You are on page 1of 8

BOND CHARACTERISTICS OF EARTH BLOCK MASONRY

By P. Walker1

ABSTRACT: Earth block construction offers a low energy alternative to conventional low-rise building materials
such as clay fired masonry units, concrete, processed timber, and steel framed construction. Used to some extent
in areas such as New Mexico, further development of the technology has been hampered somewhat by very
limited understanding of material performance. In this paper, work undertaken to study the flexural bond strength
characteristics in pressed earth block masonry is summarized. Cement stabilized pressed earth blocks were
fabricated using a manual press and masonry piers were built using both cement:lime:sand and soil:cement
mortars. Flexural strengths were determined using the bond wrench method. Bond strength is shown to be a
function of both block and mortar strength, as well as block moisture content during construction, and mortar
Downloaded from ascelibrary.org by University Of North Dakota on 11/19/14. Copyright ASCE. For personal use only; all rights reserved.

clay content. Bond strengths decreased significantly with saturation of the masonry but showed litle development
beyond 1 day after construction. Characteristic bond strengths in pressed earth masonry are typically much less
than values generally assumed in conventional masonry. In conclusion the paper outlines recommendations for
mortar usage with pressed earth blocks.

INTRODUCTION In the last 25 years there has been much work undertaken
to study the nature of bond developed between cement mortars
Forming building blocks by manual compaction of earth in and various masonry units (Hendry 1990; Baker et al. 1991;
timber molds is a very ancient technique. With development Lawrence and Page 1995). Although not fundamentally un-
of block presses such as the CINVA-RAM there has been some derstood, it appears to be largely mechanical in nature, relying
interest and further development of this technology in the last on the formation of a layer of ettringite crystals at the unit/
50 years (Fitzmaurice 1958; Middleton 1992; Houben and mortar interface. Development of this layer depends on many
Guillaud 1994; Rigassi 1995). In 1995 alone over 3,000,000 factors, including unit initial rate of absorption, unit moisture
earth blocks, including adobe and pressed earth, were manu- content, water retention properties of the mortar, mortar con-
factured in New Mexico (Smith and Austin 1996). Wider es- sistency, mortar composition, fullness of joints, cleanness of
timates suggest that unbaked earth walls provide shelter for bonding surfaces, disturbance of joint after initial construction,
more than one-third of humanity (Houben and Guillaud 1994). quality of work, unit surface characteristics, sand grading, and
Earth building, a relatively simple technology using a low applied precompression. It is likely that many of these will
cost material, will continue to find its widest application in also play an important role in the bonding between pressed
developing countries. However, at a time of increasing envi- earth blocks and mortars.
ronmental sensitivity, unbaked earth wall construction should Mehra et al. (1950) reported on adhesive strength developed
find greater application in developed countries, as it provides between cement stabilized earth blocks and similar mortars.
an attractive alternative to conventional high energy building Tensile bond strengths for mortars containing between 5 and
materials. Environmental benefits of unbaked earth blocks in- 10% cement varied between 0.007 and 0.032 MPa. Recently
clude a decreased reliance on nonrenewable resources, maxi- Venu Madhava Rao et al. (1996) reported on a limited series
mizing use of local materials thereby reducing transportation, of flexural bond strength tests involving stabilized mud blocks
waste reduction, lower energy requirements, considerable po- and cement:sand, cement:soil:sand, and cement:lime:sand
tential for recycling, and favorable insulation properties. mortars. Characteristic flexural bond strengths varied between
The physical characteristics of earth blocks can be greatly 0.004 and 0.14 MPa. Bond strength increased with cement
enhanced by mechanical compaction, often combined with the usage for the cement:sand mortars. Cement:soil:sand mortars
addition of cement, lime, or natural fibers. Minimum com- developed better bond strength than the similar cement:sand
pressive strength requirements above 2.0 MPa are readily and cement:lime:sand mixes. This improvement in bonding
achieved using soils with 5–15% clay content, stabilized with was primarily attributed to the grading of the soil based mor-
5–10% cement, and manually compacted at pressures of only tars.
2 MPa. Performance specifications for earth blocks are out- Development of national and regional standards for earth
lined elsewhere (New 1991; Middleton 1994; Rigassi 1995). construction has been hindered by a lack of reliable data on
Recent recommendations for pressed earth block construc- various physical aspects of performance. Therefore, the inves-
tion suggest using soil:cement mortars in mix proportions sim- tigation reported herein to study bond strength characteristics,
ilar to those used for block production (Middleton 1992; Mu- and ultimately mortar suitability for pressed earth blocks, was
kerji 1994). Weak cement:lime:sand mortars can also be used undertaken.
(Fitzmaurice 1958; New 1991). Contrary to general masonry
practice some proposals have promoted the use of mortars with OBJECTIVES AND SCOPE
strengths similar to or greater than the blocks. To date, there
is little published scientific data to support these recommen- The objective of the work presented in this paper was to
dations or published design values for flexural bond strength study the bond characteristics of pressed earth blocks and mor-
of pressed earth block masonry. tars. Suitability of differing mortars was assessed on the basis
of mortar workability and bond strength developed between
1
Lect., Dept. of Arch. and Civ. Engrg., Univ. of Bath, Bath, BA2 7AY, blocks. A range of pressed earth blocks and soil:cement mor-
U.K. tars were produced by varying constituent materials to repre-
Note. Associate Editor: David A. Lange. Discussion open until January sent typical unit characteristics. Characteristics of pressed
1, 2000. To extend the closing date one month, a written request must blocks and corresponding mortars are outlined. Flexural
be filed with the ASCE Manager of Journals. The manuscript for this
paper was submitted for review and possible publication on March 12,
strengths developed between these blocks and both soil:cement
1998. This paper is part of the Journal of Materials in Civil Engineering, and cement:lime:sand mortars were determined using the bond
Vol. 11, No. 3, August, 1999. 䉷ASCE, ISSN 0899-1561/99/0003-0249– wrench test. Effects of block moisture content at construction,
0256/$8.00 ⫹ $.50 per page. Paper No. 17878. pier moisture content, and age on bond strength were also
JOURNAL OF MATERIALS IN CIVIL ENGINEERING / AUGUST 1999 / 249

