Compressive strength testing of compressed earth blocks

Abstract

As with other masonry units, compressive strength is a basic measure of quality for compressed earth blocks. However, as compressed earth blocks are produced in a great variety of sizes the influence of block geometry on measured strength, primarily through platen restraint effects, must be taken into account. The paper outlines current methodologies used to determine compressive strength of compressed earth blocks, including direct testing, the RILEM test and indirect flexural strength testing. The influence of block geometry (aspect ratio), test procedure and basic material parameters (dry density, cement content, moisture content) are also discussed. Proposals for the future development of compressive strength testing of compressed earth blocks are outlined.

Full text

Compressive strength testing of compressed earth blocks
Jean-Claude Morel, Département Génie Civil et Bâtiment, URA 1652 CNRS, Ecole Nationale des
Travaux Publics de l’Etat, Rue Maurice Audin, 69518 Vaulx en Velin cedex, FRANCE.
Abalo Pkla, Département Génie Civil et Bâtiment, URA 1652 CNRS, Ecole Nationale des Travaux
Publics de l’Etat, Rue Maurice Audin, 69518 Vaulx en Velin cedex, FRANCE
and
Peter Walker, Dept. Architecture & Civil Engineering, University of Bath, Bath, BA2 7AY, UK.
Tel: 01225 386646; Fax: 01225 386691; Email: P.Walker@bath.ac.uk
Abstract
As with other masonry units, compressive strength is a basic measure of quality for compressed earth
blocks. However, as compressed earth blocks are produced in a great variety of sizes the influence of
block geometry on measured strength, primarily through platen restraint effects, must be taken into
account. The paper outlines current methodologies used to determine compressive strength of
compressed earth blocks, including direct testing, the RILEM test and indirect flexural strength testing.
The influence of block geometry (aspect ratio), test procedure and basic material parameters (dry
density, cement content, moisture content) are also discussed. Proposals for the future development of
compressive strength testing of compressed earth blocks are outlined.
Keywords: Compressive strength testing; Compressed earth blocks; Aspect ratio
 Corresponding author

1. Introduction
Plain masonry elements, such as loadbearing walls, arches and vaults, have developed to take
advantage of the material’s relatively high compressive strength. The capacity of masonry in
compression is strongly related to the compressive strength of the masonry units (stone, brick, and
block), as well as mortar strength, bonding pattern and many other factors. Though other parameters,
such as density, frost resistance and water absorption, may be specified in design, compressive
strength has become a basic and universally accepted unit of measurement to specify the quality of
masonry units. The relative ease of undertaking laboratory compressive strength testing has also
contributed to its universality as an expression of material quality.
For many centuries hand moulded unburnt mud blocks, adobes, have been used for loadbearing
masonry structures. Though adobes are most used for lightly loaded single and two-storey residential
building, adobes have also been used to construct 10-storey high buildings in Yemen [1]. Over the past
fifty years compressed earth blocks have developed and been increasingly used, especially in
developing countries such as Mayotte [2,3]. In that case, earth is a clayey soil with variable quantity
and quality of clay depending of the building site. The clay fraction is less than adobe, and usually less
than 25% of dry wet. The considerable variation of the composition of earth makes more important the
measurement of the compressive strength of compressed earth block and skilled masons to find the
optimum compsition during the block manufacture.
Compaction of moist soil, often combined with 4-10% cement stabilisation, significantly improves
compressive strength and water resistance in comparison with traditional adobe blocks. Dimensional
stability and tolerances are also much improved, allowing construction procedures similar to fired clay
and concrete block masonry, rather than the wet hand moulded method generally used for adobe.
Quality control strength testing of compressed earth blocks has often followed procedures developed
for fired clay and concrete block units [4]. However, the suitability of these procedures has largely not
been checked by scientific study. The compressive strength of compressed earth blocks can be many
times lower than similar fired bricks. Resistance is also significantly influenced by moisture content.
Previous studies have reported on the compressive strength characteristics of compressed earth blocks
[5-13]. Strength is improved by compactive effort (density) and cement content (generally linear
correlation), but reduced by increasing moisture content and clay content (cement stabilised blocks).
National and international standards have also developed for compressed earth block test procedures
[4, 14-16]. However, unlike other masonry units, there is little general consensus on test procedure for
compressed earth blocks. Should blocks be tested wet or dry? How should dimensional effects, such as
aspect ratio, and platen restraint be taken into account?
This paper reviews the current situation and seeks to inform the on-going debate on the development
of compressive strength test procedures for compressed earth blocks. A number of the different test
procedures currently in use are described and, where possible, compared. Results of experimental
studies are also presented. The compressive strength of blocks measured by differing tests is also
compared with other parameters, such as three-point bending strength.
2. Outline of compression test procedures
2.1 Background
Experimental compressive strength of materials such as concrete, stone, fired and unfired clay is a
function of test specimen dimensions. Load is normally applied uniformly through two stiff and flat
hardened steel platens. As compressive stress increases the test specimen expands laterally, however,
due to friction along the interface between the platen and test specimen, lateral expansion of the

