Professional Documents
Culture Documents
Structural Brickwork-Macmillan Education UK (1981)
Structural Brickwork-Macmillan Education UK (1981)
Structural Brickwork-Macmillan Education UK (1981)
Structural
Brickwork
Arnold W. Hendry
B.Se., Ph .D., D.Se., E LCE.,
E I.Struet.E., E R.S.E.,
M
©Arnold W. Hendry 1981
Softcover reprint ofthe hardcover 1st edition 1981 978-0-333-25748-7
Typeset in 10/11 PR by
Reproduction Drawings Ltd, Sutton, Surrey
Preface viii
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
SubjectIndex 208
PREFACE
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.
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.
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
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
8 STRUCTURAL BRICKWORK
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
STRUCTURAL DESIGN OF BRICKWORK BUILDING 9
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
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
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
10 STRUCTURAL BRICKWORK
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
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.
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
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.
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" )
(2.1)
(2 .2)
(al (b )
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>
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
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
18 STRUCTURAL BRICKWORK
(J . =0 r
lateral
tension
/
I
/
8./ Minimum lateral tension
/ A::=-_---~,-in brick (Iine Cl
/
/
/
local compression
where b is the height of the brick, and j the mortar bed thickness. Une A is
expressed by
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
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
20 STRUCTURAL BRICKWORK
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
=1+2 .91
a )0,805
(2.17)
(o
oZ
'b
x
.-
CIi
~
e'"'"
;;;
iö
a.
'ü
.s 4
Ci.
Ci
'iö
~
0 .4 0 .8 1.2 1.6
M inor principal stress 03 (p.s.i. x l 0 3 )
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
22 STRUCTURAL BRICKWORK
Co
Compression
Tension
CJ
2
)
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
IV
W
24 STRUCTURAL BRICKWORK
80
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 )
~
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 )
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
26 STRUCTURAL BRICKWORK
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
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 )
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
28 STRUCTURAL BRICKWORK
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)
(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 STRENGTH OF MATERIALS 29
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
Height
of unit 125 Q
I +Sample A
(mm) QSample B
100
75
50
25
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
THE STRENGTH OF MATERIALS 31
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.
1 .0
o • Full scale
Bri ck w ork
strength : + 1sca le
br ick strength a ~ scale
0 .8
06
0 .4
~:----_ ...a
.,. +
+
jt
Width 01 be aring plate (t is
thickness 01 w all)
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
",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 .
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.
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
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")
Table 2.6
Effect of workmanship on the compressive strength of non-reinforced
brick walls* (from Gross et al. 55)
*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.
40 STRUCTURAL BRICKWORK
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
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
(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
THE STRENGTH OF MATERIALS 45
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.
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 STRENGTH OF MATERIALS 47
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
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)
~
:>:'
~
~
~.0~"<9
'1:-.-.;\0,
a c/at
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 .
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)
THE STRENGTH OF MATERIALS 51
, /.
.-- I
r-
I
~
. ./ I" 1/
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)
52 STRUCTURAL BRICKWORK
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
4 Satt i 1 : * : 3
ä Satti 1 :1 :6
~ Lawrence
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
)
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 (%)
Fixed fram e
___- - S crew c1amp
IL-::H--" F
Screw c1amp
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-
56 STRUCTURAL BRICKWORK
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).
THE STRENGTH OF MATERIALS 57
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
o S.C.P.R.F.. U .S.Ä .
• S.C.P.R.F.. U. S.Ä.
D B.R.S.. U.K .
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.
- - U.SA S.C.P.R.F.
_ . _ .- W . Germany
<,
e 1 ......
d 3
o 6 12 18 24
Sienderness ratio
h
y ~---+--+--
p
P
(a) (b)
d' = 3 (t - y - ep )
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
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)
70 STRUCTURAL BRICKWORK
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
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
72 STRUCTURAL BRICKWORK
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.
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
~ ~ '\
~Ms
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.
(Ah)2
-p
EJv
e=4yc ~ (1 - ;J (3.8)
w=4 ~: (1 - ~) (3.9)
(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 .
THE STRENGTH OF BRICK MASONR Y 77
Table 3.1
Formulae for bearing capacity and eccentricity (Risager)
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 .
Figure 3.14 Wall and floor slab rotations as a function of applied load on structure
78 STRUCTURAL BRICKWORK
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
".
..
