Waste Manage Res 2004: 22-2: 108–114 Copyright © ISWA 2004

Printed in UK – all right reserved

Waste Management & Research
ISSN 0734–242X

Soil reinforcement with recycled carpet wastes

A root or fibre-reinforced soil behaves as a composite mate- Hossein Ghiassian

rial in which fibres of relatively high tensile strength are Gholamreza Poorebrahim
embedded in a matrix of relatively plastic soil. Shear stresses Department of Civil Engineering, Iran University of Science and
Technology, Narmak, Tehran, Iran
in the soil mobilize tensile resistance in the fibres, which in
turn impart greater strength to the soil. A research project
Donald H. Gray
has been undertaken to study the influence of synthetic Department of Civil & Environmental Engineering, University
fibrous materials for improving the strength characteristics of of Michigan, Ann Arbor, MI 48109, USA
a fine sandy soil. One of the main objectives of the project is
to explore the conversion of fibrous carpet waste into a value- Keywords: Fibre reinforcement, sand, shear strength, recycled
carpet waste, wmr 722–3
added product for soil reinforcement. Drained triaxial tests
were conducted on specimens, which were prepared in a cylin-
drical mould and compacted at their optimum water con-
Corresponding author: Hossain Ghiassian, Department of Civil
tents. The main test variables included the aspect ratio and Engineering, Iran University of Science and Technology, Narmak,
the weight percentage of the fibrous strips. The results clearly Tehran, Iran.
show that fibrous inclusions derived from carpet wastes E-mail: hossghia@iust.ac.ir

improve the shear strength of silty sands. A model developed

DOI: 10.1177/073424X04043938
to simulate the effect of the fibrous inclusions accurately pre-
dicts the influence of strip content, aspect ratio and confin- Received 4 December 2003; accepted in revised form 11 February
ing pressure on the shear strength of reinforced sand. 2004

Notation

The following symbols are used in this paper. 345 effective minor principal stress
346 consolidation pressure
Af cross-sectional area of fibrous strips 37 deviatoric stress
AR aspect ratio of fibrous strips
B pore water pressure coefficient
Bf width of strips
Introduction
Cc coefficient of gradation of sand As the world population grows, so do the amount and type of
Cu uniformity coefficient of sand waste being generated. Many of the wastes produced today
Gs specific gravity of sand grains will remain in the environment for years to come. Waste
Gsf specific gravity of strips generation is much lower in developing countries, whereas
Ip plasticity index the USA has more than double the waste generation per cap-
Nf number of strips ita than many European countries. According to the US Envi-
Wf strip content (weight ratio of strips/drysoil) ronmental Protection Agency (US EPA 1992), the munici-
Ws dry weight of sand pal solid waste generated in the US is about 200 million tons
11 axial strain per year, among them about 38% being paper products, 8%
12 volumetric strain plastics, and 3% carpets and textiles. The creation of non-
23 unit weight of water decaying waste materials, combined with a growing con-
341 effective major principal stress sumer population, has resulted in a waste disposal crisis. This

Waste Management & Research 108

Downloaded from wmr.sagepub.com at PENNSYLVANIA STATE UNIV on March 5, 2016
Hossein Ghiassian, Gholamreza Poorebrahim, Donald H. Gray

