2 C2
a = flexural compressive strength co-
aPo efficient
Pk l =
A = net area
I +~ = distances from neutral axis to outer-
C2
most fibres in maximum compression
aPo = kern eccentricities from centroid of
Pk2 = - - - - section in directions land 2,
C2 respectively
1+ - = cornpressive strength of masonry
Cl f~
under axial load
I I = moment of inertia of net section
ek l ab out its centroidal axis
AC2 = moment capacity of masonry , produc-
I ing maximum compressive stress in
ek2 = - - outer fibres on sides land 2,
ACI respectively
Po = Af:n P = axial compressive load on the cross
section
= compressive loads capa city of masonry
applied at kern eccentricities e kl and
e k2 , respectively
= axial load capacity of rnasonry
used in conjunction with the moment magnificr method to deal with sections of
th is shape , and could be extended to cellular or diaphragm walls. In both cases a
check on the shear stress at the intersection of the component walls would be
requ ired .
References
I. C. B. Monk, 'A Historical Survey and Analysis of the Compressive Strength
of Brick Masonry', Research Report No. 12 (Structural Clay Products
Research Foundation, Geneva, Ill., 1967).
2. N. Davey and F. G. Thomas, 'The Structural Use of Brickwork', Structural
Paper No. 24 (Institution of Civil Engineers , London, 1950).
3. F. G. Thomas, 'The Strength of Brickwork', Struct. Engr., 31 (1953)
35-46 .
4 . P. Haller, 'Load Capacity in Brick Masonry', in Designing, Engineering and
Constructing with Masonry Products, ed. F. B. Johnson (Gulf, Houston ,
Tex ., 1969) pp . 129-49.
5. - 'Cornpressive, Transverse and Racking Strength Tests of Four-inch Brick
Walls' , Research Report No. 9 (Structural Clay Products Research Founda-
tion, Geneva, Ill., 1965).
6. - 'Compressive and Transverse Strength Tests of Eight-inch Brick Walls',
Research Report No . 10 (Structural Clay Products Research Foundation,
Geneva, Ill., 1966).
7. - 'Compressive and Transverse Tests of Five-inch Brick Walls', Research
Report No . 8 (Structural Clay Products Research Foundation, Geneva,
Ill., 1965).
8 . S. S. Hasan and A. W. Hendry, 'Effect of Slenderness and Eccentricity on
the Compressive Strength of Walls', Proceedings 01 the Fourth International
Brick Masonry Conference (Brugge) 1976, Paper 4 .d.3 .
9. N. Royen, 'Knickfestigkeit Exzentrisch Beanspruchter Säulen Aus Baustoff
der Nur Gegen Druck Widerstandsfähig Ist', Bauingenieur, 18 (1937) 444.
10. K. Angervo, 'Uber die Knickung und Tragfähigkeit eines exzentrisch ged-
rucken Pfeilersohne Zugfestigkeit', Publication 26 (Staatliche Technische
Forschurgsen stalt , Helsinki, 1954).
11. J . C. Chapman and J . Slatford, 'The Elastic Buckling of Brittle Columns',
Proc. Instn civ. Engrs, 6 (1957) 107-25.
12. S. Sahlin, 'Diagrarns of Critical Stress for Columns of Material without
Tensile Strength', Report No . 16/25 (Swedish National Institute for Build-
ing Research , Stockholm, 1965).
13. C. B. Monk , 'Column Action of Clay Masonry Walls', in Designing, Engineer-
ing and Constructing with Masonry Products, ed. F. B. Johnson (Gulf,
Houston, Tex., 1969) pp . 161-70.
14. C. J. Turkstra, "The Capacity of Masonry Walls under Eccentric Vertical
Loads', Structural Mechanics Series No. 71-3 (Structures Laboratory,
McGill University , Montreal, 1971).
15. S. Sahlin, Structural Brickwork (Prenti ce-Hall, Englewood Cliffs, N.J. ,
1971) pp. 91-116.
16. S. Risager, 'Structural Behavior of Linear Elastic Walls having No Tensile
Strength', in Designing, Engineering and Constru cting with Masonry
Products, ed. F. B. Johnson (Gulf, Houston, Tex ., 1969) pp. 257-66.
17. J. Colville, 'Analysis and Design of Brick Masonry Walls', Dept. of Civil
Engineering, University of Edinburgh, 1977.
18. J. Colville, 'Simplified Design of Load Bearing Brick Masonry Walls', Proc.
Br. ceram. Soc., 27 (1978) 217 -34.
19. J. Colville and A. W. Hendry , 'Tests of a Load Bearing Masonry Structure',
Proc. Br. ceram. Soc. , 27 (1978) 77-84.
20. B. P. Sinha and A. W. Hendry , 'An Investigation into the Behaviour ofa
Brick Cross-Wall Stru cture ' , Proc. Br. ceram. Soc ., 27 (1978) 67-76 .
21. A. H. P. Maurenbrecher, 'Wall-Floor Slab Joint Behaviour in Brickwork ',
Ph.D. Thesis, University of Edinburgh , 1973.
22. F. Y. Yokel and R. D. Dikkers, 'Strength of Load Bearing Masonry Walls ',
J. struct. Div. Am. Soc. civ. Engrs., 97 (1971) 1593-1609.
23 . P. D. Burns,'Unreinforced Brick Masonry Wallsunder Vertical Loads',
M.Eng. Thesis, McGill University , 1972.
24. C. Turkstra and J. Ojinaga, "The Moment Magnifier Method Applied to
Brick Walls', Proceedings of the Fourth International Brick Masonry
Conference (Brugge) 1976 , Paper 4 .b.3 .
25. R. G. Drysdale and S. E. A. Sallam, 'Design of Masonry Walls and Columns
for Combined Axial Load and Bending Moment', Proceedings of the First
Canadian Masonry Symposium (Calgary) 1976, pp. 394-408.
26. S. G. Fattal and L. E. Cattaneo, 'Structural Performance of Masonry Walls
under Compression and Flexure', Building Science Series 73 (National
Bureau of Standards, Washington, D.C., 1977).
27. B. P. Sinha, A. H. P. Maurenbrecher and A. W. Hendry, 'An Investigation
into the Behaviour of a Five Storey Cavity Wall Structure', Proc. Br. ceram.
Soc., 24 (1975) 147-60.
28. S. SahIin, Structural Masonry (Prentice-Hall , Englewood Cliffs, N.J ., 1971)
pp. 183-91.
29. W. G. Curtin and G. Shaw, Brick Diaphragm Walls in Tall Stingle-storey
Buildings (Brick Development Association, London, 1977).
30. F. Sawko and W. G. Curtin , 'Effective Thickness and Structural Efficiency
of Cellular Walls and Piers', Proc. Instn civ. Engrs, 65 (1978) 893-8.
31. B. P. Sinha and A. W. Hendry , 'Cornpressive Strength ofAxially Loaded
Brick Walls Stiffened Along their Vertical Edges', Proceedings of the Fifth
International Brick Masonry Conference (Washington) 1979, Paper 4.5
(prelim . vol.) pp. 117-19.
4 DESIGN ANALYSIS OF UNREINFORCED
BRICKWORK STRUCTURES
4.1 General
Three analytical problems arise in the design of brick masonry structures, which
relate to (1) the distribution of verticalloads amongst the various walls in the
building, (2) the determination of eccentricity of loading on walls, and (3) the
distribution of lateral loads on the walls. Conventionally these problems are
resolved in design calculations by rather arbitrary assumptions, but they can be
treated more rationally by the methods described in this chapter.
4 .2 Verticalload analysis
4.2 .1 Load distribution on walls

In simple cross-wall structures the allocation of floor loads to the supporting walls
is straight forward . However, with two -way spanning slabs, and with complex wall
arrangements, the problem becomes more difficult and considerable differences in
estimated wall loads can arise according to the assumptions made. The most usual
procedure is to sub-divide the floor areas into triangles and trapeziums, as in the
case ofreinforced concrete beam design, and to allocate the loads from these areas
to the appropriate walls. With a simple reetangular slab this is probably reasonably
accurate over all, but it should be noted that the distribution of the force applied
to the walls will not be uniform along their lengths, being in fact concentrated
towards the centre. It is probable that this non-uniformity will gradually even out
down the height of a wall, and in the lower levels of a multi-sto rey building there
would be an approximation to uniformity.
An alternative approach described by Sutherland 1 is to divide the floors into
tributary areas, the load from each being allocated to a particular wall group ,
taking into account any displacement of the centroid of the loaded area from the
centroid of the wall group . This procedure is illustrated in figure 4.1 in which a
comparison is made between the wall stresses resulting from the two approaches-
the 'wall unit' method is probably more accurate, but requires more lengthy
calculations and the extent to which it is more accurate is uncertain.
91
ID
N
Neutral axis
Wall No. 46 otwall un it
1.03 N/mm 2
1.13 N/mm 2
.,----------,
- - - -- -, r-- --, 1
. - - - - - -- - - -,
- - - --, ,----, ' ~ " " ';"W'bOf"";'
I' 11 Wall No. 44 :: :: .: I varies Iinearly trom
11 L, 2
.,./ 1.0 N/mm 0 , ' .' : • 1.03 to 1.12 N/mm 2
~ ,I I' . .
I' - -{ " - .. . ", .
-----_.1,I ------1" . 1.12N/mm 2
' - , e" - - - , - ... - - -. I
• I 11 rt I , I I .-. cn
_'I _,~ -,I ...,
__ .J 1.. • r'~1 WALL UNIT 'A'
::.: C·~ '.: :.:_-:.1 :xl
"I ,
, 1 I I
1 I
R
...,
t
,1 '
-, r---·---. r I c::
- , , - 0I
r'- 0-· - - - - - -. ,-- - "\. _ J :xl
, I >
Wall No. 53 r-
101 " • 1 Wall No. 52 ttl
1.2 Nimm 2 • < I I 0.99 N/mm 2 ,, ,1 :xl
I I
,, ,I I I ()
L,.' l'
,J
, I
______,I,
~
:xl
~1..
51 r - - - - - -'".J ~-r!!. ::, _· ' _ ~- t ': I X
WALL UN IT 'B '
Wall No. 51
1.12 N/mm'l
(a ) (b)
Figure 4 .1 Wall stresses calculated by alt ernative methods of calculating loading (Sutherland ). (a) Tributary areas allo cat ed to
individual walls ; (b) loading allo cat ed to wall units
DESIGN ANALYSIS OF UNREINFORCED BRICKWORK STRUCTURES 93
50 100 150 0
At po int 1 ( x 10 6 )
1!:::::::::::::::!It=l No read ings laken
until 1SI s to rev
co m pleted
,
229 ' 1
I:'===~
"'----------~vibraling
300 mm I~========:
wir e
str ain gauges
(b) (c)
Figure 4.2 Strain measurements in a cross-wall building, (a) Recorded change in

strain in first-floor wall; (b) location of strain gauges (same layout on opposite
face) ; (c) estimated tributary floor area supported
There is some reason to believe that in rat her tall buildings, say of ten storeys
or more , there will be a tendency for the stresses in the lower sections of walls to
even out , not only in individual walls but in wall groups as well. Some evidence for
this was found by Stockbridge/ from strain measurements taken in a five-storey
cross-wall building; figure 4.2 shows arecord of the strains measured in a wall of
this building as the storeys above were constructed. On the basis of the load dis-
tribution indicated in figure 4.2c , it would have been expected that the reading at
point I would have been considerably smaller than that at point 2, whereas in
fact , they were almost equal. The stressing of the wall at point 3 was complicated
by the presence of a lintel , XY, above the end of the wall. The effect of this lintel
was initially to attract load to this area, but after construction had reached the
first storey the rate of increase of strain decreased considerably until by the time
the fifth floor was reached, the strains across the width of the wall were becoming
much more uniform than in the earlier stages.
4.2.2 Analytical models for verticalload analysis

The conventional analytical model for the design of a brickwork structure for
verticalloads is one in which the walls and floor slabs are effectively interconnected
by hinged joints. The forces from the floor slabs are transmitted to the walls
eccentrically, as indicated in figure 4.3 . These eccentricities are determined by
empirical rules as, for example, in the British Code of Practice for Structural Use
t/2
(a) [b} (c)
Figure 4.3 Typical assumed eccentricities in conventional analytical model.

(a) Slab simply supported on wall, e = t /2 - a/2 or a/3; (b) continuous tloor slab,
tloor load Pr much smaller than Pu, e = 0 ; (c) discontinuous slabs, eccentricity of
Pu = 0, ej = aj/2, e r/2, eccentricity of load on lower wall is calculated from these
assumed eccentricities
of Masonry, BS 5628, which suggests that the load from a single floor or roof may
be considered to act at one -third of the depth of the bearing area from the loaded
face of the wall or, in the case of a continuous floor slab passing over a wall, each
side of the floor may be taken as being supported on half of the total bearing area .
In this Code, the load from the floors above the wall under consideration is
assumed to be axial and correspondingly the eccentricity of loading at the lower
end of a wall section is taken as zero .
This type of assumption obviously simplifies calculations and is very widely used
in design, and although extremely crude and inherently inaccurate, being protected
by large safety factors, it has given satisfactory results in terms of structural per-
formance. A more rational analytical model for a masonry structure must take
into account the ability of the wall-floor slab joints to transmit bending moments,
and in certain cases it is possible to regard these joints as fully rigid, as discussed in
section 3.4.2.
4 .2.3 Frame analysis applied to brickwork structures

Where joint rigidity can be justified it will be appropriate, at any rate for the less
complex types of structures, to apply some form of frame analysis . Thus, a
method of calculating eccentricities based on consideration of a section of rigid
jointed structure surrounding the wall under consideration was described by
Haller. 3 The formulae for interior and exterior walls are summarised in figure 4.4 ,
and have been derived by the normal methods of structural mechanics. In order to
Interior w all
Hinge (free rot at ion) Sup erimposed loads Dead loads
pl ;x 91;X
h 1 2( 1 + r) - (l2/I, )2 M , =M2=M
M' =4 N,
(2 + 3r)11 - (1, /1, )'1
M
1 2(12 /1,)' (1 + r )- r 4N,
M2 = - ---=--,---- -
h 4 N,
N , =4(1 +f)2_ r'

h
h J ,E, h J,E,
a , =- - a2=- -
I , J ,E, I, J ,E,
Exterior wall
h 1 a +2 1 a +3
M ' +ZN M , =M2= 4 N
1 a +l
M 2+
h 4 N
M, h J,E,
a +--
I J ,E,
h N =4( a +2)2 -(a + 1)'
h/2 1
M, =M, = - -
8(a +3)
h/2
Figure 4.4 Haller's partial frame method for calculating eccentricities
apply these formulae, the EI values for the walls and floor siabs must be known .
As far as the walls are concerned, the values given by equation 3.19 may be used,
and for the floors one of the normal methods for calculating the flexural rigidity
of a reinforced concrete slab may be adopted.
Vahakallio and Makela" have developed a method for calculating eccentricities
on the basis of a simplified elastic analysis, which assumes that horizontal members
have bending strength but that vertical members have no tensile strength. The dis-
tribution of moments at a joint is calculated by considering a section of the
4
3
L,
Figure 4.5 Method of Vahakallio and Makela-notation
structure consisting of the floor siab panels and the walls intersecting at the joint
und er consideration, as indicated in figure 4.5. The end moments in the members
are then obtained from the following equations
MOl = a (Mq - aOI x M /A o) (4.1)
M02 = a (M p - a02 x M /A o ) (4.2)
M 0 3 = - a a03 x M /A o (4.3)
M 0 4 = -aa04 x M /A o (4.4)
where a the carry-over factor is 1.1 at crossed joints, 1.2 at right-angled joints, and
1.5 at joints
Mp and Mq are the fixed end moments due to u .d.!. p and q
M=Mq - M p (;;;'0) (4.5)
aQj = kjEjI;/L j (4.6)
where k, = 3 or 4 for hinged or fixed ends respectively, for i = 1 or 2 in figure 4.5
ai
and k j = for i = 3 or 4
A o = L aOj at the joint (4.7)
The value of ai,which makes allowance for cracking , is a function of the rela-
tive eccentricity, and is shown irrflgure 4 .6. The dimension Li used in calculating
3 .0 .......- -_ _
2 .5
2 .0
s,
1..5
1.0
0 .5
0
0 .2 0.3 0.4 0 .5
eld
Figure 4.6 Coefficient aj to allow for effect of cracking on wall stiffness
aOj for the walls is the distance from the joint to the point of inflection in the
member, that is, Ajh. An initial value of Aj = 0.5 may be assumed, and corrected
after a trial calculation of moments. Since the value of the eccentricity has to be
assumed in obt aining aj, it is evident that an iterative procedure will, in general,
be required.
VahakalIio and Makela extended this method of analysis to permit its applica-
tion to cases in which the floor slabs are supported in a variety of possible ways.
Equations 4.1 to 4 .7 still apply, but the fixed end moments Mq , Mp and Mare as
follows
M q = CtQb 2 (4.8)
Mp = CtPb 2 (4.9)
M = Mq - Mp (~ 0) (4.10)
The bending moment coefficient Ct for plates having various support conditions
has been calculated ," and these values are shown in table 4 .1. Values of Ct are
obtained from this table where the plate edge at 0 (figure 4.5) is fixed, and the
other edges are supported as in the structure under consideration.
Flexural stiffness coefficients for the elements meeting at the joint 0 are
given by
}
EII I _ E 313
aOI = tl - - a03 = t3r 3a 3 - -
LI L3
(4.11 )
E 212 - E 414
a02 = t2 - - a04 = t4 r 4 a 4 - --
L2 L4
where t l and t 2 are the factors relating to the floor slabs given in table 4 .2; t 3 and
t 4 have been calculated" for walls and may be obtained from figure 4.7 . The value
of aj is found from figure 4.5 ; the coefficient rj is to allow for openings in a wall
and may be taken as the ratio of the net length of brickwork to the overall length
ofthe wall.
98 STRUCTURAL BRI CKWORK
Table 4.1
Bending moment coefficients for plates
I j:n_-:lb
L ___ _ ...J
2
OL_ __ _ ..J b 3 r ---ilb
___ _ J
a a a
a Cak C bk C bt Cak Cb k Cat Cak Cbk
b (I) (2) (3) (4) (5) (6) (7) (8)
1,0 0,0555 0,0555 0,0575 0,0368 0,0429 0,0575 0,0429 0,0368
I ,I 564 628 626 364 471 632 456 435
1,2 564 694 672 355 505 684 472 494
1,3 555 754 714 343 536 728 480 518
1,4 542 807 752 331 560 761 478 597
1,5 526 854 786 320 577 786 469 643
1,6 508 894 816 310 592 800 456 686
1,7 492 928 842 301 602 807 442 726
1,8 478 958 864 293 610 807 430 762
1,9 469 982 884 286 620 804 426 796
2,0 464 996 900 280 632 800 426 828
00
1250 1250 704
4 Ob a
5 Ob a
a C at C bt Cak C bk C bt Cak C bk
-
b (9) (10) (11) (12) (13) (14) (15)
1,0 0,0415 0,0415 0,0314 0,0314 0,0479 0,0232 0,0310
I, I 440 470 317 357 510 224 330
1,2 462 522 316 392 537 217 347
1,3 479 571 311 422 558 211 361
1,4 492 617 304 448 577 206 374
1,5 503 661 294 471 594 201 385
1,6 511 702 283 489 609 196 394
1,7 517 740 272 505 622 190 409
1,8 520 775 26 1 516 635 184 405
1,9 521 808 253 529 648 177 409
2,0 520 840 250 538 660 168 414
00
833 417
DESIGN ANALYSIS OF UNR EINFORCED BRICKWORK STRUCTURES 99
6
[~= =]b
0
7 Ob 0
a
--
b
C at
(16)
Cak
(17)
C bk
(18)
C at
(19)
C bt (21 )
(20)
C ak C bk
(22)
1,0 0,0179 0 ,0310 0,0232 0,0308 0,0367 0,0212 0,0247
I,I 536 339 276 311 406 206 268
1,2 538 364 322 314 441 201 288
1,3 635 383 369 316 473 196 305
1,4 673 397 417 318 500 190 320
1,5 703 405 465 319 184 332 465
1,6 729 409 511 320 546 179 342
1,7 555 409 555 320 565 175 350
1,8 770 405 595 320 580 171 358
1,9 785 397 631 320 591 167 365
2,0 790 388 664 320 600 162 370
00
8 CJb 0
9
Ob 0
a C at C bt Cak C bk Cat C bt Cak C bk

b (23) (24) (25) (26) (27) (28) (29) (30)
1,0 0,0367 0,0308 0,0247 0,0212 0,0292 0,0292 0,0167 0,0167
1,I 396 367 258 247 300 337 108 194
1,2 420 424 265 280 305 378 168 214
1,3 440 477 268 310 308 414 167 231
1,4 454 529 268 336 310 445 164 246
1,5 579 264 264 360 310 470 160 259
1,6 473 628 258 379 310 491 156 272
1,7 480 675 250 396 310 509 152 284
1,8 487 718 241 410 310 525 148 394
1,9 494 760 230 422 310 543 146 303
2,0 500 800 224 430 310 560 146 310
Freeedge Subscript k refers to midspan moment

~ Fixed edge Subscript t referes to support moment
a Longer side of p1ate
b Shorter side
100 STRUCTURAL BR ICKWORK
Ta ble 4 .2
Values of factors t l and t z
2 3
=12==r~1 b
4 =R~r]
0
b ~~T~~l b
0 0 0
a t2 tl t l SI2 t2 S21
-
b (I) (2) (3) (4) (5) (6)
1,0 6,43 6 ,43 7,20 0 ,384 7 ,20 0 ,384

1,1 5,91 6 ,38 7,10 355 6,61 409
1,2 5,51 6 ,34 7,01 328 6 ,15 425
1,3 5,18 6,32 6,94 302 5,73 132
1,4 4 ,91 6 ,30 6,87 275 5,40 432
1,5 4 ,69 6,29 6,82 0,248 5,11 0 ,426
1,6 4 ,50 6 ,28 6 ,77 223 4 ,88 413
1,7 4 ,34 6 ,28 6 ,72 198 4,68 393
1,8 4 ,21 6 ,28 6 ,68 170 4 ,50 371
1,9 4 ,09 6,2 8 6,65 146 4 ,34 351
2,0 3 ,98 6 ,28 6 ,62 0 ,120 4 ,19 0 ,332
00 3 ,00
5
~I]l b 0
6
:1[1=3J. 0
b
7
lli b
0
[;~:~~:~I b
0
a t2 S21 tl SI3 tl S\2 t2 S21 S21
b (7 ) (8) (9) (10 ) (I I) (1 2) (13) (14) (15)
1,0 6,65 0,202 6,65 0 ,20 2 7 ,78 0 ,358 7 ,35 0 ,292 0 ,137
1,1 6 ,21 225 6 ,54 180 7 ,63 330 6 ,80 316 159
1,2 5 ,90 251 6 ,46 163 7 ,49 301 6 ,35 335 182
1,3 5,63 279 6 ,39 145 7 ,37 272 5,99 347 206
1,4 5,40 303 6 ,35 127 7 ,26 243 5,70 351 231
1,5 5,22 0,322 6,32 0 ,107 7 ,16 0,217 5,47 0 ,352 0 ,257
1,6 5,08 337 6 ,30 088 7,07 200 5,28 350 283
1,7 4 ,96 349 6 ,29 070 6,99 190 5,13 318 310
1,8 4 ,86 36 0 6,28 048 6,91 182 5,0 0 344 338
1,9 4 ,77 369 6,28 030 6 ,85 177 4,90 339 367
2,0 4 ,71 0 ,377 6 ,28 0 ,0 13 6 ,79 0 ,174 4,8 2 0,334 0 ,391
8
=EI1 b o
t!T~-i b
a
a t 1 S 12 S 12 t2 S21
b (I 6) (I7) (1 8) (19) (2 0)
1,0 7,3 5 0 ,1 37 0 ,29 2 7 ,78 0 ,358

1,1 7,23 117 266 7 ,17 381
1,2 7 ,12 102 211 6,6 7 398
1,3 7, 03 0 30 217 6 ,2 1 410
1,4 6,9 5 081 194 5,8 1 414
1,5 6, 88 0 ,07 3 0,171 5, 5 1 0,415
1,6 6 ,82 067 151 5,24 411
1,7 6 ,76 063 132 5,00 404
1,8 6,7 1 0 58 112 4 ,73 394
1,9 6, 67 05 5 0 95 4 ,58 380
2,0 6,63 0 ,0 52 0,0 80 4 ,40 0 ,35 2
9 :[O b [~~~,~:31 b
a a
t l Sl2 Sn t2 S ZI S24
(2 1) (22) (23) (24) (25) (26)
7,88 0 ,05 8 0 ,257 7,88 0 ,25 77 0 ,0 58