\..;: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
Po~ 1.0
0 .8
~~
0.6
r
P
Po 0.4
0.2
0 .8
P 0 .6
0.4
0 .2
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
THE STRENGTH OF BRICK MASONR Y 83
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
;J
rigidity of a masonry element
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
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.
Table 3.2
Formulae for interaction curves for solid reetangular brickwork section
Me = Pd (I
2 \
__ g !:-)
aPo
g =2 (1-~)
\: Ad 2
aPo
2
21
Ad
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
THE STRENGTH OF BRICK MASONRY 85
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.
\ u t ,
STRUCTURAL BRICKWORK
J r
Ia) Wall stiffened by piers
U
Table 3.3
Test results of axially lo aded strip walls and walls st iffened
along their vertical edges by returns
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
88 STRUCTURAL BRICKWORK
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
THE STRENGTH OF BRICK MASONR Y 89
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).
90 STRUCTURAL BRICKWORK
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
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)
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.
t/2
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.
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
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
96 STRUCTURAL BRICKWORK
4
3
L,
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
DESIGN ANALYSIS OF UNREINFORCED BRICKWORK STRUCTURES 97
3 .0 .......- -_ _
2 .5
2 .0
s,
1..5
1.0
0 .5
0
0 .2 0.3 0.4 0 .5
eld
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
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)
5
~I]l b 0
6
:1[1=3J. 0
b
7
lli b
0
[;~:~~:~I b
0
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
DESIGN ANALYSIS OF UNREINFORCED BRICKWORK STRUCTURES 101
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)
9 :[O b [~~~,~:31 b
a a
t l Sl2 Sn t2 S ZI S24
(2 1) (22) (23) (24) (25) (26)
4 .0
STRUCTURAL BRICKWORK
Case B
.
tt
E~-~--=ttI •
h
I r
3.0
2.0
0 .5 2
I/).h
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.
o
w
104 STRUCTURAL BRICKWORK
Stabilising ti e
Ouarr y
face
:
1
2438
13 0
~
::
+ 30
2438
/.
." =
+'30
-t
2438
102
~
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
DESIGN ANALYSIS OF UNREINFORCED BRICKWORK STRUCTURES 105
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.
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\
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
0
0
co
N
ELEVATION SECTION
Figure 4.12 Outlin e of test stru cture for wall-floor slab interaction (Germanino
and Macchi)
C4 C5
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
108 STRUCTURAL BRICKWORK
2.44 m
152 mm
L
2.46 m
- + ' - - - - - - 4.57 m - - - - - -
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"
..Actual moment at end of slab rigid frame moment comp uted on basis of full joint rigidity
o
\D
-
110 STRUCTURAL BRICKWORK
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
(4.16)
if€<cI>,or
Z = 33 .3 I cI> (I - 2€i (4.17)
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
DESIGN ANALYSIS OF UNREINFORCED BRICKWORK STRUCTURES 113
<I>
0 .7
0 .6 Double curvature
(J (J
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
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.
114 STRUCTURAL BRICKWORK
-
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)
5h
4h
3h
2h
y x
-~I
'yY
,I
'.'
, x!'J. ()
"-
," ". .
y
~
x
Figure 4.19 Translation and rotation of wall element
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
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,
DESIGN ANALYSIS OF UNREINFORCED BRICKWORK STRUCTURES 119
t
b
* ..
b
:-r------=Jt
I t
I ...,r--,~
: : Cross-sectional area A
S I
I
'---_....lI
I'
-- - -- ..
Ib) I -section (c) Closed sect ion
(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
120 STRUCTURAL BRICKWORK
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)
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.
DESIGN ANALYSIS OF UNREINFORCED BRICKWORK STRUCTURES 121
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',
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. 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 .
8rickwork f1exural I
'r: l it
strength 6 / : curve
4 /
./. .;:
2
o
---
..
: .,.
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
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)
LATERALLY LOADED UNREINFORCED WALLS 125
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
126 STRUCTURAL BRICKWORK
_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 .
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
LATERALLY LOADED UNREINFORCED WALLS 127
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 .
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)
N
\0
130 STRUCTURAL BRICKWORK
Pe Ht
work done by lateral force at failure =- - (5.3)
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
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
LATERALLY LOADED UNREINFORCED WALLS 131
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
400
lateral
pressure
(kN/m 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
134 STRUCTURAL BRICKWORK
+-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.
LATERALLY LOADED UNREINFORCED WALLS 135
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
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
LATERALLY LOADED UNREINFORCED WALLS 137
1.0
10 C:Jt
L, ~
t
~
11// , 1.
