Harakeke Reinforcement of Soil-Cement Building Materials

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 14

ARTICLE IN PRESS

Building and Environment 42 (2007) 3066–3079


www.elsevier.com/locate/buildenv

Harakeke reinforcement of soil–cement building materials:


Manufacturability and properties
Michael Segetin, Krishnan Jayaraman, Xun Xu
Department of Mechanical Engineering, The University of Auckland, Private Bag 92019, Auckland 1020, New Zealand
Received 30 June 2006; accepted 24 July 2006

Abstract

Fibre from the New Zealand flax plant, Phormium tenax, or Harakeke as it is otherwise known in Maori, has been used to reinforce
soil–cement composites in an attempt to improve the strength and ductility of the composite material. Previous investigations have found
the interfacial bond strength between the harakeke fibre and the soil–cement matrix to be an important factor for the strength of the
composite. In an effort to improve the interfacial bond strength, an enamel paint coating has been applied to the fibre surface. Fibre
lengths of 70 and 85 mm have been investigated along with fibre content levels of 0.6% and 0.8% measured as a percentage of the dry
mass of the soil in the composite. Significant improvement in the ductility of the soil–cement composite has been realised with the
addition of the harakeke fibre-reinforcement. It has been found that the specific make-up of the composite with regard to fibre length and
content is not critical for the achievement of ductility improvement. Instead, these parameters were found to be controlled more by
manufacturing issues such as the workability of the composite. It was found that 85 mm represented an upper limit on fibre length for the
manual mixing technique adopted, while a maximum fibre content level of 0.8% was achieved. In order to reduce manufacturing
difficulties, different mixing techniques were studied. A successful tumble mixing technique has been identified which is able to improve
composite uniformity and the ease of manufacture. Based on the similarity of results for the two fibre content levels investigated, it is
recommended that a level of 0.6% be adopted as this both improves the ease of manufacture and minimises the amount of fibre required.
When the fibre content level was dropped below 0.6%, it was found that the material exhibited a more brittle failure behaviour.
r 2006 Elsevier Ltd. All rights reserved.

Keywords: Flax fibre; Harakeke; Soil–cement; Strength; Ductility

1. Introduction The extent of earth construction in the developed world


may actually be a lot greater than one might expect. For
Earth has been used as a construction material for example it has been estimated that in Australia approxi-
thousands of years [1], with its most primitive form dating mately 20% of the new building market is occupied with
back to the hunter–gatherer period where humans lived a earth based construction projects [4]. The prevalence of
migratory life and often constructed temporary, seasonal earth as a construction material may be attributed to its
shelters from brush and wooden members which they then proven durability which is evidenced by the fact that
covered with mud [2]. Due to its many benefits earth is still many of the oldest buildings still standing are made from
used extensively today [1–5] and it has been estimated that earth [6].
over half of the world’s current population reside in
earthen based homes [2]. Although many of these people
live in developing nations where other materials may be
1.1. Soil stabilisation and reinforcement
unavailable or simply deemed too expensive, the use of
earth in construction is not limited to the developing world.
Soil on its own can be used to produce adequate housing;
however most often some form of stabiliser is added to the
Corresponding author. Tel.: +649 373 7599; fax: +649 373 7479. mixture to enhance the properties of the material. There
E-mail address: k.jayaraman@auckland.ac.nz (K. Jayaraman). are two main techniques used to stabilise earth for

0360-1323/$ - see front matter r 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.buildenv.2006.07.033
ARTICLE IN PRESS
M. Segetin et al. / Building and Environment 42 (2007) 3066–3079 3067