issue has received considerable public attention in recent The present study was undertaken to examine the influ-
years, particularly as it has become increasingly difficult, ence of shredded carpet waste in improving the shear strength
both financially and politically, to dispose of the steadily and volume change characteristics of a fine sandy soil under
growing volume of waste. With municipalities under pressure triaxial compressive loading conditions, and to understand the
to find acceptable alternatives to the traditional forms of mechanisms controlling the behaviour of the material and soil.
refuse disposal – landfilling and incineration – the recycling The influence of strip content (Wf), strip aspect ratio (AR), and
of waste into useful products has been advocated by policy confining pressure were assessed through drained triaxial test-
makers and some environmentalists as a possible solution to ing on specimens compacted at their optimum water contents.
the municipal waste disposal crisis.
A great amount of fibrous textile waste is discarded into Materials
landfills each year all over the world. About half of this
waste is from carpets, with the main constituents being plas- The soil samples used in this study were obtained from the
tic and polymeric fibre, which decays at a very slow rate and region of Kerman in central Iran. The soil is fine sand, classi-
which is difficult to handle in landfills. In the US, the cost of fied as silty sand (SM) according to the Unified Soil Classifi-
landfilling this waste is estimated at US$ 200 million per cation System. The specific gravity of solids (Gs) was 2.675.
year (Wang et al. 1993). The development of low-cost tech- The grain size distribution was 85.8% sand, 12% silt and 2.2%
nologies to convert textile waste into useful products could clay size material, with a uniformity coefficient (Cu) of 2.84
reduce disposal pressures on landfills and result in cost sav- and coefficient of gradation (Cc) of 0.71. The soil has plas-
ings for carpet industries. Accordingly, many carpet and tex- ticity index (Ip) of about zero and is essentially non-plastic.
tile manufacturers, fibre and chemical suppliers, recycling The reinforcing fibrous strips were prepared by cutting car-
companies, and academic institutions are actively pursuing pet waste into short elements with approximately 5 mm ×
various methods to recycle and convert fibrous waste into 5 mm square cross-section and various length of 5, 15, 25, 35
various useful products (Wang et al. 1993, 1994, Gardner and 45 mm corresponding to aspect ratios (AR, length/width)
1995, Schut 1995, Wang, 1997, Frost et al. 1999). of 1, 3, 5, 7 and 9 respectively. The specific gravity of the
One promising reuse of these wastes lies in soil reinforce- fibrous strips (Gsf) is 1.0. Tensile strength and elastic modulus
ment and construction applications. Extensive studies have were, 800 and 730 kPa, respectively, with an almost linear
been reported in the literature that indicate the use of natural strain at failure of 105%.
and synthetic fibres in soil can enhance the shear strength Various amounts of strips were added to the soil to yield strip
and volume change characteristics of soil, and improve its contents or weight fractions (Wf), defined as weight of strips/
load-bearing capacities and durability (McGown et al. 1978, dry weight of sand, of 0, 0.4, 0.6, 0.8 and 1%, respectively.
Verma and Char 1978, Hoare 1979, Gray and Ohashi 1983,
Gray and Al-Refeai 1986, Gray and Maher 1989, Maher and
Gray 1990, Al-Refeai 1991, Maher and Ho 1994, Ranjan et Preparation and testing of specimens
al. 1996, Andersland and Khattak 1997, Consoli and Prietto
1998). Embedded fibres generally act as tensile reinforcing The specimens were prepared by mixing soil and water, with
members that transfer shear stresses that develop in the soil the addition of appropriate amounts of strip to yield a speci-
matrix into tensile resistance in the fibre inclusions via fibre/soil fied strip content. Samples were compacted in eight layers
frictional interactions. The lack of adequate tensile strength into a 100-mm (4-inch) diameter and 200-mm (8-inch) high
in a soil can therefore be compensated by the fibres, and the cylindrical mould, with a target maximum dry unit weight
shear strength of soil increases. Even when a soil deforms in and optimum water content of 16.7 kN m–3 and 9.2%, respec-
shear or compression, normal strains or tensile extensions can tively. In order to produce homogenous samples, the under-
develop in other directions. Some tensile extension is required compaction method (Ladd 1978) was carried out as part of
to mobilize the strength of imbedded fibres. In this respect, the sample preparation.
use of randomly distributed fibres is more advantageous due to Drained triaxial tests were conducted to determine the
the maintenance of strength isotropy and the absence of poten- stress–strain and strength characteristics of reinforced soils with
tial planes of weakness that can develop parallel to oriented various strip contents or strip/soil weight ratios (Wf) and ARs.
reinforcement (Maher and Gray 1990). The carpet fibres typi- All specimens were fully saturated with a minimum meas-
cally consist of nylon fibres and polypropylene tape yarns, ured B value of 0.98. The use of CO2 gas with percolation of de-
both proven durable and effective in fibre-reinforced soil. It aired water and back pressure application up to 400 kN m–2
is thus reasonable to expect beneficial effects from the pres- ensured full saturation. The effective confining pressure var-
ence of fibrous strips made of recycled carpet waste in a soil. ied from 15 to 300 kPa.

