Você está na página 1de 10

874

FACTORS INFLUENCING GROUTEO MASONRY PRISM COMPRESSIVE STRENGTH

J C SCRIVENER
Professor of Building
The University of Melbourne
Parkville, Victoria, Australia 3052

L R BAKER
Director, Masonry Research Centre
Deakin lIniversity
Waurn Ponds, Victoria, Australia 3217

ABSTRACT
The paper reports results of a testing program on hollow grouted masonry
prisms subjected to axial compressive loads to failure. Concrete and clay
units, stack-honded and running-bonded, were used and the effect of
changes in the compressive strength of the grout was investigated. The
grout compressive strength was determined both by tests on grout poured
into non-absorbent moulds and on grout removed from a hollow core after
contact with the absorbent masonry. l\xi al stra i n measurements were taken
and Young's Modulus of the prisms calculated.

The results and conclusions of the test program are comparerl with
those of other researchers and the provisions of the draft masonry code in
Austral i a.

INTROOUCTION
Some early work on the compress i ve strength of grouted hollow block
concrete masonry was conducted in .l \ustralia by Isaacs [6J who made an
empirical fit to experimental results in terms of block and grout
strengths.

Later more extensive experimental programs were carried out by


Drysdale et al [3,4,5,121. The f irst tests [31 showed that the strengths
of ungrouted prisms ano grouterl portions could not be simply added to
prerlict strengths of grnuted prisms. A failure criterion [5J tookinto
account both the strength and strain di fferences of the units, mortar and
grout. Unfortunately both the tests and the theory were on half blocks,
fully bedded with al igning webs and the prisms had an aspect ratio of only
three so that platen restraint could have harl a significant influence.
Nevertheless the work clearly showed that, tak ing the strength of the
875

ungrouted prism as datum, the grout was only about 30% effective in the
grouted prism. Subsequently Drysdale and Hamid [4J observed that the
compressive strength of masonry assemblages will be less when the webs in
successive courses do not align so work was carried out on running bond
specimens [12J. Ungrouted prisms were face shell bedded while grouted
prisms were fu11y bedded. Grouted prisms failed by the splitting of the
face she11s leaving the grout undamagerl. Using the ungrouted prism
strength as datum, the writers have interpreted the grout in the grouted
prisms to be 46% effective in 2 high prisms ranging to 24% effective in 5
high prisms.

Priestley anrl Chai [81 argued that ohserved strain measurements


indicate that grout reaches approximately 90% of its peak stress when the
face shells reach their peak stress.

Using a finite element analysis Shrive [9J showed that capping and
height/width affect the stress d-istribution within a prism. His
recommendation that prisms of height/wirlth of five should be used to give
the unconfined compressive strength has heen adopted in the draft
Australian masonry code [lll.

Maurenbrecher [71 concluded that there is little rlifference in


compressive strength between stack-honrled and running-bonded specimens.

The draft Australian masonry code [111 proposes that the strength of a
grouted pri sm be expressedin terms of the strength of a face shell bedderl
ungrouted pri sm and the strength of the grouted area. The foll owi ng
program of testing was carried out hecause of the widely varying estimates
af the effectiveness of the grout and the rlesirability of tests on
Australian hollow units.

OESCRIPTION OF MATERIAlS ANO TESTS


The parameters investigaterl were:

(i) hollow concrete units and hollow clay units;


(ii) stack-bonded prisms and running-bonded prisms;
(iii) strengths of in-situ grout and of specimens constructed in non-
absorbent cylinder moulds ;
(iv) vertical load deformation characteristics;
(v) testing with full cross section plywoorl capping and with capping
covering the face shells only;
(vi) the relationship between prism strength anrl unit strength.

In order to restrict the number of parameters to be investigated the


following rlecisions were taken:

(i) One mortar mi x was used i n a 11 tests. flecause of the number of


specimens and time factors it was necessary to mix mortars on
severa 1 occas i ons and every attempt was made to achi eve
consistency. The inevitable variation in mortar batches, as
inrlicated ta some extent by the flow test results, was not
considered to be a concern as prism strengths are not highly
rlependent on mortar strength variations.
876

(ii) 8ed thicknesses were kept constant at 10mm.


(iii) Prisms of four units high were used throughout. While prism
height/least lateral dimension is a factor which influences the
compressive failure load of piers, the splitting mode of failure
always prevailed echoing the failure mode in masonry walls in
compression.

The materials and test details follow.