specimen is confined. This confinement of specimens by platen restraint increases apparent strength of
the material. As the distance between the platens, relative to the specimen thickness (aspect ratio),
increases the platen restraint effect reduces.
In materials that are readily cast, such as concrete and mortar, the enhancement in compressive
strength is accommodated by specifying a standard test specimen size and shape, usually cube or
cylinder. Though test results are not true (unconfined) values for compressive strength of the material,
by adopting a standard geometry comparison between different samples and specified requirements is
readily achieved. However, when testing preformed, rather than cast, specimens of varying size, such
as masonry units, the effects of specimen geometry on unit strength is not as easily accommodated.
The approaches adopted in testing of fired clay, concrete and compressed earth block testing are
discussed within the following section.
2.2 Compressive strength testing fired clay and concrete masonry units
The compressive strength of both fired clay and concrete masonry units is determined by load testing
single units in a compression testing device, in a manner similar to the testing of cast concrete and
mortar cubes. To accommodate surface unevenness units are either temporarily capped with either 3-4
mm thick plywood or similar sheeting, or capped with a thin layer of cementitious or gypsum based
mortar. Units satisfying dimensional requirements can generally be tested between temporary
cappings. Bricks containing frogs and other recesses are generally filled in with a suitable strength
mortar, though cellular and hollow units are usually tested with the voids unfilled and strength
expressed as a function of gross, rather than net, cross-sectional area. In countries, such as Australia,
where hollow concrete blocks are laid on two parallel thin beds of mortar along their faces (face shell
bedding), unit compressive strength is correspondingly determined by applying the test load through
the two face shell capping strips [17].
In countries where fired clay bricks are generally manufactured in one standard size, such as the UK
where nearly all bricks are nominally 215 x 102.5 x 65 mm, geometrical effects on apparent brick
strength are ignored as specimen geometry is uniform, as with concrete cube or cylinder testing.
Similarly in having a standard test geometry design values for material properties, expressed as a
function of concrete cylinder or unit brick strength, are readily obtained.
Concrete masonry blocks come in a much greater variety of sizes and formats (solid, cellular and
hollow). Consequently in the current British Standard for masonry [18] the effects of unit geometry are
catered for in determination of design compressive strength of concrete block masonry by expressing
values as a function of both apparent block strength and geometry. Alternatively in the draft Eurocode
for structural masonry, block strengths are normalised by applying a shape factor to account for aspect
ratio effects [19]. In Australia similar geometrical variations in both fired clay and concrete blocks are
also catered for by applying a geometrical correction factor. The empirically based aspect ratio
correction factor [20] seeks to remove the influence of platen restraint by converting test values to
unconfined strengths (defined as that achieved by specimen with an aspect ratio of at least 5). For
standard Australian fired clay brick, measuring 230 x 110 x 76 mm, the aspect ratio correction factor is
0.60, for example.
2.3 Compressive strength testing compressed earth blocks
2.3.1 Direct unit strength
The procedure adopted in many national standards and codes of practice is similar to that used for fired
clay and concrete blocks [4]. Individual units are capped and tested directly between platens. Block
surfaces are usually sufficiently flat and parallel that only thin plywood sheet capping is necessary. As
blocks are also typically solid preparation of recesses and voids is not necessary. Blocks are generally

