84 Vth International Brick Masonry Conference

11-9. Some Factors Affecting the Bond Strength of Brickwork

L.R. Baker
Deakin University, Geelong, Victoria, Australia, University of Calgary, Alberta, Canada

ABSTRACT
Factors that injluence the bond strength between brick and mortar include: the age of the specimen, curing
conditions, the jlow of the mortar and the composition of the mortar.
A preliminary study of these factors indicated that the bond strength did not always increase with time after
the first few days and that when it did the increase was not great. DifJerent curing conditions were found to
have a significant injluence although it was difficult to reach general conclusions. The jlow of the mortar had
a dramatic effect on bond strength. Within the wurkable range of mortars, the greater the jlow the greater the
bond strength. The magnitude of this effect was dependent upon the composition of the mortar. Of the four
compositions tested the high cement content mortar gave the highest bond strength at high jlow but nearly the
weakest bond strength at low jlow.
I t is clear from this study that further investigations are essential to the understanding of the factors injluencing
bond strength. A disturbing feature of the tests was that ultimate bond strengths less than permissible code values
were obtained for some commonly used mortars.

INTRODUCTION tars in Series I tests. Canadian materiais were used in mak-

ing stack-bonded specimens:
The bond strength of brickwork, in a strict sense , is the
"gluing" or adhesive strength of mortar to brick . This is PC . .. .. .. Portland Cement Type 10
dependent on the properties of the two materiais at the
brick-mortar interface. Practical\y, however, the quantity MC .... .. Masonry Cement Type H
is measured by tests in which failure may sometimes occur L ........ High Calcium Hydrated Lime Type S
within the mortar joint itse\f. AIso the keying effect of
Brick ..... 197 x 92.5 x 57 mm pressed, 3 cores 40
mortar in the core holes, or "frog", may make a significant
mm dia, with an I.R.A . of 3.2 kg/m 2/min .
contribution to measured strength.
Bond strength is calculated as a direct stress in a tensile-
bond test, as a shear stress in a shear-bond test, or as a Each mortar had a cementitious materiais to sand aggre-
modulus of rupture in a flexural-bond test. In the direct gate ratio of 1:2Y2. Cementitious materiaIs in the four
and shear tests an ultimate load is divided by an area, but mortars were PC:L, PC:MC, MC:L in equal parts by vol-
in the flexural test an ultimate moment is divided by an ume respective\y, and MC. For each of these different
elastic section modulus. Usual\y the effect of core holes on composition mortars four different flows were used , rang-
section properties are neglected if they constitute less than ing from the wettest workable mix to the driest workable
25% of the gross area . Any shrinkage cracks that occur in mixo These extremes were judged by the bricklayer. To
the joint are neglected. Such cracks would occur at the obtain this range of flows for each mortar the initial mix
edges of the joint and hence reduce the section modulus was made to the wettest workable consistency and suffi-
to a greater degree than the corresponding reduction in cient mortar taken from it to make three nine-high stack-
area. bonded piers. While these piers were being made, dry
Properties of the mortar that affect bond strength are: ingredients, in the correct proportion, were added to the
air content, water retention, flow, composition, sand grad- mixer to reduce the flow a desired amount. After remov-
ing, shrinkage, admixtures and aging properties. Brick ing sufficient mortar for the next set of specimens, the
properties influencing bond strength are: width of brick, process was repeated until ali four sets of specimens had
initial rate of absorbtion, surface texture, core pattern and been built.
surface coatings of silicone, dust etc. Approximately 12 minutes was required to build each
Many workmanship factors affect the bond strength set of specimens so the last sets were made 36 to 48 min-
achieved in practice: pressure applied to joint while laying, utes after initial mixing. To check that this did not have
area of contact, thickness of joint, movement of brick after a significant effect on flexural-bond strengths, a set of
initial placement, laying conditions, curing conditions and specimens were built with a fresh batch of MC mortar with
e\apsed times between mixing, spreading and laying. This a flow of 115%. Comparison was made with specimens
paper presents some experimental data on the influence made from a similar flow mortal' obtained by adding dry
of the flow of mortar, freezing and thawing, and aging, ingredients and built 27 to 40 minutes after initial mixing.
on the flexural-bond strength of brickwork . It was subsequently found that these strengths were prac-
tically the same, thus validating the mixing procedure.
After 28 days storage in the laboratory at 20°C each
FLOW OF MORTAR
joint in the prisms was tested to failure in a machine '
The effect of varying the flow of mortar on the flexural- designed by the author and built at the University of Cal-
bond strength was investigated using four different mor- gary. Each joint was c1amped between a fixed and sus-
Session lI, Paper 9, Some FactoH AJJecting the Bond Strength oJ Bnckwork 85