7,73 04 2 228 7,27 284 084
7,59 0 35 206 6 ,78 304 115
7 ,46 03 1 191 6,35 3 17 150
7,3 4 030 180 6, 00 325 189
7,23 0 ,030 0 ,171 5, 72 0 ,331 0 ,232
7, 15 0 30 165 5,49 335 276
7, 04 0 30 160 5,3 1 338 320
0 ,36 030 155 5, 15 339 362
6, 88 0 30 152 5,0 2 34 0 399
6, 8 1 0 ,0 30 0 ,14 9 4 ,91 0 ,340 0 ,428
102
4 .0
Case B
.
tt
E~-~--=ttI •
h
I r
3.0
2.0
1.0 L-._ __ + _ - - ' - _
0 .5 2
I/).h
Figure 4.7 Factors t 3 and t 4 for walls
The method described above uses stiffness coefficients for the walls which are a
function of eccentricity. It has been shown by Frisch-Fay" that these coefficients
are modified by axial load and by the relative tensile strength of the material, as
indicated in figure 4.8. The results of Frisch-Fay's analysis could be used in a
moment distribution caJculation , although, with the relatively low slenderness
ratios normally employed in brickwork structures, the modification of the
structural parameters is not likely to be very large.
4 .2.4 Experimental verification offrame action in brickwork structures
In addition to the experimental work on wall-floor slab joints discussed in section

3.4.2 , a number of investigations have been carried out on simple brickwork
structures with the object of investigating overall frame action.
Sinha and Hendry 7 carried out full-scale tests on the three-storey, twin bay
structure shown in figure 4.9. Strains, deflections and rotations were measured for
(1) all floors loaded, and (2) floors in alternate bays loaded. The resulting deflec-
tion patterns are shown in figure 4.10, where experimental and calculated results
are compared . The theoretical values were obtained from a standard frame pro-
gramme that took into account the axial and shear deformations of the columns
and beams ; in some cases the axial deformation of the wall was as much as 20 per
cent of the maximum slab deflection , so that omission of this effect could lead to
serious error . In this case there was no cracking of the elements and the moments
of inertia of the walls and slabs were calculated on the gross cross-sections; values
t:I
tTl
VI
C5
4 z
;I-
4
Cracked",,- / Elastic z;l-
column'\ col umn r-
-<
VI
5 1\/\ , ,'N \ , u;
3 Cl
\
, / o'Tl
PIPe 2 .046 1.5 >...... ,,, C
5 / Z
Cracked .... ' Elast ic :;0
tTl
2 column / column
1.0
Z
'Tl
o:;0
("J
tTl
0 .5 PIPe t:I
t:l:l
:;0
fi
~
::E
o 0 .5 1.0 1.5 2.0 o 0.5 1.0 1.5 2.0 o
:;0
PIPe ~
PIPe
...,VI
:;0
(a) (b] C
("J
...,
C
Figure 4.8 Sti ffness and carry-ove r factors für rnasonry columns with no tensile strength :;0
tTl
(Frisch-Fay). (a) Stiffn ess co efficie nts; (b) carry-over factors VI
o
w
Stabilising ti e
Ouarr y
face
:
1
2438
13 0
~
::
+ 30
2438
/.
." =
+'30
-t
2438
'--r: r- ..., f,.-

ELEVATION
. \0-
102
~
Figure 4.9 Structure tested at Torphin Quarry , Edinburgh , t ö examine frame

action in wall-floor slab systern
of the elastic moduli were determined experimentally. On the whole , the agree-
ment between theoretical and experimental deflections is good.
Figures 4.11 a and b show for the same structure and loading cases measured
and calculated strains and eccentricities. Owing to the small magnitudes of the
strains it was difficult to achieve a high degree of accuracy, consequently the
eccentricities derived from the measured strains can only be regarded as approxi-
mate. Agreement with the calculated values in these circumstances is therefore
reasonably satisfactory.
A somewhat similar test has been reported by Germanino and Macchi" of a
twin bay, two-storey structure, in the form shown in figure 4.12 . In this case the
floor spans were quite large and the loading levels were taken up to the ultimate.
An axial load of 300 kN , simulating a superimposed load of eight floors , was
applied to the left hand wall only. The elastic modulus for the masonry was
determined by tests on small walls with a mean value of 7300 kN/mm 2 being
adopted . An E value of 30 000 kN/mm 2 was taken for the floor slab concrete. A
9.6 kN U.D .L.
Deflection in mm
experiment e or-e
theory - -
Vertical deflection
scale twice
°
horizontal
0 .2 Q, 0 .2 0.2, ,0.2 0 ..2 , 0 ,.2
0 .1 0 .1 0 .1 0.1 0.1 0 .1
Figure 4.10 Measured and calculated deflections for brick wall and concrete slab
structure--alternate floors loaded (see also figure 3,17)
first analysis was carried out assuming that joints 4 and 6 in figure 4.13 were hinged ,
that the remaining joints were rigid, and that the walls and floor slabs were un-
cracked . The results, however, did not agree weIl with those obtained experiment-
ally , because joints 4 and 6 were capable of transmitting some bending moment
and joint 3 was not fully rigid. An improved representation of the structure was
obtained by assuming all the joints as rigid, and adjusting the moments of inertia
of the walls and slabs as indicated in figure 4.13, Comparison between the calculat-
ed and experimental moments for two loading cases was then found to be as
folIows.
Load in C2 Loads in C3-<:1-<:5
Theoretical 2,11 3.45 5,56 4 ,36 2,28 2,09 6.46 13,00 10,30 9,50 6,00 2.18
Experimental 1.97 3.73 5.70 3,85 1.97 1.40 5,25 14,20 13,60 10,10 8.60 2,20
These results on the whole are reasonably satisfactory although a rather corn-
plex adjustment of the moments of inertia was applied , and this might not give
equally satisfactory results for other loading cases.
.....
o
0\
- 16 ( - 3 2 ) + 1 1(3 8 ) - 6 +23 -- 19 (2 9 ) 14(36)

p8.5(10)
eid 1.06( 1.7) et ä1.1(1 .7)

+2 .5(3.0) . J, 3(3) eldO .25(0 .3)
- 18 ('3 4 .8 ) t 13(26.6) - 17(2 8) 23(34) 5(9.8 - 2 . 5(2 .8 ) CI:l
o-J
36 - 18 31(30) - - 17 (- 2 4 ) ::a
3(3 .7 )-3( -1 .5)
..J,9(6) f!ld 0 .57( 1.67)
eld 0 .58(0.83) ~::a
etä 0.45(0.6) >
r-
30.5(34 -38(25) - 17 (2 5) 37(34) l;l:l
23.4(25) - 13. 5( 12 ) 26(33) - 6 (- 12 ) - 5 (- 7 ) 10(12) ::a

22(25) n
+22 .7 (23) .1-6.5 8) + 10(11.5) + 2.5(2 .5)
eld 0 .005(6) eld 0 .25 0.41 )
~
o
et ä 0 .2 6(0.32) eldO .5 0 .63) ::a
eid 0 .029(0) eld 0 .25 0 .2) I ;.::
23) +8 8) eld 0.2(0 .16)
..J,21
--~ 10 .2(11.5)
18 27 - 2 ( + .5 ) 3
2 .4 (2 2 .5)
'rr . n, 177
"
(a) (b)
Figure 4.11 Measured and calculated strains and eccentricities in brick wall and concrete slab structure.
Theoretical values in brackets; strains in units of 10- 6 ; gauge positions indicated -; axial strain -!-; eccentricity
ratio ejd . (a) All floors loaded ; (b) alternate floors loaded
DESIGN ANALYSIS OI" UNREINI"ORCED BRI CKWORK STRUCTURES 107
0
0
<j'
0
Ma sonry w all 0
co
N
r.c. ribb ed slab
0
0
co
N
ELEVATION SECTION
~'- ----Jfurn w all tor longitudin al S~~_bi~OO

7000 I 7000 • ~
L
1 .~ ~
PLAN
Figure 4.12 Outlin e of test stru cture for wall-floor slab interaction (Germanino
and Macchi)
C4 C5
Inerti a dist ribut io n : ...... = 50000 crn"

• A ssume d hing ed co nnect io n
Figure 4.13 Assumed static scheme and moments of inertia for wall and floor
slabs in test structure of Germanino and Macchi
Germanino and Macchi also carried out an analysis of this structure at a load
level corresponding to the ultimate limit state. This requ ired an iterative pro-
cedure starting with a calculation on th e basis of the uncr acked structure
described above, but with loads corresponding to the ultimate limit state. Where
this analysis indicated that the floor slab would be cracked , the moment of inertia
was re-calculated by a method proposed by Cauvin.? which allows for the stiffen-
ing effect of concrete on the tension side of a cracked beam or slab. The moments
of inertia of wall seetions were reduced when the eccent ricit y of load exceeded
2.44 m
152 mm
L
2.46 m
- + ' - - - - - - 4.57 m - - - - - -
Figure 4.14 Test structure of Colville and Hendry
d/6. In this case, the effective thickness of the wall over the length in which
tensile stresses appear (taken as 0 .1 of the height) was 3(d/2 - e) ; if e > d/2, a
hinge was introduced. Calculations on this basis showed good agreement between
theoretical and measured slab deflections. The tes ts also indicated that at collapse
the assumption of hinges at the critical sections of floors and at the joints of the
upper floors is justifiable .
A further test on a brickwork structure from the point of view of frame
behaviour has been described by Colville and Hendry,' 0 and in this case, the
experiments related to a two -storey, single bay structure, as shown in figure 4.14.
Loads were applied to the tops of the walls and to the floor slabs by hydraulic
jacks, and measurements were made of the deflections and rotations, from which
the wall and slab bending moments were derived . Various combinations of wall
and floor slab loadings were applied , and the results of a few of these tests are
summarised in table 4.3 . The most interesting observations from these results is
that over 80 per cent of the fully rigid joint moment is obtained with a precorn-
pression of the order of 0.3 N/mm 2 . However, the increase in joint fixity with
precompression is not linear, rising only from just over 80 per cent to just under
90 per cent, between precompressions of 0 .22 and 0 .63 N/mm 2 respectively. A
furt her series of tests indicated that the bending moments in the slab were about
the same whether the compressive stress in the wall was applied before or after the
Table 4.3
Tests on a two-storey brickwork structure
Test no . Loading Slab deflections and Moments arising from Rema rks
rotations floor load
Wall precornpression Floor load Arising from floor load M Percentage joint
(N/mm 2 ) (kN/m) (mm) 8 (10- 3 ) (kNj m) rigidity"
BI 0 2.41 3.18 1.98 1773 45 .0 Upper floor loaded

B2 0 3.98 5.84 4 .04 2093 32.2 Lower floor loaded
CI 0.22 2.41 1.88 0.84 3215 81.6 Precompression applied
C2 0.43 2.41 1.78 0.71 3351 85 .1 prior to loading lower floo r
C3 0 .63 2.41 1.58 0.59 3540 89 .8
01 0.22 3.98 3.07 1.26 5395 82 .9
02 0.43 3.98 2.90 1.09 5599 86 .0 As CI to C3
03 0.63 3.98 2.77 0.98 5741 88 .2
..Actual moment at end of slab rigid frame moment comp uted on basis of full joint rigidity
o
\D
-
slab loads. This suggests that the loading sequence resulting from the construction
of a building is unimportant in relation to the sIab and wall moments.
Although knowledge of the structural behaviour of wall-floor slab combinations
is incomplete at present , the results of the tests outlined above confirm that , with
appropriate adjustments to allow for cracking , it is possible to apply frame analysis
methods to brickwork structures when the wall compressions are in excess of
about 0.3 Njmm", A lack of full joint rigidity will result in an over-estimation of
load eccentricity on the walls, and correspondingly, slab mid-span moments, as
computed by rigid frame analysis , will be under estimated, and should be
increased by a factor
(1 - p)
F= 1 + - - -- - - (4.12)
(ms/m r - 1)
where p is the joint fixity expressed as a fraction
ms the simple span moment
m; the restraining moment at the slab end assuming full joint rigidity
4.2.5 Approximate calculation of eccentricities
An alternative approach to the calculation of load eccentricities on bearing walls
has been developed by Awni and Hendry!" following the methods originally used
by SahIin 11 and by Colville.I?
Tab1e 4.4
Notation for calcu1ation of eccentricity
Os is the slab rotation

Ow is wall rotation
Oj = Os - 0w, that is, angular displacement within the joint
H
lf> = 0 - -
w t
H , t are respectively , height and thickness of wall
wL 2
M= - ---
12
w is uniform load on the slab per unit width
L is span of slab
R is a factor depending on load eccentricity, slenderness and curvature type
Pu is totalload at joint from floors above
P L is totalload at joint below slab
'I' = P~
PL 2EI
ßis joint stiffness, ß =
ßL
e is load eccentricity € = elt
(El)s and (El)w are flexural rigidities of slab and wall respectively
K = (2El)sH/(EI)wL
DESIGN ANALYSIS OF UNREINFORCED BRICKWORK STRUCTURES lll
Using the notation set out in table 4.4, Colville showed that the following
relationships for the parameter Z = PH 2 /(EI)w hold for cracked and uncracked
walls in single and double curvature
(a) Single curvature, (i) Uncracked section
2
rr i/J
Z= - - - (4.13)
4€ +i/J
(ii) Cracked section
Z=33.31i/J ~ -2€-~Y (4.14)
(b) Double curvature , (i) Uncracked section

rr2 cI>2
Z= - -- (4.15)
(cI>+€i
(ii) Cracked section
(4.16)
if€<cI>,or
Z = 33 .3 I cI> (I - 2€i (4.17)
if € > cI>, then

wL 3 ML
8 = --- - - - (4.18)
s 24 (EI)s 2 (EI)s
where
M = (PLeL + Pue u) = (P L + Pu) e
assuming eL =e u =e , thus
3
8 = _wL (heL +Pueu)L (4.19)
s 24 (EI)s 2 (EI)s
Letting 8j = M/ß and 8 w = (heLH)/[(EDwR] we have from the relationship
8 w -», -8j
PLeH = wL 3
(4 .20)
(EI)wR 24 (EI)s
Substituting 1/J and M as defined above, and solving for e
M
e = - - =- - - - -- - -- - - - - (4.21)
PL fcl~ + 1/J) (I + ?J~I)s)
ßI.
+ (~ _CE!J -~)J
RL (EI)w
or
M
e= - - - - - - - ------:=-- (4.22)
PLt [(1 + 1jt)(1 + ß) + K/R]
If the joint is rigid, that is, (Jj =0, ß =00, ß =0 and equation 4.21 reduces to
MR
e= _ (4.23)
PLt [(1 + Ijt)R +K]
The factor R is determined from the moment-rotation equations for double
and single curvature for uncracked and cracked sections, as folIows.
For an uncracked wall in single curvature, we have from equation 4.13
11'2 cI>
Z=--
4e + cI>
As previously defined
=(Jw !! = P LH
2
cI> eL ~ = Ze (4.24)
t (EDw t R R
Substituting in the expression for Z
rr2 (Z e/R )
Z= (4.25)
4e + (Ze/R)
from which
rr2 - Z
R=--- (4.26)
4
Similarly for the double curvature case
R = rrv(Z) - Z (4.27)
For cracked seetions the derivation of R in this way results in awkward cubic
expressions. To avoid this difficulty, Awni has derived relationships giving the
maximum rotation capacity cI>max at the buckling load in terms of e by differen-
tiating the various expressions for Z, and equating to zero, thus obtaining the
curves shown in figure 4.15 . From equation 4 .24
cI> Z
(4.28)
e R
that is, the slope of the cI> against e relationship, if regarded as linear, is equal to
Z/R. For the single curvature case, it will be seen from figure 4.15 that the
relationship is linear and
cI>
max =_Z =1.332
e R
<I>
0 .7
0 .6 Double curvature
(J (J
0 .5 i <t < t tan(J =0.25

t <t< t tan(J =0.16
0.4
0 .3
0 .2
0 .1
0 0 .1 0 .2 0.3 0.4 0 .5
Figure 4.15 Curves of <I> against e for double and single curvature
Substituting in equation 4.26 then gives

R = 1.85 (4.29)
for both uncracked and cracked seetions . The <l>max against e curve for the double
curvature case is not linear, but ta king appropriate values of the slope for the
uncracked and cracked cases, the same procedure results in
R = 2.345 for uncracked seetions
(4.30)
R = 1.275 for cracked seetions
Using these values of R it is then a simple matter to calculate e directly from
equation 4.23 .
Summarising, the procedure in design is to calculate the following
wL 2
M = - - 'Pu,P L , t/J =Pu/PL
12
K = (2EI)sH/(EnwL
The eccentricity is then found by substituting these quantities into equation 4.23,
that is,
MR
€= -
ht
[(1 +t/J)R +K]
using the appropriate value of R from equation 4 .29 or 4.30.
4.3 Lateral load analysis

Wind loading may be a serious factor in the design of multi -storey brickwork
buildings and appropriate methods of lateral load analysis are required . The con-
ventional procedure is to use what is often termed the cantilever method , which
-
treats the structure as aseries of cantilever walls interconnected by links capable
-r T
D
f t
.......
-
D -
D ..ao
-
Dr-- - -
/ / ""I ' 'I / . , / 11 I / / , // / 1// /
(a) Shear wall w ith (b) Cant il ever method Ic) Shea r continuum
open ing s
,
I
-- - - - - - -
r- -, --- I I I I
,
I I I
I I 1 I I
-
I I I I
I
I
I
'- -_. - J
1 I
---."I
,
I
r- --, I
I
I
-
I
,
I I I
I I I I I I
I I
L ___ I I I I
:....
---'I
I
I r I
---I I
I
I 1 1
I I
-
I I I 1
-
I I I
I
I L __J I I
.----, -'I I
I
I ,, I I
I
I
I
I
LJ --
I
I I I
I
I
1__ _ JI
I I
,,- --, ,
~ I
I I I
I I
I I
I
I
,
I I
I
I
I
,I
I
I
i,
1 //// / /// /1/// / // / / 1 / 1//// / 1 / /1 ' // 1 'I / / / I
(d) Frame analogy (eI Wide column (f) Fin ite elem ent
frame analogy
Figure 4.16 Theoretical methods for the estimation of wind stresses and
deflections
DESIGN ANALYSIS 01" UNREINFORCED BRICKWORK STRUCTURES I 15
"""""'rT'"--..---0 9 kN
.---+01-+--4_ 1.8 kN 5h
.---+ol-+-....-t_ 1.8 kN
t--+++-14--18 kN
t--~I+-i- 1.8 kN
A - cantilever method
/ B - equivalent frame
co
C - wide column frame
analogy
D - continuum method
ITJ
6 5 4 3 2 o
B ending moment (k N rn)
Figure 4.17 Bending moments in five-storey cross-wall structure caIculated by

various methods
Sh ear continuum Experimental
5h
4h
3h
2h
o 500 1000 1500 2000

3
Deflection (m m x 10 1
Figure 4.18 Cornparison of measured and caIculated deflections of shear wall

structure tested at full scale
of transmitting direct forces but not bending mornents, as shown diagrammatically

in figure 4.16a . More complex theoretical representations are possible , including
storey by storey deflection compatibility.l" distributed shear interconnection 15
and various frame analogies.P these are indicated in figures 4.16b to e. Finite
element analysis is also possible.!? although unlikely to be necessary for design
calculations (figure 4.16f).
Analyses using these methods give quite different results for bending moments
and deflections in a particular case (see figures 4.17 and 4.18), so that a compari-
son with experimental results is useful in deciding the most satisfactory method
for brick masonry structures. Such experiments have been carried out on full-
scale and model masonry stru ctures at Edinburgh University.18,19,20 The results
of these tests indicate that the simple cantilever method over-estimates the
deflections and ben ding mornents, and may therefore be used as a means of
checking whether lateralloading is likely to be critical in a given case. At the
same time, however , this method neglects the bending moments in the floor
slabs, which could, exceptionally, crack as a result of wind loading.
4.3.1 Frame analysis for lateral loads

If lateralloading calculated on the basis of the cantilever approach appears to be
critical , a more accurate analysis is possible on the basis of an equivalent frame
using any convenient method of calculation . The shear interconnection method
and the wide column frame analogy do not appear to give satisfactory results ,
presumably because of incomplete fixit y of the interconnecting elements; attempts
have been made by Michael21 and by Soane/? to allow for th is effect.
An important problem in applying frame analysis to brickwork structures is the
definition of the effective width of return walls that act as flanges to a shear wall,
and of the effective width of floor slabs which interconnect these walls. Soane'?
has suggested, on the basis of model studies on a multi-storey brickwork structure,
a value of 14tf + t w for T-section walls, and 7tf + t w for Lsection walls. The same
author found that the effective width of floor slabs could be taken as one-half of
the bay width (that is, the distance between adjoining shear walls where these are
regularly spaced parallel to one another). These figures are supported by the
results of the full-scale and model tests on brickwork stru ctures, referr ed to
above .18-20
4.3 .2 Benjamin 's method for irregular wall arrays

Many brick masonry buildings have irregular wall arrays for which frame analysis
methods for lateralloading are not readily applicable. For such cases, the approxi-
mate method described by Benjamin!" may be appropriate. In this method, the
lateral loads are allocated between the various walls in the system in proportion to
their translational and torsional rigidities.
Consider a typical wall element in plan , as shown in figure 4.19 , displaced tu
and Lly and rotated through an angle LlB by the movement of a rigid horizontal
DESIGN ANALYSIS OF UNR EINFORCED BRICKWORK STRUCTURES 117
y x
-~I
'yY
,I
'.'
, x!'J. ()
"-
," ". .
y
~
x
Figure 4.19 Translation and rotation of wall element
diaphragm . The forces F x and F y necessary to produce the translational move-

ments are then
Fx=Rx!:ix+RxyD.y (4.31)
r, = R xy D.x +RyD.Y (4.32)
where R x and R y are respectively the rigidities (force per unit displacement) of
the element parallel to the X and Y axes , and R xy is the force necessary to prevent
displacement in the Y direction with unit displacement in the X direction. If the
element is now rotated through an angle D.() , the forces Fx and F y then become
Fx =RxD.x + RxyD.y + y R x D.() - xR xyD.() (4.33)
F y = R xy D.x + R y D.y + y R xy D.() - xR y D.() (4.34)
The torsional moment on the wall due to the rotation D.() about its own axis is
T=JD.() (4.35)
where J is the relevant torsional constant. If several walls are interconnected by
the diaphragm, the total forces on the latter are
r, = ~ F x
Py = ~ Fy
Tp = ~ yFx - ~ xFy + ~ T
and, substituting from equations 4.33, 4.34 and 4.35
Px = f:lx~Rx + f:ly~Rxy + f:l() ~ yR x - f:l() ~ xR xy (4.36)
Py = f:lx~Rxy + f:ly~Ry + f:l()~yRxy - f:l()~xRy (4.37)
T p = f:lx ~ yR x + f:ly ~ y R xy + f:l() ~ y2 R; - f:l() ~ xyR xy
- (f:lx ~ xR xy + f:ly ~ xR y + f:l() ~ xyR xy - f:l() ~ x 2R y)
+ f:l() ~ J (4 .38)
Calculations are simplified if the reference axes are such that
~ yR x - ~ xR xy = 0 (4.39)
~yRxy - ~xRy =0 (4.40)
In this case
Px = f:lx~Rx + f:ly~Rxy (4.41)
Py = f:lx~Rxy + f:ly~Ry (4.42)
Tp = f:l() (~y2 R x + ~X2 R y - 2~xyRxy + ~J) (4.43)
The total torsional rigidity of the structure is
Jp = ~ y2 R; + ~ x 2R y - 2 ~ xyR xy +~J (4.44)
The procedure for finding the forces F x ' F y and the torsional moment T on an
individual wall is thus as follows
Calculate s;R y , R xy and J for each wall .
Find the position of the reference axes to satisfy equations 4.39 and 4.40.
Solve equations 4.41, 4.42 and 4.43 for f:lx, f:ly and f:l().
Substitute these displacements in equations 4 .33 and 4 .34 for F x and F y ,
and the torsional moment on the wall is given by TpJjJp .
Wall rigidities taking account of bending and shear deformations may be calculated
from
R = (h 3 /3Ei~11 .2h/AG) (4.45)
Note that the area A is the area resisting shear, so that the 'flange' areas of an 1-
section , for example, are omitted .
Values of E and G are also required for this calculation. As indicated in section
2.9, the value of E depends on the strength of the brickwork and on the stress
level. An average value of 600 to 700 times the crushing strength of the brickwork
would, however, be reasonable. The value of G is also rather uncertain, but taking
Poisson's ratio as 0.1 would give a shear modulus of E/2.2, which has been found/"
to give deflections in reasonable agreement with experimental results,
t
b
* ..
b
(a) Open sect ions
:-r------=Jt
I t
I ...,r--,~
: : Cross-sectional area A
S I
I
'---_....lI
I'
-- - -- ..
Ib) I -section (c) Closed sect ion
Figure 4.20 Torsion of various cross-sections
Calculation of torsion constants should be by the norm al methods for the

strength of matertals.P The J value for a particular section depends on its type ;
thus , referring to figure 4.20 , an 'open' seetion has a torsion constant equivalent
to a reetangular seetion of the same tot allength and thickness and is generally
relatively small. For a thin reetangular seetion
J=tbt3 (4.46)
The torsional stiffness of an I-section , on the other hand, results from the flexure
of the flanges, and an equivalent torsional constant should be calculated on this
basis. Closed seetions, that is, of tubular form , are very much stiffer in torsion
than open ones and, in general terms, the torsional constant for a non-circular
section of uniform thickness is
(4.47)
s
where S is thc perimeter of the section.
Location of reference axes to satisfy equations 4.39 and 4.40 is achieved by
taking trial axes such that the coordinates of point (x, y) are (x t , Y 1)' Then
X=XI -X Y=YI -Y
where X and y are the coordinates of the desired origin related to the trial axes .
Substituting in equations 4.39 and 4.40 gives
X "LR xy - Y "LR x = "LxlR xy - z v,«, (4.48)
x"LRy - y"LR xy ="LxIRy -"LYIRxy (4.49)
which can be solved simultaneously for and y. x
If the principal axes of the wall section are parallel to the X and Yaxes ,
R xy =0 and in this case
"LyR x =0 "LxR y =0 (4.50)
Äx = !:-=.- Äy = -.!.L_ (4.51)