0 .9
L
L,
0 .8
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)
.. L
"
Figure 5.13 Approximate arehing theory (Hodgkinson er al.)
140 STRUCTURAL BRICKWORK
Table 5.3
Summary of arehing tests on horizontally spanning walls3 !
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
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
LATERALLY LOADED UNREINFORCED WALLS 141
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.
Thesis, University of Edinburgh, 1972.
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.
142 STRUCTURAL BRICKWORK
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
..
~
~ ' .
:= (iil poc ke t in brickwork
P;.)., ·,9
\5L.-.__ ~ 11
E3p1l1
BE l horizonta l reinforcement
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.
or
(6.4)
and
(6.5)
REINFORCED AND PRESTRESSED BRICKWORK 147
k,x
f,
Figure 6.3 Intersection of equation 6,6 with stress-strain curve for steel
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
Figure 6.4 Stress-strain curves for various types of brickwork (PoweIl and
Hodgkinson)
REINFORCED AND PRESTRESSED BRICKWORK 149
0 .27x O.4x
E=0 .0 0 2 2
x
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
}
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
REINFORCED AND PRESTRESSED BRICKWORK 151
Ta ble 6 .1 cont
}
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.
}
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
*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
1.0
o 2 . 4 6
ald
Figure 6.7 Comparison between shear strength of reinforced concrete, grouted
cavity and bed reinforced brickwork beams
154 STRUCTURAL BRICKWORK
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 .
(6.16)
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
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
156 STRUCTURAL BRICKWORK
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)
0 .2
~
;:l:l
;.::
brickwork seetions
158 STRUCTURAL BRICKWORK
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.
REINFORCED AND PRESTRESSED BRICKWORK 159
(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
References
1. A. Dinnie and R. Beard, 'Reinforced Brickwork Silos for Grain Storage',
Proc. Br. ceram. Soc., 17 (1970) 121-36.
162 STRUCTURAL BRICKWORK
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
168 STRUCTURAL BRICKWORK
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
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.
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
172 STRUCTURAL BRICKWORK
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
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 •
RESISTANCE OF BRICKWORK STRUCTURES TO ACCIDENTAL DAMAGE 173
9176
i t
Room 1 Room 3
Room 2 Room 4
....... '----- .
c;::::::::;J c;:::=:::;J
r-----"1
t:;==;:l
"1'<t0
"
t;:::=::;J c;:::::::;:J
R R
Cl)
I I I I
Section AA
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).
RESISTANCE OF BRICKWORK STRUCTURES TO ACCIDENTAL DAMAGE 175
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
»:"" ......
/ '\
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 .
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 .
178 STRUCTURAL BRICKWORK
k x/L C
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
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
BRICK MASONRY WALLS IN COMPOSITE ACTION 179
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)
(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)
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
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
Figure 8.4 Param eters a, ß and 'Y for co rnposit e wall beam in Davies and Ahmed 's
soluti on
182 STRUCTURAL BRICKWORK
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.
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 =
BRICK MASONRY WALLS IN COMPOSITE ACTION 183
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.
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)
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)
184 STRUCTURAL BRICKWORK
I,
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)
BRICK MASONRY WALLS IN COMPOSITE ACTION 185
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)
0 .0 1 ,,~
..
0 .02 '(.. ~ , . •.,-... . • - - . - . - ....... '--=:-.
-, '. -- .--
.. \:~~.'::a~ ~
:.: ~.:._~::.- --- - - - - -- - --...........:.:":'...
':' ~
- _.':-":~'-
0 .03
0 .04
Figure 8.9 Bending moments in wall-beams with op enings (Davies and Ahmed )
Table 8.2
Comparison of calculated and experimental vertical stress
concentration factors ' ?
*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)
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
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.
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
192 STRUCTURAL BRICKWORK
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
(8.28)
where
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
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)
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~
~
"
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
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)
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
o Experimental
2 .25
2 .00
1.75
1.50
1.25
Rectangu lar (I/h =2)
•
1.00
3 4 5 6 7 8
). ,I
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
(8.40)
Table 8.4
Infilled frames with operrings
(Frame members 1.5 X 0 .75 ins section in a11 tests)
hf -h~ ±.
2 T'
+.
h~
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
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.
202 STRUCTURAL BRICKWORK
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 .
BRICK MASONRY WALLS IN COMPOSITE ACTION 203
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
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
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