tested in the direction in which they have been pressed which is also the direction in which they are
generally laid. Test samples generally comprises between 5 and 10 blocks.
There are a few internationally recognised standard compressed earth block sizes, such as 295 x 140 x
90 mm, corresponding to the type of block press in use. However, in general block sizes vary widely
[1,15]. The method of production, in general non-industrial, enables the manufacturer to vary block
size, and shape, to suit requirements by using mould inserts.
Geometrical effects on individual block compressive strength are generally treated in one of two ways.
In many cases standard test procedures make no attempt to correct test result for platen confinement.
Average or characteristic compressive strength is simply expressed following statistical manipulation
of individual test results [4]. In an alternative approach, used in both Australia [15,21] and New
Zealand [14], platen restraint effects are catered for by factoring test values with an aspect correction
factor. Correction factors used, Table 1, are generally the same as derived for fired clay units, though
other work has suggested alternatives believed to be more appropriate to compressed earth blocks [22].
In some cases cubes cut from solid blocks have been tested in direct compression instead. However,
comparative strength testing of blocks and cubes of same material show poor direct correlation, though
in this case cubes were pressed separately rather than cut from the blocks [13]. By testing cubes effects
of geometry on compressive strength might be readily accommodated. However, the effects of
material non-uniformity arising from the manufacturing process require further investigation.
2.3.2 RILEM test
In an attempt to directly measure unconfined compressive strength of compressed earth blocks RILEM
Technical Committee 164 has proposed the test set-up shown in figure 1 [23]. To double the
slenderness ratio of the test specimen, blocks are halved and stacked bonded using an earth mortar bed
joint. The earth mortar joint replicates masonry construction and enables even and uniform stress
transfer between stacked blocks. To enable even transfer of stress between platens and blocks the
specimens are capped with a layer of neoprene. A sheet of Teflon is also placed between the platen and
specimen at each end to minimise friction. Half blocks may be prepared following splitting strength
test, an indirect tensile strength test similar to the Brazilian test performed on concrete cylinders.
In development of this test results have been compared with those from cylinder tests of similar
material. The test seeks to replicate compressive strength developed by cylinder of aspect ratio 1.5:1,
which is seen as giving unconfined strength value [9,23]. Compressive strength test results using the
RILEM procedure have been independently checked by three research laboratories in France and
North Africa [9,12,24]. Compressive strengths using this procedure are compared with values obtained
from cylinder tests or testing half blocks (with Teflon sheeting in place to reduce friction) in figure 2.
All the CEB are made with manual press (most popular) and that is why the compressive strength is in
a range of 2-3MPa, higher values need hydraulic press or higher content of cement which is often too
expensive.
Results are from materials with and without cement stabilisation, each point representing an average of
between 2 and 13 repeat tests. On average the RILEM test under-estimates the unconfined
compressive strength of blocks or cylinders. Variation between the RILEM and cylinder test result is
in part due to variations in material dry density between the two different methods of manufacture.
The inclusion of a mortar joint in the test specimen alters the specimen format and behaviour. The test
is no longer simply on an individual masonry unit, but effectively on a simple stacked bonded masonry
prism. The mortar joint, even if made from identical material, is weaker and less stiff than the blocks,
due to higher initial moisture content and lack of compaction. In compression greater lateral expansion
of the mortar joint places the blocks in a state of compression and biaxial lateral tension [25], whereas