J. Mater. Civ. Eng. 1999.11:249-256.


TABLE 1. Soil Characteristics

Soil I Soil II Soil III Soil IV Soil V


Characteristic Building sand Clay soil (0.8:0.2)a (0.6:0.4)a (0.4:0.6)a (0.2:0.8)a (0.1:0.9)a
(1) (2) (3) (4) (5) (6) (7) (8)
Grading (by mass)
Fine gravel fraction (2–6 mm) (%) 4 25 8 13 17 21 23
Sand fraction (0.06–2 mm) (%) 89 25 77 63 51 38 33
Silt fraction (0.002–0.06 mm) (%) 4 4 4 4 4 4 4
Clay fraction (<0.002 mm) (%) 3 44 11 20 28 37 40
Atterberg limits
Liquid limit — 69 22 27 33 40 44
Plasticity Index — 38 Nonplastic 7 16 20 22
a
Ratio of building sand:clay soil by dry mass.
Downloaded from ascelibrary.org by University Of North Dakota on 11/19/14. Copyright ASCE. For personal use only; all rights reserved.

TABLE 2. Earth Block Characteristics

Series
Characteristic A B C D E F
(1) (2) (3) (4) (5) (6) (7)
Constituent soil type I III IV I II III
Cement content (%) 10 10 10 5 5 5
Average dry density (kg/m3) 1,775 1,620 1,520 1,733 1,659 1,584
Average total water absorption (%) 14.8 19.6 24.9 14.6 19.8 22.7
Initial rate of absorption (kg/m2/min) 8.2 9.9 11.7 11.5 11.5 12.2
Average drying shrinkage (%) 0.031 0.063 0.093 0.011 0.033 0.067
Dry compressive strength (MPa) 8.40 4.44 3.11 3.32 2.66 2.33
Wet compressive strength (MPa) 4.26 1.95 1.67 1.62 1.08 0.80
Wet modulus of rupture (MPa) 1.02 0.229 0.141 0.288 0.179 0.057
Average wetting/drying durability (mass reduction after
12 cycles) (%) 0.7 5.5 14.2 4.9 15.5 38.2

TABLE 3. Mortar Characteristics


Water:cement ratio Slump Initial flow Water retention 28-day cube strength
Mix designation (by mass) (mm) (%) (%) (MPa)
(1) (2) (3) (4) (5) (6)
1:1/4:3 (cementa:lime:sand) 0.70 20 110 58 34.0
1:1:6 (cement:lime:sand) 1.3 20 110 66 9.3
1:2:9 (cement:lime:sand) 1.9 15 110 79 4.2
1:3:12 (cement:lime:sand) 2.6 15 110 77 2.2
Soil I ⫹ 5% cement 3.9 15 101 76 1.2
Soil II ⫹ 5% cement 4.6 5 83 82 1.1
Soil III ⫹ 5% cement 5.2 10 82 91 1.6
Soil IV ⫹ 5% cement 6.1 5 75 86 1.0
Soil I ⫹ 10% cement 2.1 20 110 72 3.4
Soil III ⫹ 10% cement 2.9 10 88 87 3.0
Soil IV ⫹ 10% cement 3.1 10 84 90 2.9
Soil V ⫹ 10% cement 3.4 5 87 87 3.0
a
High-early strength cement.

considered. Test results and conclusions are presented in this 2). Soil:cement mortars were formed using the blended soils
paper, together with recommendations for mortar usage and also stabilized with 5 and 10% cement by mass. Cement:lime:
flexural bond strength values for design. sand mortars were formed using the building sand, cement,
and a hydrated building lime. Mix details of both the soil:
MATERIALS cement and cement:lime:sand mortars are outlined in Table 3.

Combining, in five different mix proportions, a building TEST SPECIMENS


sand and a residual kaolinite clay soil of high plasticity formed
soils used to manufacture both blocks and mortars. The grad- The clay soil was thoroughly air dried, clods broken-down
ing and Atterberg limits of these two natural and five blended using a ‘‘wacker-plate,’’ and finally passed through a 5-mm
soils (soils I–V) are given in Table 1. Using a blend of these screen sieve prior to mixing. The building sand was also air
two soils allowed full control over important experimental pa- dried and screened before use. Blended soils were mechani-
rameters, such as clay content, and avoided the mineralogical cally dry mixed together for a minimum of 2 min in a pan
and organic matter variance that is likely to occur when using mixer before addition of cement. Once dry ingredients had
a range of natural soil deposits. Stabilization of clay soils with been thoroughly mixed, water was added slowly to achieve
sand is a commonly employed technique in earth wall con- the appropriate consistency. In block production, this was the
struction (Houben and Guillaud 1994). The five blended soils standard Proctor optimum moisture content of the soil:cement
were stabilized with portland cement. mix (ASTM D 558). After wet mixing for 2 min, blocks were
Pressed earth blocks were fabricated using blended soil compacted within 30 min. Wet mixing of all mortars was
mixes I–IV stabilized with 5 and 10% cement by mass (Table stopped after only 60 s, the mix was then left to stand for 10
250 / JOURNAL OF MATERIALS IN CIVIL ENGINEERING / AUGUST 1999