construction purposes: chemical stabilisation and fibre other soil–cement composites may utilise entire bamboo
reinforcement. stalks as a form of fibre reinforcement [13]. Steel fibre-
reinforcements found in concrete structures are also used
1.1.1. Chemical stabilisation for the reinforcement of soil–cement composites [3,22,23].
The most popular form of chemical stabilisation is the Steel fibres can themselves be found in a variety of forms,
addition of cement to a soil. It has been shown in many the most common of which is the standard reinforcing bar.
studies [7–10] that the addition of ordinary Portland Other forms of steel fibre reinforcement include mesh and
cement to earth building materials can improve their various geometries of shorter fibre inclusions [22].
mechanical properties. As soil for earth construction is
often obtained from the building site or nearby [7,11–13],
1.1.4. Fibre-reinforcement
there will always be some variability in the amount of
As the focus of the present study is on natural fibres it is
cement needed to sufficiently improve the strength of the
important that their interaction with the soil–cement
material. It has generally been found [3,7,8,10,12,14,15]
matrix be understood. Adhesion at the fibre–matrix inter-
that the amount of cement required to be added to a soil is
face has been found to be governed by the following three
in the range of 3–10% by dry weight of soil, with a higher
factors [18]:
cement content tending to give greater strength to the
composite. It has also been suggested that cement contents
in excess of 7.5% may increase costs too much for low-cost (a) The shear resistance of the soil due to the surface form
housing projects [7]. It is also known that the strength of and roughness of the fibre.
the earthen material increases with the curing time allowed (b) The compressive friction forces on the surface of the
for the cement and for the cement range identified above fibre due to shrinkage of the soil.
compressive strength values obtained after 7 days of cure (c) The cohesive properties of the soil.
only reach 70% of that observed after 21 or 28 days [14]. It
is therefore common practice for soil–cement materials to Further, each of these three factors is affected by
be cured for at least 21 days prior to their use or testing dimensional changes of the natural fibre which can occur
[1,10–12,14,16,17]. Along with the improvements in com- due to changes in moisture and temperature [18]. Such
pressive strength associated with the addition of cement, an changes in fibre dimension can occur during the curing
increase in soil cohesiveness is often found [12,15,18]. It can stage of the earth–fibre composite material and this then
therefore be assumed that the compressive strength of a determines a possible mechanism resulting in a poor
soil–cement composite is related to its cohesiveness. interfacial bond. During the mixing and drying stages of
production, the hydrophilic nature of the natural fibre can
1.1.2. Fibre-reinforcement lead it to absorb water and effectively push out on the
From the beginnings of earthen construction, fibre- soil–cement matrix. Then near the end of the curing
reinforcement has been used to improve the properties of (drying) period the fibre loses the water it has absorbed,
this versatile material [2,5]. Fibre-reinforcement can take causing it to shrink back. Because the matrix is now set, a
on many forms from the introduction of large fibres such void is formed around the periphery of the fibre and a
as steel or bamboo [13], also commonly found in concrete weakened interfacial bond can result [10,13,18]. This
construction, to much smaller ones such as sisal and flax mechanism is illustrated in Fig. 1.
[17–19]. It has been reported that strength, toughness and Therefore, in looking for an additive to improve the
ductility all improve with fibre content up to 3% by mass bonding characteristics of fibre-reinforced soil–cement
of soil [10]. It is also widely accepted that the greatest effect composites, it is required that the treatment reduce the
fibre addition has on a soil–cement composite is to give a transfer of water between the matrix and the fibre. Possible
vast improvement in ductility [3,8–10,18]. This improve- additives identified from the literature include water
ment in ductility can be explained by the way that the fibres resistant coatings such as asphalt emulsion, rosin–alcohol
act to suppress the development of cracks within the mixture, paints [13], bituminous materials [13,17,18], a
soil–cement matrix [3,10]. In some situations the gains in water soluble acrylic and a polystyrene coating [17].
ductility can have quite significant impacts. For example, Khazanchi et al. [17] with their study of wall panels
in the case of a significant earthquake the extra ductility consisting of soil, 2.5% cement and polystyrene coated
provided by fibre-reinforcement could provide the vital wheat straw, reported a fibre bond strength of 1.3 MPa.
extra seconds between getting out alive or being trapped With the same material but using bitumen and acrylic
inside when the building falls down [20]. coatings, they also achieved bond strengths of 0.1–0.12 and
0.08–0.125 MPa, respectively. In this study untreated fibres
1.1.3. Fibre-reinforcement gave bond strengths in the range of 0.07–0.08 MPa.
Fibre-reinforcement of soil–cement composites can take In addition to bonding agents, the length of the fibre also
on a number of forms with fibre lengths and diameters plays a role in the interfacial bond strength. The longer a
varying over a large range. For example, flax fibres have an fibre is, the greater a surface area it has to form a bond with
average diameter of approximately 23-mm [21] whereas the matrix material. For sisal and flax fibres, which can be
ARTICLE IN PRESS
3068 M. Segetin et al. / Building and Environment 42 (2007) 3066–3079

Fig. 1. Schematic of the effect of fibre deformation due to moisture changes.

deemed to be comparable, it has been found that fibre fibre-reinforcement in the composite will further enhance
lengths in the range of 60-85 mm are preferable [3,10]. local Maori acceptance of the project as Harakeke is a
plant that is of particular significance to their culture due to
1.2. Mixing of fibre-reinforced soil–cement composites the many uses they have found for it throughout their time
in this country.
The mixing of natural fibres through soil–cement In addition to having cultural significance to the local
composites is not well discussed in the literature. Some community, the use of flax reinforced earth as a housing
information is provided by Allen [24] that folding fibres material also presents the environmental benefits asso-
through a soil matrix is the most effective method ciated with the use of natural materials. Earthen building
of mixing. This can be done with the use of a front- materials are also able to regulate heat well, keeping the
end-loader, bobcat or similar device with a bucket interior climate comfortable and minimising the need for
attachment [24]. additional heating and cooling of the living space.
The lack of available information relating to the mixing
of natural fibres in soil–cement composites has motivated 3. Experimental details
the investigation into different mixing methods in the
current research. 3.1. Materials

2. Scope of the present research The soil used in this investigation was sourced from a
local quarry and can be classified as a very silty SAND
The work presented in this paper focuses on the containing approximately 20% silt and 77% sand particles.
development of soil–cement building materials appropriate Fibres from the New Zealand Flax plant Phormium tenax
for the construction of habitable dwellings within New were used in this study and were extracted from the leaf of
Zealand. In particular the use of fibre extracted from the the harakeke plant by means of a mechanical stripping
New Zealand flax plant Phormium tenax, or Harakeke as it device. Ordinary Portland cement was used for the
is traditionally known to Maori, is investigated as a means stabilisation of the soil. The paint used for coating the
of improving the mechanical properties of the soil–cement flax fibres was an enamel paint which was diluted using
composite. mineral turpentine.
The current research is part of an ongoing investigation
within the University of Auckland, aimed at developing 3.2. Fibre mixing
low cost housing through the use of earthen-based
construction technology. The primary focus of the The mixing of flax fibres through a soil–cement matrix is
investigation is the provision of new technologies which a problematic task which requires much attention in order
can be used to develop Whareuku (earth based buildings) to make flax fibre-reinforced soil–cement composite
on Papakainga (communally owned Maori land) by panels a viable construction material. The major area of
making use of local resources and traditional materials. It concern is the tangling of fibres, which often makes it very
is intended that the local community be involved as much difficult to obtain a homogenous mixture. If adequate
as possible in the construction process, giving them the mixing techniques cannot be developed, large scale
opportunity to acquire the skills necessary for the production of fibre-reinforced soil–cement structures will
construction of further housing. This community involve- not be feasible.
ment is a key element for the success of the project as it will There are two avenues which can be taken when
instil a level of pride in the housing produced, ensuring that investigating the mixing of fibre through a soil–cement
the developed construction materials and techniques will composite. Fibres can either be mixed through the
continue to be used in the future. The selection of flax as a soil–cement matrix material manually or a mechanical
ARTICLE IN PRESS
M. Segetin et al. / Building and Environment 42 (2007) 3066–3079 3069