"LRx "LRy
J p = "Ly 2R x + "Lx 2R y + "LJ (4.52)
and the actions on individual walls are
_ Ry yR
Fx - Px --.. _. + T - _.- y (4.53)
"LRx p J
p
Ry xR y
F = P - - _. - - T -- - (4.54)
y Y "L R p J
y p
J
M=T -- (4.55)
PJp
The validity of this method of calculation has been confirmed by comparison with
the test results from experiments on model/ ' and full-scale structures.P A more
rigorous study of torsional effe cts has been presented by Keskin and Davies.v'
References
1. R. J. M. Sutherland , 'Design Engineer 's Approach to Masonry Construction',
in Designing, Engineering and Constructing with Masonry Products , ed . F. B.
Johnson (Gulf, Houston, Tex ., 1969) pp . 375-85 .
2. J. G. Stockbridge, 'A Study of High-Rise Load Bearing Brickwork in Britain',
M.Arch . Thesis , University of Edinburgh, 1967 .
3. P. Haller, 'Load Capacity of Brick Masonry', in Designing, Engineering and
Constructing with Masonry Products , ed . F. B. Johnson (Gulf, Houston,
Tex ., 1969) pp. 129-49.
4. P. Vahakallio and K. Makela, 'Method for Ca1culating Restraining Moments
in Unreinforced Masonry Structures',Proc. Br. ceram. Soc., 24 (1975)
161-73.
5. - 'Massiva betangsplattar', Statens Betongkommittee, SUR : S Forlags AB
(Stockholm , 1966).
6. R. Frisch-Fay , 'Stability Functions for Structural Masonry ', Int. J. Solids
Struct. , 13 (1977) 381-93.
7. B. P. Sinha and A. W. Hendry, 'An Investigation into the Behaviour of a
Brick Cross Wall Stru cture', Proc. Br. ceram. Soc., 27 (1978) 67-76.
8. G. Germanino and G. Macchi, 'Experimental Research of a Frame Idealisa-
tion for a Bearing Wall Multi-storey Structure', Proc. Br. ceram. Soc., (in
press) .
9. A. Cauvin, 'Analisi non lineare di telai piani in cemento armato', Technical
Report 1st Scienza e Technica delle Construzioni Pavia, 1977 .
10. J. Colville and A. W. Hendry, 'Tests of Load Bearing Masonry Structure',
Proc. Br. ceram. Soc., 27 (1978)77-84.
11. S. Sahlin, Strnctural Masonry (Prentice-Hall, Englewood Cliffs, NJ ., 1971)
pp .91-116.
12. J. Co1ville, 'Analysis and Design of Brick Masonry Walls', Research Report ,
Dept. of Civil Engineering, University of Edinburgh, 1977.
13. A. Awni and A. W. Hendry, 'A Simplified Method for Eccentricity Calcula-
tion ', Proceedings 0/ the Fifth International Brick Masonry Conference
(Washington) 1979, Paper 5.3 (prelim. vol.) pp. 243-6.
14. J . R. Benjamin , Statically Indeterminate Strnctures (McGraw-Hill, New
York , 1959) pp . 205-71.
15. R. Rosman, 'Approximate Analysis of Shear Walls Subjectedto Lateral
Loads',J. Am. Concr. Inst ., 61 (1964)717-33.
16 . W. W. Frischmann, S. S. Prabhu and J. F. Toppler, 'Multi-Storey Frames and
1nterconnected Shear Walls Subjected to Lateral Loads', Concr. constr.
Engng, 58 (1969) 227-34 .
17. 1. A. McLeod, 'New Reetangular Finite Element for Shear Wall Analysis',
J. Struct, Div. Am. Soc. civ. Engrs, 95 (1969) 399-409 .
18. B. P. Sinha, A. H. P. Maurenbrecher and A. W. Hendry, 'Model and Full-
scale Tests on a Five-storey Cross Wall Structure under Lateral Loading',
Proceedings of the Second International Brick Masonry Conference (Stoke-
on-Trent) 1971, ed. H. W. H. West and K. H. Speed (British Ceramic Research
Association, Stoke-on-Trent, 1971) pp . 201-8 .
19. M. Rostampour, 'Aspects of the Design of Multi-storey Buildings in Light-
weight Concrete Blockwork', Ph.D. Thesis, University of Edinburgh, 1973 .
20 . U. C. Kalita and A. W. Hendry, 'An Experimental and Theoretica1 Investiga-
ti on of the Stresses and Deflections in Model Cross-Wall Structures',
on-Trent) 1971, ed . H. W. H. West and K. H. Speed (British Ceramic Research
Association, Stoke-on-Trent, 1971) pp. 209-14.
21 . D. Michael, 'The Effect of Local Wall Deformations on the E1astic Interaction
of Cross Walls Coupled by Beams', in Symposium on Tall Bu ildings, ed . A.
Coull and B. Stafford Smith (Pergamon, Oxford, 1967) pp. 253-70.
22. A. J. M. Soane, 'Interaction of Brickwork Walls and Concrete Floors under
Lateral Load', in Designing, Engineering and Constrncting with Masonry
Products, ed. F. B. Johnson (Gulf, Houston, Tex. , 1969) 178-274.
23 . J. Case and A. H. Chilver, Strength 0/ Materials (Edward Arnold, London,
1959) pp. 282-7 .
24 . O. Keskin and S. R. Davies, 'The Effect of Torsion on Multi-Storey Struc -
tures', Proc. Br. ceram. Soc., 24 (1975) 127-37 .
5 LATERALLY LOADED UNREINFORCED
WALLS
5.1 General
It is possible to distinguish two categories of wall in relation to lateral strength:
firstly, wall panels, the resistance of which depend primarily on the flexu ral
strength of brickwork and, secondly, those whose resistance depends on the action
of in-plane forces . The first category of walls are those found in low-rise buildings
and in the upper floors of multi-storey buildings. The lateralloading on these
usually arises from wind pressure, although they have to be sufficiently robust to
withstand relatively small incidentalloads, for example from the movement of
people and equipment in a building. The second category includes walls having the
degree of precompression to be expected some two or more storeys below roof
level in a loadbearing brickwork structure, and walls whose location in a concrete
or steel structure is such that in-plane forces would be generated in them if they
were subjected to lateral deflections . These walls have lateral strengths greatly
exceeding the order of wind loading , and the problem of estimating their resistance
is likely to be associated with that of accidental damage .
5.2 The strength of brickwork panels without precompression
5.2.1 Experimental studies

A number of experimental studies of laterally loaded panels supported on three or
four sides without precompression have been reported . Losberg and Johansson ,'
and Hallquist.? drew attention to the development of a crack pattern at failure
very similar to the yield line pattern in laterally loaded concrete slabs. Several
investigators'"? have carried out tests on model-scale panels, all of which have
confirmed that failure takes place along a definite pattern of lines, dependent on
the support conditions and the ratio of height to width. The patterns of these
'fracture' lines resemble 'yield' lines in corresponding laterally loaded reinforced
concrete slabs, but on account of the brittle nature of brickwork the resemblance
is geometrical only .
A very extensive programme of tests on laterally loaded panels was undertaken
by the British Ceramic Research Association.8 - 1 1 The work was carried out at
full-scale on walls supported on three and on four sides, supplemented by tests on
122
LATERALLY LOADED UNREINFORCED WALLS 123
Lateralla ilure 8
pre ssure
8rickwork f1exural I
'r: l it
strength 6 / : curve
4 /
./. .;:
2
o
---
..
: .,.
0 .2 0.4 0.6 0 .8 1.0 1.2 1.4 1.6 1.8

H eigh tll ength 01 pan el
Figure 5.1 Laterally loaded panelsvrelationship of failing pressure to aspect ratio

for walls supported on three sides (West and Haseltine)
small specimens for the determination of the flexura1 strength of the brickwork
used for the various panels. The results are described in references 10 and 11 from
which figure 5.1 has been derived for panels built in various brick-mortar corn-
binations and supported on three sides. There is considerable scatter of results, but
the relationship between failing strength, brickwork flexural strength and aspect
ratio may be represented empirically by
!j = 1.40 - 0 .82 -hi- + 2.27 (hiJ 2

(5.1)
where p is the failing pressure in kN/m 2

f flexura1 strength of the brickwork in the stronger direction, in N/mm 2
h, L are the height and length of panel respectively (M)
Points from model tests by Kheir" , if plotted as in figure 5.1 , are in general
agreement with the results of West et al.5 on full -scale walls.
In a review of experimental work on laterally loaded panels, Baker'? has
pointed out that secondary effects in lateral load testing of brickwork panels can
significantly affect the observed failing pressure. These effects arise mainly from
unintentional, or indeterminate, restraint or yielding at the supports. Furthermore,
in comparing results from diverse sources, apparent inconsistencies can arise from
the use of different methods for determining the flexural tensile strength of the
material from small specimens.
5.2.2 Calculation of strengtb of laterally loaded panels

In view of the difficulties in ensuring that support conditions in lateralloading
tests are in accordance with the idealised representations, and because of the un-
certainty surrounding the failure criteria for brickwork in flexure , in addition to
that arising from the measurement of brickwork flexural strength, great precision
in correlating the calculated strength of laterally loaded panels with experimental
results is not to be expected.
Two approaches have been considered: firstly, by the application of elastic
theory and secondly , by various modifications to yield line analysis. On the face
of it , elastic theory would appear to be the more promising since the load-deflection
relationship for laterally loaded panels is reasonably linear ; however, allowance
must be made for the orthotropic properties of brickwork, but this presents no
difficulty in principle . The main problem relates to the criterion of failure for
brickwork subjected to biaxial bending in the presence of vertical compression.
Baker 1 2 , 1 3 has suggested that the relevant failure criterion is qualitatively as
indicated in figure 5.2 . This suggests that there is an interaction between the
moments of resistance in orthogonal directions, and further , that failure is
influenced by the compressive stress arising from self weight. In any given case,
the moments M h and Mv at the critical point in the panel will increase along a
path OA with increase in lateral pressure until OA inter sects the appropriate failure
line. Such interaction curves, however, have still to be established .
'v
IN m m') 0 .6
A xial comor ession 0 . 1 N mm
2
04
02
o 0 .5 10 1.5
' , IN .mm ' J
Figure 5.2 Possible criterion of failure of brickwork in biaxial ben ding (Baker) .
Dashed lines indicate 'no interaction' criteria
Exper imental
16 .. /
failure 14
pressure
(kN /m 2 )
12
10
0 2 4 6 8 10 12 14 16
Calculated pressure
(kN/m 2 )
Figure 5.3 Comparison between calculated and experimental results for laterally
loaded panels (Haseltine and West)
Table 5.1
Lateral resistance of brickwork panels - 'fracture' line method (Sinha 14) ; m =
ultimate mornent/unit length along bed joint ; IJm = ultimate mornent/unit length
normal to bed joint ; k = ratio of elastic moduli parallel to and at right angles to
bed joint
Frccture Iines
fJ-rn
I- - r n
T
aL
1 f-- L-l
~L
Table 5.1 cont
T1= }ßaL m= wa'L: ('ß -IJ.ß2ß'

12 2?
) 0
1+ - - -
aL k
_L ß .»: -[ jC·25IJ.Q~ + 1) - 1]
]paL 1.5IJ. Q2 k
The failure pattern of brickwork panels , resembling the yield Iine pattern in
reinforced concrete slabs, has prompted the application of yield line analysis to
this problem, although it is obvious that the basic assumption of constant moment
along a failure line cannot occur in a brittle material. Haseltine et al. 11 have shown
that strength calculations based on yield line theory , in which the orthogonal
ratio, IJ., is taken as being equal to the strength ratio of the brickwork, gives results
in good agreement with tests on walls supported on three or four sides, provided
that due allowance is made for the boundary conditions. In this method also the
effect of vertical compressive stress is allowed for by adding this stress to the
flexural stress and modifying the orthogonal ratio. A comparison between calcu-
lated and experimental results is shown in figure 5.3 .
A modification of the yield line theory has been proposed by Sinha.!" in which
it is assumed that the load is distributed in proportion to the stiffness in the two
principal directions. This results in the formulae shown in table 5.1 for various
edge conditions, and good agreement is demonstrated between the calculated and
experimen tal results for model brickwork panels . k is typically 1.25.
To avoid confusion, it would probably be more appropriate to refer to these
methods as 'fracture' line rather than 'yield' line analyses . It is undoubtedly true
that they offer acceptable methods of strength calculation but do not, strictly
speaking, represent the structural behaviour of the panels .
5.3 Lateral strength of walls with precompression
5.3.1 Experimental studies
The lateral strength of brick masonry walls with precompression has been quite
thoroughly investigated. An extensive series of tests by West et al. 15 demonstrated
experimentally the relationship between lateral strength and precompression of
storey height strip walls of various thicknesses and materials. The results of a
220mm
walls
150
lateral
resistance
(kN /m 2 )
+ +
100 )( 1 :t :3 mortar
a1 :t :3 fondu
+ 1 :1 :6
- - Simple theory
50
o 1.0 2 .0 3.0 4 .0
2
Precompression (N /mm )
Figure S.4 Lateral resistance of brick walls with precompression (after West et al.)
number of these tests are summarised in figure 5.4 ; they show that there is a more
or less linear increase in lateral resistance for all walls up to about a precompression
of 2 N/mm 2 • Above this level of precompression, the strength tends to fall away
from the linear relationship as a resuit of loeal eompression at the lines of failure .
Lateralloading tests on storey height eavity walls built within a five-storey
brickwork strueture were reported by Hendry et al. 1 6 These experim ents showed
the same strength charaeteristics as those indieated in figure 5.4 for strip walls and
also established experimentally the effect of returns (figure 5.5) . The lateral
resistanee of walls of this type is usually of interest in relation to aecidentalload-
ing, typically as a result of gas explosion. Morton and Hendry , 17 therefore investi-
gated the strength of strip walls with preeompression subjected to adynamie load ,
and demonstrated that at rates of loading equivalent to a gas explosion in a
building there was no significant difference as compared with the resistance to a
slowly increased load .
5.3.2 Theoretical treatment

The mechanism of failure of laterally loaded strip walls with precompression is
indicated in figure 5.6. If the tensile bond of the mortar to brick is neglected, and
No return e
5 One return • a
T~o returns s
B.C.R.A.
A
P
Po
Bat'
4 Po =--;T
P = ultima te transverse
pressure
3
• .'
D
No return
o 2 3
Ratio of length to height of wall
Figure 5.5 Effect of returns on the strength of laterally loaded walls with
precornpression
if no loeal erushing takes plaee at the 'hinges', the system ean be represented
statically by a three-hinged arch so that the lateral pressure at failure is given by
8a
Pe = S2 (5.2)
where Pe is the ultimate lateral pressure

o is the preeompression
S = Hit, which is the slenderness ratio
As previously observed this relationship agrees closely with experimental results
up to preeompressions of the order of 2 N/mm 2 , but above this level a more
elaborate analysis is required taking into aeeount the eompression of the brick-
work and possible local failure . The following approach was developed by
Morton .l" based on the following assumptions.
(1) The tensile bond between the bricks and rnortar, and the self weight of
the wall are negleeted.
(2) The top of the wall is supported against lateral movement without
impeding vertical movement.
(3) The preeompression remains eonstant.
Referring to figure 5.6
at t H
P /2 ,at B =- (l -cosB) +- sin B
Pe H/ 2 eH
2 2
tot i"
r-
~
Pe H/2 tT1
::tl
>
r-
H
Pe ~
-1
I 5>
H
P e H/ 2
i=t°
I
I 2
I
I
t:l
tT1
t:l
I I c:
Z
P e H/ 2 ::tl
'"-=t= '~ I . tT1
at at Z
"r'J
(a) (b) (c) o
::tl
otT1
Figure 5.6 (a) Failure mechanism of laterally loaded wall with precompression . (b) Lift of half t:l
~
height of wall at failure . (c) Rotation geornetry of half height of wall >
r-
r-
CI:l
N
\0
Pe Ht
work done by lateral force at failure =- - (5.3)
4
work done in lifting a mass above top of wall = 28 (ot + : ) (5.4)
where 28 is the maximum lift at the top of the wall taking into account elastic
deformations
w is the weight per unit length of wall
equating 5.3 and 5.4
Pe = ~ (ot +~)28 (5.5)

Ht 2
The vertical distance through which the mass above the wall is lifted is equal to
the geornetric distance calculated on the basis of a rigid material, less the shorten-
ing due to the elastic deformation. To evaluate the shortening, it is necessary to
relate 8, the angle of rotation of the wall, to the lateral deflection at the point of
instability, to the precompression 0, the compressive strength of the brickwork oe
and the slenderness ratio S. Defining u = Bit , where B is the horizontal deflection
of the centre line of the wall at rnid-height, it can be shown that, if it is assumed
u = 1 - oloe' the following equation for cos 8 are obtained
2
cos 8 (I +S2)+2 (1 -~:) cos8 + [(1 -~:Y _ S 2 ] =0 (5.6)
The positive square root of equation 5.6 gives the value of 8 at which the restering
moment becomes zero . The various stress conditions at failure shown in figure 5.7
can be identified, and from these Morton has calculated the elastic shortening /j. .
Thus in figure 5.7, cases 2 and 3
o
/j.=~ H (5.7)
E
and in case 4
H _/ 2
/j. = - V(omax - 20av t tan 8) (5.8)
2E
where 0max is the maximum compressive stress in the material and 0av the average
stress. The lift is then
(5.9)
If the compressive stress at the extreme fibre equals the ultimate stress, it is
assumed that the mass above the wall has been displaced to its limit and further
rotation causes the corner of the brickwork to be progressively crushed while the
displacement remains at its maximum. On this basis, the curves shown in figure
5.8 have been calculated for the following materials properties
Case 1 I I ~~O
Case 2
~ ~KL ~
TanO / -, Tan 0
b::::j
~ K
> K
> L Wcase3 ~ L
~ M
J
~
> L > K
> M Wcase4 > L
~ N ~ M
-, /
ltd
> K
> L
> M
> N
2 2
K=-(u,-u) L=-u
t t
2 2
M =,(u+u,) N =-(u~-u~)
tu
Figure 5.7 Failure stress patterns related to angle of rotation of half wall in
Morton 's theory
Ultimate strain : 0.001 Density of brickwork 1700 kg/m"

Ultimate tensile strength : 0.35 Njrnm? E : 7000 Njrnm?
These curves show the relationship between lateral strength and slenderness ratio
for various precompressions. An alternative presentation of the results given by
this theory is given in figure 5.9, which shows the lateral resistance of storey
height walls (2.6 m) plotted against precompression for various values of E, which
may in turn be related to brickwork strength. These results have been found to
correlate weIl with experimental values.
5.4 The lateral strength of infill panels
5.4.1 Arehing theories for strip walls

A brickwork panel built into a steel or concrete frame can develop very high
resistance to lateral pressure as a result of 'arching' effects in the wall. A solution
for the lateral strength of such panels subject to blast loads was produced by
McDowell et al. ;19,20 further experimental and theoretical studies of the resistance
of masonry walls to blast loading have been reported by Wilton, Gabrielson and
others.2 1 - 2 6 These extensive investigations, carried out on full-scale walls in a
large shock-tunnel, are summarised in reference 27.
W
IV
0 =4 N/mm 2
400
lateral
pressure
(kN/m 2 )
300 E-:7000 N/mm 2
CI>
...,
:;.:l
c::
~
200 c::
:;.:l
>
r-
t:l:l
:;.:l
n
;.:
100 ~
:;.:l
;.:
o
o 1-- :::::::-- 3 ,
5 I
10 15 20 25
Siend erness ratio
Figure 5.8 Relationship between lat eral strength and slende rness ratio for various precompressions
105 mm storey height wall
60
lateral t'"
pressure ~
(kN/m 2 ) tT1
:xl
>
e-
t'"
40 -e
r-
o
>
tl
tT1
tl
C
20 Z
:xl
E = 7000 N/mm 2 tT1
Z
'"rj
o
:xl
(')
tT1
tl
o 2 4 6 8 10 12 ~
Precompression (N/mm 2 ) >
t'"
f;;
Figure 5.9 Lateral strength of storey-height (2 .6 m) walls against precompression for various
values of E
.....
w
w
+-8-+
~l'
---+ ---r-
p(u) _______
r(u)
Figure 5.10 Analysis of laterally loaded strip wall with rigid boundary condition
(McDowell et al.)
McDowell et al. derived equations for the resistance of a laterally loaded wall
deflecting between unyielding supports, as indicated in figure 5.10. The theory is
based on the following assumptions as to material properties
(l) the tensile resistance is negligible
(2) the material has an elasto-plastic stress-strain relationship
(3) there isno strength recovery beyond the elastic range, that is, a slight
decrease in strain in the plastic range results in an instantaneous drop in
stress to zero and a permanent set equal to the plastic strain in the material.
The wall is assumed to def1ect in such a way that each half-wall rotates about
the first point in contact with the support, as indicated in figure 5.10 . Referring
to this diagram
H I - cos e
a= -
4 sin e
and
B=H
I - cos e
sin e
Thus
B
a= -
4
Putting u = Bit' and S = Hit', then
2u
sine= - - - - -
S [I + (ulS)2 ]
1 - (U/S)2
cos e = ----'--'--'--
1 + (U/S)2
The fraction of the half-depth in contact with the support is
1 + (U /S)2
Q = (I -u/2) (5.10)
1 - (U/S)2
The shortening of the material at any position y is then
ut' (I - 2g/t' - u/2)

o = (5.11)
y S [1 - (U/S)2 ]
The average strain along a fibre of the beam at a distance y from the bottom
surface is
2
€av= - O y (5.12)
H
Each fibre of the half-wall is unstressed at one end where a crack develops, and
the assumption is made that the strain varies lineariy to zero at this end . The strain
at the contact end is then given by
4
€y = 2 €av = - 0y
H
= 4u (I - 2y /t' - u/2)
(5.13)
S2 [1 -- (U/S)2]
Table 5.2
Solution for strip walllaterally loaded between unyielding supports"
8 16
Range Range ofu Stress patterns - -, P(u) --M(
of R t oc(t')2 u)
°c
R-;::'.1 u-;::.o
2 __ ~ ~u ~ _~y ~~ (I _~u) ~ _~ y
R<2..2 o~ u ~ I - V(l - 2R)
2 2
1 )
R<2 I - __ 1 4 ( I + -R + -3u - 2u - -R
2
v(I - 2R) ~ u< V(2R) 14~ -- ~ - :u)
2 4 3u
u
V(2R) ~u<I ~ 4 (I - u) + -~ [2V(2R) - u)2 4(1 - u)2 +- [2V(2R) - U)2 (5u - 4V2R)
-- 2R 6R
u
1/8 ~ R I ~u < 2 V(2R) _~ I ~ (2 V(2R) - U)2 -[2V(2R) _ u ) 2 (Su -- 4 y 2R )
<.12 2R 6R
u -;::. 2 V(2R) _ _--------.Jl o
u
V(2R) ~ u < 2V(2R) --= r l 4 (1 - u) + !!..- [2V(2R) - u)2 4(1 - u)2 + - [2V(2R) - U)2 (5u - 4V2R)
I 2R 6R
R < 1/8 2 V(2R) ~ u <I - - -0 4 (1 - u ) 4(1 - u)2
u-;::'I _ _ 10 o
Introducing the non-dimensional parameter R = (€c/4) S2 where €c is the maxi-

mum compressive strain in the material at failure, the fol1owing expression gives,
to a close approximation
U€c ,
€ = - (1 - 2y/t - u/2) (5.14)
Y R
This equation expresses the distribution of strain along the contact area at the
supports and at rnid-height of the wal1, and hence the distribution of stress . The
arehing force is then evaluated as the resultant of the stress distribution along the
contact area by integrating across the depth of the wall. On the basis of this
equation and the assumed material properties, the stress patterns shown in table
5.2 have been derived in terms of the deflection parameter u. The corresponding
values of the arehing force P(u) are also shown . When the force P(u) is known ,
the resistance moment is given by
M(u) = p(u)r(u) (5.15)
where r(u) is the lever arm , given approximately by
r(u) = t(1 - u - 2y/t') (5.16)
From these equations and a knowledge of the maximum compressive strain at
failure it is possible to calculate the resistance of the wal1 to lateral pressure.
McDowel1 et al. used these relations to calculate the resistance of panels to blast
from atomic explosions. Morton 18 produced a similar solution for the pressure
pulse resulting from a gas explosion, and Wilton et al. 24 modified the original
theory in applying it to the results of their shock tunnel experiments.
5.4 .2 Walls supported on four sides