restraint of the blocks places the mortar joint in a state of triaxial compression. Inclusion of mortar
joint introduces a further variable into the test set-up, with performance of specimens also dependent
on the quality of work in combining half blocks and mortar joint.
2.3.3 Indirect tests
A small number of indirect compressive strength tests have been developed, primarily in order to allow
in-situ quality control testing of materials in the absence of laboratory testing facilities. The most
widely quoted indirect test methodology is the three-point bending test. Blocks are subject to single
point loading under simply supported conditions through to failure. Forces required to induce failure in
this manner are typically 80-150 times lower than that required to induce failure under uniform
compression and as such are normally quite achievable under site conditions, without resort to
sophisticated equipment. Flexural failure stress is calculated assuming pure bending (maximum
moment divided by elastic section modulus), ignoring the other potentially significant effects such as
shear and compressive membrane action (arching). Correlation between compressive and three-point
bending strength has been established experimentally by a number of workers; results show
considerable scatter but there is widely considered to be sufficient evidence to enable lower bound
prediction of compressive strength based on flexural strength [26]. Design guidelines and standards
have adopted this approach. Disadvantages of the test method include susceptibility to defects in the
blocks (shrinkage cracks). Another, less widely accepted, indirect test method is the splitting test, akin
to the Brazilian test used for concrete, in which the block is loaded in compression through two thin
steel bars along opposing faces. This induces indirect tensile stress, causing the block to split along the
line of the load. The advantage of this methodology is the greatly reduced forces required to induce
failure. Blocks from this test can also be used in the RILEM compression strength test, enabling direct
correlation between the two measured results.
3. Compressive strength characteristics of compressed earth blocks
3.1 Influence of specimen geometry
As previously discussed the geometry of test blocks has a significant influence on the value of
measured compressive strength using the standard test methodology described in section 2.3.1. The
apparent strength enhancement due to platen restraint depends on the ratio of height to thickness
(aspect ratio) of the block. As previously outlined one approach adopted is to correct measured
strength by a single aspect ratio correction factor. The distinct advantage of this approach is that it
enables a variety of different block sizes to be used, but of course it relies on accurate correction
factors.
To date, the correction factors in use were established for fired clay masonry rather than weaker and
non-uniform compressed earth blocks. Geometric effects on compressive strength of compressed earth
blocks stem not only from platen restraint, but also influence of friction during block manufacture.
Density of blocks produced using single acting ram presses is not constant, but reduces with height
away from the ram face due to friction along the mould sides. Experimental studies have confirmed
that the apparent unconfined compressive strength value is achieved when the aspect ratio reaches 5
[11,20]. However, beyond an aspect ratio of 1.5 the compressed earth block material is unlikely to be
homogeneous, due to friction during manufacture [23,27]. Though confined strengths have shown
significant scatter at lower aspect ratios, the correction factors proposed by Krefeld [20] would seem to
provide a reasonable improvement of the data.
Walker [28] has also reported on the influence of block geometry on RILEM test results. For varying
sized blocks, made from the same material, results of the RILEM test procedure do not correspond to
the results of direct block tests. Under direct (confined) compression block strength increased from 8.5
N/mm2 (aspect ratio 125/140) to 16.0 N/mm2 (45/140) despite a 3% reduction in density of the thinner

block. The experimental skew in apparent strength due to geometry is at least 88% of the measured
performance. When the same blocks were tested using the RILEM test the thinnest block produced the
least compressive strength, 2.26 N/mm2 compared to 3.14 N/mm2 for the 125 mm high block. In this
case the geometric effect is reversed (lowest strength for the thinnest block), and much reduced in
comparison with the direct strength test, with the experimental skew only around 28% of the measured
data. Geometric effects are least evident when aspect ratio correction factors are applied to the
confined direct block values, yielding strengths of 5.7 N/mm2 (125/140) and 6.4 N/mm2 (45/140)
respectively. However, considering all dry densities for these corrected data still leads to a coefficient
of variation of 26%.
The influence of block geometry on RILEM test strength is expected from classical masonry behaviour
[25], as the single mortar bed joint remained approximately 10 mm thick throughout. For thinner
blocks the 10 mm mortar joint has had a significantly greater effect on prism strength, figure 3. The
geometric effect could, perhaps, be mitigated by adjustment of mortar joint thickness in accordance
with varying block height, and warrants further investigation. It should also be noted that in practice
the variation in compressed earth block geometry is not as extreme as described above, but it is
possible to extend this work to adobe where the variation in block geometry can be even greater.
3.2 Influence of test procedure
Compressive strengths derived from differing test procedures or specimens have been compared in
experimental studies. Correlation between the RILEM test and adjusted strength values from direct
testing whole blocks is shown in figure 4. The correlation between the adjusted block strengths and
prism test results is similar to that shown in figure 2 above, though prism strengths are around 300%
lower than the corresponding adjusted block strength. Unlike the previous correlation there is no direct
parity between adjusted block strength and the RILEM test strength. This disparity might suggest,
together with the results in figure 2, that the Krefeld aspect ratio correction factors are incorrect.