means of mixing can be used. Both manual and mechanical 3.2.2. Mechanical mixing
mixing techniques were investigated in the current study There are a number of considerations which need to be
and details of the range of tests conducted are presented taken into account when searching for appropriate
below. mechanical mixing techniques. The selected technique must
be able to give a homogenous mix to the composite,
3.2.1. Manual mixing ensuring that fibres are evenly dispersed throughout the
Manual mixing of flax fibres is by far the simplest matrix material and are not clumping together. Fibre
method and can be used to produce a high-quality mixture damage is also a key concern as this could lead to
with good fibre dispersion. This method of mixing involves premature fibre failure and inefficient use of the material.
simply taking the soil and cement in a container and mixing Three different mechanical mixing techniques have been
it with hands and a trowel, to obtain an even mixture as trialled in the current study, namely a cultivator or rotary-
shown in Fig. 2a. Fibres are then added in small amounts hoe device, a concrete mixer and a tumble mixer.
and mixed into the matrix material until the desired (i) Cultivator or rotary-hoe: A mechanical cultivator
content level is reached. Although this task is simple, it is device (usually used for aerating soils before the planting of
very labour intensive; average mixing times for test seeds) was trialled in an effort to ease the task of mixing
specimens produced by this method would lie in the range flax fibres into the soil–cement composite. Soil was spread
of 40–50 min. If this method were to be used to produce onto the laboratory floor for the cultivator to be pushed
larger structures, many workers would be required in through. The constraints of the testing program restricted
order to produce the volume of material required in a usage to the production of only two flexural specimens and
short enough time span so as to prevent the water content this presented some problems given the large size of a small
of the soil from dropping significantly during the mixing cultivator. Trying to push the cultivator over the spread
process. soil proved to be a cumbersome task with the relatively

Fig. 2. Mixing methods trialled; (a) manual mixing, (b) cultivator mixing, (c) concrete mixer, (d) tumble mixer.
ARTICLE IN PRESS
3070 M. Segetin et al. / Building and Environment 42 (2007) 3066–3079

small amount of soil and the mix was sprayed out behind during baking. On a large scale this can be done effectively
the cultivator meaning it would need to be swept up and re- with the use of a front-end-loader or a bobcat with a
gathered for mixing to continue. In order to overcome this bucket attachment [24]. This type of mixing technique
problem, the machine was mounted in one place with the could also be extended to a manual method by making use
blades just off the ground. A small box was then built of a shovel or similar implement.
around the cultivator to contain debris and a chute was A tumble mixing technique was developed making use of
created to feed the mixture into the blades of the device. a small manual clothes washing device as shown in Fig. 2d.
This setup can be seen in Fig. 2b. Although this setup did This technique follows the basic principle of folding the
not reflect accurately how the cultivator device would flax fibre soil–cement mixture which is known to be
behave when used on a larger amount of soil, it did give an effective [24]. This device was only capable of processing
indication as to the effectiveness of such a device at mixing a small amount of fibre-reinforced soil–cement; however
flax fibres into a soil–cement matrix. the concept could easily be extended to a larger scale device
The device did not perform the mixing task as well as if the process proved to be successful.
intended, fibres tending to be separated from the soil after Initial testing of the tumble mixing device revealed that
being fed through a number of times. This seemed to only this device could be used to quickly and easily obtain a
occur when fibres became dirtier and could suggest that a mixture with good fibre dispersion. The relative success of
state of over-mixing was occurring. Over-mixing is known the tumble mixing device over other mixing techniques has
to be an issue which can lead to non-homogeneous justified further, more detailed investigation. It has also
mixtures with fibres tending to separate from the soil been found through experience in earlier investigations that
[24]. The separating fibres congregated in front of the fibres would separate more easily from each other once
blades, while the soil that was thrown behind the blades did they had become slightly dirtied by the soil–cement matrix
appear to have an even mixture of fibres through it. The material. This finding contradicts reports that fibre is best
overall failure of this method could be attributed to the mixed through a soil when both soil and fibre are in a dry
way in which the machine was used. There is reason to state [24]. Further investigation of this observation was
believe that on a larger scale, with fibres spread evenly over therefore incorporated into the investigation of the tumble
a larger area of soil, the cultivator device would be able to mixing concept, by attempting to mix fibres in a slurry of
mix the fibres through the soil. No tangling of fibres soil and water before the remainder of the composite
around the cultivator blades was noted and the blades did constituents were added.
not appear to cut or damage the fibres. Due to the quality of the results of a mixing trial being
(ii) Concrete mixer: A concrete mixer device (Fig. 2c) subjective, this investigation was conducted on a trial and
with four mixing vanes was used in several trials to produce error basis where successful trials were built upon and
successful mixtures which were then used to produce unsuccessful experiments were documented and discarded.
flexural specimens. In general however, fibres showed a A total of eight different trials were run with different
tendency to ‘float’ above the soil rather than mix through methods of adding the composite constituents being
it; again this is an indication that over-mixing has occurred. examined. The quantities of the composite constituents
The fibres were also observed to tangle around the mixer used for these experiments are outlined in Table 1.
vanes which required regular clearing. There are several A detailed investigation into the use of a tumble mixing
factors that may be contributing to the poor performance device has revealed a mixing technique that would be
of the concrete mixer. The drum of the mixer was on a high suitable for adoption on a larger scale. This method
angle and this may be contributing to the soil staying at the involved pre-mixing the cement through a dried soil in the
bottom of the drum. The fibres are also light in comparison tumble mixing device. A slurry was then made with a small
to the soil and so do not tend to get dragged down into the amount of this mixture and the full amount of water
soil at the bottom of the mixer drum. For this device to required for the formation of the composite; the water
work more effectively modification to the mixing vanes content of this slurry mixture was approximately 350%
may also be necessary to create more of a folding action (measured as a percentage of the dry mass of the soil in the
and reduce tangling problems. slurry). Fibre was mixed through this slurry before the
It should also be noted that fibres treated with paint were remaining soil–cement mixture was added as a single
found to mix more easily in the concrete mixer than those
that were left untreated. This is thought to be due to the Table 1
stiffening effect that the paint has on the fibre. Stiffer fibres Composite constituent levels for tumble mixing trials
do not appear to wrap around each other and tangle as
Constituent Content level Mass (g)
easily. This finding may help to highlight potential fibre
(% of dry mass of soil)
treatments which could be used to improve the mixing
behaviour of the fibres in the future. Soil (dry mass) 100% 2000
(iii) Tumble mixer: It has been reported [24] that the Cement 10% 200
Fibre 0.7% 14
most effective method of mixing fibre through a soil matrix
Water 26% 520
is to fold it in the same manner that ingredients are folded
ARTICLE IN PRESS
M. Segetin et al. / Building and Environment 42 (2007) 3066–3079 3071