Masonry Uni ts
Mediumweight and heavyweight hollow concrete units 200mm high x 190mm wide
x 390mm 10n9 with two rectangular cores 122mm x 145mm and hollow clay
units 150mm, high x 127mm wide x 263mm long with two rectangular cores 68mm
x 83mm were Ilsed.

Table 1 gives the unconfined compressive strengths of units averaged


from six specimens. No adjustment was necessary from the failure loads of
the units to account for height/thickness effects according to Table 2,
which gives the aspect ratio factors of the draft Austral ian code [l1l.
However adjustments were made for the grouted prisms.

TARLE 1
Unconfined compressive strengths of units

Materi al Full Capping Strip Capping

Co-eff Co-eff.
Load Stress of Load St ress of
(kN) (MPa) variation (kN) (MPa) variation

Medium
weight
concrete 490 o.1i 0.10 371 15.8 0.09
Heavy
weight
concrete 1120 15.1 0.19 738 31.5 0.08
f:lay 512 15.3 0.08 387 23.3 0.09

"
877

TABLE 2
Aspect Ratio Factor

Height/thickness
ratio o 0.4 1.0 5.0 or more
Aspect ratio
factor o 0.50 0.70 1.00

Notes: The thickness used for evaluating the height/thickness ratio is:
(a) for solid or cored masonry units, or for pr i sms made from such units -
the overall width of the unit.
(h) for hollow masonry units or for prisms made from such units - the
mi nimum thickness of the face shell of the unit.
This table is taken directly from the draft Australian masonry code [111.
Grouts
Four different strength Portland cement grouts of high fluidity were
supplied by a ready mix concrete planto The prism cores and 100mm
di ameter x 200mm high cylinders were filled one day after the prisms were
laid. The grout strengths are given i n Table 3.

TABLE 3
Unconfined compressive strengths of grouts

Core Strength (MPa)


Cylinder Mediumwe i ght Heavyweight
Strength Concrete Concrete Clay
Desi gnation UWa) Units Units Units

A 0.3 0.4 0.4 0.3


S 3.0 3.4 3.0 3.1
C lli.4 13.3 12.3 17.7
O 22.0 20.8 18.8 20.3

Note: Cylinder strengths are the average from six tests on 100mm diameter
x 200mm high specimens formed in non-absorhent moulds. Six specimens were
obta i ned by knock i ng out or cutt i ng from grouted cores of the unit S. The
strengths recorded are the averages adjusted for equi val ence with
specimens of aspect ratio two using ASTM correction factors [IJ.
Specimens were cured as the prisms and released from the 'moulds' just
prior to testing at 21 days age.
Mortar
A mix of 1 part Portland cement to 1 part lime to 6 parts mortar sand was
used throughout. The six separate mortar batches gave flows (see [10J) of
198,200,201,201,202 and 202% with average compressive strengths from three
cylinders at 21 days of 6.3,6.8,7.6,6.8,7.3 and 7.6 MPa respectively.
878

Unit Reddi ng
Un it 5 we re face she 11 bedded except that the end webs were s 1 i ght 1Y
buttererl to contain the grout. The bottom unit of each prism was lairl in
mortar over the full cross sectional area.

Pri sm Bondi ng
The runni ng-bonderl pri srns had a full unit on the base and thi rd courses
with two half units on the second and top courses. The stack-bonrlerl
pr -isrns were constructed norrnally.

Pri sms
Five prisms of each unit and of each grout were const r ucted and cured in
t he arnhient conditions of the laboratory.

Prism Testing
At 21 rlays age the prisms were testedin a Universal Test machine with
hall seats at top anrl bottom. The grouterl pri Srns were pl aster capped and
testerl with 3mm plywood over the full cross section while strips over the
face shells wer,e used for the ungrou t ed prisms (thisis the stanrlarrl test
procedure of the draft Australian corle [111). The compressive load,
measured with a load cell, was applierl axially at the unifonn r ate of
lS0kN per minute.

Strain Measurement
On some pr -isms longitudinal strains were recorded using linear
rlisplacernent transdrlcers over the height of a unit inclurling a central
mortar joint (see Figure 1). Four transducers were placed symmetrically
on the faces of a pr i sm anrl the deformations averaged to eliminate any
eccentricity effects.

PRISM TEST BEHAVIOUR

Types of fail ure


The pri sm fai 1 ures were predomi nant 1y i nducerl by ve rt i ca 1 sp 1 it ti ng in the
central two courses with occasional spalling in a unit adjacent to a
platen. While most fa i lures originated from a crack on a unit end or ends
(see Figures 2 and 3) there were some originating from cracks on a unit
face particularly with prisms of higher grout strength. Of the 45
runn i ng-bond prisms tested, the four face shell failures originated in the
perpend between the two half units.
The first cracking usually occurred between 0.9 and 1.0 of the failure
1oarl. However with one cl ay pri sm the fi rst crack was at O.R1 of the
failure load.