109 Waste Management & Research

Downloaded from wmr.sagepub.com at PENNSYLVANIA STATE UNIV on March 5, 2016
 Soil reinforcement with recycled carpet wastes

Fig. 1: Triaxial compression test results at 3'c = 100 kPa for reinforced Fig. 2: Triaxial compression test results at 3'c = 100 kPa for reinforced
specimens with AR = 1, (a) deviatoric stress versus axial strain; (b) specimens with AR = 3, (a) deviatoric stress versus axial strain; (b)
volumetric strain versus axial strain. volumetric strain versus axial strain.

although this effect is more pronounced at higher Wf values.

Results and analysis
Other researchers (Gray and Ohashi 1983, Shewbridge and
The stress–strain curves obtained in triaxial compression drained Sitar 1989, Ranjan et al. 1996). have also reported similar
tests at an effective confining pressure of 100 kPa are presented findings.
in Figs 1 to 5 corresponding to AR values of 1, 3, 5, 7 and 9 It can be seen in Fig. 7 that the strength increases (at any
for different strip contents. As can be seen in these figures, constant Wf) with AR. Considering the inverse relationship
the overall soil behaviour is greatly influenced by the presence between the number of strips (Nf) and AR from the following
of reinforcing strips as Wf and AR values change. Consistent expression:
results are seen in all five figures as explained below.
Wf Ws
By adding strips to soil, the peak and residual strengths N f = 8888888888888888888888888888888
8 (1)
A R A f B f G sf 2 w
increase, the maximum modulus decreases, and the post strength
behaviour changes from that of a ‘brittle’ to a ‘ductile’ type where Nf is the number of strips, Ws is the dry weight of sand
material with an increase of strip content (at constant AR). Af is the cross-sectional area of strips (0.25 cm2), Bf is the
The same type of behaviour can be observed as the AR is width of strips (0.5 cm) and Gsf is the specific gravity of the
increased at constant strip weight ratios (Wf). These results strips. It seems that the contribution of fibrous strips to soil
can more clearly observed in Figs 6 and 7 for the variation of reinforcement is more controlled by the length (aspect ratio)
peak strength versus Wf and AR, respectively. At any con- of strips rather than strip quantities. This means the mode of
stant AR, the peak strength increases with Wf while the rate failure with short strips (low AR values) should be governed
of increase (the slope of curve) reduces with increasing strip more by pullout rather than breakage. This behaviour should
content. This means that optimum values for strip content however change gradually at higher strip lengths (i.e., higher
and aspect ratio should exist consistent with economical and AR values) where more strips would fail in breakage due to
practical considerations or constraints. Similarly in Fig. 7, at more mobilized frictional resistance along the length of the
any constant Wf, the peak strength still increases with AR embedded strips. As a result, the rate of increase in strength

Waste Management & Research 110

Downloaded from wmr.sagepub.com at PENNSYLVANIA STATE UNIV on March 5, 2016
Hossein Ghiassian, Gholamreza Poorebrahim, Donald H. Gray

Fig. 3: Triaxial compression test results at 3'c = 100 kPa for reinforced Fig. 4: Triaxial compression test results at 3'c = 100 kPa for reinforced
specimens with AR = 5, (a) deviatoric stress versus axial strain; (b) volu- specimens with AR = 7, (a) deviatoric stress versus axial strain; (b) volu-
metric strain versus axial strain. metric strain versus axial strain.