addition and mixed further in the tumble mixer for a short into a large open container. Fibres were then added in large
period of time. bundles to the paint mixture, soaked thoroughly and then
One detriment of using a slurry technique for mixing is the excess paint was squeezed off before placing them in a
that there is a requirement for the soil being used to be drying basket. Coated fibres were then left to dry for a
dried to a level well below its optimum water content period of 2–3 days before being used for specimen
before the slurry is made. This should not hinder the production.
adoption of this method however as it is an issue which can This method of coating fibres, although effective, was
easily be managed in the planning phase of a construction not ideal as it resulted in an excessively heavy coating and
project. fibres were found to stick together after coating. The
Although a successful method for mixing fibre through coating work also involved spending extended periods
the soil–cement matrix has been found using the tumble crouched over the coating container in the presence of toxic
mixing technique, the only available apparatus for doing fumes. For future use a quicker method of application that
this for the current study was not of sufficient size for the gives a finer coating will be required. This will likely come
production of flexural specimens. This meant that the in the form of a spray on coating method, which would
mixing of the test specimens had to revert back to a manual have been investigated for the current study however the
hand mixing technique which required high labour and required facilities were not available. It should be noted
time inputs but could reliably produce a high quality that a spray-can method of coating used in an initial
mixture with good fibre dispersion. coating feasibility investigation produced a good quality
coating without fibres sticking together in large bundles.
3.3. Specimen preparation This method of using canned spray-paint would not be
practical on a large scale however due to the expense
3.3.1. Specimen designation involved. A similar method using commercial paint
In order to keep track of the variables used for each spraying techniques may be more viable.
specimen, a meaningful specimen naming method had to be
adopted which could quickly and easily give full informa- 3.3.3. Specimen compaction
tion about the composition of a specimen without the need Each specimen was compacted in two layers of equal size
for referral to experimental notes. As there were three using a heavy wooden mould and a 21 kg steel hand-
factors being investigated, fibre coating, fibre length and tamper with a 100 mm  100 mm square face. The
fibre content level, it was decided that a three character composite mixture was spread into the mould taking care
name would be assigned to each specimen followed by a to ensure that large clumps of fibre were not present. Once
number representing which repeat of the experiment a half of the mixture had been transferred to the mould it
specimen belonged to. The first character of the specimen was levelled out and pressed firm by hand before
name could take the value of ‘‘C’’ for a coated fibre compaction with the tamper. The soil was tamped
specimen or ‘‘N’’ for a non-coated fibre specimen. The continuously until a ringing sound could be heard as the
second character of the specimen name related to the tamper rebounded from the surface of the specimen. Care
length of the fibres with ‘‘S’’ being used for short, 70 mm was taken to ensure that an even amount of compaction
fibres and ‘‘L’’ being used for the longer 85 mm fibres. was given across the entire surface of the specimen. Once
Fibre content level was identified in the third character of compaction of both layers of the specimen was completed
the specimen name as being either low, ‘‘L’’ (0.6%) or high, the mould was dismantled and the specimen was carefully
‘‘H’’ (0.8%). Using this convention a specimen named removed.
CSH_03 would then be identified easily as a coated fibre
specimen containing 70 mm fibres at a content level of 3.3.4. Specimen cure and drying
0.8% and belonging to the third experimental run. By this Upon removal from the mould specimens were bagged,
method the nine different specimen combinations are then labelled and left to cure for a period of 3 days out of direct
NLH, NLL, NSH, NSL, CLH, CLL, CSH, CSL and NIL, sunlight. This curing period was required to ensure that the
where NIL refers to a control specimen containing no fibre- specimen did not dry too quickly as this can often result in
reinforcement. cracking. Following this cure period the bags were opened
and the brick was left to dry for a further period of 18 days
3.3.2. Fibre coating before testing could commence, giving a total period of at
In order to limit the transfer of moisture between the least 21 days between the time of specimen manufacture
harakeke fibres and the soil–cement matrix, fibres were and the time of testing.
coated with an enamel paint using an immersion method. It
was hoped that by controlling moisture transfer this 3.4. Testing
coating may be able to improve the interfacial bond
strength between the fibre and the soil–cement matrix by Flexure and compressive testing followed testing stan-
limiting dimensional changes in the fibre. The paint was dards set out by the American Society for Testing and
diluted with turpentine to a watery consistency and poured Materials (ASTM). The test standards used were ASTM
ARTICLE IN PRESS
3072 M. Segetin et al. / Building and Environment 42 (2007) 3066–3079