J. Mater. Civ. Eng. 1999.11:249-256.


min, and then mixing recommenced for a further 60 s prior to BOND WRENCH TESTING
use (‘‘Methods’’ 1984).
All blocks were produced using a manual constant volume Stack-bonded piers, two blocks high, were prepared in the
press manufacturing units with nominal dimensions 295 mm laboratory by an experienced mason. All blocks were laid in
(length) ⫻ 140 mm (width) ⫻ 120 mm (height). The blocks their as-pressed aspect. Joints were fully bedded over the 295
were solid, without perforations or recesses, and with uni- ⫻ 140 mm bed face area, struck flush with vertical faces, and
formly flat surfaces on all faces. The compaction pressure was maintained at a uniform 10-mm thickness. Blocks were gen-
maintained at 1.5–2.0 MPa throughout production. Once com- erally air dried in the laboratory before construction, although
pacted the fresh blocks were extruded from the mold and moist in some cases blocks were prepared by oven drying and/or
cured under plastic sheeting for 28 days. Around 1,500 blocks soaking in water to a prescribed moisture content for at least
were manufactured for the study. ASTM or Australian Stan- 48 h before laying. After construction the piers were moist
dard methods were followed wherever possible to determine cured for a minimum of 5 days under plastic sheeting.
dry density (ASTM C 140), unconfined characteristic dry and The test program was devised to investigate the effects of
wet compressive strengths (Middleton 1992), characteristic block characteristics, mortar type, block moisture content at
Downloaded from ascelibrary.org by University Of North Dakota on 11/19/14. Copyright ASCE. For personal use only; all rights reserved.

modulus of rupture (‘‘Concrete’’ 1984), wetting/drying dura- construction, pier moisture condition at testing, and age on
bility (ASTM D 559), total water absorption (ASTM C 140), flexural bond strength. Piers were constructed using the range
initial rate of absorption (‘‘Masonry’’ 1997), and drying of blocks and mortars outlined in Tables 2 and 3, respectively.
shrinkage (‘‘Concrete’’ 1984). Physical characteristics were To assess influence of moisture content at construction, blocks
determined using five randomly selected blocks for each pa- were laid oven dry, saturated, and at two other intermediate
rameter and from each mix. moisture contents. To assess effect of moisture content at test-
Pressed earth blocks having a range of typical physical char- ing, a number of piers were saturated by 24-h immersion in
acteristics were produced by varying the blended soil grading water. In general, bond testing was undertaken when piers
and cement content (Table 2). Minimum compressive strength were 7 days old, but one series was undertaken in which bond
requirements for pressed earth blocks vary between 1.0 MPa strengths at 1, 3 or 4, and 28 days were also recorded. Unless
(wet) and 2.0 MPa (dry) (New 1991; Houben and Guillaud saturated, all piers were air dried in the laboratory for 48 h
1994). Blocks produced using a manual press rarely exceed before testing.
10 MPa. By having drying shrinkage <0.10%, the experi- Bond wrench testing was conducted in accordance with rec-
mental blocks satisfy an upper limit proposed for cement sta- ognized standard procedures (‘‘SAA 1988; ASTM C 1072).
bilized blocks (Fitzmaurice 1958). Test equipment is comprised of a bond wrench and retaining
The ASTM D 559 wetting/drying test was used for accel- frame to clamp the test pier (Fig. 1). The lower block of each
erated assessment of durability performance. In this test, pier was clamped in the retaining frame, and the wrench at-
blocks are subject to 12 48-h cycles, comprising a 5-h im- tached to the upper block. To prevent damage, thin plywood
mersion in water, followed by a 42-h oven drying at 70⬚C. sheets were placed between the contact faces of the pier and
After drying, and prior to immersion, each surface of the block test equipment. The bond wrench was initially set horizontal
is abraded with a wire scratch brush to remove loose material. before load application. An out-of-plane moment was applied
Durability is considered satisfactory if the mass reduction after to the joint using dry sand added gradually to a container sus-
the 12 cycles is <10%. As this test is considered to be very pended from the wrench arm. Flexural bond strengths of in-
severe, the durability of all blocks, perhaps with the exception dividual piers fsp were determined from
of Series F, is therefore considered satisfactory (Table 2). Msp Fsp
Total water absorption and initial rate of absorption values fsp = ⫺ (1)
are also given in Table 2. Total water absorption is closely Zd Ad
correlated with clay content, as a greater proportion of water where Ad = cross-sectional area of test joint; Fsp = total com-
is adsorbed by clay minerals. While the comparatively high pressive force on test joint; Msp = bending moment about cen-
water absorption values are unlikely to be detrimental to block troid of joint; and Zd = section modulus of test joint.
performance, high initial rate of absorption values suggest that Ten piers were tested as part of each series. The ‘‘95%’’
mortars with high water retention characteristics might be re- characteristic flexural bond strengths f ⬘mt for each series were
quired. Alternatively, some limited prewetting of blocks before determined in accordance with accepted methods for masonry
laying might be necessary. and earth construction (‘‘SAA’’ 1988; Middleton 1992) as fol-
Workability, retentivity, and strength characteristics of all lows:
test mortars are summarized in Table 3. Slump testing, using
a 150-mm-high and 50–100-mm-tapered-diameter cone f ⬘mt = fsp(av) ⫺ 1.65(␴n⫺1) (2)
(‘‘Methods’’ 1984), has proven to be the most reliable means
of monitoring workability of soil based mortars (Walker and
Stace 1997) and was used throughout this investigation. Irre-
spective of mortar type, slump values of 5–20 mm proved
most suitable. Water retention was determined by comparing
flow table consistency before and after the samples had been
subjected to a suction pressure of 50 mmHg (ASTM C 110).
Minimum retention values of 60–65% have been proposed
(‘‘Mortar’’ 1963). Water retention values of cement:lime:sand
mortars, although generally lower than usual, improved with
lime usage as expected. The relatively coarse grading of the
fine aggregate used is considered the most likely cause for
these low water retention values. Typically the soil:cement
mortars exhibited better water retention characteristics—water
retention this time improving with clay content. Mortar
strengths were determined from three 50-mm cubes that were
tested saturated in uniaxial compression, directly after removal
from immersion curing in water. FIG. 1. Bond Wrench Test