versus AR should decrease as can be seen in Fig. 7. If strips composite materials. Thus, the strength mobilization in a soil
with lower modulus are used and/or higher confining pres- matrix with strip inclusions may not occur simultaneously.
sures are applied, the plots in Fig. 7 would be expected to This means that the soil mass first reaches its ultimate shear
become asymptotic at smaller AR values. This behaviour can strength when the rate of dilation becomes a maximum,
be seen in Fig. 8 for the case of Wf = 1% as an example. The whereas the strip contribution to load-bearing capacity still
plots are presented in terms of93'1/3'3 ratio such that the influ- continues to increase. At higher axial strains in the speci-
ence of effective confining pressures on strip contribution to men, which also correspond to more tensile strains in the
the shear strength of reinforced sand specimens can be depicted embedded strips, the strip contribution to the total strength
more clearly. It is seen that as the confining pressures decrease, reaches to its ultimate value after the rate of dilation has passed
the plots tend to become asymptotic at smaller AR values. its maximum value.
Visual observations of the specimens after the tests con- It can also be seen in the above figures that as the strip
firmed that the strips indeed failed mainly by being pulled content increases, the amount of volumetric strain increases.
out from the soil on the rupture surface. However, the influence of aspect ratio variation on the volu-
The volume change result for the unreinforced specimen metric strain seems negligible.
agrees well with the expected behaviour of dense sand as Figures 9 and 10 summarize the variation of volume change
shown in Fig. 1. After the occurrence of a small initial com- versus the strip content and aspect ratio. Similar trends to
pression at the beginning of loading, the specimen dilates and those observed in Figs 6 and 7 are seen in these figures as well.
the rate of dilation becomes greatest at the time of failure. At any constant AR, the total volume change increases with
This trend, however, changes slightly with the presence of Wf whereas the rate of increase (the slope of the curve) slightly
strips in the soil (Figs 1 to 5). In reinforced specimens, the reduces with increasing strip content. Furthermore, at any
maximum dilation rate does not coincide with the onset of constant Wf, the total volume change still increases with AR
failure but instead occurs earlier. This could be attributed to and this effect is more pronounced at lower aspect ratios.
the fact that the reinforced specimens seem to behave as Shewbridge and Sitar (1989) have reported that the shear

111 Waste Management & Research

Downloaded from wmr.sagepub.com at PENNSYLVANIA STATE UNIV on March 5, 2016
 Soil reinforcement with recycled carpet wastes

Fig. 7: The influence of aspect ratio on triaxial compression strength of

reinforced specimens at 3'c = 100 kPa for different strip contents.

Fig. 5: Triaxial compression test results at 3'c = 100 kPa for reinforced
specimens with AR = 9, (a) deviatoric stress versus axial strain; (b) volu-
metric strain versus axial strain.
Fig. 8: The influence of aspect ratio on triaxial compression strength of
reinforced specimens for strip content of Wf = 1% and different confin-
ing pressures.

Fig. 6: The influence of strip content on triaxial compression strength of

reinforced specimens at 3'c = 100 kPa for different aspect ratios.

band thickness increases in proportion with the strip content. Fig. 9: The influence of strip content on volume change of reinforced
specimens at 3'c = 100 kPa for different aspect ratios.
This in turn can increase the amount of dilation in the shear
zone. In addition, visual observation of the reinforced speci- regression analysis of all data including those for the unrein-
mens during and after the tests revealed that several shear forced samples (130 triaxial compression tests) was performed
zones tended to develop in the specimens due to the pres- based on a mathematical power model proposed by Ranjan et
ence of strips in the soil. This shearing behaviour is also con- al. (1996). The model has the following general form:
sistent with volume change results in Figs 9 and 10.
2 1
Figure 11 presents the failure envelopes for the case of Wf = 341 = k3436 W f7 A R f * f 9 (2)
1.2 and different strip aspect ratios. Other results correspond-
ing to different Wf values exhibit similar trends for the varia- where914564574524584599are constants obtained from a regression
tion of 3'1 versus 3'3; therefore they have been omitted. A analysis, and9 2 39 and9 29 are coefficients corresponding to the