D1635-00 [25] for flexure and ASTM D1634-00 [26] for specimens. This means that each specimen produced can be
compression. Specimen size was determined in accordance used for both flexural and compressive testing, saving both
with fibre-reinforced concrete standard ASTM C1018-97 time and raw materials.
[27]. This called for a flexural specimen width and depth of Compressive specimens were cut from the ends of the
at least three times the maximum fibre length and a broken specimens used for flexural testing to a length of
specimen length of at least 50 mm greater than three times 175 mm, this being 25 mm longer than their 150 mm depth
the depth. For practical purposes however, the width and as required. Care was also taken to select specimens which
depth of the specimens were kept at 150 mm, this is were relatively free from cracks, chipped surfaces and other
permitted in the fibre-reinforced concrete standard ASTM obvious defects such as excessive amounts of fibre
C1018-97 for fibre lengths up to 75 mm. A specimen length clumping.
of 600 mm was chosen for the flexural specimens. Specimens were tested on their side with respect to their
moulded position. They were placed in a central location
3.4.1. Flexural testing on the testing machine (Grade A Contest Concrete Test
Flexural testing standard ASTM D1635-00 [25] makes Machine). The load was applied continuously at a rate of
use of simple soil–cement beams tested on their side relative 140 kPa/s in accordance with testing standard ASTM
to their moulded position. The beams were loaded under D1634-00 [26] and care was taken to avoid any shock
third-point loading conditions with the lower supports loading.
placed apart at a distance of 450 mm, corresponding to
three times the depth of the beam. The flexural testing 4. Results and discussions
setup can be seen in Fig. 3. An Instron Universal Testing
Machine, equipped with a HP 3852A data acquisition and 4.1. Flexural strength
control unit was used for flexural testing, with a crosshead
advance rate of 0.02 mm/s. It is a requirement of the New Zealand Earth Building
One of the major effects of including fibre into a standard NZS 4298 [28] that a flexural strength exceeding
soil–cement composite is the improvement given in 0.25 MPa be achieved for any earth building material. This
ductility [3,8–10,18]. Therefore the ductility of the flexural means that this level of flexural strength must be achieved
specimens was of interest in the current study. To obtain a by the composite technology being developed in the current
measure of the ductility of the specimens, the displacement study for it to be used in this country.
of the crosshead of the testing machine was measured and
plotted against the flexural stress of the specimen. 4.1.1. Specimens without fibre reinforcement
Although this does not give a measure of the strain All control specimens created without fibre reinforce-
experienced by the specimen, it does give an indication of ment were found to show a similar stress development
ductility which can be used for comparison. profile when tested, with a maximum flexural strength of
0.427 MPa being attained with specimen NIL_11. This
level of flexural strength exceeds that required by NZS 4298
3.4.2. Compression testing
[28] however it can be seen from Fig. 4 that this specimen
The compressive strength of specimens was determined
reached an unusually high strength level. Specimens
following the compressive testing standard ASTM D1634-00
NIL_08 and NIL_10 appeared to give more consistent
[26]. This standard makes use of portions of broken flexural
results however giving strengths of 0.258 and 0.26 MPa,
beams as tested under ASTM D1645-00 [25] for compressive
respectively. These results are comparable with those
obtained in an earlier study [29] and although they do
meet the requirements of the building standard, they do so
only marginally and so it is required that an improvement
to the strength of the soil–cement composite be made. It
can also be seen from Fig. 4, that these non-reinforced
specimens did not exhibit ductile behaviour. This was also
noted during testing in the catastrophic nature by which
these specimens failed.

4.1.2. Specimens with fibre reinforcement


Fig. 5 shows that in general specimens produced with
non-coated fibre-reinforcement showed good recovery after
first-cracks occurred. This finding indicates that loading
was being successfully transferred to the fibre reinforce-
ment during testing and that ductility had definitely been
improved. This improvement in ductility is further
Fig. 3. Flexural testing setup. reinforced by the fact that crosshead displacements of
ARTICLE IN PRESS
M. Segetin et al. / Building and Environment 42 (2007) 3066–3079 3073

Fig. 4. Flexural strength comparison for specimens with no fibre-reinforcement.

Fig. 5. Flexural strength comparison for non-coated, fibre-reinforced specimens.