Stress-Strain 8ehaviour
Strain was reasonably l i near with stress up to an average of 77 't. and 68%
failure load for the concrete masonry and clay prisms respect i ve l y. The
clay prisms showed greater non-linearity than the concrete prisms.

The Young's Modulus figures quoted in Table 4 are for secant modulus
from zero loarl to the load where the strain departed significantly from
1 i nearity.
00
--.I
\O

Figure 1. Prism under test shawing Fi gure 2. Vertical splitting failure Figure 3. Vertical crass-sectian
transducers in pasitian af prism af prism at splitting crack
880

Unconfined Compressive Strengths of Prisms


Tahle 4 gives the uncanfinerl campressive strengths of prisms averaged from
six tests.

TARLE 4
Unconfinerl Compressive Strengths of Prisms

Cylinder
Grout Prism Co-eff Young's E/Prism
Strength Loarl af Modulus E St re s
Material Ronding (MPa) (kN) Variation (MPaxI0 3 ) (xl0 3)

Medium- Stack 389 0.09


weight 0.3 337 0.14 9 2.0
concrete 3.0 541 0.02 9 1.2
16.4 593 0.12 9 1.1
?2. O 582 0.10 6 0.8
Running 347 0.11
16.4 512 0.08 8 1.1
22 .0 577 0.08 11 1.4
Heavy- Stack 614 0.16
weight 0.3 435 0.08 15 2.5
concrete 3.0 602 0.04 14 1.7
lfi .4 1040 0.12 ;10 1.5
2;1.0 846 0.20 19 1.6
Running 481 0.09
16.4 779 0.13 15 1.5
22 .0 799 0.11 19 1. 8
Clay Stack 350 0.12
0.3 286 0.10 I) 0.7
3.0 393 0.06 7 0.1)
16.4 360 0.20 7 0.6
22 .0 464 0.12 7 0.5
Running 360 0.07
11;.4 392 0.03 8 0.7
22.0 402 0.07 8 0.7

Note: The pri sm l oads have been adjusted for aspect ratio facto r
according to Table 2.
881

DISCUSSION OF RESULTS

Pri sm St rengt h
In Figure 4 the experimental results for unconfined compressive strengths
of mediumweight concrete prisms are plotted against grout strengths. The
results show that only a small proportion of the grout strength is
effective in the determination of the prism strength and generally support
the tests of Drysdale et al [3,4,5,121. The results compare favourably
with the draft Australian masonry code [111 proposal that

Grouted prism strength Fg ungrouted prism strength Fug +


grout area Ag x square root of grout stress fg

which recogn ises that the higher the grout strength the less effective as
a percentage of its own strength the grout is on the prism strength.
Figures 5, 6 and 7 for prisms of mediumweight concrete, heavyweight
concrete and cl ay units respecti vely compare the Austral i an code proposal
and the experimental results for both stack-bonded and running-bonded
prisms.

While running-bonded prisms in concrete masonry have 75 to 99% of the


strength of comparable stack-bonded prisms (see Figures 5 and 6), with
c 1ay pri sms there are on 1y mi nor st rength diffe rences due to a bond i ng
change (see Figure 7). The clay results confirm those of Maurenbrecher
[71. With concrete rnasonry it appears that the perpe nd joint weakens the
prisrn. On the basis of so few tests and for the sake of such small
di fferences a change from the simpl e stack-bonderl pri sm test to the more
comp l icated running-bonded prisrn test i s not warranterl.

Table 4 and Figures 5, 6 and 7 show that with very weak grout (0.3
MPa) the grouted pri sm strength i s reduced below the ungrouted pri sm
strength which confirms the findings of Brown and Whitlock [2J.

The strength of ungrouted prisrns is more closely indicated by tests on


units with face shell capping than units with full capping .

Prism Behaviour
While the Young's Modulus results show considerable scatter it can be seen
that, apart from the concrete prisms with very low grout strength,
concrete masonry prisms exhibit an E slightly higher than 1000 x prism
strength, the generally recognised val ue. On the other hand, for brick
prisrns E/prism strength tends to be much less than 1000.