Waste Management & Research 112

Downloaded from wmr.sagepub.com at PENNSYLVANIA STATE UNIV on March 5, 2016
Hossein Ghiassian, Gholamreza Poorebrahim, Donald H. Gray

Fig. 10: The influence of aspect ratio on volume change of reinforced Fig. 12: Comparison between experimental and predicted strength of
specimens at 3'c = 100 kPa for different strip contents. carpet waste reinforced fine sand.

sition or break occurring at a well-defined (critical) confin-

ing stress 3crit9or conversely, curvi-linear with no obvious break
or abrupt transition in the failure envelope. Well-graded, angu-
lar, fibre-reinforced sands exhibited the bi-linear form of
behaviour, whereas uniform, rounded sands exhibited the curvi-
linear form. The results of the present study indicate that this
transition is fairly smooth with no sharp break9or abrupt change
in the failure envelope. It seems, therefore, that in addition
to the effects of granulometry (grain shape and gradation) the
use of more flexible fibres causes the transition to become
smoother, such that the failure envelope can more reasona-
Fig. 11: Principal stress envelopes for carpet waste reinforced fine sand
for strip content of Wf = 1.2%.
bly be modelled as a curved or curvi-linear envelope without
any well-defined critical confining stress.
shear characteristics of the strips and soil, respectively, which
in turn are functions of effective confining pressure9345. Since
only one type of soil and strip was examined in the present
Conclusions
study, these coefficients were not necessary in the model and, This study has demonstrated the benefits of using fibrous strips
instead, their influences were accounted for and lumped made from carpet waste for soil reinforcement. The presence
together in the9 3436 9term. In order for the model to be appli- of these randomly distributed synthetic strips had a signifi-
cable to plain soil (i.e. Wf = 0, AR = 0), the terms Wf and AR cant influence on the shear characteristics and constitutive
were replaced by (1 + Wf) and (1 + AR) respectively. The behaviour of fine sand. The following are the major conclu-
following relationship was finally obtained as a result of the sions of this study.
regression analysis:
• Inclusion of strips increased the peak and residual compres-
341 = 13.243410.732
1 + Wf  0.1712
1 + AR  0.0792 (3) sive strengths, maximum modulus, and ductility of fine sand.
• At any constant aspect ratio (or constant strip content),
The accuracy of this model was examined by comparison the peak strength and total volume change increase with
between predicted and measured strengths as shown in Fig. strip content (or aspect ratio) whereas the rate of increase
12. Predicted values were computed using equation (3) and reduces with increasing strip content (or aspect ratio).
measured values were obtained from the triaxial tests on both • At any constant strip content, the influence of reinforce-
plain and reinforced specimens. Good agreement was obtained ment on strength and volume change is more pronounced
between predicted and measured results particularly for the at higher aspect ratio.
reinforced samples. • Unlike plain soil specimens, the maximum dilation rate in
In past tests performed by Maher and Gray (1990) on ran- reinforced specimens does not coincide with the onset of
domly distributed, fibre-reinforced sand in which fibres with failure time but instead occurs earlier.
a relatively high modulus were used, the failure envelopes of • An empirical equation based on a probabilistic model and
the soil–fibre composites were either bi-linear with the tran- regression analysis of all the test data can be used to char-

113 Waste Management & Research

Downloaded from wmr.sagepub.com at PENNSYLVANIA STATE UNIV on March 5, 2016
 Soil reinforcement with recycled carpet wastes

acterize and accurately describe the influence of different Acknowledgement

variables such as strip content, strip aspect ratio and con-
This study was funded via a cooperative support provided by
fining stress on the shear strength increase.
Iran National Sub-Committee for Earthquake & Landslide
Hazard Reduction and Kerman Carpet Company. This
financial assistance is gratefully acknowledged.