pended frame and a moment applied via the suspension Of the group of specimens built outside, some we.·e left
shaft. Results are shown in Table I and Fig. I. lt is elear outside to go through the 30 freezelthaw cyeles before
from the graphs that within the practical range of the four being tested. Others were moved into the laboratory after
mortars, increased strength was obtained by increasing the the first 24 hours to go through only one freezelthaw
flow of the mortar. This result supports some codes2. 3 that cyele. Others were moved at 24 hour intervals to go
require "the maximum amount of water to produce a through two freezelthaw cyeles.
workable consistency." Direct tensile-bond tests by Isber- On the testing day, ali joints except those broken in
ner 4 using Masonry Cement ranging in flow from 100% handling, were tested in flexural-bond as described in the
to 135% also indicated that bond increased with increasing previous section. Results are shown in Table II and Fig.
mortar flow. Curves he obtained were convex-up, indicat- 11. lt can be seen that the mortar with calcium chloride
ing a relatively minor drop in strength as flow was reduced added produced higher flexural-bond strengths for both
slightly below the maximum workable consistcncy. Ali the specimens built in the laboratory at 20°C and those
curves in Fig. 1 (except MC:L), however, are concave-up built at O°C. This result confirms research by the Struc-
and indicate a drastic reduction in flexural bond as flow tural Clay Products Institute 8 that coneluded that this
is reduced from the maximum workable consistency. The admlxture in small quantities does not significantly reduce
common MC mortar used at a flow of 125%, considered bond strength. The amount of calcium chloride used in
by the bricklayer to be normal practice, produced a flex- the present tests (I Y2%) is probably insufficient to effec-
ural strength only about a quarter of that obtained with tively lower the freezing point of the mortar. Its significant
mortar of maximum workable consistency. effect is to accelerate initial set and liberate heat of hydra-
For the materiais used the flow of the mortar was an tion at a faster rate during this initial period, resulting in
important and very sensitive variable determining flex- increased strength.
ural-bond strength. One composition mortar did not con- An unexpected result was that, for both mortars, a
sistenly give the highest bond strength for ali flows. The higher flexural-bond strength was achieved by those spec-
MC mortar had a much lower strength than the PC:L imens that initially set at O°C, than those that set at 20°C.
mortar which is often assumed to be its equivalent. A dis- Repeated freezelthaw cyeles led to a reduction in flexural-
turbing feature is that the measured ultimate flexural bond strength for specimens built with mortar containing
bond strength of the commonly used MC mortar was less calcium chloride . Although the detailed picture is not
than the permissible value in most codes of practice. elear, these results are in general agreement with tests by
the Structural Clay Products Institute. 8 They reported that
FREEZING AND THA WING although compressive strength of brickwork is adversely
affected by freezing, bond strength remains unaffected.
The National Building Code of Canada 5 requires that
"when the mean daily temperature at the job site falls
AGEING
below 4Y2°C (40°F), mortar, water and masonry units shall
be maintained at a temperature not less than 4Y2°C during lt is usually assumed that the flexural-bond strength of
laying ... and shall be protected from freezing for 48 brickwork increases with time as the hydration of the
hours after laying." This requirement in substance is sup- cement becomes more complete. Some previous tests 9
ported by the Brick Institute of America 6 and other codes. indicated that greatest strength occurred as early as three
The British 7 and Canadian codes also prohibit the use of to seven days after building.
calcium chloride as a frost inhibitor in the mortar. Two subsequent series of tests have been carried out.
A pilot study of the effects of laying brickwork at freez- Series III tests used the mortars PC:L, PC:MC, MC:L as
ing temperatures and of subsequent freezelthaw cyeles on specified in Series I with flows of I 15%, 140% and 120%
the flexural-bond strength was investigated by Series 11 respectively. Pressed bricks 149 x 67 x 50 mm containing
tests. Two mortars were used. Each contained Masonry three 40 mm dia cores were used. The initial rate of
Cement Type H and sand in the proportions 1: 3 by vol- absorption of the bricks averaged 3.2 kg/m 2 /min .
ume. One of the mortars had Anhydrous Calcium Chlo- Series IV tests used three mortars composed of Ordi-
ride added at the rate of 1 Ib/bag of cement. Both mortars nary Portland Cement, Hydrated Lime and Sand. The
had a flow of 125%. Extruded brick 250 x 80 x 70 mm proportion of cementitious material to sand aggregate was
with five 25 mm dia cores were used. Seven-high stack- 1: 3 for ali mortars. Cementitious materiais for each of the
bonded piers were made with these materiais. Ali speci- three mortars were PC:L in equal volumes, PC and L.
mens were tested at an age of 45 days. One group of Flows were 90%, 95% and 90% respectively. Extruded
specimens was built inside the laboratory at 20°C and bricks 230 x 110 x 75 mm with 5 cores 64 x 20 mm were
another group built outside at just below freezing. Of the used, having an initial rate of absorption of AO kg/m 2 /min.
former group, some were left in the laboratory until tested In Series IH tests, two seven-high stack-bonded piers were
and hence did not go through any freezelthaw cyeles. used and in Series IV, ten single joint specimens were
After the first 24 hours the remaining specimens were used to establish flexural-bond strengths at each age. Ali
placed outside in below freezing temperatures and some specimens in both series were cured in the laboratory in
returned to the laboraory after a further 24 hours. These air at 20°C.
specimens went through one freezelthaw cyele. Those Results are shown in Table lU and Fig. lIl. No elear
specimens that remained outside until testing day went relationship exists between flexural-bond strength and
through approximately 30 freezelthaw cyeles. age. In the Fig. IIIb the 28 day strength has been stan-
86 Vth lnternational Brick Masonry Conference