JOURNAL OF MATERIALS IN CIVIL ENGINEERING / AUGUST 1999 / 251

J. Mater. Civ. Eng. 1999.11:249-256.


where fsp(av) = average flexural bond strength of test series; and similar or weaker strength blocks. Typically, fracture depth
␴n⫺1 = unbiased standard deviation for test series. was limited to 1–5 mm of the block face. However, in some
cases the split occurred much deeper in the section (Fig. 4).
RESULTS In a number of cases a fourth combined interface and block
type failure was observed, in which significant amounts (up to
Mode of Failure 90%) of block material adhered to the mortar. Failure in this
case was often along the interface of the tension face.
As might be expected joint failure in flexure was sudden
and brittle. Three quite distinctive primary modes of failure
were observed during testing. The first of these, termed inter- Characteristic Strengths
face failure, comprised a ‘‘clean’’ separation of the joint along
the interface between the block and mortar (Fig. 2). This type Characteristic flexural bond strengths varied between 0.006
of failure occurred in nearly half the samples tested. The least and 0.252 MPa. Commonly bond strengths were <0.10 MPa
common failure was where the mortar joint failed in flexure, in all except the strongest blocks used in the investigation.
Downloaded from ascelibrary.org by University Of North Dakota on 11/19/14. Copyright ASCE. For personal use only; all rights reserved.

splitting horizontally and remaining adhered to both block These values compare with minimum characteristic bond
faces (Fig. 3). For a significant number of joints, out-of-plane strength of 0.20 MPa assumed by the Australian Standard 3700
flexural failure of the block face was observed. Often this oc- (‘‘SAA’’ 1988). The experimental values are similar to those
curred when relatively stronger mortars were combined with outlined in limited previous work reported by Mehra et al.
(1950), reporting tensile strengths between 0.007 and 0.032
MPa, and Venu Madhava Rao et al. (1996), reporting char-
acteristic flexural strengths between 0.004 and 0.14 MPa. Co-
efficients of variation varied between 13.6 and 52.3%, al-
though were generally between 15 and 35%, which is typical
for this type of masonry test.
In light of the low values and recorded variation, it is sug-
gested that any characteristic flexural bond strength be ignored
in design unless verified by test. For many low-rise domestic
single-story earth buildings in situ verification is unlikely and
indeed generally unnecessary. A flexural bond strength of 0.01
MPa is equivalent to ⬃30% of the dead-weight stresses de-
veloped at midheight in a 3-m-high earth wall. The vertical
bending moment capacity of a 295-mm-thick solid earth block
wall, with zero flexural strength, is comparable to that devel-
oped by a 110-m-thick solid brick wall with a bond strength
FIG. 2. Interface Failure of 0.2 MPa, assuming simple restraint conditions along the top
and bottom edges and no restraints along the vertical edges of
the wall.

Influence of Block Type

Influence of block characteristics was considered by com-


paring performance of piers built with six different blocks us-
ing both 1:3:12 and 1:1/4:3 cement:lime:sand mortars (Table
4). With both mortar grades, joint failure was, except where
block type A was used with the weaker mortar, caused by
block face splitting. It is not surprising therefore that resultant
bond strengths demonstrate a close linear correlation with
block transverse modulus of rupture (Fig. 5). With the one
already noted exception, average bond strengths for each block
type were very similar to the two cases despite a 1,600% in-
crease in mortar cube strength (Table 4). Bond strengths using
FIG. 3. Mortar Failure soil:cement mortars were also closely related to block strength
(Table 5).
Although bond strength is clearly limited by block strength,
it is possible for bond values to increase. For example, average
bond strengths with block type B increased by over 70% by
using 1:3:12 mortar laid at the joint’s optimum moisture con-
tent (Table 6) compared with the bond developed when they
were laid air dry with a 1:1/4:3 mortar; although both failure
modes were due to horizontal splitting of the block face. Im-
proved penetration of the mortar into the block face is the most
likely reason for this increased bond strength. Electron mi-
croscopy was applied in an attempt to observe the physical
nature of the bond, unfortunately with little success due to
weak and friable nature of the materials.
Where bond is ultimately limited by block strength there
seems little benefit in increasing cement content of a given
mortar type. Selecting a mortar appropriate to the block’s char-
FIG. 4. Block Face Failure acteristics far better optimizes bond strengths.
252 / JOURNAL OF MATERIALS IN CIVIL ENGINEERING / AUGUST 1999