The theory described in section 5.4 .1 relates to a strip wall. Ifthe wal1 is similarly
restrained on its vertical edges , the problem becomes considerably more compli-
cated ; McKee and Sevin 2 0 have proposed the empirical curve shown in figure 5 .11
which permits the substitution of an equivalent 'one-way' panel for one of side
and base dimensions LI and L 2 • Taking values of 0c = 7 N/mm 2 , €c = 0.001 in
Morton 's solution and using the equivalent wall coefficient, gives the curve of
lateral pressure at failure against slenderness ratio S shown in figure 5.12 . Also
shown in th is diagram are curves indicating the limitation of lateral resistance
which results from the wal1 being pushed out of the restraining structure by shear
failure around its perimeter, which is liable to occur at low values of S .
A few experimental test results have been plotted on figure 5.12. Three ofthese,
reported by Thornas.i" were 'static' tests in which panels were built into a con-
crete encased steel frame and loaded by hydraulic jacks. Two others were obtained
as part of an extensive investigation into the effects of gas explosions, carried out
by the British Ceramic Research Assoclation ."? In these cases , the walls were
built into the end of a heavy concrete bunker and loaded by exploding agas-air
mixture in the bunker. The remaining result is from the tests reported by
Gabrielson and Kaplan.I?
138
//
1.0
10 C:Jt
L, ~
t
~
11// , 1.
0 .9
L
L,
0 .8
O.7 L- ......... ~---~::_--__:""_:_
o 0 .25 0 .5 0 .75 1.0
Figure 5.11 Eq uivalent one-way spanning panel for wall simply supported on
four sides
In the case of walls having a slenderness ratio of 10 or less, there was evidence
of shear failure around the perimeter. The other walls showed very clearly that
failure followed the 'yield line' pattern of fracture lines also noted in the tests on
simply supported panels. It would appear from these results that the lateral
resistance of fully restrained panels can be estimated with fair accuracy, even
though there is some uncertainty about the ultimate strength and strain values
that should be used in the calculation.
In the foregoing it has been assumed that the brickwork panel is tightly built
into the surrounding structure. If there is a gap between the wall and the restrain-
ing structure it is still possible for arehing to develop after initial cracking of the
brickwork . Gabrielson and Kaplan!" have investigated this case and have shown
that a wall with a small gap between the top edge and the support frame developed
about 16 per cent of the strength of a fully restrained wall. This , however, was
still some three times the strength of a corresponding wall tested with simple
support conditions.
LAT ERALLY LOADED UNR EINrORCED WALLS 139
Lateral
pres sure
IkNl m ')
300
\
200 \ B.R.S . 343 mm A rch ing theory
\ -" , -".+ (a, = 7 NImm ' . 1: , ~ 0 . 0 0 1 1
"0J> ,, ·qs~ B.C.R.A . bunker test 2 20 mm
v"'""
;Y:
<,1<..-;" .......
<, Cl Gab riels on and Kap lan 203 mm
100
Sh ear failure --....... B C RA bunk er
on pe rim ete r "+
BR.S .
te st 10 5 m m
220 m m
----" BRS 105 mm
o 5 10 15
Slendern ess rat io . S
Figure 5.12 Lateral strength of infill panel s restrained on four sides (after Morton)
5.4 .3 An approximate theory for infill panels

The analysis described in section 5.4.2 is rather complex, and for approximate
calcul ations Hodgkins on et al. 3 1 have suggested the simple anal ytical model
shown in figure 5.13. This assumes tha t , at failure , the bearing width at the
supports and at the centre of the wall is a and the lateral pressure at failure is
given by
BP{t - a - 0)
Pa = -- .--.--- .-- - - .- (5.17)
L2
where P is the arch thrust
t is the wall th ickness
L is the span
o is the deflection at mid-height .
.. L
"
Figure 5.13 Approximate arehing theory (Hodgkinson er al.)
Table 5.3
Summary of arehing tests on horizontally spanning walls3 !
Brick Characteristic Failure Deflection Thrust stresslfk

type strength of pressure at failure
brickwork, fk of wall (mm) Including Excluding
2 2
(N/mm ) (kN/m ) deflection deflection
Assumed a
tll 0 tlS tilO us
A 16.0 32.9 20 2.S 1.5 2.0 1.I
B 6.4 21.0 27 4.6 2.7 3.3 1.8
Appropriate values of a and of the compressive strength fk under the conditions

in an arehing situation have been suggested on the basis of experimental results. A
number of walls were tested, spanning horizontally between rigid supports 2.72 m
apart, with the results summarised in table 5.3 .
The thrust has been calculated from equation 5.17 for two assumed values of a
and taking the deflection either as found experimentally or zero . The thrust stress
has been related to the characteristic strength for the brickwork as given in
Restraining
force
(kN/m)
600
400
105 mm wall
200
o 5 10 15 20 25
Sienderness ratio
Figure 5.14 Forces required to develop full strength of 105 mm and 220 mm
walls by arehing action
BS 5628 :1978 . As a result, it is suggested that for design purposes the bearing
width , a should be taken as t/lO , the deflection omitted and the thrust stress
limited to 1.5 times the value of /k ' These proposals are of a tentative nature , but
could be applied with reasonable confidence to panels spanning between two rigid
supports, and in particular to those spanning horizontally which are unsupported
on their top edges, and which rest on a damp-proof course at their bases.
Attention should be drawn to the magnitude of the restraining forces generated
by arehing action ; by way of illustration, the curves of figure 5.14 have been
derived from Morton's theory. From this figure it will be seen that considerable
forces are required to develop the full strength of an infill panel and, conversely,
large forces may be imposed on the surrounding structure if a brickwork panel is
loaded to failure by lateral pressure.
References
1. A. Losberg and S. Johansson, 'Sideways Pressure on Masonry Walls of Brick-
work' , CIB Symposium on Bearing Walls(Warsaw) 1969.
2. A. Hallquist, 'Lateral Loads on Masonry Walls', Reprint 172 (Norwegian
Building Research Institute , Oslo, 1970) .
3. K. M. H. Satti, 'Model Brickwork Panels under Lateral Loading, Ph.D.
4. A. M. A. Kheir, 'Brickwork Panels under Lateral Loading' , M.Phi1. Thesis,
University of Edinburgh, 1975.
5. H. W. H. West, H. R. Hodgkinson and W. F. Webb, 'Lateral Loading Tests
on Walls with Different Boundary Conditions', Proceedings 0/ the Third
International Brick Masonry Conference (Essen) 1973 , ed. L. Foertig and
K. Gobel (Bundesverband der Deutschen Ziegelindustrie, Bonn, 1975)
pp. 180-6 .
6. L. R. Baker, 'Flexural Strength of Brickwork Panels', Proceedings of the
Third International Brick Masonry Conference (Essen) 1973, ed. L. Foertig
and K. Gobel (Bundesverband der Deutschen Ziegelindustrie, Bonn, 1975)
pp. 378-83 .
7. B. P. Sinha, M. D. Loftus and R. Temple, 'Lateral Strength of Model Brick-
work Panels', Proc. Instn civ. Engrs, 67 (1979) 191-8.
8. H. W. H. West and H. R. Hodgkinson, 'The Lateral Load Resistance of
Brickwork without Precompression', Proc. Br. ceram. Soc., 24 (1975)
101-13.
9. H. W. H. West and B. A. Haseltine, 'The Design of Laterally Loaded Walls',
Proceedings 0/ the Fourth International Brick Masonry Conference (Brugge)
1976 , Paper 4.b.10.
10. H. W. H. West, H. R. Hodgkinson and B. A. Haseltine, 'The Resistance of
Brickwork to Lateral Loading, Part 1, Exper imental Methods and Results of
Tests on Small Specimens and Full Sized Walls', Struct. Engr, 55 (1977)
411-21.
11. B. A. Haseitine, H. W. H. West and J. N. Tutt, 'The Resistance of Brickwork
to Lateral Loading, Part 2 , Design of Walls to Resist Lateral Loading' , Struct.
Engr, 55 (1977) 422-30.
12. L. R. Baker, 'The Lateral Strength of Brickwork-An Overview', Proc. Br.

ceram Soc., 27 (1978) 169-88.
13. L. R. Baker, 'A Failure Criterion for Brickwork in Bi-axial Bending' , Proceed-
ings of the Fifth International Brick Masonry Conference (Washington) 1979,
Paper 2.7 (prelim. pp. 37-40).
14. B. P. Sinha , 'A Simplifled Ultimate Load Analysis of Laterally Loaded
Model Orthotropic Brickwork Panels of Low Tensile Strength', Struct. Eng.,
56 (1978) 81-4.
15. H. W. H. West, H. R. Hodgkinson and W. F. Webb, 'The Resistance of Brick
Walls to Lateral Loading',Proc. Br. ceram . Soc ., 21 (1973) 141-64 .
16. A. W. Hendry , B. P. Sinha and A. H. P. Maurenbrecher, 'Full Scale Tests on
the Lateral Strength of Brick Cavity Walls', Proc. Br. ceram. Soc., 21 (1973)
165-80.
17. J. Morton and A. W. Hendry, 'An Experimental Investigation ofthe Lateral
Strength of Brickwork Panels with Preeompression under Dynamic and
Statie Loading', Proceedings of the Third International Brick Masonry
Conference (Essen) 1973, ed. L. Foertig and K. Gobel (Bundesverband der
Deutschen Ziegelindustrie, Bonn, 1975) pp. 362-9.
18. J. Morton , 'A Theoretical and Experimental Investigation of the Static and
Dynamic Lateral Resistanee of Briekwork Panels with Reference to Damage
by Gas Explosion', Ph.D. Thesis, University of Edinburgh, 1970.
19. E. L. McDowell, K. E. McKee and E. Sevin, 'Arching Action Theory of
Masonry Walls',J. struct. Div. Am. Soc. civ. Engrs, 123 (1958).
20. K. E. McKee and E. Sevin, 'Design of Masonry Walls for Blast Loading',
Trans. Am. Soc. civ. Engrs., 124 (1959)457-71.
21. A. B. Willoughby, C. Wilton, B. L. Gabrielsen and J. V. Zaccor, 'Loading
Structural Response and Debris Characteristics of Wall Panels, URS 680-5
(URS Research Co., San Mateo , CaHf., 1969).
22. C. Wilton, B. Gabrielsen, J. Edmunds and S. Bechtel, 'Loading and
Structural Response of Wall Panels' , URS 709-4 (URS Research Co., San
Mateo , CaHf., 1969).
23. C. Wilton and B. L. Gabrielsen, 'Shock Tunnel Tests of Preloaded and Arched
Wall Panels', URS 7030-10 (URS Research Co., San Mateo, CaHf., 1973).
24. C. Wilton, B. Kaplan, B. L. Gabrielsen and J. V. Zaccor, 'Blast/Fire Inter-
action, Blast Translation and Toxic Gases', SSI 743-8 (Scientific Services
Inc. , Redwood City, CaHf., 1976).
25. B. L. Gabrielsen and C. Wilton, 'Shock Tunnel Tests of Arehed Wall Panels',
URS 7030-19 (URS Research Co., San Mateo , CaHf., 1974).
26. B. L. Gabrielsen, K. Kaplan and C. Wilton, ' A Study of Arehing in Non-
Reinforced Masonry Walls', SSI 748-1 (Scientific Services Inc., Redwood
City , our. 1975).
27. B. L. Gabrielsen , C. Wilton and K. Kaplan, ' Response of Arehing Walls and
Debris from Interior Walls Caused by Blast Loading', URS 7030-23 (URS
Research Co., San Mateo, CaHf., 1975).
28. F. G. Thomas, 'The Strength of Brickwork', Struct. Engr., 31 (1953) 35-46.
29 . N. F. Astbury, 'Gas Explosions in Load Bearing Brickwork Structures',
Special Publication No. 68 (British Ceramic Research Association , Stoke-
on -Trent, 1970) .
30 . B. L. Gabrielsen and K. Kaplan , 'Arching in Masonry Walls Subjected to
Out-cf-plane Forces . Earthquake Resistant Masonry Construction : National
Workshop', NBS Building Science Series No . J 06 (Boulder, Colo., 1976)
pp. 283-313.
31 . H. R. Hodgkinson , B. A. Haseltine and H. W. H. West, 'Preliminary Tests on
the Effect of Arehing in Laterally Loaded Walls', Proceedings 0/ the Fourth
International Brick Masonry Conference (Brugge) 1976, Paper 4 .a.5.
6 REINFORCED AND PRESTRESSED
BRICKWORK
6.1 The application of reinforced and prestressed elements

The brittle nature of brick masonry is no great disadvantage in situations where the
load to be transmitted is vertical and compressive stresses predominate. It does,
however, put severe restrictions on its use for elements in which significant tensile
stresses are developed, and to overcome this limitation it may be useful to use
reinforced or prestressed members. Special problems also arise in brickwork
structures in seismic areas and in relation to resistance to accidental damage which
may necessitate the use of reinforcement.
There are three basic ways of introducing reinforcing steel into brickwork
construction : by placing it
(a) Within the mortar joints
(b) In specially formed pockets
(c) In a grouted cavity between skins ofbrickwork
These methods are illustrated in figure 6.1 , and examples of their application will
be found in references 1 to 8.
The flexurallimitations of brickwork can also be overcome by the use of pre-
stressing techniques'v'" although practical application ofthis method has so far
been limited.
6.2 Reinforced brickwork flexural elements

Early research work 1 1 indicated that the principles underlying the design of
reinforced concrete could be applied also to reinforced brickwork, provided that
suitable adjustments were made for the differences in material properties. This
was substantiated by the results of investigations 1 2 - 1 8 carried out in the United
States in the early 1930s and which form the basis of current practice in that
country. Experiments on reinforced brickwork carried out in the United King-
dom 1 9 ,2 0 about the same time subsequently found expression in the British Code
of Practice CP 111 :1964 for Loadbearing walls. Research in the mid-1970s on
reinforced brickwork was concerned particularly with shear strength.i!"? 3 the
effects of cyclic loading/" and laterally loaded reinforced walls.2 s - 3o
144
REINFORn:o AND I'Rt-:STRESSEO BRICKWORK 145
ta) Reinforcem ent pla ced within th e mort ar jo ints
(il in bed joi n ts

(ho rizon ta l rein forcernent)
RIR
BB (111 100011" '0'0'
(ve rtical relntorcern entl
.
:.
g§D . , . : .'
. •... ~
lb ) Reinfo rcement placed in sp eci ally fo rm ed po ckets
(i) Ou ett a bond
..
~
~ ' .
:= (iil poc ke t in brickwork
~ 11111"" 01 so ecial bricks
P;.)., ·,9
\5L.-.__ ~ 11
E3p1l1
BE l horizonta l reinforcement
(i) ve rt ic al re in forcem ent
(c) Grouted cavitv
Figure 6.1 Method s of reinforcing brickwork
Following the same principles as reinforced concrete, reinforced brickwork may

be designed either on the basis of linear elasticity, or in relation to ultima te
strength. Calculations relating to the serviceability limit state of deflection will be
based on elastic behaviour, while those relating to the ultimate limit state will
rcquire assumptions as to the stress-strain relationship up to ultimate load . Values
for the elastic modulus, and discussion of actual stress-strain curves may be found
in chapter 2; idealised stress-strain curves for brickwork and reinforcing steel will
be discussed in section 6 .2 .2 . It is usual to assurne that in flexural elernents, plane
sections remain plane after ben ding and that the tensile strength of brickwork may
be negleeted.
6.2.1 Flexural strength ofreinforced brickwork

On the above assumptions it is possible to derive expressions for the moment of
resistanee of a reinforced brickwork seetion in the same way as for reinforced
concrete. Thus, on the basis of linear elastic theory , for a reetangular section
M=tbxfbC (d -i)=Asfs (d - -j) (6.1)
where b is the width of section

x is the depth of neutral axis
fbc is the permissible compressive stress in brickwork
d is the effective depth of section
A s is the area of steel reinforcement
f s is the permissible tensile stress in steel
The depth of the neutral axis can be calculated from the equation
bx? + 2QexAs -- 2QedAs =0 (6.2)
where Qe is the modular ratio.
If the steel area is such that fbc and f~ are reached simultaneously
x 1
- - - - --_ .._ --- (6.3)
d I + (fs jQefbc)
Vallies of permissible stresses are specified in design codes for various brick-
mortar combinations.The modular ratio is a function of the brickwork compressive
strength, and willlie in the range 10 to 40 for streng to relatively weak masonry.
Design codes are now tending to be based on limit state prineiples, and thus
for consideration of the ultimate limit state a non-linear stress block has to be
considered. In this case the section analysis again follows the precedent of rein-
forced concrete. Thus, referring to figure 6 .2 and following the theory described
by Kong and Evans 31 for a singly reinforced beam, we have
or
(6.4)
and
(6.5)
REINFORCED AND PRESTRESSED BRICKWORK 147
k,x
x ...-,....-f-~:....C, co m pressive force
~---t~ T, tensile force
Figure 6 .2 Stress distribution in singly reinforced beam
f,
<, Equa tion 6 .6
Figure 6.3 Intersection of equation 6,6 with stress-strain curve for steel
Combining equations 6.4 and 6.5 gives

k fk €k
f s = - t-- (6.6)
P €k + €s
where pis the steel ratio A s /bd.
At the ultimate limit state, the values of fs and €s must satisfy equation 6 .6,
and also define a point on the stress-strain curve for the steel, as indicated in
figure 6.3. The moment of resistance is then
M u = Asfs (d -- k 2 x )
=Asfs ( I -p -k2- -fs) d (6 .7)

k , !k
The term within the brackets in equation 6 .7 is the lever arm factor for the
beam .
As in reinforced concrete, the mode of failure of a brickwork beam , reinforced
with steel having adefinite yield point , depends on thc steel ratio p . If the steel
and brickwork strains reach €y and €k simultaneously this is termed a 'balanced'

section, in which
x
(6.8)
d €k + €y
Substituting in equation 6.6 gives the steel ratio for this condition as
!k
P = k , -'- - -
€k
(6.9)
ty €k + €y
If the steel ratio is less than this, failure of the beam, described as 'under-
reinforced', is initiated by yielding of the steel, but the beam will continue to
resist increased bending moment until the strain in the brickwork reaches €k.
Conversely, in an over-reinforced beam, failure will follow the attainment of €k
in the brickwork. In this case it can be shown that
k::k (~y + €k (~) - €k =0 (6.10)
The neutral axis factor xld can be calculat.ed from this equation and the moment
of resistance from
(6.11)
In practice, over-reinforced brick masonry beams will usually fail in shear before
the full compressive strength of the brickwork is attained.
In the above discussion the stress-strain relationship for brickwork has been
characterised.by the two ratios k. and k 2 • There is not a great deal of experi-
mental data on which to base numerical values of these parameters for brickwork
but analysis of the results of Powell and Hodgkinson discussed in chapter 2
suggests that the maximum strain and the shape of the overall stress block for
Stress
(N /mm 2 )
20
10
o 0 .002 0.004 0.006

Strain
Figure 6.4 Stress-strain curves for various types of brickwork (PoweIl and
Hodgkinson)
-t 1.5 '. !)'mm t-

,c u" = 0.00 3
0 .27x O.4x
E=0 .0 0 2 2
x
Figure 6.5 Idealised stress-strain curve for reinforced brickwork
brickwork are rather variable; this may be seen in figure 6.4 . However, three
of the four brickwork types covering a considerable strength range gave average
values of 0.76 and 0.46 for k 1 and k 2 respectively.
It is usual for design purposes to use an idealised stress block, and that shown
in figure 6.5 has been suggested P for reinforced brickwork following the corres-
ponding curve used in the British Code of Practice CP 110 for the structural use
of concrete. The values of k 1 and k 2 corresponding to this stress block are 0.73
and 0.39. The design moment of resistance, after introducing the partial safety
factors for material strength, that is, 'Ymm for brickwork and 'Yms for steel, is as
follows
0.3 bd? 1.5!k
A1d = .- (6.12)
'Ymm
or
(6.13)
where
z= ~ _ 0 .53A s fy_ 'Ymm) d (6.14)
\: bd 'Yms 1.5!k
For a balanced section
A = 0.38 1.5fk 'Y ms bd (6.15)
s
'Ymm fy
6.2.2 Shear strength ofreinforced brickwork beams

Thc shcar strength of reinforced beams of various types has been studied by a
nurnber of investigators and, although the general pattern of behaviour is similar
to that of reinforced concrete, there are a number of significant differences arising
150 STR UCTURAL BRICKWORK
Table 6.1
Summ ary of shear str ength test results
Ref. Beam vc * af d P fk Mortar by

no . (N/mm 2 ) (%) (N/mm 2 ) volume
5 CA-I 0 .90 3 3.27 1.03 10.83

CA-3 0.896 3.20 1.0 1 10 .83
CB-I 0.821 3.67 1.20 10.83
CB-2 0.724 3.67 1.17 10.83
CB-3 0.765 3.67 1.16 10.83
CC-2 0.46 9 4 .23 1.31 10.83
CC-3 0.4 96 4 .16 1.31 10.83 1:0.15:3
PA-l 0.5 24 3.0 1 0.97 12.09
PA-2 0.510 3.08 0 .57 12.09
PA-3 0 .641 3.2 1 1.02 12.09
PB-l 0.517 3.46 1.07 12.09
PB-2 0.455 3 .46 1.10 12.09
PB-3 0.560 3.46 1.05 12.09
6 C3 0 .545 3 .65 0 .62 7.24 1:0 .78 :4 .7

W2 0 .655 3.61 0 .70 24 .30
7 5 1.055 1.79 1.39 11.87 1:0.25:3.5

8 0.95 2 1.79 1.39 8.07
}
8 A-l 0.717 1.2 1 1.43 5.10
A-2 0.572 1.21 1.43 5.10
A-3 0. 738 1.21 1.4 3 5.10 1:0. 16 :3
s-i 0.924 1.21 1.43 7.65
B-2 0.9 17 1.2 1 1.43 7.6 5
B-3 1.062 1.2 1 1.43 7.6 5
}
AA-l 0.96 9 1.2 1 1.39 4 .62
AA-2 1.103 1.24 1.47 4 .62
AA-3 1.04 8 1.23 1.4 5 4. 62 1:0 .3: 4. 5
BB-l 1.434 1.20 1.41 6.83
BB-2 1.234 1.2 1 1.43 6.83
BB-3 1.165 1.2 1 1.43 6.83
AF-l 0.807 1.32 1.42 7.65
AF -2 0.965 1.32 1.42 7.65
AF-3 1.117 1.34 1.47 7.6 5
BF-I 0.855 1.20 1.29 7.65
BF-2 0.876 1.18 1.27 7.6 5 1:0. 16: 3
BF-3 1.234 1.20 1.3 1 7.65
n-t 1.3 24 1.21 1.05 7.6 5
H-2 1.338 1.2 1 1.05 7.65
H-3 1.214 1.23 1.06 7.65
Ta ble 6 .1 cont
Ref. Beam vc * a/d P !k Mortar by

no . (N/mm 2 ) (%) (N/mm 2 ) volume
}
9 0.67 4 .4 1.05 8 .0
0.85 4 .9 1.47 8 .0
0.66 3.9 0.73 8 .0 1: 1:3
0.59 4 .5 1.05 8.0 appro x.
10 B6 0.641 2.5 0.52 14.7 6

B8 0.634 2.5 0 .52 14.76
B5 0.345 1.7 0.23 14.76
B7 0.559 1.7 0 .23 14.76
2C3 0.752 2.5 0.92 14.76
2C4 0.793 2.5 0.92 14.76
2Cl 0.965 1.7 0 .92 14.76
2C2 1.151 1.7 0.92 14.76
2B3 0.648 2. 5 0.92 14 .76 1:0.15 :3
2B4 0.662 2.5 0.92 14.76
2BI 0.965 1.7 0.9 2 14.76
2B2 0.95 2 1.7 0.92 14.76
2A3 0.917 1.9 0.92 14.76
2A4 0 .958 1.9 0 .92 14.76
2A I 1.16 5 1.3 0.92 14.76
2A 2 1.379 1.3 0.92 14.76
4C I 0.731 3.2 0.92 14.76
4C2 0.786 3.2 0.92 14.76
}
11 9 1.324 2.5 2.33 -10
10 1.3 10 2.5 2.33 - 10
II 1.06 9 2.5 1.89 - 10 1:0 .25 :3
12 1.14 5 2.5 1.89 - 10
13 0.993 2.5 2 .30 - 10
14 0.986 2.5 2.30 - 10
12 I 1.056 1.0 0.24 9.42