5–7 mm were reached before the composites strength fell 4.1.3. Specimens with coated fibre reinforcement
back below the 0.25 MPa level required by the NZS 4298 For the coated fibre specimens displayed in Fig. 6, it can
[28]. It was also observed that with the exception of be seen that although all of the specimens except CSH_02
specimen NLH_03, all specimens reached strength levels in and CSL_03 managed to exceed the 0.25 MPa requirement
the range of 0.35–0.5 MPa, comfortably exceeding the with their first-crack strength, only specimens CLH_01 and
0.25 MPa minimum requirement. Upon comparison with CSH_03 showed enough recovery strength to reach this
the other non-coated fibre-reinforced specimens, specimen level again. It must also be noted that the recovery strength
NLH_03 can be considered to be an outlier and of specimen CLH_01 only marginally exceeded the
disregarded. 0.25 MPa requirement. The addition of coated fibres has
ARTICLE IN PRESS
3074 M. Segetin et al. / Building and Environment 42 (2007) 3066–3079

Fig. 6. Flexural strength comparison for all coated, fibre-reinforced specimens.

also improved the ductility of the composite, but not to the the required strength improvements and a virgin fibre being
same degree as the addition of non-coated fibres. In the used to provide the much desired improvements in
case of the coated fibres the ultimate recovery strength ductility.
generally occurred in the ductility range of 1.3–2.4 mm of
crosshead displacement with the greatest amount of 4.1.5. Effect of fibre coating
ductility at the 0.25 MPa minimum strength requirement The differences noted above between coated and non-
level being recorded for specimen CSH_03 as 2.5 mm. coated fibre-reinforced specimens should be treated with
A large drop in strength was also noted for all coated fibre caution. Late in the testing programme it was realised that
specimens at the first-crack point. This indicates a slightly the fibre coating method being used was applying a layer of
more brittle nature for this material than what was found paint that was too thick. This resulted in a misrepresenta-
with the non-coated fibre-reinforced specimens. tion of the fibre content levels used for the coated fibre
specimens. It has been determined that for the coated fibre
4.1.4. Use of Duracem cement specimens the mass of the paint attached to the fibres made
Inspection of the flexural strength comparison plots will up 47% of the total measured mass of the fibres. This
reveal that a number of the experimental specimens are not finding means that the actual fibre content levels for the
represented. The results obtained from these specimens low and high fibre content coated fibre specimens were
have been omitted as they were produced using a different 0.32% and 0.43%, respectively, approximately half of the
type of cement to the rest of the specimens. The cement fibre content that was intended. Specimen *NSH_00 was
used was a product known as Duracem and is designed for therefore created with a fibre content level of 0.43% to
use in harsh environments. The specimens produced with allow a meaningful comparison between a non-coated
this cement were found to have both flexural and fibre-reinforced specimen and its corresponding coated
compressive strength values greatly exceeding those ob- fibre specimens. This comparison is represented in Fig. 8
tained with ordinary Portland cement. The Duracem where it can be seen that the coated fibre specimens
product has a price premium of 5–15% over ordinary performed comparatively well. Specimen *NSH_00 dis-
Portland cement [30]. However the gains obtained in played the greatest first-crack strength, 0.332 MPa, of the
composite strength, which can be seen for specimen three specimens portrayed in the comparison plot. How-
NSL_01 in Fig. 7, may mean that this product is worth ever, with a recovery strength of 0.293 MPa specimen
investigating in future trials. The gain noted in strength for CSH_03 was the only one of the three specimens to show a
specimen NSL_01 in Fig. 7 is typical of that shown for all recovery strength exceeding the minimum requirement of
specimens produced with Duracem cement. The gains in 0.25 MPa. Although only one non-coated fibre-reinforced
strength obtained through the use of this product may also specimen has been produced at this fibre content level,
allow for the exclusion of fibre treatments meaning that a these findings indicate that the addition of paint to the fibre
composite could be produced with the cement providing surface has not had a detrimental effect on composite
ARTICLE IN PRESS
M. Segetin et al. / Building and Environment 42 (2007) 3066–3079 3075

Fig. 7. Typical effect on flexural strength of use of Duracem cement product.

Fig. 8. Flexural strength comparison for adjusted fibre content specimen.

strength as may otherwise have been assumed upon differences in fibre length and content over the ranges
examination of the main experimental results. tested are not likely to have much impact on the strength
Unfortunately due to the lack of experimental specimens performance of the composite material. Fibres at these
no further information about the interaction between fibre lengths have been shown to be pulling out of the
length and content levels can be provided from the testing soil–cement matrix at the failure surface for both coated
information gathered. From general observations made and non-coated fibre-reinforced specimens as can be seen
during manufacture and testing of flax fibre-reinforced in Fig. 9. With 85 mm being found to represent a limit to
soil–cement specimens however it has been noted that fibre length based on mixing requirements, the above
ARTICLE IN PRESS
3076 M. Segetin et al. / Building and Environment 42 (2007) 3066–3079

findings suggest that fibre length will be determined by It is clear from Fig. 10 that only specimens CSL_03 and
manufacturing constraints rather than its effect on NIL_11 fall below this value, each with a compressive
composite strength. strength of 1.26 MPa, which is very close to the minimum
requirement of 1.3 MPa. It can also be shown that the
4.2. Compressive strength average compressive strength taken over all of the speci-
mens trialled is 2.14 MPa, comfortably exceeding the
In order for the composite technology to be adopted in requirements of the earth building standard NZS 4298 [28].
New Zealand it must meet the requirements of the New The specimens without fibre-reinforcement have an
Zealand Earth Building standard NZS 4298 [28] which average compressive strength of 1.79 MPa with a standard
state that soil used for building purposes must have a deviation of 0.5 MPa, while the coated and non-coated
compressive strength of at least 1.3 MPa. fibre-reinforced specimens have compressive strengths of

Fig. 9. Fibres shown to be pulling out of matrix material for (a) coated and (b) non-coated fibre-reinforced specimens.