J. Mater. Civ. Eng. 1999.11:249-256.


TABLE 4. Influence of Blocks

Mortar Details Flexural Bond Strength


28-day cube strength Average Characteristic Coefficient of variation
Block Mix (MPa) (MPa) (MPa) (%) Mode of failure
(1) (2) (3) (4) (5) (6) (7)
A 1:3:12 2.0 0.209 0.137 20.9 Interface or mortar
B 1:3:12 2.2 0.080 0.037 32.6 Block
C 1:3:12 2.2 0.056 0.011 48.7 Block or combined
D 1:3:12 2.0 0.095 0.063 20.4 Block
E 1:3:12 1.9 0.075 0.040 28.3 Block
F 1:3:12 2.2 0.022 0.012 27.5 Block
A 1:1/4:3 34.0 0.373 0.252 19.7 Block
B 1:1/4:3 34.0 0.078 0.032 35.7 Block
C 1:1/4:3 34.0 0.050 0.010 48.5 Block
Downloaded from ascelibrary.org by University Of North Dakota on 11/19/14. Copyright ASCE. For personal use only; all rights reserved.

D 1:1/4:3 34.0 0.099 0.067 19.6 Block


E 1:1/4:3 34.0 0.036 0.025 18.5 Block
F 1:1/4:3 34.0 0.033 0.025 14.7 Block

attained with 1:1/4:3 mortar. This might be ascribed to more


compatible retention properties and grading of the mortar.
Therefore, there would seem some marginal benefits in using
soil:cement rather than cement:lime:sand mortars with weaker
blocks. However, as mortar clay content increases, excessive
shrinkage is likely to become a problem.

Influence of Moisture Content at Construction


Bond strength is very clearly dependent on block moisture
content at the time of construction (Table 6), as has also been
widely observed in conventional masonry (Hendry 1990) and
earth block masonry (Venu Madhava Rao et al. 1996). When
the block is dry, water is rapidly sucked out of the mortar
preventing good adhesion and proper hydration of the cement.
In general, when the block is very wet the mortar tends to
FIG. 5. Relationship between Block Modulus of Rupture and float on the surface without gaining proper adhesion. Experi-
Bond Strength mental bond strengths were generally least when blocks were
laid very dry or nearly saturated (Figs. 7 and 8). Relative var-
Influence of Mortar Type iation in bond strength with moisture content was proportion-
ately less with the weaker blocks. For each block and mortar
Where the mortar bond was insufficient to produce block combination, there is a unique optimum moisture content cor-
face failure, bond strength was closely correlated with mortar responding to maximum bond strength. The optimum moisture
cube strength (Table 5). In these cases bond strength can content depends on each block and mortar combination.
clearly be improved by increasing mortar strength, while main- Optimum moisture contents using 1:3:12 mortar varied be-
taining block/mortar compatibility, up to the limiting value as tween 50 and 65% of the blocks’ total water absorption values
dictated by the block. (Table 2). Similarly, optimum moisture content for block A,
The relationship between mortar and bond strengths is var- using soil I ⫹ 5% cement mortar, was around half the block’s
iable (Table 5). For example, a 110% increase in strength of total absorption moisture content. However, as the mortar clay
the cement:lime:sand mix 1:2:9, compared with mix 1:3:12, content increases (soil V mortar), the optimum is around 80%
resulted in a 39% increase in average bond strength. Similarly, of the total absorption value. Observations of the soil V joint
for soil I based mortars, using 5 and 10% cement, there was after testing showed very little mortar penetration into the
35% increase in bond in response to 150% increase in mortar block surface when blocks were laid dry. High suction rates
cube strength. However, bond strength of piers using soil I caused the mortar to dry out rapidly along the interface. Higher
mortar (⫹5% cement) and soil V mortar (⫹10% cement), moisture contents ensured more efficient bonding, with much
when the blocks were laid air dry, decreased by 69% despite better mortar penetration into the block.
an increase in mortar cube strength of 150%. As block moisture content varied, then mode of failure also
Bond strengths, when using soil based mortars, were also changed. This was most apparent in the weaker mixes, where
closely correlated to the clay content of the mortar mix. Where block face failure changed to interface at low and high mois-
failure is not governed by block strength, there is a strong ture contents.
linear relationship between mortar clay content and bond (Fig. For blocks combined with 1:3:12 mortar, the optimum ini-
6). Increasing clay mineral content most probably interferes tial rate of absorption for maximum bond strength, as deter-
with the bonding along the joint interface. Even when interface mined during preparation, reached as high as 4.7 kg/m2/min,
failure predominates, mortar strength alone is clearly not in- which is significantly higher than the upper limit of 1.5 kg/
dicative of final bond strength. m2/min widely recommended for fired clay bricks (Hendry
For the majority of different blocks, failure was often gov- 1990).
erned by block strength irrespective of mortar type (Table 5). Block type A piers built with three quite different mortars
However, bond strength was not entirely independent of mor- all exhibited similar maximum bond strengths (Fig. 7), al-
tar type, as it was improved by using more water retentive though optimum moisture contents varied significantly. Soil I
mortars. For these weaker blocks, bond strength using soil ⫹ 5% cement mortar, with both blocks A and D, would seem
based mortar with 10% cement was generally higher than that the most preferred option as it requires least cement.
JOURNAL OF MATERIALS IN CIVIL ENGINEERING / AUGUST 1999 / 253