dardized and the ratio of the flexura l strength at age t, FI> The apparent erratic behaviour of the flexural-bond
to the strength at 28 days, F2s , has been plotted. Here, strength of brickwork as it ages points to a complicated
there is no fixed value for this ratio in the early stages of interaction of shrinkage cracking, hydration of cement,
curing. For example, the ratio of seven day strength to 28 and carbonation of lime that needs more detailed inves-
day strength, F 7/F 2s , varies from 0.43 to 1.25. Also it can tigation.
be seen that reductions in flexural strength sometimes The varying values of flexural-bond strengths measured
occur as the age of the specimens increase. in the experimental work described here indicates that
Apart from the variability of the material, there are sev- present knowledge of this property is insufficient to pre-
eral possible reasons for this erratic behaviour. Firstly, dict it from a few simple parameters. Direct measurement
shrinkage cracking of the outer portions of the joints may is desirable for ali structural brickwork, particularly in
reduce the effective section modulus. If shrinkage cracks view of the low values obtained with commonly used
extend 5 mm in from the surface of the joints in a normal materiais.
sized stack-bonded pier, an apparent reduction in strength
of about 28% would be measured.
Secondly, as pointed out by Isberner,4 hydration of the ACKNOWLEDGMENT
cement continues only while the relative humidity of the
The work reported in this paper as Series I, 11 and 111
mortar exceeds about 85%. He found that the relative
tests was performed in the masonry research laboratory
humidity at a depth of 'li! the thickness of the brickwork
of the Department of Civil Engineering, University of Cal-
reduced to 80% after three days of drying at 28°C and
gary, Alberta, Canada, while the author was there on study
50% relative humidity. Hence hydration in the outer por-
leave. Work was financed jointly by I. X.L. Industries Lim-
tion of the joint ceased after three days. Other measure-
ited, Medicine Hat Alberta and the Department of Civil
ments showed that hydration at the quarter-points and
Engineering. Special acknowledgement is accorded the
centre of the brickwork underwent 12 and 15 days hydra-
Technical Staff of the Department, particularly H. john-
tion, respectively. As flexural strength is largely deter-
son and R. Rodney.
mined by the strength of the material in the outer fibres,
The remaining work was performed at Deakin Univer-
the drying rate of the outer portion of the mortar joints
sity, Victoria, Australia, financially supported by the Brick
is most important. In previous tests 9 specimens cured in
Deveiopment Research Institute, Melbourne, Victoria.
air for seven days showed a marked increase in strength
Grateful acknowledgment is made to R. Randall and R.
when water cured. Another consideration is that hydrated-
johnson for their valued experimental work.
lime does not harden in water but requires carbon dioxide
in the air for carbonation to occur. This was indicated in
Series IV tests where some specimens made with lime
mortar were placed under water after initially setting in REFERENCES
air for 24 hours. When removed from the water after
seven days, they had practically zero strength but then I. Baker, L.R., "Measurement of the Flexural Bond Strength of
gained strength at the same rate as the specimens that had Masonry" paper presented at this conference.
been air cured from the time of building. This is shown 2. Brick Institute of America Standard Specification for Portland
Cement-Lime Mortar for Brick Masonry BIA-MI-72.
in Fig. IlIa. Shrinkage cracking, hydration of cement and
3. American Society for Testing and MateriaIs "Tentative Speci-
carbonation of lime, appear to have a complicated inter-
fications for Mortars for Unit Masonry" Part 12, 1967 ASTM
action in the development of flexural-bond strength in C270-G4T.
brickwork that is not well understood. 4. Isberner, A.W., "Properties of Masonry Cement Mortars"
Designing Engineering and Constructing with Masonry Products.
Gulf Publishing Company, Houston, Texas, 1969.
CONCLUSIONS 5. National Building Code of Canada, 1965, Part 4.
The flow of mortar is a sensitive and important para meter 6. Brick InsLitute of America "Recommended Practice for Engi-
influencing the flexural-bond strength of brickwork. Max- neering Brick Masonry" Virginia, 1975.
7. British Standards Institute "Code of PracLice for Structural
imum strengths are obtained with mortars of wettest work-
Use of Masonry" BS 5628 : Part I : 1978.
able consistency. 8. "Cold Weather Masonry ConstrucLion", Technical Notes I, IA,
The flexural-bond strength of brickwork made with Structural Clay Products Institute, December 1967, January 1968.
masonry cement morta r was not greatly influenced by con- 9. Baker, L.R. and Franken, G.L. "Variability Aspects of the Flex-
struction at freezing temperatures, freezelthaw cyeles, or ural Strength of Brickwork", Proceedings Fourth International
small additions of calcium chloride to the mix o Brick Mãsonry Conference, Brugges, April 1976.
Session lI, Paper 9, Some Factors Affecting the Bond Strength of BTickwork 87