4.0

3.5

3.0
Compressive Strength (MPa)

2.5

2.0

1.5

1.0

0.5

0.0
NIL_08

NIL_09

NIL_10

NIL_11

NLH_02

NLH_03

NLL_03

NSH_02

NSH_03

NSL_02

NSL_03

CLH_01

CLH_02

CLH_03

CLL_02

CLL_03

CSH_02

CSH_03

CSL_01

CSL_02

CSL_03

*NSH_00

Specimen ID

Fig. 10. Compressive strengths for soil–cement specimens using ordinary Portland cement.
ARTICLE IN PRESS
M. Segetin et al. / Building and Environment 42 (2007) 3066–3079 3077

2.0 and 2.49 MPa, respectively, with standard deviations of dispersion. On a large scale however they would not be
0.43 and 0.5 MPa. The standard deviations reported here suitable as they require a large labour input and are too
are relatively high compared to the average values and this time consuming. It is for these reasons that mechanical
can be attributed to the low number of experimental methods of mixing were investigated. It has been found
specimens produced. The average compressive strengths that a tumble mixing device is the most suitable solution,
are all seen to exceed the minimum requirement of 1.3 MPa offering a good quality mixture for a minimal time and
set by the earth building standard NZS 4298 [28] indicating effort input. A suitable method of adding the composite
that in terms of compressive strength the material may be constituents to a tumble mixing device has also been
suitable for building construction. It is also interesting to identified which involves mixing fibres through a slurry of
note that the compressive strengths for both the coated and pre-mixed soil and cement before adding the remainder of
non-coated fibre-reinforced specimens are greater than for the mixture.
the non-reinforced specimens. This result was not expected The immersion method used for applying the enamel
as the fibre-reinforcement is primarily intended to enhance paint coating to fibres has proved to be inadequate with
tensile or flexural behaviour and ductility. Due to the large excessive amounts of paint being added to the fibres. It is
standard deviations involved however these results can suggested that if work is to continue with the investigation
only be considered to be indicative and further experi- of paint as a fibre treatment, a new method of application
mentation is required to provide conclusive findings. will need to be used. A spray on method will most likely
The dramatic improvement in the compressive strength give the best results as a preliminary investigation into fibre
of the soil–cement composite provided by the use of the coating treatments found that this type of method gave a
Duracem cement product can be seen in Fig. 11. The use of good quality coating.
this product has increased the average compressive The investigation into the improvement of flax fibre-
strength from 2.14 to 4.55 MPa, easily exceeding the reinforced soil–cement composites has found that the
required 1.3 MPa of the New Zealand Earth Building addition of fibre to a soil–cement matrix can significantly
standard NZS 4298 [28]. enhance the ductility of the material and eliminate the
catastrophic failure pattern displayed by specimens without
5. Conclusions fibre-reinforcement.
Through the testing of control specimens it has also been
Investigations into the mixing of flax fibres through a found that the flexural strength of non-reinforced soil–
soil–cement matrix material have found that hand mixing cement composites is at present insufficient. Although the
techniques can give a high quality mixture with good fibre specimens produced in the present research managed to

6.0

5.0
Compressive Strength (MPa)

4.0

3.0

2.0

1.0

0.0
H_01NS
NLH_01

L_01NS
NLL_01

NLL_02

CLL_01
NIL_01

NIL_02

NIL_03

NIL_04

NIL_05

NIL_06

NIL_07

Specimen ID

Fig. 11. Compressive strengths for soil–cement specimens using Duracem cement.
ARTICLE IN PRESS
3078 M. Segetin et al. / Building and Environment 42 (2007) 3066–3079