3.3 Influence of dry density
Compressive strength of compressed earth blocks is strongly related to dry density achieved in
compaction. Compressive strength of individual blocks consistently increases as dry density increases,
figure 5. This relationship between strength and density has been consistently proven by test data over
the past 20 years [1]. In India block compressive strength is controlled through density [13]. Prior to
production the density and compressive strength of prototype blocks are determined in the laboratory.
Subsequently block density, for given a compactive effort, is ensured by carefully measuring, by mass,
the quantity of material added to the mould.
3.4 Influence of cement content
Cement is added to compressed earth blocks to improve durability and, indirectly, wet compressive
strength. Data produced by various researchers show strong, often linear, correlation between
compressive strength and cement content. Data shown in figure 6 is typical of the relationship between
direct compressive strength and cement content.
3.5 Influence of moisture content
Moisture content of blocks at testing has a significant influence on resultant compressive strength.
Blocks are typically tested at oven dry or ambient air dry moisture conditions, reflecting that under
service conditions. Strength reduces as moisture content increases due to the softening of binders by
water and development of pore water pressures. For plain soil, unstabilised, blocks compressive
strength when saturated is zero. Though there is some variation, depending on soil properties and
cement content, compressive strength of cement stabilised blocks following water saturation is
typically around 50% of that measured under dry conditions [10]. Moisture contents of unstabilised
materials at testing should ideally reflect in-service conditions. Testing cement stabilised blocks
following saturation allows minimum strength to be determined under easily controlled and replicable
moisture conditions, though conditions unlikely to be experienced in practice. The inclusion of mortar
joint in the RILEM test makes strength determination under saturated conditions difficult, and more
typically testing is undertaken under ambient air-dry conditions.
4. Three-point bending test
The three-point bending test has been recommended and used as a simple indirect means to measure
compressive strength of compressed earth blocks [26, 29]. Flexural modulus of rupture is determined
assuming simple, pure, bending. However, recent research has proposed alternative formulation in
recognition of the arching action that is postulated to occur in the blocks as a result of the small span to
depth ratios that inevitably occur [30]. The equivalent compressive strength σcif is given by:
elh
e
L
PL
o
cif 24
12
2



(1)
where P is failure load of the three-point bending test, L span between two roller supports, e the height
of the block, l the width of the block and ho a characteristic height, taken as 23 mm for typical sized
compressed earth blocks [30].
In both cases, classical bending [26] or with formula (1) [29], there is a linear relationship between
direct compressive strength and the strength given by the three-point bending test. Figure 7 based on
both the RILEM test procedure and that derived indirectly from three-point bending test using equation

(1). Though for only a few test results the correlation may be considered encouraging. Whereas the
correlation is poor with unconfined direct unit compression test.