2 0.556 1.5 0.24 9.42
3 0.402 2.0 0.24 9.42
5 0.364 3.0 0.2 4 9.4 2
6 1.51 2 1.0 1.46 9.4 2
7 0.788 1.5 1.46 9.4 2 1:0 .25 :3
8 0.596 2.0 1.46 9.42
9 0.4 37 2.5 1.46 9.42
10 0.491 3.0 1.46 9.42
11 0.410 4 .0 1.46 9.42
12 0.330 5.0 1.46 9.4 2
*vc = Vc/b d where Vc is the shear force carried by the reinforced brickwork beam due to
ultimate loads, b is the breadth of the section, and d is the effective depth
150
STR UCTURAL BRICKWORK
o 2 3 4 5
a/d
Figure 6.6 Collected test results for shear strength of reinforced brickwork beams plotted against shear
span ratio (Suter and Hendry)
REINFORC ED AND PRESTRESSED BRICKWORK 153
from the physical nature of the two materials. It is known that the shear resist-
ance of reinforced concrete beams is influenced by the ratio of the shear span to
effective depth , the amount of tensile reinforcement and the concrete compressive
stress. The effect of these factors on the strength of reinforced brickwork beams
can be dedu ced from figure 6 .6 and table 6.1 , which summarise theresults of a
large number of tests by various investigators on beams of a range of cross-sections,
brick patterns , etc . of type (a) in section 6.1 .
Examination of th is data indicates that, although there is a considerable
amount of scatter, there is a clear trend of increasing shear strength with decreas-
ing ratio of shear span to effective depth (a/d). This effect is marked at a/d ratios
of less than 2 and results from the mode of failure. Thus , at higher a/d ratios shear
failure follows the development of a typical diagonal crack, whereas in beams with
a low ald ratio cracking is followed by the development of a tied arch effect-the
lower the shear span ratio the greater the arehing strength and thus the apparent
shear resistance.
The shear strength of reinforced concrete beams increases with the amount of
tensile reinforcement , but the results of tests on reinforced brickwork beams,
however, do not indicate that this relationship holds in their case. Figure 6.6
differentiates between beams having a steel percentage greater or less than 1 per
cent, but there is no evidence from this that the amount of steel has any influence
on the ultimate shear stress in the beams.
It would be expected from the parallel with reinforced concrete that shear
strength would increase with the compressive strength of the brick masonry, at
least to a limited extent. Unfortunately, it is difficult to examine this point on the
basis of available test results , since the compressive strengths repo rted have been
... rei nforced co ncrete (Su te r & Kel ler)
., g rou ted cav it y b/work (Suter & K ell er)
2.0 Q grou t ed cavi ty b/work (Sinha)
• b ed joint reinforced b/work (Suter & H endry)
1.0
o 2 . 4 6
ald
Figure 6.7 Comparison between shear strength of reinforced concrete, grouted
cavity and bed reinforced brickwork beams
determined by a variety of methods which do not necessarily give the same result .
Inspection of table 6.1 , however , does not suggest that correlation between the
shear strength and compressive strength of the masonry is very high.
Figure 6 .6 shows a lower bound curve on shear strength as related to shear
span ratio, which approximates to the characteristic shear strength for this type
of reinforced brickwork beam .
Suter and Keller23 have studied grouted cavity reinforced brickwork beams and
have shown that, as regards shear strength, these are intermediate between rein-
forced concrete and reinforced brickwork beams in which the reinforcement is
embedded in the mortar joints. This is illustrated in figure 6.7, and it is suggested
that the shear capacity of a grouted cavity beam can be derived from 'the separate
shear capacities of the grouted core and the brickwork sections according to their
relative widths. The results of further experimental studies of grouted cavity beams
by Sinha?" are also plotted in figure 6.7 and are in substantial agreement with
those of Suter and Keller. It may be added that the shear strength of grouted
cavity beams increases with the steel percentage .
6.2.3 Calculation ofdeflection of reinforced brickwork beams

There is very little data available on the deflection of reinforced brickwork beams
and related methods of calculation . However, to the extent that reinforced brick-
work behaves in the same manner as reinforced concrete, the methods and formu-
lae which have been developed for the latter material may be applied. Thus, the
following relationships" for the neutral axis depth and second moment of area
of a cracked section with tensile reinforcement are relevant to reinforced brickwork
(6.16)
and
(6.17)
In practice, the effective second moment of area will vary along the length of a
beam as the extent of cracking changes along the span, in the manner suggested in
figure 6.8. Furthermore, tensile stresses will be developed in the material below
the neutral axis between the cracks, and result in a stiffening effect . In reinforced
concrete beams this effect can be allowed for by assuming the existence of a
limited tensile stress in the concrete below the neutral axis that reduces the
moment on the cracked section used in calculating deflection by an amount
equal to
I b(h _ X ) 3 '1 .
3" x tensi e stress m concrete
(d -x)
where h is the overall depth of the beam. In principle this could be applied to
reinforced brickwork beams, but experimental confirmation is lacking.
REINfoORCED AND PRESTRESSED BRICKWORK 155
Figure 6.8 Variation of crack depth along the length of a beam
The value of Young's modulus for brickwork is discussed in section 2.9 . An

approximate value of 900 /k has been suggested 3 2 for medium-strength reinforced
clay brickwork, and this would be applicable to short-term loading but should be
reduced by a factor of 2.0 when calculating long-term deflections. This is based on
results reported by Wyatt et al. 33 and is confirmed by long-terrn loading experi-
ments on grouted cavity cantilever walls carried out by Sinha. 3 4 , 3 5 If, in the case
of a cantilever wall, there is a non-rigid damp-proof course near the base the
creep deflections may be doubled.
6.3 Reinforced brickwork compression elements

The strength of reinforced brickwork compression elements has not been exten-
sively investigated, presumably because in most practical situations it is possible
to meet design requirements with unreinforced walls or columns. With axial
loading the addition of a low percentage of reinforcing steel will result in only a
small increase in the load capacity of the member but , if more extensive use is
made of reinforced brickwork, the use of beam columns or walls transmitting
substantial bending moments in addition to axial loads may become more wide-
spread and a knowledge of the strength of such elements will be required .
Tests 3 6 , 3 7 on axially loaded reinforced brickwork columns have confirmed
that their strength could be regarded as the sum of the compressive strength of
the brick masonry and of the steel at its yield point. The presence of lateral ties
was found to add slightly to the strength of the column and to prevent complete
collapse at the maximum load . Thus in limit state terms, the design capacity of a
short compression member would be
N = ikA b + 0.8/yA se (6.18)

'Ymm 'Yms
where Ab and Ase are, respectively, the areas of brickwork and steel, and 'Ymm and
'Yms are partial safety factors for these materials which have, respectively,
characteristics ik and /Y'
Consideration of the strength of reinforced brickwork members under com-
bined axial load and bending moment will be facilitated by referring to direct
load-moment interaction diagrams, similar in principle to those referred to in
section 3.5 for plain brick masonry. For this case, assuming for simplicity a
reetangular stress block, and using the notation shown in figure 6.9a
l~-b--+
(e) h
l~
.!!
1
(b) h
....10- •• A ' 2.
d2
fr------L
Figure 6.9 Strain and stress distributions in (a) plain seetion, (b) with reinforee-
ment A s2 and (e) with reinforeement Ast'
(6.19)
(6.20)
or, in dimensionless form

N bwk dc
abwk = - - = - (6.21)
Abh h
ßb k = M bwk = 0.5 d c (1 - dhc) (6.22)

w fkbh2 h \'
By calculating abwk and ßbwk at various values of dc/h between 0 and 1.0, the
axial load-moment interaction curve shown in figure 6.10 is obtained.
N
Cl = -
f,bh
;:l:l
/ t'l1
Z
'Tl
1.0 o;:l:l
/ o
t'l1
t:l
>-
Z
0 .8
t:l
-e
;:l:l
t'l1
CI>
...,
0 .6 ;:l:l
t'l1
CI>
CI>
t'l1
/ t:l
0.4 t:l:I
;:l:l
n
;.::
0 .2
~
;:l:l
;.::
o 0 .04 0 .08 0 .12 0 .16 0 .20 0 .24 0 .28 0 .32

2
ß =M/f,bh
.-
Vl
Figure 6.10 Axial force{b ending moment interaction diagrams for plain and reinforced -.l
brickwork seetions
If now an area of reinforcement A S2 is introduced at a depth of d 2 from the

lower face of the beam as shown in figure 6 .9b, this will contribute to the axial
force by an amount
(6 .23)
where [a is the stress in the steel corresponding to astrain €s2. The moment about
the mid-depth contributed by the steel is
M (A s2) = - A sd s2( %-d 2) (6.24)
In dimensionless terms
(6 .25)
(6 .26)
These parameters can be calculated for a given section using the relationship
€s2 d 2 - ( h -d e) d e/h-(1 -d 2 /h)
- = = (6.27)
€k a; de/h
Thus, [a = Es €s2 ' and hence a s2 and ßS2 can be evaluated for various values of
d.fh , The effect of the reinforcement on the interaction diagram is represented by
a vector v'[(aS2)2 + (ßS2)2] as shown in figure 6.10.
If reinforcement is introduced near the upper face of the beam, as shown in
figure 6 .9c, corresponding values of asl and ßS1 are as follows
(6 .28)
bh fk
= (t - ~~)aSl (6 .29)
Also
de/h - d'/h
(6.30)
de/h
The effect of this reinforcement in addition to A S2 is then represented orr the inter-
action diagram, figure 6.10, by the vector v'[(as d 2 + (ßs2i].
Figure 6.1 0 is based on illustrative values of the various terms; although the
characteristics of these curves are fully discussed in reference 31, the following
points may be noted
(1) Considering the interaction diagram for A s2 steel when de/h = 1 - d 2/h,
0 .85 for the example shown, the steel strain is zero, both a s2 and ßs2 are
zero and the steel is inactive.
(2) For the steel to reach the yield stress I y in this case, Es = 0.00205. This
occurs at a value of dc/h = 0.505 for the example shown. This point defines
the maximum value of ß on the interaction diagram and corresponds to a
'balanced' failure , in which the steel and brick reach their limiting values
simultaneously. In load-rnornent (N-M) combinations above this point, the
brickwork will reach its limit before the steel and vice versa below this
point.
(3) When both A s l ' and A s2 steel reinforcement is present the maximum ß
again corresponds to the balanced failure condition.
(4) Load-moment combinations corresponding to specific eccentricity ratios
lie on straight lines, as indicated in figure 6.1 O.
The interaction diagrams described above relate to short compression members.
As mentioned in section 3.5, these curves can be modified using the moment
magnifier method or otherwise to allow for the second order effects from
slenderness.
Experimental confirrnation of the calculated interaction diagrams for reinforced
brickwork is limited, but aseries of tests by Anderson and Hoffman 3 8 showed
P
(MN)
5
-: +-
e/h =0.3
~300mm
2
*' '. 111 mm
#~-J--
--
e :: .
h =406 mm
"
elh~ 111 mm
o 50 100 1 50 200 250

M(kN rn l
Figure 6.11 Results of tests on eccentrically loaded reinforced brickwork columns
compared with calculated interaction curve (Anderson and Hoffman). Brick
strength : 93 N{mm2 ; mortar: ASTM 270 type S; grout : 25 N{mm2 av.; masonry
strength : 36 N{mm2 ; steel: ASTM AI5 -4, f v = 275 N{mm2 , 4 No. 16 mrn dia. bars
good agreement with experimental results up to an eccentricity ratio of 0.34, as

may be seen from figure 6.11. Such diagrams are, however , well established for
reinforced concrete and there is no obvious reason why Ehey should not be
equally valid for reinforced brickwork.
6.4 Reinforced brickwork shear walls

The strength of reinforced masonry shear walls has been extensively st udied by
Scrivenerj'" although most of his work relates to concrete block masonry, his
general conclusions would apply to clay brickwork having corresponding
reinforcement.
It was confirmed that reinforced masonry behaved in the sarne manner as
reinforced concrete . Thus the str ength of a wall loaded in its plane as a vertical
cantilever .rnay be closely predicted by assuming that all the vertical reinforce-
ment has yielded and that the compression zone is located at die 'leeward' toe of
the wall. If, however, flexural stresses were suppressed by applying a simultaneous
verticalload to produce a shear condition in the masonry, it was found that the
location of the reinforcement had little effect on the failing load, although an
even distribution increased the initial crack ing load . Scrivener reported a shear
strength for walls, having 0.3 per cent -or more reinforcement , of 1.19 N/mm 2 •
Higher values can be expected in walls carrying high verticalloading.
Scrivener extended his werk to examine the strength and ductility of reinforced
masonry walls under cyclic loading, which is relevant to the development of
masonry structures resistant to seismic action. Walls failing in shear behaved
similarly under cyclicalloading as under static loading, but those failing in flexure
showed reduced strength owing to the crushing and dislodgrnent of the material at
the lower corners caused by the reversing movements.
6.5 Prestressed brickwork

Work on prestressed brickwork up to the present has been very limited. However,
there would appear to be no theoretical difficulties in applying the principles of
prestressed concrete to this medium-the main difficulties are likely to arise from
practical problems, such as the accommodation of cables and anchorages .
Thomas 3 9 has given an account of three separate series of tests on prestressed
clay brick and hollow block beams. The first experiments described were carried
out on beams consisting of brick 'soldiers' having 7 mm high tensile steel wires
threaded through the lowermost of three perforations in the bricks and on a
larger beam of 228 x 456 mm cross-section. Both beams had a concrete block at
each end to provide for the cable anchorages, and when loaded cent rally both
failed due to the development of tensile stresses towards the ends of the beams.
The magnitudes of the principal tensile stresses at failure in the two tests were
0 .83 N/mm 2 and 0.62 Njrnrn" . A more extensive series of tests on 'solider' brick
beams led to the conclusion that such beams behave in the same way as corres-
ponding concrete beams, and that there seem to be no technical reasons why post-
tensioned ceramic units should n öt be used for beams and floor slabs. However ,
an amount of development work is required to evolve suitab1e units , jointing
materials and anchorage blocks as well as to examine long-term behaviour. The
final series of tests described by Thomas were on beams built from 216 x 100 mm
hollow clay blocks , jointed with an epoxy-china-clay mortar and prestressed with
two 3 mm wires, and the end anchor ages bore against steel plates fitted on the
ends of these beams. Three of these beams were tested and failures occurred along
the resin-mortar joints, which it was thought may have been too thick for optimum
strength. As a result of these tests a suitable cross-section for an extruded clay
block for flooring beams was evolved.
A few tests on cavity wall panels prestressed with 12 mm rods have been
reported by Curtin et al. 4o in connection with the use of this type of element in
certain system-built schools. The prestressing was carried out by tightening a nut
on each rod against a steel plate bridging the two leaves of the wall. The tests were
sufficient to establish an approximate method of calculation and to give assurance
that the form of construction would be acceptable in terms of safety and
practicability on site. '
Prestressed brickwork shear walls have been tested by Hinkley'" and , as may
be expected , showed increased strength as compared to corresponding plain
brickwork walls.
One of the outstanding problems with prestressed brickwork is the evaluation
of the necessary allowance for loss of prestress owing to creep. The designers of a
particular prestressed flooring system''? use the following formula to calculate the
final prestress in the tendons used in their flooring planks
ope -- 093
• 0i Ap
A p +aeA w
where ope is the final stress in the plank
0i is the initial stress in the prestressing tendon
A w is the area of the tendon
A p is the area of the plank
a e is the modular ratio
This formula assurnes 4 per cent relaxation of stress in the wire and 3 per cent
creep and shrinkage in the plank . The clay tiles are assumed to be stable and not
to contribute to the loss of prestress. In th is case the area of grout in the cross-
section was stated to be approximately 25 per cent, so that the overallloss of
prestress would be less than in a concrete section of the same area.
Consideration of Lenczner's work would suggest that the loss in a typical brick
masonry cross-section, as opposed to the type of clay tile element referred to
above, might be somewhat higher and that a furt her 3 to 5 per cent creep
allowance might be appropriate. However, no direct experimental data appears
to exist at this time on this topic.
References
1. A. Dinnie and R. Beard, 'Reinforced Brickwork Silos for Grain Storage',
Proc. Br. ceram. Soc., 17 (1970) 121-36.
2. W. L. Dickey, 'Seismic Resistant Reinforced Masonry Naval Barracks', in

Designing, Engineering and Constrncting with Masonry Products , ed. F. B.
Johnson (Gulf, Houston, Tex., 1969) pp . 395-9.
3. G. C. Hanson, 'Park Mayfair East', in Designing, Engineeringand Construct-
ing with Masonry Products, ed. F. B. Johnson (Gulf, Houston, Tex ., 1969)
pp . 400-5 .
4. C. R. Abel and M. R. Cochran, 'A Reinforced Brick Masonry Retaining
Wall with Reinforcement in Pockets', Proceedings of the Second Inter-
national Brick Masonry Conference (Stoke-on-Trent) 1971, ed. H. W. H.
West and K. H. Speed (British Ceramic Research Association, Stoke-on-
Trent, 1971) pp. 295-8.
5. J . C. Grogan, 'Miscellaneous Reinforced Brick Masonry Structures', Pro-
ceedings ofthe Second International Brick Masonry Conference (Stoke-on-
Trent) 1971, ed. H. W. H. West and K. H. Speed (British Ceramic Research
Association, Stoke-on-Trent, 1971) pp. 327-30.
6. J . W. Sallada and D. A. Wakefield, 'Park Lane Towers Complex, Denver,
Colorado', Proceedings of the Second International Brick Masonry Con-
ference (Stoke-on-Trent) 1971, ed. H. W. H. West and K. H. Speed (British
Ceramic Research Association, Stoke-on-Trent, 1971) pp. 343-52 .
7. C. A. Abel, 'Prefabricated Reinforced Brick Masonry Arch Bridge', Pro-
ceedings of the Third International Brick Masonry Conference (Essen)
1973 , ed. L. Foertig and K. Gobel (Bundesverband der Deutschen Ziegel-
industrie, Bonn, 1975) pp . 544-8 .
8. A. H. P. Maurenbrecher, 'A Reinforced Brickwork Retaining Wall', Proceed-
ings of the Fourth International Brick Masonry Conference (Brugge) 1976,
Paper 4 .c.5.
9. D. Foster, 'Design and Construction of a Prestressed Brickwork Water Tank',
on-Trent) 1971, ed. H. W. H. West and K. H. Speed (British Ceramic Research
Association , Stoke-on-Trent, 1971) pp. 287-94.
10. W. C. Curtin, S. Adams and M. Sloan, 'The Use of Post-Tensioned Brickwork
in the S.C.D. System', Proc. Br. ceram. Soc., 24 (1975) 233-45 .
11. A. Brebner, 'Notes on Reinforced Brickwork Tests, Theory and Actual
Construction of Reinforced Brickwork in India', Technical Paper No. 38
(Government Public Works Dept., London, 1923).
12. Reports of Committee on Reinforced Brick Masonry of the National Brick
Manufacturers Research Foundation:
No. 3 'Method of Testing Reinforced Brick Masonry Demonstration Struc-
true',J. Am. ceram. Soc., 15 (1932) 300-5 .
No.4 'Results of Tests on Ten Reinforced Brick Masonry Demonstration
Structures with Summary Covering Tests on Thirteen Structures',J. Am.
ceram. Soc., 15 (1932) 273-99 .
No. 6 'Preliminary Outline of Program of Comprehensive Research in Rein-
forced Brick Masonry',J. Am. ceram. Soc., 15 (1932) 305-13 .
No. 7 'Tentative Specifications for Reinforced Brick Masonry', J. Am. ceram.
Soc., 15 (1932) 313-20.
No. 8 'Results of Tests on Seven Demonstration Structures with Summary
Covering Tests on Twenty Structures' , Bull. Am. ceram. Soc., 12 (1933)
166-97.
13. H. Filippi, Brick Engineering-Reinforced Brick Masonry Principles of
Design and Construction , vol. 3 (Brick Manufacturers Association of America,
Cleveland, Ohio, 1933).
14. L. B. Lent, 'Reinforced Brick Masonry',J. Am. ceram. Soc., 14 (1931)
469-81.
15. J . W. Whittemore and P. S. Dear, 'An Investigation ofthe Performance
Characteristics of Reinforced Brick Masonry Slabs', Bull. Va polytch. Inst.
Engng Exp. Stn, 9 (1932).
16. J . W. Whittemore and P. S. Dear, 'A Comparison of the Performance
Characteristics of Reinforced Brick Masonry Slabs and Reinforced Concrete
Slabs',Bull. Va polytech. Inst. Engng Exp. Stn, 15 (1933).
17. M. O. Withey, 'Tests of Brick Masonry Beams',Proc. Am. Soc. Test Mater.,
33 (1933) 651-65 .
18. D. E. Parsons, A. H. Strong and J. W. McBurney, 'Shear Tests of Reinforced
Brick Masonry Beams',J. Res. natn. Bur. Stand., 9 (1932) 749-68 .
19. F. G. Thomas and L. G. Sims, 'The Strength of some Reinforced Brick
Masonry Beams in Bending and in Shear', Stru ct. Engng, 17 (1939) 330-49.
20. C. W. Hamann and L. W. Burridge, 'Reinforced Brickwork' , Struct. Engng,
17 (1939) 198-250.
21 . G. T. Suter and A. W. Hendry , 'Shear Strength of Reinforced Brickwork
Beams', Struct. Engng, 53 (1975) 249-53.
22. G. T. Suter and A. W. Hendry, 'Limit State Shear Design of Reinforced
Brickwork Beams',Proc. Br. ceram. Soc., 24 (1975) 191-6.
23 . G. T. Suter and H. Keller, 'Shear Strength of Grouted Reinforced Masonry
Beams', Proceedings of the Fourth International Brick Masonry Conference
(Brugge) 1976 , Paper 4.c.2.
24 . J . C. Scrivener, 'Reinforced Masonry in a Seismic Area-Research and Con-
struction Development in New Zealand' , Proceedings of the First Canadian
Masonry Symposium (Calgary) 1976, pp. 371-82.
25. A. Cadjert and A. Losberg, ' Laterally Loaded Light Expanded Clay Block
Masonry. The Effect of Reinforcement in Horizontal Joints' , Proceedings
of the Third International Brick Masonry Conference (Essen) 1973, ed. L.
Foertig and K. Gobel (Bundesverband der Deutschen Ziegelindustrie, Bonn,
1975) pp . 245-51 .
26. A. Cadjert and A. Losberg, 'Lateral Strength of Reinforced Brick Walls.
Design for Wind Loads', Proceedings of the Fourth International Brick
Masonry Conference (Brugge) 1976, Paper 4.c.4.
27. - 'Reinforced Brickwork : Vertical Cantilevers 1', S.C.P.lO (Structural Clay
Products Ltd, Hertford , 1976) .
28. - 'Reinforced Brickwork: Vertical Cantilevers 2' , S.C.P. 11 (Structural Clay
Products Ltd, Hertford, 1976) .
29 . - 'Reinforced Brickwork: Pocket Type Retaining Wall', S.C.P.l3 (Structural
Clay Products Ltd, Hertford, 1977) .
30 . B. A. Haseltine and J. N. Tutt, Brick Retaining Walls (Brick Development

Association , London , 1977).
31. F. K. Kong and R. H. Evans, Reinforced and Prestressed Concrete (Nelson,
London, 1975) pp . 54-63 .
32. - 'Design Guide for Reinforced and Prestressed Clay Brickwork', Special
Publication No. 91 (British Ceramic Research Association.Stoke-on-Trent,
1977).
33. K. Wyatt, D. I..enczner and J. Salahuddin, 'The Analysis of Creep Data in
Brickwork' ,Proc. Br. ceram. Soc ., 24 (1975) 11-20.
34. B. P. Sinha, 'Reinforced Brickwork : Grouted Cavity Shear Tests', S.C.P.16
(Structural Clay ProductsLtd, Hertford, 1978).
35. B. P. Sinha, ' Long Term Tests on Reinforced Clay Brick Masonry Grouted
Cavity Vertical Cantilever Walls' , S.C.P.l4 (Structural Clay Products Ltd,
Hertford,1979).
36. I. Lyse, 'Tests of Reinforced Brick Masonry Columns',J. Am. ceram. Soc.,
16 (1933) 584-97 .
37 . M. O. Withey, 'Tests of Reinforced Brick Masonry Columns', Proc. Am. Soc.
Test. Mater. , 34 (1934) 387-405.
38 . D. E. Anderson and E. S. Hoffman, 'Design of Brick Masonry Columns' , in
Designing, Engineering and Constrncting with Masonry Products , ed. F. B.
Johnson (Gulf, Houston, Tex. , 1969) pp. 94-100.
39. K. Thomas, 'Current Post-tensioned and Prestressed Brickwork and Ceramies
in Great Britain' , in Designing, Engineering and Constructing with Masonry
Products, ed. F. B. Johnson (Gulf, Houston, Tex ., 1969) pp. 285-301.
40. W_ C. Curtin, S. Adams and M. Sloan , 'The Use of Post-Tensioned Brickwork
in the SCD System', Proc. Br. ceram. Soc., 24 (1975) 235-45.
41. A. T. Hinkley , 'Tests on One-Storey Prestressed Brickwork Shear Walls',
N.Z. Engng; 21 (1966) 245.
42 . R. J. Wass and D. J . Turner, 'A Prestressed Clay Masonry Floor', in Design-
ing, Engineering and Constructing with Masonry Products , ed . F. B. Johnson
(Gulf, Houston, Tex ., 1969) pp . 200- 9 .
7 THE RESISTANCE OF BRICKWORK
STRUCTURES TO ACCIDENTAL DAMAGE
7.1 Abnonnalloading incidents