exceed the 0.25 MPa requirement of the New Zealand earth References
building standard NZS 4298 [28], they did so only
marginally and due to the low volume of specimens tested [1] Hall M, Djerbib Y. Rammed earth sample production: context,
they cannot be relied upon to be consistently able to meet recommendations and consistency. Construction and Building
Materials 2004;18(4):281–6.
the required strength. [2] McHenry PG. Adobe and Rammed Earth Buildings: Design and
Fibre pull-out was still witnessed in flexural testing Construction. New York: Wiley; 1984.
specimens with coated fibres. This suggests that an [3] Ali HK. Sisal Fibre Reinforced Soil Cement. Master of Engineering
insufficient interfacial bond still exists between the fibre Studies Project Report, The University of Auckland, 1998.
surface and the soil–cement matrix, indicating that the [4] Easton D. The Rammed Earth House. White River Junction: Chelsea
Green Publishing Company; 1996.
paint coating is not offering any significant improvement in
[5] Houben H, Guillaud H. Earth Construction: A Comprehensive
bonding. Guide. London: Intermediate Technology Publications; 1994.
The 85 mm fibre length represents an upper length limit [6] King PE. Buildings of Earth and Straw: Structural Design for
with regard to the mixing of fibre and the findings that Rammed Earth and Straw-Bale Architecture. California: Ecological
ductility was improved for the 70 mm fibres and fibre pull- Design Press; 1996.
[7] Akinmusuru JO. Lateritic soil–cement bricks for rural housing. The
out was noted for both fibre lengths suggests that fibre
International Journal of Cement Composites and Lightweight
length is not a critical factor in the reinforcement of Concrete 1984;6(3):185–8.
soil–cement composites. A greater fibre length may be able [8] Consoli NC, Prietto PDM, Ulbrich LA. Influence of fiber and cement
to offer an increased fibre bond strength, but it would be addition on behavior of sandy soil. Journal of Geotechnical and
too difficult to mix through the soil–cement matrix Geoenvironmental Engineering 1998;124(12):1211–4.
meaning that its use is unrealistic. [9] Kaniraj SR, Havanagi VG. Behaviour of cement-stabilized fiber-
reinforced fly ash soil mixtures. Journal of Geotechnical and
Compressive testing of the soil–cement specimens has
Geoenvironmental Engineering 2001;127(7):574–84.
revealed that almost all specimens can comfortably meet [10] Medjo Eko R, Riskowski G. Effects of fibers and cement on the
the 1.3 MPa compressive strength requirement of the New mechanical behavior of soil–cement reinforced with sugar cane
Zealand earth building standard NZS 4298 [28]. The only bagasse. Housing Science 1994;18(2):79–89.
two specimens not meeting this requirement both displayed [11] Khedari J, Watsanasathaporn P, Hirunlabh J. Development of fibre-
a compressive strength of 1.26 MPa which is very close to based soil–cement block with low thermal conductivity. Cement and
Concrete Composites 2004;27(1):111–6.
the 1.3 MPa requirement. It was also found that the [12] Lasisi F, Osunade JA. Effect of grain size on the strength of cubes
addition of fibre-reinforcement to the soil–cement compo- made from lateritic soils. Building and Environment 1984;19(1):55–8.
site did not have a detrimental effect on specimen [13] Subrahmanyam BV. Bamboo reinforcement for cement matricies. In:
compressive strength as was found in a previous study [29]. Swamy RN, editor. New Reinforced Concretes. Glasgow: Surrey
The accidental use of a cement product designed for University Press; 1984.
[14] Bahar R, Benazzoug M, Kenai S. Performance of compacted
harsh environments, known as Duracem, has found that
cement-stabilised soil. Cement and Concrete Composites 2004;26(7):
this product gives significant improvements to both the 811–20.
flexural and compressive strengths of the material. [15] Olotuah AO. Recourse to earth for low-cost housing in Nigeria.
Although this product does have a price premium in the Building and Environment 2002;37(1):123–9.
range of 5–15% over that of ordinary Portland cement, it [16] Aggarwal LK, Singh J. Effect of plant fibre extractives on properties
may still be worth incorporating into soil–cement compo- of cement. Cement and Concrete Composites 1990;12(2):103–8.
[17] Khazanchi AC, Saxena M, Rao TC. Material science of natural
sites due to the extent of the gains in composite strength. organic fibres reinforced composites in polymer-cement-mud matrix
The additional cost of the Duracem cement may also be for construction engineering. Proceeding of the international
offset by the fact that its use may remove the need for fibre symposium of textile composites in building construction, France;
coating and shift the purpose of the fibre-reinforcement 1990. p. 69–76.
from a strength providing role to a role of providing only [18] Ghavami K, Filho RDT, Barbosa NP. Behaviour of composite soil
reinforced with natural fibres. Cement and Concrete Composites
ductility.
1999;21(1):39–48.
[19] Scheele S, Harris W. The cultivation, harvesting and processing of
harakeke (Phormium tenax) as a source of fibre for inclusion in earth
building composites. Manaaki Whenua Contract Report: LC9798/
Acknowledgements 134; 1998.
[20] Giesecke A. Strengthening existing Adobe housing in earthquake-
prone regions in the Andean countries. Journal of Contingencies and
The authors gratefully acknowledge the generous sup- Crisis Management 1996;4(4):241–4.
port of the UKU Sustainable Housing group at the [21] Baley C. Analysis of the flax fibres tensile behaviour and analysis of
University of Auckland and the funding provided by the the tensile stiffness increase. Composites: Part A 2002;33(7):939–48.
Foundation for Research Science and Technology of New [22] Boominathan S, Senathipathi K, Jayaprakasam V. Field studies on
Zealand for this project. dynamic properties of reinforced earth. Soil Dynamics and Earth-
quake Engineering 1991;10(8):402–6.
The authors are also thankful to the School of
[23] Murray RT, Farrar DM. Temperature distributions in reinforced soil
Engineering at the University of Auckland for the financial retaining walls. Geotextiles and Geomembranes 1988;7(1-2):33–50.
assistance provided through the Guaranteed Financial [24] Allen M. Out of the Ground: Earthbuilding in New Zealand.
Support Scheme. Palmerston North: The Dunmore Press Ltd; 1997.
ARTICLE IN PRESS
M. Segetin et al. / Building and Environment 42 (2007) 3066–3079 3079

[25] ASTM D 1635-00. Standard Test Method for Flexural Strength of With Third-Point Loading). American Society for Testing and
Soil–Cement Using Simple Beam with Third-Point Loading. Amer- Materials: Philadelphia; 1997.
ican Society for Testing and Materials: Philadelphia; 2000. [28] NZS 4298: 1998. Materials and Workmanship for Earth Buildings.
[26] ASTM D 1634-00. Standard Test Method for Compressive Strength Standards New Zealand: Wellington; 1998.
of Soil–Cement Using Portions of Beams Broken in Flexure [29] Segetin M, Lescher P, Jayaraman K, Xu X. Flax fibre-reinforced
(Modified Cube Method). American Society for Testing and soil–cement composites for low cost housing. Proceeding of the 3rd
Materials: Philadelphia; 2000. international workshop on green composites, Kyoto; 2005, p. 39–45.
[27] ASTM C 1018-97. Standard Test Method for Flexural Toughness [30] Engineered Concrete Solutions. Retrieved December 7, 2005 from http://
and First-Crack Strength of Fiber-Reinforced Concrete (Using Beam www.holcim.com/Upload/NZ/Publications/ECS_Marine&Coastal.pdf.

You might also like