.7 .4r-------------------------------------,
In ltlal s e t a t oOC

.6

/
:. . 3
>:
..."
\ HC wl t h Calc lum

'"
Z
LU
a: .- -. - -- - - - - - - - - - - - - - - - - - ---
.5

o/O
:;; . 2
------------- - ----------
. .4
I Q
z
o
'"
~ .1
HC wl t ho ut Cal c lum

I n i t I a 1 set at 20°C

/
Q.
>: PÇ : L ::>
x

._-/-_.
LU
"
t; ......J
Z
LU
a:
o L-~ __ ~ ____________________________ ~

-~

.- -
~ .3
Q t"\~~-- . 30
z
o
'"
-'....- . ~ . / No . o f Freeze!Thaw Cyc 1es
..J
c(
a:
.2
/' Figure 2. Effect of Freezing and Thawing

~/
::>
x
LU
......J ,/
~. /
---
.1

~~./ Figure 1.

o -.
_-- . . _0Y Effect of Flow
90% 100% 11 0% 120% 130% 140% 150%
Fl ow

T ABLE I -Effect of Mortar Flow-Series I Tests

I Portland I Portland I Masonry I Masonry

Cement ) Cement ) Cement ) Cement )
I Lime ) PC :L I Hasonry PC:MC I Lime ) MC:L 2V2 Sand ) MC
5 Sand ) Cement ) 5 Sand )
Mortar 5 Sand )