It is not generally possible to carry out anormal structural analysis relating to
abnormalloading arising from such causes as gas explosion, vehicle impact or bomb
blast since the location and magnitude of the load effect resulting from such
incidents are unpredictable, and in some cases it would in any case be impractic-
able to design elements to resist the forces involved. It is necessary, however, to
design buildings so that any damage resulting from such an incident remains
localised . Concern about this problem arose as a result of a gas explosion on the
eighteenth floor of a block of flats in London in 1968. This building , called
Ronan Point, was of large panel precast concrete construction. The explosion
blew out one of the external wall panels and , as a result, part of the five floors
immediately above the explosion collapsed on to the eighteenth floor . This in
turn , was unable to sustain the debris load and collapsed on to the floor below and
so on almost to ground level. The tribunal' which investigated the accident gave
this sequence of events the name 'progressive collapse'.
Prior to the Ronan Point accident , no specific consideration was given to
accidental forces and associated liability to collapse in the design of masonry bear-
ing wall structures, although the problem had been indire ct1y recognised in relation
to large panel concrete structures.i -? However, Ronan Point drew attention to the
problem in a forcible manner , and consideration had to be given to the possibility
of damage to brickwork structures on ascale disproportionate to the magnitude
of any accidental forces likely to be experienced.
In the decade following Ronan Point considerable attention was devoted to the
problem through accident statistics," field stud ies," experiments on structurest :7,8
and structural elements." :' 0 ,11 On the basis of this work , various design strategies
for minimising the effects of accidental damage have been put forward, and these
have been categorised as follows
(1) Event control
(2) Direct design
(3) Indirect design
By event control is meant the reduction of the risk of structural damage by
eliminating exposure to a particular hazard (for example , by excluding potentially
165
explosive materials from buildings), providing specific protection (for example,

bollards to prevent vehicle impact) or by incorporating features in a building which
will limit the severity of a hazard (for example, venting to limit the pressure
developed by a gas explosion). While ordinary prudence may encourage the adop-
tion of some of these precautions, they are generally outside the control of
structural designers. The alternatives, however, are matters of structural design:
the direct approach is to design a structure to resist certain specified forces or
amounts of damage without collapse, while the indirect approach is to achieve
this objective by specifying minimum levels of strength and continuity.
7.2 Direct design for accidental damage

As previously stated it is not possible to specify the loading associated with
structural accidents with any certainty, so that in applying the direct design
approach the forces to be resisted are of an idealised nature and arbitrary magni-
tude. Thus the Building Regulations in the United Kingdom require that any
element on which the stability of a structure is to depend following an accident
must be able to withstand apressure of 34 kN/m 2 from any direction. Various
other abnormal load criteria have been proposed, and have been compared by
Leyendecker and Ellingwood ."
Resistance of brick masonry walls to accidental damage is related to their
lateral strength, and may be calculated using the methods discussed in chapter 5.
In the second direct design procedure, sometimes referred to as the 'alternate
path method', specified elements, or parts of elements, are assumed to have failed
and the stability of the remaining structure is investigated . It is not generally
possible to assess the amount of damage that will result from a particular hazard
and, as in the case of pressures arising from accidental causes, some kind of
idealisation has to be adopted . The British Code of Practice for the Structural Use
of Masonry, BS 5628 :1978, requires that the effect of removing vertical and
horizontal elements one at a time should be considered. It has then to be demon-
strated that the 'damaged' structure has adequate residual stability and that
collapse of any significant portion of the structure is unlikely to occur.
Examination of a wide variety of existing designs by Morton et al. 12 has identi-
fied three particular situations which may be critical in relation to accidental
damage. These are
Case A Where there is an outside wall without returns, or only one internal
return
Case B Where there is an internal wall without returns
Case C Where the removal of a section of a wall imposes high local bearing
stresses on a return wall or walls.
In case A (figures 7.1a and b) the removal of a panel of brickwork willleave the
remaining sections of the wall supported by the floor slabs above . In case B
(figures 7.1c and d) the walls above the damaged wall will have to be carried by
the floor slabs supported around their perimeter by the remaining walls. Case C
(figure 7.1 e) arises where the unsupported brickwork spans or cantilevers across
RESISTANCE OF BRICKWORK STRUCTURES TO ACCID ENTAL DAMAGE 167
A I
GASE A
>
(a)
I. I I.
~ ~ ~ ~ ~ ~ ~
= == ~
=. . === = == ==
== = = ~
= == ~
=== = :: :
.. .. ,. = .
~
= ~ ~ =- ~
., 0= :~ :: ~ ::' :;;. "' -
.- =, = _.
::<
~ ~
= ==-, --
f-- W all panel
i rem oved
(b )
GASE B
U nsupp ort ed
I
~
inte rn a1w all
to be carried by
Il oor slabs
(c ) (d )
GASE G
Eigure 7.1 Situ ations requ iring spe ciai atte ntio n in relat ion s to accidental damage
the opening left by the removal of a wall panel, resulting in the application of a
concentrated load on the return wall. It may also be necessary to examine the
stressing of a brickwork cantilever, such as that shown in figure 7.1e, to ascertain
whether or not it can be assumed to act as a unified element; if not, it must be
considered under case A.
In cases where the support of the walls above an opening is dependent on a
remaining return wall, it is necessary to calculate the bearing stress in the brick-
work . The assumptions in the previous paragraph are based on normal methods of
structural mechanies. Thus in case A, it may be necessary to assess whether a
storey height length of outside wall can be supported by a floor slab when the
wall directly below is assumed to have been removed. In a typical case the situa-
tion may be as represented as in figure 7.2, and using the notation shown in this
diagram, yield line analysis may be applied as follows
E=P(ßL
2
+!!!-2) _P(2L - C)C (7.1)
6 3 2L
where E is the work done by the external loads per unit displacement of the
mechanism
p is the weight of the floor slab plus the uniformly distributed imposed
load
P is the weight of wall per unit length .
The internal work done in the yield lines is given by
D=m(4a+ -~ +4ila+-~-) (7.2)

o:-ß o:-ß
Equating D and E, and substituting known values of 0:, u; i 1 and i2 gives an expres-
sion for m, the minimum designed moment for the slab, in terms of ß. The value of
ß corresponding to the minimum value of m can then be found by putting
dM/dß =O. Finally m is calculatedand compared to the actual moment of resist-
ance of the slab. Other cases can be dealt with in a similar manner.
In case C, the bearing stress on areturn has to be considered, as indicated in
figure 7.3. The force transmitted to the return wall resulting from the dead and
superimposed loads on the structure above it is calculated by statics . The bearing
stress in the brickwork is then determined, allowing for spread of the load through
m
-11 m ....
t:l
~
~ L
Figure 7.2 Yield line analysis of floor slab carrying wall load on edge
RESISTANCE OF BRICKWORK STRUCTURES TO ACCIDENTAL DAMAGE 169
Concentrated
load on
1~~~~~lr
th is area of
return wall
removed
Wall'
I
,"
Figure 7.3 Wall bearing on areturn
the floor slab and, in a design situation, compared to the maximum bearing stress
for the material under accidental damage conditions.
Consideration of the removal of floor slab elements gives rise to rather intract-
able conceptual problems, for example, it is difficult to see how a reinforced
concrete floor slab in a brickwork structure could be removed without there
being serious damage to the brickwork. It is, therefore, more practical to design
floors whose removal does not have to be considered, either by direct design to
meet the hypothetical accidental damage loads, or by indirect design by incorpor-
ating a sufficient degree of continuity.
7.3 Indirect design for accidental damage

The indirect approach4 , 13 , 14 to the control of accidental damage does not depend
on the representation of the forces likely to be encountered, or on the assessment
of the effect of limited damage. Instead, the idea is to design the building on an
empirical basis to possess certain minimum levels of strength, continuity and
ductility as will enable it to survive accidental forces in an acceptable manner. The
terms 'robustness' and 'st ructural integrity' are sometimes applied in this context,
and these were satisfactorily achieved in traditional forms and methods of con-
struction. In terms of modern brick masonry structures, the first step in securing
general stability is the adoption of a satisfactory wall layout, as discussed in
chapter 1. Following this , comes the need to ensure continuity of the floor
system. In this respect, in situ, two-way spanning reinforced concrete slabs, con-
tinuous over internal walls, afford the highest degree of protection, Least
satisfactory would be simply supported precast planks without lateral.ties. The
British Code of Practice BS 5628: 1978 sets out requirements for ties in floors to
meet all reasonable continuity criteria and, in terms of this Code, avoids the
requirement to consider removal of such elements in a direct design approach.
The corresponding provision of vertical ties is a more doubtful procedure.
Firstly, it is in practice difficult to build in ties in unreinforced masonry walls
and , secondly, their effect is uncertain. Indeed, in an experimental study of a
large panelloadbearing wall structure of model-scale, Beak"! showed that con-

tinuous vertical ti es could promote progressive collapse by pulling out wall panels
at levels above and below that at which an explosion took place. Thus , if vertical
ties are to be used at all they should not be continuous from one storey to the
next.
Indirect design considerations further require '" that walls should be adequately
stiffened by returns and piers, as appropriate, and that roof structures as well as
floors should be constructed in such a way as to promote the spreading of lateral
loads amongst the walls. These elements must also be adequately tied, or strapped,
to the walls to provide lateral support.
The indirect approach is in fact applicable to all building design and is likely to
be sufficient for low-rise buildings. Tall buildings on the other hand may require
more detailed consideration on a semi-quantitative basis by direct design methods
to ensure that no weaknesses exist in their resistance to accidental forces .
7.4 Experimental studies of accidental damage

A great deal of information about the behaviour of brick masonry structures can
be obtained from a systematic study of the buildings damaged in gas explosions
and other accidents.! However, it is only possible to infer the forces involved and
the material properties in such cases; other relevant factors may be difficult to
ascertain. The results of controlled experiments are therefore of value in verifying
the basis for assessment of new designs, the effectiveness of structural precautions
and in affording information on the response of building structures to accidental
damage .
A number of tests 7 were carried out at Edinburgh University on the full-scale
experimental structure shown in figure 7.4 with the object of confirming that a
cross-wall structure of this type could survive the removal of a major bearing wall.
The walls were of 50 mm precast slabs containing the main reinforcement with
75 mm of in situ concrete topping, giving an overall slab depth of 125 mm. In the
tests, the outer cross-wall (A in figure 7.5) was first removed by lateralloading
applied by hydraulic jacks . This wall was then rebuilt and a second test carried
out by removing one of the internal panels (B in figure 7.5) in the same way .
Deflections were measured, and these showed that the maxirnum deflection of
the floor slab after the first test was 4 mrn, and .after the second test 5.4 mm . No
damage was observed anywhere in the structure except immediately above the wall
that had been removed , where the joint between the first floor slab and the wall
was broken and the wall appeared to be partly hanging from the second floor slab
and partly supported by the first floor slab. The structure appeared to be stable
after the removal of a main bearing wall, and this was confirmed by calculation
which indicated that there would be a load factor of 1.94 on the dead load plus
1.9 kN/m 2 superimposed load after the removal of the centre cross-wall, and
2.44 after the removal of the end cross-wall. It should be emphasised that this
structure was not designed with progressive collapse in mind and represented
what was considered to be sound construction in accordance with the current
Code of Practice. Naturally, no general conc1usions could be based on a single
Figure 7.4 Wall removal test on five-storey brickwork structure simulating

accidental darnage
case, but the inference for future designs was that brick masonry structures could
without difficulty be made resistant to progressive collapse as a result of acciden-
tal damage.
The experiment described was intended as a demonstration of feasibility. It
was on the lines of the direct design approach described in section 7.2, in which
6638 mm
LI~ LI~
oll
o
oll
N
Cl>
11- 11
N
N
'"
1811~
oll
o
oll
N
~ 814
laI walliavout before (bJ after removal (e) after removal
test of wall A ofwall8
Figure 7.S Wall removal tests on brickwork structure
stability following the removal of a section of bearing wall was considered but,
of course, it did not afford any information about the behaviour of a masonry
building subjected to accidental forces and in particular to a gas explosion. An
extensive series of tests was undertaken by the British Ceramic Research Associa-
tion and the Brick Development Association in order to explore this problem and
to obtain information which would have a bearing on structural design. The
investigation has been described in detail in references 6 and 8, and included gas
explosion experiments in a specially designed building representing the top three
stories of a tall cross-wall structure . These tests were supported by aseries of tests
in a bunker, the front of which was closed by brick walls or cladding panels of
various kinds. Details of the experimental building are shown in figure 7.6, from
which it will be seen that it consisted of three loadbearing cross-walls at approxi-
mately 3.8 m centres with 127 mm reinforced concrete floor slabs. The outer walls
were of 280 mm brick cavity construction, with various window and cladding
arrangements on the non-loadbearing walls. Internally the space was subdivided by
105 mm brick walls with communicating doors between two pairs of rooms.
Gas was introduced into one or more of the rooms for each test and ignited to
sirnulate a domestic gas explosion. Town gas was used in most experiments since
it has a high proportion of hydrogen, and would be thus expected to produce
higher explosion pressures than natural gas, which is mostly methane. In some
cases a gas-air mixture was contained in a polythene balloon before ignition, in
others the gas was mixed with air in the room or introduced in such a way as to
produce a layered concentration from floor to ceiling. It was expected that an
explosion originating in one room and spreading to an adjoining gas-filled room
would result in higher pressures in the latter, as a result of the turbulent mixing of
burning gas with the unignited gas in the second stage of the explosion. A number
of tests were therefore arranged in which this effect took place.
A large number of explosion tests were carried out, in most of which gas
pressures of less than 14 kN/m 2 were generated by quantities of gas, which had
they been fully confined would have resulted in pressures of up to 119 kN/m 2 •
9176
i t
Room 1 Room 3
Room 2 Room 4
Ground floor plan
....... '----- .
c;::::::::;J c;:::=:::;J
r-----"1
t:;==;:l
"1'<t0
"
t;:::=::;J c;:::::::;:J
R R
Cl)
I I I I
Section AA
Figure 7.6 Details of B.C.R.A. explcsion test structure

These 10w pressures were, of course, observed because of the venting effect of the
glazing or cladding, which blew out and thus limited the maximum pressures. In
these tests no damage to the brickwork was observed. In a further experiment,
however, in which two rooms (1 and 2 in figure 7.6) were filled with gas, an
explosion initiated in room 1 resulted in a maximum pressure of 22.7 kN/m 2 in
room 2 and the outer loadbearing cavity wall in room 2 was damaged but did not
collapse. In fact , considerable difficulty was subsequently experienced in
demolishing it prior to reconstruction.
As this was a significant result the experiment was repeated after rebuilding
the damaged wall. In this repeat case the average pressure recorded in room 2 was
slightly lower at 18.47 kN/m 2 , and only minor damage was caused to the brick-
work. A similar test was subsequently carried out in rooms 3 and 4, which had
smaller venting areas than rooms 1 and 2. The gas was ignited in room 3 and
maximum measured pressures of 16.9 kN/m 2 and 15.84 kN/m 2 were produ ced
in rooms 3 and 4, respectively . It was considered that local pressures may have
been as high as 24 kN/m 2 in room 4 and considerable damage to the brickwork
resulted . The outer cross-wall showed a characteristic roof-shaped fracture pattern
similar to the yield line pattern in a laterally loaded concrete slab but, in spite of
this, the building did not collapse and the damaged brickwork was subsequently
demolished only with difficulty.
From these, and some subsidiary experhnents.!" it may be concluded that it is
possible under rather exceptional circumstances for apressure of the order of
34 kN/m 2 to be generated in a domestic gas explosion. In the vast majority of
such incidents, however, the venting effect of doors, windows and other circum-
stances will preclude such a possibility, and pressures are unlikely to exceed
24 kN/m 2 • The provision of the Building Regulations requiring that elements
which are necessary for the continued stability of a building after it has been
subjected to accidental forces is thus seen to be weH on the safe side, at least as
far as gas explosions in residenti al buildings are concerned. Both the University
of Edinburgh and the B.C.R.A. tests confirm that multi-storey brickwork
structures, as normally designed in the United Kingdom, can withstand severe
structural damage without it resulting in progressive collapse.
References
1. - Report of the Inquiry into the Collapse ofFlats at Ronan Point , Canning
Town (H.M.S.O., London, 1968).
2. - 'International Recommendations for the Design and Construction of Large
Panel Structures', Bulletin No. 60 (C.E.B., Paris, 1967).
3. J. Despeyroux, 'L'Effondremont de L'immeuble de Ronan Point et ses
Consequences en Matiere du Codification, Annls Inst. tech. Biitim., No. 263
(1969) 1800-3.
4. E. V. Leyendecker and B. R. Ellingwood, 'Design Methods for Reducing the
Risk of Progressive CoHapse in Buildings', NBS Building Science Series No .
98 (Washington, 1977).
5. R. J. Mainstone, 'Accidental Explosions and Impact : Some Lessons from
Recent Incidents', Proceedings 0/ the Symposium on Stability 0/ Low-rise
Buildings 0/ Hybrid Construction (Institute of Structura1 Engineers, London,
1978) pp. 13-23 .
6 . N. F. Astbury , H. W. H. West, H. R. Hodgkinson , P. A. Cubbage and R.
C1are, 'Gas Explosions in Load-bearing Brickwork Structures', Special
Publication No. 68 (British Ceramic Research Association, Stoke-on -Trent ,
1970).
7. B. P. Sinha and A. W. Hendry, 'The Stability of a Five-storey Brickwork
Cross-wall Structure Following the Removal of a Section of a Main Load-
bearing Wall', Struct. Engr, 49 (1971) 467-74.
8. N. F. Astbury , H. W. H. West and H. R. Hodgkinson, 'Experimental Gas
Explosions : Report of Further Tests at Potters Marston', Special Publication
No. 74 (British Ceramic Research Association, Stoke-on-Trent, 1972).
9. A. W. Hendry , B. P. Sinha and A. H. P. Maurenbrecher, 'Full Scale Tests on
the Lateral Strength of Brick Cavity Walls with Precompression', Proc. Br.
ceram. Soc .• 21 (1974) 165-79 .
10. H. W. H. West, H. R. Hodgkinson and W. F. Webb, 'The Resistance of Brick
Walls to Lateral Loading',Proc. Br. ceram. Soc. , 21 (1974) 141-64.
11. J. Morton and A. W. Hendry, 'An Experimental Investigation of the Lateral
Strength of Brickwork Panels with Precompression under Dynamic and
Static Loading', Proceedings 0/ th e Third International Brick Masonry
Conference (Essen) 1973, ed. L. Foertig and K. Gobe1 (Bundesverband der
Deutschen Ziegelindustrie, Bonn, 1975) pp . 362-9.
12. J. Morton, S. R. Davies and A. W. Hendry, 'The Stability of Load-bearing
Brickwork Structures Following Accidenta1 Damage to a Major Bearing
Wall or Pier', Proceedings 0/ the Second International Brick Masonry Con-
ference (Stoke-on-Trent) 1971, ed. H. W. H. West and K. H. Speed (British
Ceramic Research Association, Stoke-on-Trent, 1971) pp . 276-81.
13. R. J. A. Sutherland, 'Principles for Ensuring Stability', Proceedings 0/ the
Symposium on Stability 0/ Low-rise Buildings 0/ Hybrid Construction
(Institute of Structura1 Engineers, London, 1978) pp. 28-33.
14. J. F. A. Moore, 'The Stability of Low-rise Masonry Construction', Proceed-
ings 0/ the Symposium on Stability 0/ Low-rise Buildings 0/ Hybrid
Construction (Institute of Structural Engineers, London , 1978) pp. 38-46 .
15. R. Beak, 'Explosion Tests on a \4-scale Model of an 18-storey Large Panel
Structure', BRE Note N149/76 (Building Research Establishment, Watford,
Herts., 1976).
16. D. J. Rasbash, K. N. Palmer, Z. W. Ragowski and S. Ames, Gas Explosions
in Multiple Compartments (Directorate General of Development , Dept. of
th e Environment , London , 1971).
8 BRICK MASONRY WALLS
IN COMPOSITE ACTION
8.1 Composite wall-beam elements
8.1.1 Structural action ofwall-beams

It has long been recognised that structural interaction takes place between a
masonry wall and a supporting steel or concrete beam . In simplest terms this has
been represented by assuming that the beam supports only part of the brickwork
represented by a triangular load intensity diagram with zero ordinates at the
»:"" ......
/ '\
I \ (a)
\
/ \
/ \
Vertical stress
t{
-, «::;
d istribution ~~
I
I \ Shea r st ress I I
, \. {distribution -. r ~
~~' Ib)
Figure 8.1 Structural action in composite wall beam. (a) Arehing forces in wall;
(b) vertical and shear forces in beam
176
BRICK MASONRY WALLS IN COMPOSITE ACTION 177
supports and maximum loading at midspan. The loading from the remainder of
the brickwork was assumed to be transmitted to the support points by arehing
action as indicated in figure 8.la. While this is essentially a correct reflection of
the structural behaviour of the system as far as it goes, it is of limited quantitative
value because it fails to give any indication of the concentrated compressive stresses
in the wall, which may be critical, or of the actual bending moments in the beam .
A number of theoretical and experimental studies of the problem have shown
quite clearly that the vertical and shear stresses at the wall-beam interface are
concentrated towards the supports, as indicated in figure 8.1 b. Both the shear and
vertical stress distributions in these areas can be approximately represented by a
triangular diagram, and the more flexible the beam, the more concentrated these
stresses are towards the supports. Although the shear force tends to counteract
the downward deflection of the beam, there is a tendency for this element to
deflect downwards away from the wall, with the possible development of a crack
between the top of the beam and the bottom of the wall. The shear force also
induces an axial tension in the beam , the magnitude of which varies across the
span.
Within the wall an arehing action is developed and the vertical stresses are
heavily concentrated towards the supports.
The type of behaviour described in the previous paragraphs has been found to
take place with walls having a ratio of height to length greater than about 0.6.
Below this ratio, the shear becomes greater than can be resisted as the wall-beam
interface and, aIthough composite action is still possible, the element must even-
tually be treated as a purely flexural rnember .
8.1.2 Theoretical solutions
Using essentially the conceptual model described above , and against a background
of theoretical and practical research,' Wood and Sirnms? put forward a simple
method for th e calculation of composite wall beams . Instead of the triangular
distribution of vertical compressive stress in the vicinity of the supports, a rectan-
gular stress block was assumed, extending a distance x into the span from each
end of the beam . Thus the ratio of the average compressive stress in the wall to
the maximum in the area of the reetangular stress blocks is C = L/2x . The bending
moment in the central section of the beam is then
WL Wx Wx
M=-- - = - - = (8.1)
k 2 2 4
from which C = k/8 and xll: = 4/k. Taking possible values of the bending moment
coefficient, k ; leads to the corresponding values for the extent of the stress block
and the stress concentration factor .
k x/L C
8 1/2 I No composite action

12 1/3 1.5
24 I/.6 3
48 1/12 6
100 1/25 12.5 Maximum allowable composite action
This indicates that for composite action to be possible , the average compressive
stress in the wall must be relatively smalI. On the basis of these values of C, it is
possible to derive values of k which are consistent with the design stresses pre-
scribed by any particular code of practice .
Thus in terms of BS 5628 : 1978, the design strength per unit area of a wall will
be ßfk /rm ' where ß is a reduction factor for slenderness, fk and rm respectively
the characteristic strength and partial safety factor for the material. If the average
stress in the wall is less than the design stress by a factor F , and if the design
strength may be increased by 50 per cent in the region of concentrated stress at
the support , then
CFß f~ » 1.50 fk_ (8.2)

rm rm
Since C = k/8, this equation leads to the bending moment factor for the beam
k» 12/Fß (8.3)
whe n M = WL/k . This will have the effect of Iimiting the interactive effect accord-
ing to the compression in the wall as a proportion of the design strength.
This simple analysis has been elaborated by Wood and Simms to allow for the
effect ofaxial tension in the beam on the assumption that a parabolic line of
thrust is developed in the wall. A limitation on the tensile stress in the reinforcing
steel was also suggested as a means of limiting its extension when acting as a tie to
the arehing forces in the brickwork.
This approximate solution is useful as a means of obtaining a quantitative feel
for the problem and, regulated by detailed analysis and experimentation, as a basis
for simple design rules, Theoretical solutions for composite wall-beam systems
have in fact been produced,' ,3-7 based on a variety of elastic analysis techniques.
Finite element methods have also been applied'"'!? and, together with experi-
mental studies, have resulted in a satisfactory understanding of the problem and
in proposals for suitable methods for design analysis.
Stafford Smith and Riddington1o,11 developed a finite element program for
the problem using a four-node reetangular element , with two degrees of freedom
at each node and with linearly varying displacement functions along the boundaries.
The program also allows for tensile cracking at the wall-beam interface . This work
confirmed that the total behaviour of the system remains unchanged when the
height to length ratio exceeds 0.7 . These investigators pointed out that the com-
posite wall-beam is the same type of problem as the beam on an elastic foundation
and the infilled frame in so far as the distribution of stresses between the elements
depends on their relative stiffness. Also, in these problems separation of the ele-
ments is possible, the lengths remaining in contact being a function of the relative
stiffness. It is therefore essential that this parameter should enter into the analysis.
Thus representing the length of contact between wall and beam as cd.
(8.4)
where EI is the flexural rigidity of the beam

Ew is the elastic modulus of the wall material in compression
t is the thickness of the wall
L is the length of the wall
From this
(8.5)
or
Q B
(8.6)
L K
where K = y(EwtL 3 lEI) and Bis a constant, found as a result of experimental
investigation to have an average value of unity, in which case
L
Q= - (8.7)
K
It will be seen from the above that the stiffer the beam relative to the wall, the
longer the length of contact x, and this in turn increases the bending moment in
the beam and reduces the wall stresses.
Finite element analysis permitted the definition of the vertical compressive and
horizontal shear stress distributions over the contact lengths, and these were found
to be approximated reasonably weIl by triangular diagrams. The investigation also
indicated that the peak compressive stress in thc wall could be represented by
(8.8)
where Ww is the total verticalloading at the wall-beam interface.