5. Summary and conclusions
Compressed earth blocks are produced in a greater variety of unit sizes than many other masonry
blocks. If compressive strength is to remain a meaningful and general characteristic defining quality
and suitability of compressed earth blocks, the influence of unit geometry on performance needs to be
accommodated in a reliable and consistent manner.
To date, the most recommended compressive strength test procedure used for compressed earth blocks
undertake direct, confined, tests on single units. To accommodate geometric variation aspect ratio
correction factors, developed for fired clay masonry, have been adopted. Test results show wide
variation and suggest that at the low aspect ratios typical for most blocks, the current aspect ratio
correction factors may not be suitable. Cutting standard shaped specimens from solid blocks, such as
cubes, might be an alternative solution to this problem, though the effects of material non-uniformity
needs to be further evaluated.
RILEM Technical Committee 164 has proposed an alternative test method that represents a radical
departure for testing masonry units. Blocks are tested together with a mortar joint in a prism. Though
results have shown that the test performance is less dependent on variation in block geometry, the test
procedure may provide an unconfined masonry strength rather than block strength. The RILEM test is
also dependent on mortar performance and quality of work in preparation of the prism.
Indirect testing, such as the three-point bending test, can provide an indication of relative strength.
However, the test results are subject to considerable scatter. Recent developments in stress analysis of
three-point test performance could improve reliability of this simple test, but further work is required
to assess its generality, including the effects of material type, block geometry and method of
manufacture on performance.
In conclusion, further research work is required to investigate the influence of geometric effects on
compressive strength performance if a generalised test procedure is to be developed and widely
accepted. Direct testing of blocks needs to correlate unconfined performance with confined for a
variety of block sizes and materials. As effects such as manufacture procedure are likely to have a
significant influence, a single universal relationship may not be forthcoming. The significance of
mortar properties (materials, thickness) and block height on RILEM test needs further investigation
before the test can be universally accepted. However, as fundamentally the direct block test and
RILEM test measure two different parameters, it might be that in the future the two test procedures
will co-exist alongside each other.
References
1. Houben, H. and Guillaud, H. Earth construction: a comprehensive guide, IT Publications,
London, 1994.
2. Guillaud, H., Joffroy, T. and Odul, P. Compressed earth blocks: Volume II. Manual of design
and construction, vieweg, Eshborn, Germany, 1995.
3. Heathcote, K. An investigation into the erosion of earth walls, PhD Thesis, University of
Technology, Sydney, Australia, 2002.
4. Walker, P. Specifications for stabilised pressed earth blocks, Masonry International, 1996,
10(1), 1-6.
5. Lunt, M.G. Stabilised soil blocks for building, Overseas Building Notes, Building Research
Establishment, Garston, 1980.
6. Olivier, M. and Mesbah, A. Le matériau terre: Essai de compactage statique pour la fabrication
de briques de terres compressées, Bull. Liaison Labo. P. et Ch., 1986, 146, 37-43.
7. Heathcote, K. Compressive strength of cement stabilised pressed earth blocks, Building
Research and Information, 1991, 19 (2), 101-105.

8. Venkatarama Reddy, B.V. and Jagadish, K.S. Influence of soil composition on strength and
durability of soil-cement blocks, The Indian Concrete Journal, 1995, 69 (9).
9. Olivier, M., Mesbah, A., El Gharbi, Z. and Morel, J.C. Test method for strength test on blocks
of compressed earth, Materials & Structures, 30, November, 1997, 515-517 In French.
10. Walker, P. Strength, durability and shrinkage characteristics of cement stabilised soil blocks,
Cement & Concrete Composites, 1995, 17(4), 301-310.
11. Walker, P. Characteristics of Pressed Earth Blocks in Compression, Proc. 11th International
Brick/Block Masonry Conference, Shanghai, China, 14-16 October, 1997, p 1-10.
12. Pkla, A. Caractérisation en compression simple des blocs de terre comprimées (btc):
application aux maçonneries btc-mortier de terre, PhD, INSA, Lyon, 2002.
13. Venkatarama Reddy, B.V., Sudhakar M. Rao and Arun Kumar M.K. Characteristics of
stabilised mud blocks using ash-modified soils, The Indian Concrete Journal, February 2003,
903-911.
14. New Zealand Standard 4298:1998. Materials and workmanship for earth buildings. Standards
New Zealand.
15. Standards Australia Handbook 194. The Australian earth building handbook. Standards
Australia, Sydney, Australia, 2002.
16. Centre for the Development of Enterprise. Compressed earth blocks testing procedures, CDE,
Brussels, Belgium, 2000.
17. Standards Australia Standard 2733. Concrete masonry units, Standards Australia, Sydney,
Australia, 1984.
18. British Standard 5628-1: 1992. Code of practice for use of masonry – Part 1: Structural use of
unreinforced masonry. BSI.
19. British Standard prEN 1996-1-1:2003. Eurocode 6. Design of masonry structures. BSI.
20. Krefeld, W.J. Effect of shape of specimen on the apparent compressive strength of brick
masonry, Proceedings of the American Society of Materials, 1938, Philadelphia, USA, 363-
369.
21. Middleton, G.F. (revised by Schneider, L.M.) Earth-wall construction, Bulletin 5, CSIRO
Division of Building, Construction and Engineering, 4th Edition, Sydney, 1992.
22. Heathcote, K. and Jankulovski, E. Aspect ratio correction factors for soilcrete blocks,
Australian Civil Engineering Transactions, Institution of Engineers Australia, 1992, Vol. CE34,
4, 309-312.
23. Olivier, M. Le matériau terre, compactage, comportement, application aux structures en blocs
de terre, PhD, INSA, Lyon, 1994.
24. Hakimi, A., Yamani, A., and Ouissi, H.. Rapport: Résultats d’essais de résistance mécaniques
sur échantillons de terre comprimée, Materials & Structures, 1996, 29, 600-608.
25. Hendry, A.W.. Structural Brickwork, Macmillan, London, 1981.
26. Venkatarama Reddy, B.V. and Jagadish, K.S. Field evaluation of pressed soil-cement blocks,
in Proceedings of the 4th International Seminar on Structural Masonry for Developing
Countries, 1990, Madras, 168-175.
27. Mesbah, A., Morel, J.C. and Olivier, M. Comportement des sols fins argileux pendant un essai
de compactage statique: détermination des paramètres pertinents, Materials & Structures, 1999,
32, 687-694.
28. Walker, P. Strength and durability testing of earth blocks, in Proc. 6th International seminar on
Structural Masonry for Developing Countries, Bangalore, India, 2000, 111-118.
29. Morel, J.C. and Pkla, A. A model to measure compressive strength of compressed earth blocks
with the three point bending test, Construction & Building Materials, 2002, 16, 303-310.
30. Morel, J.C., Pkla, A. and Di benedetto, H. Interprétation en compression ou traction de l’essai
de flexion en trois points, Revue Française de Génie Civil, 2003, 7, 221-237.