Flow% 145 125 115 100 150 130 11 5 95 140 125 100 90 140 125 11 5 95

N 24 24 24 24 24 24 24 24 24 24 24 24 24 24 24 24
J .70 .51 .37 .36 .6 1 .24 .11 .04 .32 .31 .24 .2 1 .26 .07 (.049 O
(.052*
V .09 .20 .23 .29 . 15 046 .55 .81 .19 .20 .29 .24 .29 .74 .67 -

N = No. of Tests . J = Mean Joint Strength (MPa) . V = Coefficient of Variation . 'Value oblained using fresh ly mixed morlar

TABLE 2-Effect ofFreeze/Thaw Cycles-Series 11 Tests

Specimens Built in Specimens Buil t outside

Laboratory at 20°C at O°C
No. of Freezerrhaw Cycles
O I 30 I 2 30

I Masonry Cement ) N 7 8 12 10 12 11
) Flow 125% J .15 . 18 .19 .22 .21 .20
'- 3 Sand ) V .34 AO .20 .43 .31 .35
~
'-
o
:2: I Masonry Cement ) N 10 12 12 12 12 12
) J .26 .26 .2 1 .37 26 .25
Flow 125%
3 Sand ) V .55 .32 .2 1 .2 1 .16 .17
Plus Calcium Chloride
88 Vth lnternational Brick Masonry Conference

TABLE 3-Effect of Ageing

Age in Days
Mortar 1 2 3 7 14 28 35

1 Portland Cement ) N 4 9 9 9
PC:L
1 Lime ) J .11 .23 .27 .227
Flow 115%
5 Sand ) V .18 .32 .23 .49

--
'"QJ
'C
QJ
1 Portland Cement
1 Masonry Cement
)
)
PC:MC
Flow 140%
N
J
9
.48
9
.47
9
.43
VJ 5 Sand ) V .18 .15 .28

I Masonry Cement ) N 9 9 9
MC:L
I Lime ) J .16 .17 .27
Flow 120%
5 Sand ) V .08 .25 .33

I Portland Cement ) N lO 10 10 lO \O 10
PC:L
I Lime ) J .32 .22 .33 .27 .25 .61
Flow 90%
6 Sand ) V .16 .31 .31 .27 .42 .38

I Portland Cement ) N 10 10 lO 10 lO
PC
3 Sand ) J .32 .61 1.55 .98 1.34

-
>
'C
'QJ"
QJ I Lime )
Flow 95%
V

N
.29

10
.28

10
.38

10
.26 .40

lO
VJ L
3 Sand ) J .02 .03 .11 .23
Flow 90%
V .25 .1 8 .12 .14

I Lime ) N lO
L
3 Sand ) J Cured in Water for 7 Days .12
Flow 90%
V .39

I Portland Cement ) N .30 30 20 20

PC:L
1 Lime ) J .60 .77 .70 .56
Flow 95%
Q') 6 Sand ) V .19 .17 .25 .24
....;
QJ
~ I Portland Cement N
) 14 16 15 8 16
PC:L
1 Li me ) J .36 .3 1 .37 .41 .41
Flow 95%
6 Sand ) V .33 .37 .19 .39 .39

N = No. ofTests,j = MeanJoint Strength (MPa), V = Coefficient of Variation

 89
Session lI, Paper 9, Some Factors AJJecting the Bond Strength of Brickwork

1.6

1.5

1.4

1.3

1.2

1.1

1.0

10
a.. ·9
:c
:I:
~
I- .8
Z
w
cr:: .7
l-
_RE~ 9
V>

o --= ____p,.
~ .L SERIES IV
z .6 --... - _~. PC:L
o
co
...J
./ .
«
cr:: .5 ./
./

:::>
x
•_SERIES - -PC:
. _ 111 -- Me.
7--------
w
...J .4 .-". _ _ _ _ • ~ ~ REF 9 PC:1-
LI..

.3 \ . ,
.~
./
\! .~: - - ___ . <. SERrES
./

,,/ PC:L
I
re'

.2 ~RI~ l\l. Mf.·\.._ =:: .:::-= :-_ .

. .- -: _ -sER'€.SlV C - .
..1
. __ ."'" "",,.. . SER\éS '" \..
O. ~~----~~~--+---------------~------~
14 28 35
1 2 3 7
Time days

F
t
F
28

Time days
Figure 3. Effect of Ageing