J. Mater. Civ. Eng. 1999.11:249-256.


TABLE 5. Influence of Mortar Characteristics

Mortar Details Flexural Bond Strength


28-day cube Average Characteristic Coefficient of variation
Block Mix (MPa) (MPa) (MPa) (%) Mode of failure
(1) (2) (3) (4) (5) (6) (7)
A 1:1:6 9.3 0.323 0.180 26.8 Interface ⫹ block
1:2:9 4.2 0.291 0.188 21.5 Interface
1:3:12 2.0 0.209 0.137 20.9 Interface
Soil I ⫹ 5% cement 1.2 0.226 0.162 17.2 Mortar
Soil II ⫹ 5% cement 1.1 0.136 0.062 33.0 Combined
Soil III ⫹ 5% cement 1.6 0.074 0.057 13.9 Interface
Soil I ⫹ 10% cement 3.4 0.306 0.192 22.6 Interface
Soil III ⫹ 10% cement 3.0 0.143 0.023 50.9 Interface
Soil IV ⫹ 10% cement 2.9 0.085 0.048 26.4 Interface
Downloaded from ascelibrary.org by University Of North Dakota on 11/19/14. Copyright ASCE. For personal use only; all rights reserved.

Soil V ⫹ 10% cement 3.0 0.070 0.031 33.8 Interface


B 1:3:12 2.2 0.080 0.037 32.6 Combined
Soil III ⫹ 5% cement 1.6 0.045 0.014 41.8 Combined
Soil III ⫹ 10% cement 3.0 0.129 0.037 43.2 Block
C 1:3:12 2.2 0.056 0.011 48.7 Block
Soil IV ⫹ 5% cement 1.0 0.018 0.012 20.2 Interface
Soil IV ⫹ 10% cement 2.9 0.034 0.006 49.9 Combined
D 1:3:12 2.4 0.095 0.063 20.4 Block
Soil I ⫹ 5% cement 1.2 0.079 0.042 28.4 Block
Soil I ⫹ 10% cement 3.4 0.115 0.085 15.8 Block or interface
E 1:3:12 1.9 0.075 0.040 28.3 Block
Soil II ⫹ 5% cement 1.4 0.095 0.054 26.2 Block
Soil II ⫹ 10% cement 2.8 0.151 0.093 23.3 Block
F 1:3:12 2.2 0.022 0.012 27.5 Block
Soil III ⫹ 5% cement 1.3 0.027 0.009 40.4 Combined
Soil III ⫹ 10% cement 3.0 0.034 0.007 48.1 Block

TABLE 6. Influence of Block Moisture Content at Construction


Block Details Mortar Details Flexural Bond Strength
Moisture Coefficient of
content IRA 28-day cube Average Characteristic variation
Block (%) (kg/m2/min) Mix (MPa) (MPa) (MPa) (%) Mode of failure
(1) (2) (3) (4) (5) (6) (7) (8) (9)
A 1.7 6.6 1:3:12 2.4 0.170 0.117 18.9 Interface or mortar
6.2 4.7 1:3:12 2.4 0.202 0.148 16.2 Interface or mortar
10.4 0.9 1:3:12 2.4 0.168 0.111 20.6 Mortar
15.2 0.0 1:3:12 2.4 0.111 0.063 26.2 Interface
0.3 6.6 Soil I ⫹ 5% cement 1.1 0.205 0.134 21.0 Mortar
6.3 4.7 Soil I ⫹ 5% cement 1.1 0.209 0.162 13.6 Mortar
9.2 1.1 Soil I ⫹ 5% cement 1.1 0.185 0.139 15.1 Mortar
14.9 0.0 Soil I ⫹ 5% cement 1.1 0.115 0.077 20.0 Interface
0.6 6.6 Soil V ⫹ 10% cement 3.0 0.079 0.042 28.4 Interface
2.8 4.8 Soil V ⫹ 10% cement 3.0 0.070 0.031 33.8 Interface
10.2 1.1 Soil V ⫹ 10% cement 3.0 0.189 0.121 21.8 Mortar
14.9 0.0 Soil V ⫹ 10% cement 3.0 0.184 0.087 31.9 Block or combined
B 2.5 6.9 1:3:12 2.3 0.044 0.006 52.3 Interface or combined
9.9 4.5 1:3:12 2.2 0.080 0.037 32.6 Block
12.7 2.0 1:3:12 2.3 0.133 0.075 26.4 Block
22.1 0.0 1:3:12 2.3 0.073 0.041 26.6 Interface or combined
C 3.4 8.9 1:3:12 2.0 0.036 0.017 32.0 Combined
11.8 5.6 1:3:12 2.2 0.056 0.011 48.7 Block
19.4 1.7 1:3:12 2.0 0.086 0.027 41.6 Block or combined
28.8 0.0 1:3:12 2.0 0.060 0.030 30.3 Combined
F 1.0 7.1 1:3:12 2.3 0.017 0.010 25.0 Combined
9.2 4.3 1:3:12 2.2 0.022 0.009 35.8 Block
11.0 2.0 1:3:12 2.3 0.035 0.014 36.4 Block
21.7 0.0 1:3:12 2.3 0.026 0.015 25.6 Block