The effect of the axial stiffness of the beam was also considered, and it was
found that an increase in this property increased the tie force, thereby reducing
the bending moment in thc beam and increasing the peak wall stresses.
The results of Stafford Smith and Riddington 's study covering a wide range of
wall-beam combinations are summariscd in figure 8.2. It was found that conserva-
tive estimates of the stresses in the wall and in the beam could be calculated from
the following formulae
Tie force T
-
Wall load Ww 0.5
and Bending moment
Beam moment 0.4
Uniformly dist- Tie force
ributed moment 0.3
-M-
WwL/8
0.2
0 .1
0 5 10 15
Relative stiffness parameter K
Figure 8.2 Maximum bending moments and tie force in beam (Stafford Smith and
Riddington)
W
Maximum stress in wall 1.63 ~ (EwtL 3 /El)° ·28 (8.9)
Lt
W L
Maximum ben ding moment in beam -~--- (8.10)
4(EwtL 3/El)I/3
Maximum tie force in beam Ww/3.4 (8.11)
The effect of extending the beam into the surrounding brickwork was also examin-
ed , and this led to the conclusion that the stresses in both elements would be
reduced in this case, although negative bending moments could be induced in the
beam near its supports. Design procedures have been derived 12 from this investiga-
tion applicable to walls resting on steel and concrete beams .
Extensive finite element studies have also been carried out by Ahmed.P on the
basis of which an approximate solution for the composite wall-beam problem has
been derived by Davies and Ahmed.!" In this solution a relative stiffness parameter
Rf =./ (E~H3) (8.12)
is introduced. This is similar to the corresponding parameter K of Stafford Smith

and Riddington, but has been preferred by Ahmed on the grounds that it directly
represents the ratio of the wall to the beam stiffnesses. The ratio C of the maxi-
mum to the average compressive stress in the brickwork, as derived for different
H/L ratios by finite element calculations, was then plotted with the result shown
in figure 8.3 . This indicates a linear correlation, giving
C= 1 + ßR f (8.13)
BRICK MASONRY WALLS IN COMPOSITE ACTIO N 181
C
18
H/L > 1
16
14
12
10
8 a m =a.v ( l +ßRI
o 2 4 6 8 10 12 14
Figure 8.3 Vert ical stress con centration factor plo tted against pararneter Rf
(Davies and Ahm ed)
0 .14
Cl ß }'
0 . 12
0 .5 2 0.1
0 .4 1.6 0 .08
ß
0 .3 1.2 0 .06
Cl
y
0.2 0 .8 0 .04
0.1 0 .4 0 .02
0 .2 0 .4 0 .6 0 .8 1.0 1.2 1.4 1.6

H/L
Figure 8.4 Param eters a, ß and 'Y for co rnposit e wall beam in Davies and Ahmed 's
soluti on
The coefficient ß, derived from figure 8.3 , can be plotted against H/L as shown in
figure 8 .4. The maximum vertical stress in the wall is then
(8.14)
The distribution of the vertical stress at the interface depends on the parameter R r ,
as shown in figure 8.5.
Figure 8.5 Yertical stress distribution at wall-beam interface
A similar procedure is used to calculate the beam axial force , and in this case,
an axial stiffness ratio parameter
(8.15)
is used, where A is the area of the beam. From finite element results the ratio
T/Ww is plotted against Ra ' as shown in figure 8.6 , resulting in a relationship of
the type
T
(8.16)
Ww
The coefficients a and 'Y are plotted against H/L in figure 8.4 .
Assuming a triangular distribution of the vertical stress at the supports, of
length Iv, it will be seen that
(8.17)
or from equation 8.14
L
Iv =
0.5
T
T= Ww(a-i'K )
0.4
ui;e;
R. = - -
A bEb
0 .3
0 .2
0 .1
o 2 3 R.
Figure 8.6 Ratio of tie force to total verticalload related to parameter Ra
The finite element analyses showed that the shear stress acts over a length two to
three times that of the vertical stress , that is
2L
Is = -- - - (8.18)
CI + ßRf)
Again assuming a triangular stress distribution
-} T mist =T (8.19)
and , since T= Ww (0: - rRa) ' we have

Ww (0: - rRa) CI + ßRf )
T
m
= Lt
(8 .20)
The bending moment at any section in the beam results from the vertical load -
ing and the horizontal shear at the interface , which is eccentric to the axis. Thus ,
referring to figure 8.7, the bending moment over the central region of the beam
due to the verticalload is given by
(8.21)
Shape of stress dist ribution curve

/' related to value of R,
I,
Figure 8.7 Verticalloading on beam
where rl; is the distance of the support reaction to the centroid of the stress
diagram. Also
(8.22)
where A is a coefficient which depends on the shape of the stress diagram. From
equations 8.14, 8.20 , 8.21 and 8.22
(8 .23)
Davies and Ahmed have found that the axial tension varies along the length of the
beam , approximately as shown in figure 8.8 , so that the force T x at a distance x
from the support is
2xT
t;» L (8.24)
and the bending moment due to this force is, on substituting for T from equation
8.16
Figure 8.8 Assumed variation of tie force across span of beam (Davies and
Ahmed)
Wwdx .
M H = - - (a - 'YRa) (8.25)
L
The total bending moment at x = Iv, assumed to approximatc to the maximum
value on the span, is then
WwLr - 2Wwd (a - 'YRa)
Mm = - - -- -- - (8.26)
4(1 + ßRr)'A
and at midspan
WwLr - 2Wwd (a - 'YRa) (1 + ßRr)
Me = -- - - - - - - - - - - -- (8.27)
4(1 + ßRr ) 'A
The values of rand 'A depend on the shape of the stress distribution diagrams,
which in turn depend on R r . Appropriate values and the corresponding bending
moment formulae are shown in table 8.1.
Table 8.1
Forrnulae for beam bending moments''f
Case I
R r <' 5 stiff beam
r= 0.2 and 'A = 0 .25
WL - 10 Wd (a - -yR )
M m = - - - --- ---_._ -- -- -- -- - a-
5(1 + ßRr)
WL - 2.5 Wd (o - 'YRa) (1 + ßRr)
and Me =
5(1 + ßRr)
Case 2
5 <R r<7 flexible beam
r = 0.25 'A = 0.33
WL - 8 Wd (a - 'YRa)
Mm =
5.33 (1 + ßRr)
WL - 2.66 Wd (a - 'YRa) (I + ßRr)
and Me =
5.33 (1 + ßRr)
Case 3 Rr ~ 7 very flexible beam

r = 0.33 and 'A = 0.5
WL - 6 Wd (a - -yRa )
Mm =
6 (1 + ßRr)
WL - 3 Wd (a - 'YRaHI + ßRr)
and Me =
6 (1 +ßRr)
,
186 STRUcrURAL BRICKWORK
Span
o 0 .2 0.3 0.4 0 .6 0 .8 L
0 .0 1 ,,~
Bend ing ~~\

\"\.
momen t \ •.~. ....::-0.z:.. :': :::..:.: '-
..
0 .02 '(.. ~ , . •.,-... . • - - . - . - ....... '--=:-.
-, '. -- .--
.. \:~~.'::a~ ~
:.: ~.:._~::.- --- - - - - -- - --...........:.:":'...
':' ~
- _.':-":~'-
0 .03
0 .04
0.05 W all 5 _._.- So lid

WL W all 6 _ . _ .- Central w indow
W all 7 Central door
W al18 Door op ening at 0 .25L trom r.h. suppor t
W al19 Central w indow
Figure 8.9 Bending moments in wall-beams with op enings (Davies and Ahmed )
Davies and Ahmed 15 also investigated a range of wall-beam structures in which

the walls were perforated by window or door openings. This work showed that
cent rally placed openings have only a small effect on the strength of wall-beam
structures, and this is illustrated in figure 8.9 , which shows the bending moments
in the supporting beam of a typical structure calculated by a finite element analysis.
On the other hand, an opening close to a support results in considerably greater
bending moments in the beam , and much higher stresses in the wall. In the latter
case Davies and Ahmed suggested that the columns of brickwork between the
opening and the end of the wall should be designed to carry half of the loading on
the wall-beam structure.
8.1.3 Experimental results and verification ofwall-beam theories

A considerable amount of experimental investigation of wall-beam structures has
been reported ,' , 1 7 - 2 1 the rcsults of which confirm the conclus ions of the theo-
retical analyses described in section 8.1. 2 and contribute information concerning
the behaviour of wall-bcams at ultima te load.
Burhouse/? showed that in the majority of cases failure took place by crush -
ing at the lower corners of the panel, followed by failure of the supporting beam .
Similar results were reported by Stafford Smith et al. 21 for structures having
'light' to 'medium' supporting beams , that is, as shown by analysis , where the
wall arehing forces are heavily concentrated at the ends of the beam. With a very
heavy support beam, local damage to the brickwork in the vicinity of the supports
was much less severe and, at the ultimate load, failure took placc by the develop-
ment of vertical and diagonal cracks away from the supports. Model experiments
by Davies and Ahmed 14 showed cracks radiating from the support points, typical
of the failure patterns observed in experiments on concentrated loading on brick-
work.
It is clear from the foregoing that the critical condition for failure will be, in
most cases, the concentrated vertical stress distribution around the supports. This
can be estimated by one or other of the methods described in the preceding
section. If the beam is of exceptionally heavy section the stress concentration will
be greatly reduced, and in such a case overall compressive failure of the wall may
be the critical condition.
Some comparisons are possible between experimental results and the values
calculated by the theories referred to in the previous paragraph-thus table 8.2
Table 8.2
Comparison of calculated and experimental vertical stress
concentration factors ' ?
Test H/L K* Rat Stress concentration factor

no. S.S. & R. D. & A. W. & S.§ Exp .
6 0 .58 8.48 5.05 1.5 17.86 8.58 6.4 8.33

7 0 .58 8.48 5.05 1.5 17.86 8.58 6.4 9.18
8 0.83 8.48 6.83 2 .25 17.86 9.89 8.85 8.72
9 0.33 8.48 3.0 0 .75 17.86 7.6 8 .5 9.39
10 0 .81 8.48 6.69 2.19 17.86 9.7 4 .75 10.89
*K J ~E;;e)
= -Stafford Smith and Riddington
3
•(E tH
tRf =~ \ ~
)
-Davies and Ahmed
EwtH
tR a = ~ -Davies and Ahmed
§CaIculated on basis of BS 5628 values for fk and slenderness reduction factor ß (Wood and
Simms)
shows comparisons of vertical compressive stress concentration factors for five

wall-beams tested by Burhouse.i? In this case, the beams were reinforced concrete
sections of 3658 mm span, and the walls were 105 mm thick single-leafbrickwork.
Various parameters for the wall-beam combinations are shown in table 8.2, along
with the ratios of maximum to average compressive stress as calculated by the
approximate methods of Wood and Simms, Stafford Smith and Riddington, and
Davies and Ahmed. These may be compared with the reported experimental
values-the Davies and Ahmed method appears to give the most consistent agree-
ment with experimental results, while the formula given by Stafford Smith and
Riddington gives a very high result for these beams. The very approximate method
of Wood and Simms gives values of the correct order of magnitude, but agreement
Table 8.3
Comparisons be t ween calculated and ex perime nta l values for vertical stress conce ntratio n fac tors,
bending moment and tie fo rce coefficients/ '
Vert ical st ress conce ntra tio n facto rs Bend ing moment coefficients* Tie force coefflcients]
Test HIL K Rf Ra S.S . & R . D. & A. W. & S.:\: Ex p. S.S . &. R. D. & A. W.&S. Exp . S.S. &R. D. & A. W. &S . Exp .
no .
1 0. 8 7. 9 5.4 3.49 10.1 2 8.02 7. 33 8.4 62.8 49 .9 58.7 156 6.8 22 8 .8 11.5
2 0.8 6. 2 4.33 1.97 5.16 6.63 5.13 5.3 45 .5 39 .6 41.1 56 6.8 10.8 8 .8 9.1
3 0.8 3.7 2. 59 0.87 2.9 4 .36 2.8 5 22.8 28 22 .8 20 6.8 7.8 8.8 7.8
*M = WwLIB .M. Coeff.

t T = Ww/T.F. Coeff. Coefficients quoted are for average values of tie force
:j:Values calculated on ratio of brickwork stress at failure/b rickwork strength and slenderness redu ction coefficient from BS 56 28: 1978
or otherwise with the experimental results may be somewhat fortuitous , as the
calculation depends on values of material strength and slenderness reduction factor
from a particular code of practice, in the present case BS 5628: 1978.
Some further comparisons are given in table 8.3 for three wall-beams tested by
Stafford Smith et al. 2 1 where the beams were steel sections encased in concrete of
2500 mm span . These results also indicate that the vertical stress concentration
factor can be estimated with a fair degree of accuracy by the various methods.
Comparisons are also shown in this table for beam bending moment coefficients,
and again the three methods of calculation give similar results . The experimental
values, however, were of rather doubtful accuracy for the light and medium
support beams since the experimentally derived bending moment diagrams were
irregular and markedly unsymmetrical; the coefficients quoted refer to the maxi-
mum values from these diagrams.
Calculated and experimental values are also shown for the tie force in the beam
in table 8.2, where agreement between the various methods is less satisfactory,
with only the Davies and Ahrned solution reflecting differences in the tie force , as
a fraction of the applied load , between the light , medium and heavy beams. This,
however , is borne out by comparing the average tie forces shown by the experi -
mental results, and agreement between these and the Davies and Ahmed values is
good for the medium and heavy beams.
8.2 Infilled frames
8.2.1 Structural action of infill panels

Many steel and reinforced concrete frames are built with a brickwork infill which
adds considerably to the strength and rigidity of the structure as a whole. A con-
siderable amount of research work has been carried out on the stiffening effect of
infill panels although very little use of this appears to be made in practical design,
possibly because of the uncertainty of the composite action when, as is usual, the
brickwork is not completely bonded to the frame and there is the fear that infill
panels may be removed at some stage in the life of a building. Despite these diffi-
culties there may weil be circumstances in which designers may wish to take
account of the effect of infilling as, for example , in resisting exceptional forces
where some degree of cracking would be unimportant.
The general nature of the structural interaction between frame and infill has
been clarified, and is illustrated in figure 8.10. On first application of a racking
load there may be full composite action between the frame and wall if these are
bonded together. At a comparatively early stage, however, cracks will develop
between the two cornponents, except in the vicinity of two of the corners where
the infill panel will lock into the frame and there will be transmission of com-
pressive forces into the brickwork. At this stage, it is convenient to consider that
the brickwork is acting as a compression diagonal within the frame, the effective
width of which depends on the relative stiffness of the two components and on
the ratio of the height to the width of the panel. This action continues until the
shear resistance of the brickwork is overcome and a crack, slightly inclined to the
horizontal, is developed. Several more or less parallel cracks of this type may
(a) First shear crack (b] Propagation of (c) Crushing and

cracks spalling
Figure 8.10 Failure of infilled frames
develop with further increase in load and failure may finally result from the loss of
rigidity of the infill as a result of these cracks, or from local crushing or spalling
of the brickwork in the region of the concentrated loads . In some cases the strength
of the structure may be limited by the strength of the frame members or joints.
8.2.2 Calculation of strength and stiffness of infilled frames

Although the analysis of infilled frame structures has been attempted using the
theory of elasticity 2 2 , 2 3 and by finite element analyses.i" -26 the rather uncertain
boundary conditions between the brickwork and the frame suggest that an
approximate solution would be appropriate. Various approximations 2 7 - 29 have
been proposed, the most highly developed being that based on the concept of an
equivalent diagonal strut, which was originally proposed by Polyakov'? and
subsequently developed by other investigators. 31 -34
The essential problems in this approach are (1) to determine the contact
p
+ 1\
I
\
t,
Beam
x
M Brickwork
x (b]
(a)
Figure 8.11 Beam on elastic foundation analogy for contact length between infill
and frame
1.0
Experimental results
... square
0 .8 • rectangular 1 : 1.6
ah lh
er • rectangular 1 : 2 tc
::0
a It fi
0.6 ah n a, 7f
;:><:
h =2A hh ;, =J;j :::
>
CIl
o
Z
0.4 ::0
-<:
A
• . =::
+ >
r-
r-
CIl
0 .2
Z
o
o
:::
"'0
o
9 10 11
2 3 4 5 6 7 8 oCIl
:::i
t'T1
Ahh er A,I
~
Figure 8.12 Parametrie plots for infilled frarnes Ö
z
\0
lengths between the frame and the brickwork, (2) to find an effective width for
the equivalent strut and (3) to establish the mode of failure and strength of the
brickwork. The contact length depends on the relative stiffness of the frame and
infill, and on the geometry of the panel. Stafford Smith 3 4 first developed the
analogy with a beam on an elastic foundation, by which the column of an unfilled
frame under lateral load may be regarded as one half of a beam on an elastic
foundation (figure 8.11a) which under a central concentrated load P, remains in
contact with the foundation over a length le known as the characteristic length.
The general solution for the beam on an elastic foundation is
y =eNc (A cos Ne +B sin Ne) + e- Nc (C cos Ne + D sin Ne)

where A = {I (kiEl), k is the foundation modulus, and A, B, C andD are constants
depending on the loading and end conditions. The characteristic length is defined
when Ale = 1T in the general solution and the contact length in the case of the
column of an infilled frame is half of this, leading to a parameter
(8.28)
where
Ah =! (Ewt sin 2(J)

4Elhh
(8.29)
in which E w ' E are the elastic moduli of the wall and frame materials respectively
t is the wall thickness
I h is the second moment of area of the column
h is the height of the wall
I is the length of the beam
(J <tan" (h//)
Similarly the beam member of the infilled frame may be represented as a beam
on an elastic foundation loaded by a moment M, figure 8.11 b, giving rise to a
corresponding parameter
(8.30)
I (Ewt
where
Al = sin 2(J) (8.31 )

4EI l i
in which 11 is the second moment of area of the beam. The parameters Ah and Al
have the dimension (length) -l , and Qh and Ql have the dimension of length, so
that it is convenient to multiply the first two and to divide the second two by I
or h as appropriate to obtain the dimensionless graphs shown in figure 8.12,
from which the contact lengths can be found for any given system.
A value for the effective width of the equivalent diagonal strut can be obtained
BRICK MASONR Y WALLS IN COMPOSITE ACTION 193
Figure 8.13 Effective width of equivalent diagonal strut
on the basis that the compression band is defined by the lengths O:h and 0:1 ' as
shown in figure 8.13 . The distribution of compressive stress between the two
limiting points will not be uniform, and on the assumption that it is triangular
with a maximum on the diagonal, the effective width may be taken as
w= tV(0:\2 +O:h 2 ) (8.32)
= ~ j [(:J 2 (2~J 2J + (8.33)
As previously noted, failure of the brlckwork takes place by shear cracking

along the joints between the bricks , which may be followed by crushing near the
loaded corner . The cracking load of the brickwork can be estimated on the basis
of the following relationship for shear strength
7=70 +/JOn (8.34)
where 70 is the shear bond strength
/J is the apparent coefficient of friction
0h is the normal stress on the bed joint
If R cr is the force acting on the equivalent diagonal, then the shear stress and
normal stress at the centre of the panel are approximately
7 = Ra cos () (8.35)
lwt
(8.36)
lwt lw
The latter relationship has been suggested by Kadir 2 6 following work by Seddorr"
on partially loaded concrete walls. Substituting in equation 8.34 and putting 0:1 /lw
=7T/"'l'llw gives for the horizontal load on the wall at the point of cracking
Tolwt
Pw = R cr cos () = -----:..:..---
1- /J tan () (_--.!!._) (8.37)
AI i:
........
\0
~
00
~
?"S
~
c::
">
~
~
n
"~
o~
~
"
Figure 8.14 (Pwan/Ptotat) per cent as a function of Ahh

Figure 8.15 Distribution of lateral stress at ultimate load
Kadir has carried out an elastic analysis to obtain a relationship giving the per-
centage of the total lateral force applied to the frame-wall system carried by the
wall, in terms OfAhh. The result is shown in figure 8.14 and, from this and equa-
tion 8.37, it is possible to calculate the total cracking load on the structure,
assuming that the frame remains elastic up to this point.
Following initial cracking of the brickwork, and again assuming that the frame
does not fail, the brickwork will resist increased lateralloading on the frame by
friction, wedging and arehing actions within the frame. It is not possible to apply
rigorous methods of analysis to the structure in this state, but as an approxima-
tion it might be assumed that crushing failure of the equivalent diagonal strut
takes place over the effective width, w. If the lateral load Pwu is applied along a
length of the column equal to w cos 8 and it is assumed that the loading varies
linearly, as shown in figure 8.15, then the ultimate load carried by the infill is
Pwu = 1fwt w cos 8 (8.38)
where f w is the ultimate compressive strength of the brickwork. To this load must
be added the resistance of the frame at the deflection, 0f , corresponding to the
maximum load in the brickwork . If the frame remains elastic, its stiffness can be
calculated by normal methods of analysis; thus for a reetangular frame
(8.39)
Estimation of Of is difficult but, as an approximation based on experimental results,

it could be calculated on the basis of an assumed brickwork strain at failure of
0.003.
The analysis described above is based on many simplifications, but as experi-
mental results show considerable variation in terms of stiffness, initial cracking
load and strength the method is possibly as accurate as the situation allows.
Comparisons between Kadir's approximate theory and the results of tests on
one-third scale brickwork panels in steel frames are shown in figures 8.16, 8.17
and 8.18 . Kadir also found that the equivalent diagonal strut method could be
applied to multi-storey frames with reasonable accuracy .
\0
0\
40
Stiffness
P/t>,
(kN /mm)
35
- - Approximate theory
• Exper imental results (square) tI:l
""i
30 o Exper imental results (rectangular) :;.:l
~
c:::
:;.:l
• >
l'
25 t:ll
:;.:l
('5
• ~
o:;.:l
20 ~
15
3 .0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7 .0

A. h/h
Figure 8.16 Lateral stiffness as a function of Ahh
2.50
o Experimental
2 .25
2 .00
1.75
1.50
1.25
Rectangu lar (I/h =2)
•
1.00
0 .75 '--_ _........._ _

~ L...-_ _......... _ _~
3 4 5 6 7 8
). ,I
Figure 8.17 Cracking strength as a function of All
8.2.3 Infill panels with openings

Infill panels frequently contain door or window openings, which will obviously
reduce their effectiveness in stiffening the surrounding frame to an extent
dependent on their size. Experiments by a number ofinvestigators 2 2 , 3 1 ,3 6 , 3 7
have indicated that cent rally located openings may reduce the strength and stiff-
ness of an infilled frame by about 50 per cent and 70 per cent respectively. The
results of a number of model-scale tests by Kadir 2 6 are summarised in table 8A,
and these illustrate the effect of various sized centrally placed openings in a square
infilled panel. The load at first cracking was reduced by approximately 50 to 80
per cent, and the ultimate load by 0 to 40 per cent, as compared to a correspond-
ing frame without openings . However, because of the variability in the test results
and the small number of experiments no defmite conclusions could be drawn.
Kadir suggested an approximate method of analysing infilled panels with open-
ings in which the panel is replaced by a diagonal member of equivalent stiffness, and
0.25 o
iO
~
0.25
0.25
0.25
Approximate theory
0.25 Experimental results (square)
Experimental results (rectangular)
C/.l
1-3
iO
c:
("1
1-3
0.20 c:
iO
Rectangular (I/h=2) >
r-
t::C
iO
......
("1
0.15
~o
Square iO
~
0.10
0.05
3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0
Ah/h
Figure 8.18 Ultimate strength as a function of Ahh

the stiffness of this diagonal can be calculated by considering the brickwork as a
frame from the relationship
(8.40)
Table 8.4
Infilled frames with operrings
(Frame members 1.5 X 0 .75 ins section in a11 tests)
Strength Exp. stiffness Est,