Table 1. Aspect ratio correction factors
Aspect ratio
0
0.4
0.7
1.0
3.0
≥5.0
Krefeld’s correction factor
(use linear interpolation)
0
0.50
0.60
0.70
0.85
1.00
Heathcote & Jankulovski’s correction factor
(non linear)
0
0.25
0.40
0.58
0.90
1.00

Figure 1 RILEM test set-up

y = 1,1339x
R2 = 0,7332
0
1
2
3
4
0,0 1,0 2,0 3,0 4,0
RILEM prism compressive strength (N/mm2)
Compressive strength (N/mm2)
data from pkla 2002, Half block test
data from hakimi 1996, Cylinder test
data from olivier 1994, Cylinder test
Figure 2 Comparison of unconfined block or cylinder strengths and RILEM prism compressive
strength, manual compaction press.

R2 = 0.7029
R2 = 0.985
0
0.5
1
1.5
2
2.5
3
3.5
4
40 60 80 100 120 140
CEB height (mm)
Compressive Strength (Mpa)
corrected= unconfined value
procedure (a)
Figure 3 Effect of block height on RILEM test compressive strength, the correction is made
following sample aspect ratio with Heathcote data (table 1)

y = 3.0338x
R2 = 0.5957
0
2
4
6
8
10
12
14
16
18
20
0.0 1.0 2.0 3.0 4.0
RILEM compressive strength (N/mm2)
Block compressive strength (N/mm2)
Figure 4 Comparison between RILEM prism strength and direct block strength

1
2
3
4
5
1.7 1.8 1.9 2.0
Dry density (Mg/m3)
Compressive Strength (N/mm2)
soil with bentonite and 4% cement
soil with kaolinite with 4% cement
soil with kaolinite, no cement
Figure 5: Relationship between dry density and compressive strength

0
2
4
6
8
10
12
14
16
18
20
0 2 4 6 8 10 12 14
Cement content (%)
Compressive Strength (N/mm2)
Guettala 1997
Walker 2000
Figure 6 Effect of cement addition on CEB compressive strength.

R2 = 0.8193
R2 = 0.5507
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
16.0
0 2 4 6 8
Compression strength from 3 point bending
test (Mpa)
Compression strength (MPa)
Rilem test
Direct unit compression corrected
with Heathcote
Figure 7 Comparison between RILEM test and bending strength estimation of compressive
strength from formula (1)