Figures 2 and 3 give the vertical splitting failure pattern most


common 1y obse rved in these test s. It i s i nte rest i ng to observe that the
vertical split continues through the grouted cores, which is in contrast
with failures of Wong and Drysda le [121 where though the fa i lures
originated "in the masonry shells the grout remained undamagerl.

Grout St rengt hs
Overall, the strength of grout in the cores of the units, Table 3, tended
to be slightly lower than the strength of grout in non-absorbent
cylinders, though this did not apply in all cases. This is a surpnslng
result as it has been presurned by many that the suction of the masonry
882

)(
1000 1000

Fg=Fug + f'g Ag

~
G
, - \
.... po~\e
.... '- ~!\).!\q
)'l\J.'>

Fg=Fug + 1.2 [f7; Ag


500 500
- - F g=F ug + 0.3 f' g Ag
Fg=Fug + ~ Ag
z z
:::. :::.
..c:
...: ..c:
1J' +'
1J'
C C
...
Q)
...
Q)

+' +' Experimental points


(fJ [J)

E Stack bonded prisms E X Stack bonded


...,Ul ...,
Ul
... ...
p.,
, p.,
I
o Running b o nded

O 10 20 O 10 20
Grout Strength (MP a ) Gro ut Strength (MPa)

Figure 4. Effectiveness of grout Figure 6. Prism strengths in heavy-


in prisms of mediumweight weight concrete units
concrete units
700 700

)C

)(

z
./

~ -- o 0
~ Fg=Fug + ~ Ag
..c:
)c

+'
..c: 1J'
+'
1J'
C
C
(l)
... ~Ag
Q)
... Experimental points
+'
(fJ
+'
(fJ E
E X Stack bonded ..,
.Ul
..,
Ul
. ...
p.,
...
p.,
o Running b o nded
L
o 10 20 O 10 20

Grout Strength (MP a ) Gro ut Strength (MPa)


Figure 5. Prism strengths in Figure 7. Pr i sm strengths in clay
med i umwe i ght concrete units
units
883

unit had the effect of reducing the water content of grout in - situ with
the consequent rising of the grout strength. It is a convenient result as
the strengths obtained from simple cylinder tests would appear to reflect
in-situ grout strength most adequately.

CONCLUSIONS
From the results of the prism tests reported, the fo11owing conclusions
are reached.

1. The experi menta 1 resu lts con fi rm that on 1y asma 11 proport i on of grout
strength is effective in the determination of prism strength.

2. The ratio of Young's Modulus to prism strength appears to be slightly


greater than 1000 for concrete unit prisms and somewhat less than 1000
for clay prisms.

3. The strength of grout within the cores of ho11ow units is slightly


lower than the strength of grout poured in non-absorbent moulds.

4. The strength of running-bonded prisms of grouted concrete masonry is


slightly below equivalent stack-bonded prisms. With clay prisms the
difference is minoro

REFERE"CES
1. American Society of Testing Materials. C42-84a, ASTM 1984.
2. Rrown, R.H. and Whitlock, A.R. Compressive strength of grouted hollow
brick prisms. ~~asonry: Materials, Properties and Performance,
ASTM STP 778, 1982.
3. Orysdale, R.G. and Hamid, A.A. Rehaviour of concrete block masonry
under axial compression. ACI ,Journal, Proc. V.76, June 1979, 707-
21.
4. Dry sdale, R.G. and Hamid, A.A. Influence of the characteristics of
the units on the strength of block masonry. Proc. Second North
American Masonry Conf., Maryland, Aug. 1982, 13p.
5. Hamid, A.A. and Drysdale, R.G. Suggested failure criteria for grouted
masonry under axial compression. ACI Journal, Proc. V.76, Oct.
1979, 1047-6l.
6. Isaacs, H. The ultimate strength of grouted ho11ow concrete block
masonry. Constructional Review, May 1975, 36-47.
7. Maurenbrecher, A.H.P. Effect of test procedures on compressive
strength of ITlasonry prisms. Proc. Second Canadian Masonry Symp.,
Ottawa, June 1980, 119-32.
8. Priestley, M.J.N. and Chai, Y.H. Prediction of masonry compression
strength from constituent properties. New Zealand Concrete
Construction, V.28, March and April 1984.
9. Shrive, N.G. The prism test as a measure of masonry strength. Proc.
Rritish Ceramic Society, A3A-12.
Ir). Standards Association of Austral ia. Masonry cement. AS 1316-1972.
11. Standards Association of Australia. Draft SAA Masonry Code, 1987.
12. Wong, H.E. and Orysdale, R.G. Stress-strain characteristics of
concrete hlock masonry. Source unknown.

Você também pode gostar