There is clearly a need to select a block moisture content barriers that will generally ensure that earth walls remain very
appropriate to the mortar for optimum bond strength. In gen- dry. However, in extreme weather events (e.g., storms and cy-
eral, a midrange moisture content for low clay content and clones), walls may be considerably wetter. It is necessary to
cement:lime:sand mortars is recommended. High clay content assess under these circumstances, when peak lateral loads may
mortars require higher block moisture contents to maximize also be experienced, the influence of moisture on flexural bond
bond strength. strength. Flexural performance of piers using blocks A and B
built with cement:lime:sand and soil:cement mortars were as-
Influence of Pier Moisture Content at Testing sessed. All piers were prepared under air-dry conditions and
Satisfying durability and dampness requirements through cured under plastic sheeting. Half of the piers were tested air
the shelter of an eaves’ overhang, render coatings and moisture dry and the remainder after overnight immersion in water.
254 / JOURNAL OF MATERIALS IN CIVIL ENGINEERING / AUGUST 1999

J. Mater. Civ. Eng. 1999.11:249-256.


In all cases there was a significant reduction in bond
strength with saturation of the joint. This reduction, compared
with air-dry performance, varied between 15.8 and 57.8% (Ta-
ble 7). The greatest reductions were recorded with soil:cement
mortars. In one series (block A with soil:cement mortar), sat-
uration of the joint also altered the mode of failure (Table 7).
Reduction in bond strength may be attributed to a number of
factors, including development of pore-water pressures in both
the mortar and blocks and liquefaction and swelling of unsta-
bilized clays in both the soil:cement mortars and blocks.

Influence of Age
A limited series of bond tests were undertaken to assess
Downloaded from ascelibrary.org by University Of North Dakota on 11/19/14. Copyright ASCE. For personal use only; all rights reserved.

development of flexural bond with time. Piers using blocks A


and D and soil I ⫹ 5% cement mortar were tested at 1, 3 or
FIG. 6. Influence of Mortar Clay Content on Bond Strength 4, 7, and 28 days after construction. While mortar cube
strength increased with age as expected, there was little change
in bond strength after 1 day (Table 8). This might be attributed
in part to failure mode and importance of block strength in
Series D. In Series A, where mortar splitting failure occurred,
mortar shrinkage on curing is likely to have contributed to the
observed behavior. The growth and strength development of
the mechanical bond between the block and mortar would
seem to cease after the initial set of the cement. Similar bond
strength development has also been observed in both concrete
and fired clay masonry (Sise et al. 1988), although other tests
have shown an increase in bond with age (De Vitis et al.
1995).

SUMMARY AND CONCLUSIONS


Three distinct primary modes of failure in flexure, all brittle,
were observed depending on whether block strength, block/
FIG. 7. Influence of Moisture Content on Bond Strength— mortar interface strength, or mortar strength governed. Block
Block A strength governed resultant bond strength in a significant num-
ber of joints studied. A fourth combined interface and block
face failure mode was also observed.
Characteristic flexural bond strength between pressed earth
blocks and both soil:cement and cement:lime:sand mortars is
often <0.1 MPa. Typically, in any series there is significant
variation in bond strength of identical samples. Therefore, un-
less confirmed by in situ testing it is recommended that char-
acteristic flexural bond strength be ignored in design ( f ⬘mt = 0
MPa).
The strong correlation between flexural bond and block
strength in a number of series suggests that little benefit is to
be gained from using comparatively high strength mortars with
most pressed earth blocks. In general, this study recommends
using soil:cement mortar based on the same soil mix as the
block and stabilized with 5% cement. However, higher cement
mortars may be warranted when greater shrinkage is antici-
pated, for example, with mortars using soils with clay contents
FIG. 8. Influence of Moisture Content on Bond Strength exceeding 15–20%.

TABLE 7. Influence of Sample Moisture Content During Testing


Mortar Details Flexural Bond Strength
Moisture con- Coefficient of
28-day cube tent at testing Average Characteristic variation
Block Mix (MPa) (%) (MPa) (MPa) (%) Mode of failure
(1) (2) (3) (4) (5) (6) (7) (8)
A 1:3:12 1.8 6.8 0.202 0.148 16.2 Interface
1.8 14.0 0.170 0.088 29.2 Interface
A Soil I ⫹ 5% cement 1.1 7.1 0.226 0.162 17.2 Mortar
1.1 13.2 0.129 0.094 16.4 Combined
B 1:3:12 2.2 9.4 0.080 0.037 32.6 Block or combined
2.2 21.1 0.050 0.017 40.0 Block
B Soil III ⫹ 5% cement 1.6 13.5 0.045 0.014 41.8 Combined
1.6 18.6 0.019 0.014 15.9 Combined

JOURNAL OF MATERIALS IN CIVIL ENGINEERING / AUGUST 1999 / 255

J. Mater. Civ. Eng. 1999.11:249-256.


TABLE 8. Influence of Age at Testing
Mortar Details Flexural Bond Strength
Coefficient of
Cube Age at test Average Characteristic variation Mode of
Block Mix (MPa) (days) (MPa) (MPa) (%) failure
(1) (2) (3) (4) (5) (6) (7) (8)
A Soil I ⫹ 5% cement 0.45 1 0.129 0.089 18.8 Mortar
0.66 3 0.158 0.116 16.1 Mortar
0.94 7 0.144 0.091 22.3 Mortar
1.2 28 0.147 0.069 32.2 Mortar
D Soil I ⫹ 5% cement 0.43 1 0.085 0.029 39.9 Block
0.67 4 0.094 0.015 50.9 Combined
0.74 7 0.073 0.053 16.6 Block
1.1 28 0.088 0.039 33.7 Combined
Downloaded from ascelibrary.org by University Of North Dakota on 11/19/14. Copyright ASCE. For personal use only; all rights reserved.