Test h w X lw Opening stiffness
no. (in .) a X b (in.) Crack Ultim. Initial Post-er. (approx)
(ton) (ton) (ton/in) (ton/in) (ton/in)
No opening 1.05 4 .02 55 .5

WW5 3.375 X 3.126 0.35 4 .52 64 22 .7 97.0
WW6 3.375 X 3.126 0.375 4.32 80 25.0 97.0
15.75 2.75 19.6 31.7
WWl 6.75 X 6.25 0.35 28 .5
X
WW3 6.75 X 6.25 0.25 2.85 33 18.5 31.7
15.75 0.20 5.4 6.06
WW2 11.25 X 9.50 2.025 6.3
WW4 11.25 X 9.50 0.15 1.98 9.7 5.7 6.06
No operring 1.27 5.28 68.0

WWA 4 .5 X 6 .5 0.45 4.45 83 .0 29 .5 73 .0
WWAI 4 .5 X 6.5 0.40 5.22 98.0 45 .0 73 .0
15.75 20 .7 31.6
WWB 6.875 X 9.45 0.30 4 .00 57 .0
X
WWBl 6.875 x 9.45 0.20 3.94 82 .0 18.0 31.6
25 .25
WWC 9 .0 x 15.75 0.125 2.20 18.75 13.0 11.0
WWCl 9.0 x 15.75 0.20 2.275 38.4 14.0 11.0
15.75 No opening 1.05 4.02 55 .5

WDI X 11.25 x 6 .25 0.40 2.50 20.5 7.90 28.5
WD2 15.75 11.25 x 6.25 0 .35 2.20 21.5 8.60 28.5
hf -h~ ±.
2 T'
+.
h~
Figure 8.19 Dimensions of equivalcnt brickwork frame

Frame
I
A" "
_Rigid arm
Figure 8.20 Analysis of infilled frame with opening (Liauw and Lee)
where J h and J 1 are the moments of inertia of the vertical and horizontal sections
of the brickwork 'frame' shown in figure 8.19. The stiffness of the frame plus
infill is then
K = 12EIh (6 hh +lhl ) + 48Ew (Jt!h ) (8.41 )

h 3f 31th + 21hl h w ' (h w ' +h f ) Jth w ' +Jhlw '
in which the first term is the frame stiffness .
Liauw '? has put forward a method for the calculation of the stiffness and
strength of inftlled frames with openings that uses astrain energy method to
establish the area of the equivalent diagonal strut. Referring to figure 8.20, an
infill with a centrally placed door opening is replaced by two members , one hori-
zontal and one vertical, of effective lengths LI and L 2 respectively and connected
by a rigid joint. Interactive forces between the surrounding frame and the inftll are
assumed to be concentrated at two diagonally opposite corners and transmitted to
this 'frame' by rigid arms. The effective length of the beam and height of the
column are
LI =B +C I
L 2 = H + C2 < L2 '
where L 2 ' is the distance from the bottom of the wall to the centroidal axis of the
beam, and CI is taken as half the depth of the beam . Then from consideration of
the total strain energy , the deflection in the direction of the load is
2
_ P {sin 8 3 3 cos" 8 3 3]
~- - '-- [(L-md - m i ]+ - - [(L2 - m2) +m2
3E 11 ~
+ 1.2E (LI sin 8 + _~~OS2_~) + LI cos ~ + L 2 sin

2 2 2
8_} (8.42)
G Al A2 AI A2
where ml :; (h 1/2) cot (J and m2 :; (h 2 /2) tan (J.
The diagonal stiffness of the infill is then equal to the reciprocal of the deflection
t. when p « I. The stiffness of the equivalent diagonal strut is thcn given by
(8.43)
and its cross-sectional area A e :; Ld/Et.. The infilled frame is then represented by a
diagonally braced frame using these equivalcnt struts.
Uauw suggests calculating the strength of the frame , when the opening lies
below thc compression diagonal , with reference to the equivalent arca A e and the
ultimate compressive stress of the infill material. If the opening extends above the
comprcssion diagonal the infill is subjected to bending, shear and compression, and
it is nccessary to assess the strength of thc brickwork under diagonalloading. This
will usually be limited by the bending or shear strength of one of the beam
elements and may be calculated approximately on this basis. Liauw reported
rcasonable agreement between the results calculated in this way and those obtained
experimentally.
References
I. R. H. Wood, 'Studies in Composite Construction. Part 1, The Composite
Action of Brick Panel Walls Supported on Reinforced Concrete Beams',
National Building Studies Research Paper 13 (HMSO, 1952).
'2 . R. H. Wood and L. G. Simms, 'A Tentative Design Method for the Cornpo-
site Action of Heavily Loaded Brick Panel Walls Supported on Reinforced
Concrete Beams', BRS CP26/69 (Building Research Station, Watford, Herts .,
1969).
3. S. Rosenhaupt, 'Stresses in Point Supported Composite Walls',J. Am.
Concr. Inst., 61 (1964) 795.
4. A. Coull, 'Composite Action of Walls Supported on Beams', Bldg Sci., 1
(1966) 259.
5. J. R. Colbourne, 'Studies in Composite Construction : An Elastic Analysis of
Wall-Beam Structures', BRS CP15/69 (Building Research Station, Watford,
Herts ., 1962).
6. A. L. Yettram and M. J . S. Hirst , ' An Elastic Analysis for the Composite
Action ofWalls Supported on Simple Beams', BldgSci. , 6 (1971) 151-9.
7. M. Levy and E. Spira , 'Analysis of Composite Walls with and without
Openings', Rep. wkg. Commn Int. Ass. Bridge Struct. Engng, 33-1 (1973)
143-66.
8. D. J. Male and P. F. Arbon, 'A Finite Element Study of Composite Action
in Walls', Second Australasian Conference on Mechanics of Structures and
Materials, University of Adelaide, August 1969.
9. D. R. Green, 'The Interaction of Solid Shear Walls and their Supporting
Structures', Bldg Sci., 7 (1972) 239-48.
10. B. Stafford Smith and J. R. Riddington, 'The Composite Behaviour of
Masonry Wall on Steel Beam Structures', Proceedings of the First Canadian
Masonry Symposium (Calgary) 1976, pp . 292-303.
11. B. Stafford Smith and J. R. Riddington , 'The Composite Behaviour of
Elastic Wall-Beam Systems', Proc. Instn civ. Engrs, 63 (1977) 377-91.
12. J. R. Riddington and B. Stafford Smith, 'Composite Method of Design for
Heavily Loaded Wall/Beam Structures', Proc. Instn eiv. Engrs, 64 (1978)
137-51.
13. A. E. Ahmed, 'A Study of the Composite Action between Masonry Panels
and Supporting Beams', Ph.D. Thesis, University of Edinburgh, 1977.
14. S. R. Davies and A. E. Ahmed, 'An Approximate Method for Analysing
Composite Wall/Beams', Proc. Br. ceram. Soc., 27 (1978) 305-20 .
15. S. R. Davies and A. E. Ahmed, 'Composite Action of Wall-Beams with
Openings', Proceedings of the Fourth International Brick Masonry Con-
ference (Brugge) 1976, Paper 4.b.6.
16. C. B. Saw, 'linear Elastic Finite Element Analysis of Masonry Walls on
Beams',Bldg ss..s (1974) 299 -307.
17. C. B. Saw, 'Composite Action of Masonry Walls on Beams', Proc. Br. ceram.
Soc., 24 (1975) 139-46.
18. S. Rosenhaupt, 'Experimental Study of Masonry Walls on Beams', J. struct.
Div. Am. Soc. eiv. Engrs, S73 (1962) 137-66.
19. S. Rosenhaupt and Y. Sokal, 'Masonry Walls on Continuous Beams',J.
Struct. Div. Am. Soc. civ. Engrs, 91 (1965) 155-71.
20 . P. Burhouse, 'Composite Action between Brick Panel Walls and Their
Supporting Bearns' , Proc. Instn. civ. Engrs, 43 (1969) 175-94.
21 . B. Stafford Smith , M. A. H. Khan and H. G. Wickens, 'Tests on Wall Beam
Structures', Proc. Br. ceram . Soc., 27 (1978) 289-303 .
22 . S. Sarchanski, 'Analysis of the Earthquake Resistance of Frame Buildings
Taking into Consideration the Carrying Capacity of the Filling Masonry' ,
Proc. 2nd Wld Conf Earthq. Engng, 111 (1960) 2127-41.
23. T. C. liauw, 'Elastic Behaviour of Infilled Frames', Proc. Instn. civ. Engrs,
46 (1970) 343-9 .
24 . T. Karamanski, 'Calculating Infilled Frames by the Method of Finite Ele-
ments', in Symposium on Tall Buildings, ed. A. Coull and B. Stafford Smith
(Pergamon, Oxford, 1967) pp . 455-61 .
25. D. V. Mallick and R. T. Severn, 'The Behaviour of Infilled Frames under
Static Loading', Proc. Instn eiv. Engrs, 38 (1967) 639-56.
26. M. R. A. Kadir, 'The Structural Behaviour of MasonryInfill Panels in
Framed Structures', Ph.D. Thesis, University of Edinburgh, 1974.
27. R. H. Wood, 'The Stability of Tall Buildings', Proc. Instn. civ. Engrs, 11
(1958) 69-102 ; discussion, 12 (1959) 502-22.
28 . J. R. Benjamin and H. A. Williams, 'The Behavior of One-Storey Brick Shear
Walls ',J. struct. Div. Am. Soc. eiv. Engrs, 84, ST4 (1958) Paper 1728.
29. T. C. Liauw, 'An Approximate Method of Analysis for Infilled Frames with
or without Openings', Bldg Sei. , 7 (1972) 233-8 .
30. S. V. Polyakov, 'Masonry in Framed Buildings', Gosudalst-Vennoe Izdatel'
stvo literature po Straitel'stvu i Arkitecture, Moskva, 1956 , trans. G. L.
Cairns (Building Research Station, Watford, Herts., 1963).
31. M. Holmes, 'Steel Frames with Brickwork and Concrete Infilling', Proc.
Instn. eiv. Engrs, 19 (1961)473-8 .
32 . B. Stafford Smith , 'lateral Stiffness of Infilled Frames',J. Struct. Div. Am.
Soc. civ. Engrs, 88 ST6 (1962) 183-9.-
33 . R. J. Mainstone, 'On the Stiffness and Strengths of Infilled Frames', Proc.
Instn civ. Engrs, Supplement IV, Paper 73605 (1971) 57-90 .
34 . B. Stafford Smith , 'Behaviour ofSquare Infilled Frames',J. Struct. Div.
Am. Soc. civ. Engrs, 92, S'I'l (1966) 381-403 .
35 . A. E. Seddon, 'The Strength of Concrete Walls under Axial and Eccentric
Walls', Symposium on Strength of Concrete Structures (Cement and Con-
crete Association, London, 1956) pp. 445-73 .
36. J. R. Benjamin and H. A. Williams, 'Behaviour of One-Storey Walls Contain -
ing Openings',J. Am. Concr. Inst., 30 (1958) 605-18.
37. T. C. Liauw and S. W. Lee, 'On the Behaviour and the Analysis ofMulti-
Storey Infilled Frames subject to lateral Loading', Proc. Instn civ. Engrs,
63 (1977) 641-56 .
AUTHORINDEX
Abe1, C. R. 162 Cubbage, P. A. 175

Adams , S. 162,164 Curtin, W. G. 90,161 ,162,164
Ahmed, A. E. 180 ,181 ,184,186,187,
189,202 Davenport, S. T. E. 57
Akroyd , T. N. W. 56 Davey , N. 89
Albrecht,W.57,58 Davies,S.R.120,121,175,180,181,
Ames, S. 175 184,186,187 ,189,202
Anderson,D.E. 159,164 Dear,P.S. 163
Anderson, G. W. 56 Despeyroux, J . 174
Angervo ,K .67 ,70,89 Dickey ,W.L. 162
Arbon, P. F. 201 Dikkers, R. D. 59 ,90
Armstrong, A. C. 58 Dinnie, A. 161
Astbury, N. F. 14,55 ,143,175 Drysdale, R. G. 90
Awni , A. 110 ,121
Edmunds, J . 142
Baker, A. L. L. 9, 11 Ellingwood, B. R. 1.66, 174
Bake~L . R. 55,60,61,123,124, Evans, R. H. 146, 164
141,142
Ba1mer, G. G. 56
Fattal, S. G. 83 ,86,90
Beak, R. 170, 175
Fillipi, H. 163
Beard,R.45,59,161
Fisher, K. 30,31 ,32,33 ,58
Bechte1, S. 142
Florin, G. 58
Beech, D. G. 5,11 ,57 Foster, D. 162
Benjamin, J . R. 116, 121, 202, 203
Francis, A. J. 14,55
Bradshaw, R. E. 32,58
Frisch-Fay,R. 102 ,120
Brebner,A. 162
Frischman, W. W. 121
Burhouse, P. 186 ,187,202
Bums, P. P. 90
Burridge, L. W. 163 Gabrie1son, B. L. 131,137,138,142,
143
tacovic, F. 57,43 Germanino, G. 104, 107, 121
Cadjert , «. 163 Grandet, B. 51, 52, 60
Catteneo, L. E. 83,86,90 Green, D. R. 201
Cauvin, A. 107,121 Grimm, T. 28,29 ,57
Chapman,J .C. 67 ,68 ,69,70,89 Grogan,J.C. 59,162
Chinwah, J. C. G. 47 ,60 Gross, J . G. 39,59
Clare, R. 175
Cochran, M. R. 162 James, J. A. 31 ,32,54,57,61
Colboume , J . R. 201 Jerems , L. E. 55
Colville,J. 76 ,79,90,108,121 Johansson , S. 122,141
Coull , A. 201 Johnson, F. B. 60
205
206 AUTHORINDEX
Kadir , M. R. A. 193, 195, 197 ,202Samarasinghe, W. 50 ,61
Kalita, U. C. 121 Sarchanski, S. 202
Kaplan, K. 137,138,142,143 Saw,C.B . 202
Keller , H. 154 ,163 Sawko, F. 90
Keskin, O. . 120 , 121 Schneider, H. 47,57,58,60
Khan, M. A. N . 202 Scrivener, J . C. 160, 163
Kheir,A.M.A. 123 ,141 Seddon , A. E. 193,203
Khoo,C. L. 19,20,21,26,56 Severn, R. T. 202
Kong, F. K. 146 ,164 Sevin, E. 137,142
Shaw, G. 90
Lawrence, S. J . 60 Simms, L. G. 57,85,89,90,163,
Lee, S. W. 203 177 ,178,187,201
Lenczner, D. 15,45,56,59 ,161,164 Sinha , B. P. 45,51 ,58,60,79,90,
Lent , L. B. 163 102,121 ,126,141,142 ,154,155,
Levy, M. 201 164, 175
Leyendecker, E. V: 166,174 Slatford , J. 67,68,69 ,70,89
Liauw, T. C. 200, 201, 202, 203 Sloane, N. 162 ,164
Loftus, M. D. 141 Soane ,A.J.N. 116,121
Losberg , A. 122 , 141, 163 Spira, E. 201
Lyse, I. 164 Stafford Smith, B. 178, 179, 180,
186 , 189 , 192 , 201,202,203
McBurney , J. W. 163 Stockbridge, J. G. 93,120
Macchi,G. 7,11,104,107 ,121 Strong, A. H. 163
McDowell ,E.L. 131 ,134,137 ,142 Suter, G. T. 154 ,163
McIntosh,1. D. 35,59 Sutherland,R.J.N.91,120,175
McKee,K.E.137,142 Svendsen, S. D. 59
McLeod , I. A. 121 Szabo , P. 57
McNeilly , T. N. 59
Mainstone, R. J . 175,203 Temple, R. 141
Makela, K. 95, 120 Thomas, F. G. 28,29,89,137,142,
Male,D.J. 201 163
Mallick, D. V. 202 Thomas, K. 160, 164
Maurenbrecher, A. W. P. 79,90,121, Thorogood, R. P. 4
162,175 Toppler,J .F. 121
Michael, D. 116 , 121 Turkstra, C. J . 72, 87 , 89, 90
Milner, R. M. 59 Turner, D. J . 164
Monk,C.B. 14,30,56,72,89 Turnsek,V.43 ,47,57
Moore, J . F. A. 175
Morgan , T. W. 60 Vahakellio,P. 95,120
Morsy,E.H . 14,56
Morton ,J . 127,128,137,141,142, Wakefie1d, D.'A. 162
166 , 175 Waldum, A. 59
Murthy, C. K. 58 Ward, N. 57
Wass,R.J. 164
Saffi , K. M. H. 141 Webb , W. F. 141 ,142,175
Sahlin,S. 27 ,45 ,57,59,70,74,76, West, H. W. H. 14,30,32,54,55,
77 ,85,89,90 57 ,60,126,141,142,143 ,175
Salahuddin,1. 59,164 Whittemore, J . W. 163
Sallada,J.W. 162 Wickens, H. G. 202
Sallam, S. E. A. 90 Williams, H. A. 202, 203
AUTHORINDEX 207
Willoughby, A. B. 142 Yettram , A. L. 201
Wilton, C. 131,137,142 Yokel, F. Y. 90
Withey, N. O. 163,164
Wood , R. H. 177 ,187 ,201 ,202
Wyatt , K. 59, 164 Zaccor, J . V. 142
SUBJECT INDEX
Aceidental damage 165-74 empirical studies 26-34

Analysis.Jateral load 114-20 factors affecting 12, 26-34
vertical load 91-4 theories of 15-26
Analysis ofstructures 10,91-1 21 walls 62-90
Arehing theory , walls 131-41 workmanship factors 34-43
Concentrated loads 32
Bending momen t coefficien ts, plates , Core wall structure 2
98 Cross wall structure 2
Biaxial ben ding I 24 Curing, effect on strength 41
Biaxial strength, brick 20
brickwork 50 Deflection, reinforced brickwork 154
Brick characteristics 29 Diaphragm wall 85
Bricks , disturbance after laying 40
Brickwork , chases in 33 Eccentricity , calculation of 77,95,
compressive strength 12-43 , 110-13
63-90 effect of 63, 65, 70
deforrnation properties 43 Elastic rnodulus, compression 43
effect of joint thickness 24 ,30, shear 49
39
plain 4 Failure, compression 13
prestressed 160 Failure theory , elastic 15-17
prism strength 25 strength 17- 26
reinforced 4,144-60 shear 49
shear strength 45-9 Fin iteelementanalysis 49,114 ,116 ,
tensile strength 49-55 179 ,190
workmanship factors 34-43 Flexural tensile strength, 53-5
Brickwork bond 31 Fracture lines 124
Brickwork curing 41 Frame action 102-10
Brickwork laid out of plumb 40 Frame analysis 94-110
lateral loads 116
Cavity walls 83-5
Cellular wall layout 2 Gas explosion 165 , 172
Characteristic load 5 Grouted cavity construction 145
Characteristic strength 5
Chases in brickwork 33 Infill panels , lateral strength 13 I - 4 1
Columns, brittle, stiffness coefficients Infilled frames 189-201
102 Interaction, bed materials 12
theory 67 wall-floor slab 72-81 , 102-13
Composite wall beams 176-89 Interaction curves, plain brickwork
Compressive strength, brickwork 12- 81,84,88
43 reinforced br ickwork 155-60
208
SUBJECT INDEX 209
Joint, unfilled bed 38 Robustness 3
unfilled perpend 36 Ronan Point 165
wall-floor slab 77-80
Joint fixity 105 ,108 Safety factor, global 8
Joint materials, effect of 14 partial 6
Joint th ickness, effect of 24 ,30,39 Shear strength, plain brickwork 45-9
reinforced brickwork 149-54
Lateral load analysis 114-20 Shear walls, analysis 114-20
Lateral strength, flexural 123-6 Slendemess, effect of 62,65 , 70
Lateral strength of infill panels 131- Stability, lateral 3
41 Strength, brick, biaxial 20
Lateral strength with precompression brickwork , biaxial 50
126-31 compressive 12-43
Limit state design 4 flexural 53-5
Load, characteristic 5 prisrn 25
shear 45-9
Moment-magnifier method 81 tension 49-53
Mortar, effect on brickwork strength theories 15-26
24 walls 62-90
incorrect proportioning of 35 workmanship 34-43
mortar, trixial 21
Orthogonal strength ratio 52 , 126 Stress-strain relationship 43 , 148-9
Suction rate, incorrect adjustment 36
Panels, infill, lateral strength 131-41
laterally loaded 123-6 Tensile strength, brickwork 49-55
Panels with precompression 126-31 Torsion of sections 119
Piers, cornpressive strength 62
Plain brickwork 4 Wall end rotation 74
Plain brickwork interaction curves 84, Wall-floor slab joints 77-80
88 Wall layout 2-3
Prestress, loss of 161 Wall stiffness 97
Prestressed brick work 160 Wall types, special 83-8
Walls, cavity 83-5
compressive strength 62-90
Quetta bond 145 diaphragm 86
fin 85
Reinforced brickwork, compression laterally loaded 122-40
155 T section 88
deflection 154 verticalloading on 91-4
flexure 146-9 Workmanship factors 34-43
shear 149-54
shear walls 160 Yie1d 1ine analysis 124, 168
Returns, effect of in compression 86 Young's modulus 43-5

Structural Brickwork-Macmillan Education UK (1981)

Uploaded by

Copyright:

Available Formats

You might also like

Structural Brickwork-Macmillan Education UK (1981)

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Structural Brickwork-Macmillan Education UK (1981)

Uploaded by

Copyright:

Available Formats

STRUCTURAL BRICKWORK

Professor 0/ Civil Engineering,

All rights reserved. No part of this publication may be

First published 1981 by

ISBN 978-1-349-81441-1 ISBN 978-1-349-81439-8 (eBook)

Structural design of brickwork buildings I

2 The strength of brickwork 12

4 Design analysis of unreinforced brickwork structures 91

5 Laterally loaded unreinforced walls 122

7 The resistance of brickwork structures to accidental damage 165

8 Brick masonry walls in composite action 176

Author Index 205

It is commonplace to observe that brick masonry is one of man's oldest building

Figure 1.1 Cellular wall layout

(a) Simple cross-wall structure

Figure 1.2 Cross-wall structure

Figure 1.3 Core-wall structure

1.4 Limit state design of brickwork

1.5 Derivation of partial safety factors

Figure 1.4 Ultimate limit state condition

Unit strength coefficient n Coefficient of variation

Assumed probability of failure 10 - 6

Unit strength coefficient n Coefficient of variation

Category of construction control

Category of manufacturing: special 2.5 3.1

1.6 Analysis of brickwork structures

2.1 Compressive strength: general

2.2 Factors affecting compressive strength

2.3 Failure in compression: indications from standard tests

2.4 Interaction of brick and bed materials

Joint material Compressive strength Ratio to brick strength

Aluminium sheet 106 0 .96

Joint material Compressive strength Ratio to brick strength

Steel 56.48 1.4

2.5 Failure theories based on elastic analysis

Figure 2.1 Stresses in brick-mortar composite

Figure 2.2 Failure envelope for brick material in biaxial compression-tension

Figure 2.3 Effect of joint thickness on brickwork compressive strength

Figure 2.4 Hilsdorf's failurc theory

multiplied by a 'non-uniformity' factor U. Referring to figure 2.4, linc Ais the

This leads to the minimum lateral eonfinement of the mortar joint of

in whieh Oxj is the lateral eompressive stress in the mortar joint

0x = .t -(0 - f';') (2.12)

°x = Oz = fbt (1 - ;f) (2 .13)

where Uu is a coefficient of non-uniformity, which Hilsdorf established experi-

__: _ ) = 1 _ (~)0 .546 (2 .16)

Figure 2.5 Biaxial compression-tension failure envelopes

where al is the major principal stress

O~ _ _""" _ _-L._ _"""" ..I..-_ _"""_ _- I ._ _--I_ _ ~

Figure 2.6 Principal stress relationship for mortars in triaxial cornpression

Figure 2.7 Failure envelope for brick in biaxial compression-tension

!-. = 0.9968 - 2.0264(-~)+ 1.2781 -0 .2487 (2.18)

Figure 2.8 Axial stress-lateral strain relationship für I :~ :3 mortar

Failur e en velopes for

Figure 2.9 Graphical solution for brickwork prism compressive strength

where x = c = al represents the compressive strength of the prism .

Figure 2.10 Brickwork prism compressive strength (S.C.P.R.F. National Testing

M ort ar/bri ck th ick ness rat io , Cl

Figure 2.11 Effect of mortar{brick thickness ratio on brickwork compressive