Increasing mortar strength does not always directly improve Lawrence, S., and Page, A. W. (1995). ‘‘Mortar bond—A major research
bond strength as other factors such as mortar type, soil mortar program.’’ Proc., 4th Australasian Masonry Conf., University of Tech-
nology Sydney, Sydney, Australia, 31–37.
clay content, and block moisture condition, and strength may ‘‘Masonry units and segmental pavers.’’ (1997). AS4456, Standards Aus-
govern performance. tralia, Sydney, Australia.
Block moisture content during construction was perhaps the Mehra, S. R., Uppal, H. L., and Labhu Ram, (1950). ‘‘Use of stabilized
most important of all factors on resultant bond strength. Op- soil in engineering construction, Section VI—Adhesion of plasters on
timum moisture content, for maximum bond strength, for most stabilized soil walls.’’ J. Indian Roads Congr., 15, 483–491.
mortars may be taken as approximately half the blocks’ total ‘‘Methods of sampling and testing mortar for masonry construction.’’
(1984). AS2701, Standards Australia, Sydney, Australia.
water absorption value. However, as soil mortar clay content Middleton, G. F. (Revised by Schneider, L. M.) (1992). Bulletin 5: Earth-
increases, optimum moisture may approach the block’s satu- wall construction, 4th Ed., Commonwealth Scientific and Industrial
rated moisture content. Research Organization, Sydney, Australia.
Saturation of earth block masonry significantly decreases ‘‘Mortar for masonry construction.’’ (1963). ASA123, Standards Australia,
flexural bond strength, with the most marked decrease when Sydney, Australia.
using soil:cement mortars. This reduction in bond strength is Mukerji, K. (1994). Stabilizers and mortars for compressed earth blocks.
GATE-ISAT, Eschborn, Germany.
attributed to pore-water pressure effects as well as swelling New Mexico Building Code, Section 2413. (1991). Construction Industries
and liquefaction of clays in both the blocks and soil mortars. Division, Santa Fe, N.M.
There was little recorded improvement in flexural bond Rigassi, V. (1995). Compressed earth blocks volume 1. Manual of pro-
strength after the initial 24-h set of the cement. While this can duction. Vieweg, Eschborn, Germany.
partly be attributed to limitations of block strength on bond, ‘‘SAA masonry code.’’ (1988). AS3700, Standards Australia, Sydney,
it seems that mechanical bonding reaches full strength in this Australia.
Sise, A., Shrive, N. G., and Jessop, E. L. (1988). ‘‘Flexural bond strength
initial period. of masonry stack prisms.’’ Proc., British Ceramic Soc., Vol. 2, 103–
107.
ACKNOWLEDGMENTS Smith, E. W., and Austin, G. S. (1996). ‘‘Adobe, pressed-earth, and
rammed-earth industries in New Mexico.’’ Bull. 159, New Mexico Bu-
The writer wishes to gratefully acknowledge the financial support of reau of Mines and Mineral Resources, Socorro, N.M.
the Australian Research Council and the assistance of staff, in particular, Venu Madhava Rao, K., Venkatarama Reddy, B. V., and Jagadish, K. S.
T. Stace and B. Whan, in the Division of Environmental Engineering, (1996). ‘‘Flexural bond strength of masonry using various blocks and
University of New England, New South Wales, Australia. mortars.’’ Mat. and Struct., 29(3), 119–124.
Walker, P., and Stace, T. (1997). ‘‘Properties of some cement stabilized
compressed earth blocks and mortars.’’ Mat. and Struct., 30(11), 545–
APPENDIX I. REFERENCES
551.
Baker, L. R., Lawrence, S. J., and Page, A. W. (1991). Australian masonry
manual, 2nd Ed., Deakin University Press, Victoria, Australia. APPENDIX II. NOTATION
‘‘Concrete masonry units.’’ (1984). AS2733, Standards Australia, Sydney, The following symbols are used in this paper:
Australia.
De Vitis, N., Page, A. W., and Lawrence, S. J. (1995). ‘‘Influence of age Ad = cross-sectional area of test joint;
on the development of bond strength.’’ Proc., 4th Australasian Ma- Fsp = total compressive force on test joint;
sonry Conf., University of Technology Sydney, Sydney, Australia, 299–
f m⬘ t = characteristic flexural bond strength;
307.
Fitzmaurice, R. (1958). Manual of stabilized soil construction for hous- fsp = flexural bond strength;
ing. United Nations, New York. f sp(av) = average flexural bond strength of test series;
Hendry, A. W. (1990). Structural masonry, 2nd Ed., Macmillan, London. Msp = bending moment about centroid of joint;
Houben, H., and Guillaud, H. (1994). Earth construction—A comprehen- Zd = section modulus of test joint; and
sive guide. Intermediate Technology Publications, London. ␴n⫺1 = unbiased standard deviation for test series.

256 / JOURNAL OF MATERIALS IN CIVIL ENGINEERING / AUGUST 1999

J. Mater. Civ. Eng. 1999.11:249-256.

You might also like