References

Al-Refeai, T. (1991): Behavior of granular soils reinforced with discrete ran- Maher, M.H. & Ho, Y.C. (1994): Mechanical properties of kaolinite/fiber
domly oriented inclusions. Journal of Geotextiles and Geomembranes, soil composition. Journal of Geotechnical and Geoenvironmental Engi-
10, 319–335. neering, ASCE, 120, 1380–1393.
Andersland, O.B. & Khattak, A.S. (1997): Shear strength of kaolinite/fiber Maher, M.H. & Gray, D.H. (1990): Static response of sand reinforced with
soil mixtures. In: Proceedings of the International Conference on Soil Rein- distributed fibers. Journal of Geotechnical and Geoenvironmental Engi-
forcement, Paris, France. neering, ASCE, 116, 1661–1677.
Consoli, N. & Prietto, D.M. (1998): Influence of fiber and cement addition McGown, A. et al. (1978): Effect of inclusion properties on the behavior of
in sandy soils. Journal of Geotechnical and Geoenvironmental Engineering, sand. Geotechnique, 28, 323–346.
ASCE, 124, Ranjan, G., Vasan, R.M. and Charan, H.D. (1996): Probabilistic analysis of
Frost, J.D., Wang, Y., Murray, J. & Jones, A. (1999): Utilization of carpet, randomly distributed fiber reinforced soil. Journal of Geotechnical and
textile & apparel waste for soil reinforcement. In: Proceedings of the Geoenvironmental Engineering, ASCE, 122, 419–426.
Fourth Annual Conference on Recycling of Fibrous Textile and Carpet Schut, J.H. (1995): Big plans for carpet. Plastics World, December, 25.
Waste, Dalton, GA. Shewbridge, S.E. & Sitar, N. (1989): Deformation characteristics of rein-
Gardner, H.C. (1995): Carpet recycling technology. International Fiber forced sand in direct shear. Journal of Geotechnical and Geoenvironmen-
Journal, August, 36. tal Engineering, ASCE, 115, 1134–1147.
Gray, D.H. & Maher, M.H. (1989): Admixture stabilization of sands with US Environmental Protection Agency (1990): Report PB90-215112,
discrete, randomly distributed fibers. In: Proceedings of XIIth Interna- Springfield, VA.
tional Conference on Soil Mechanics and Foundation Engineering Vol.2, Verma, B.P. and Char, A.N. (1978): Triaxial tests on reinforced sand. In:
Rio de Janeiro, Brazil, pp. 1363–1366. Proceedings of a Symposium on Soil Reinforcing and Stabilizing Technique,
Gray, D.H. Ohashi, H. (1983): Mechanics of fiber reinforcement in sand. Sydney, Australia, pp. 29–39.
Journal of Geotechnical and Geoenvironmental Engineering, ASCE, 108, Wang Y. (1997): Carpet waste for soil stabilization. In: Proceedings of Second
335–353. Conference on Recycling of Fibrous Textile & Carpet Waste, May,
Gray, D. & Al-Refeai, T. (1986): Behavior of fabric versus fiber reinforced Atlanta, GA.
sand. Journal of Geotechnical and Geoenvironmental Engineering, ASCE, Wang, Y., Cho B.S. &. Zureick A.H. (1993): Fiber reinforced concrete using
112, 804–820. recycled carpet industrial waste and its potential use in highway con-
Hoare, D.J. (1979). Laboratory study of granular soils reinforced with ran- struction. In: Proceedings of the Symposium on Recovery & Effective
domly oriented discrete fibers. In: Proceedings of the International Con- Reuse of Discarded Materials & By-Products for Construction of Highway
ference on Use of Fabrics in Geotechnology, Paris, France, pp. 47–52. Facilities, Denver, CO, pp. 4.111–4.117.
Ladd, R.S. (1978): Preparing test specimens using undercompaction. Geo- Wang, Y., Zureick, A.H., Cho, B.S. & Scott, D.E. (1994): Properties of fiber
technical Testing Journal, ASTM, 1, 16–23. reinforced concrete using recycled fibers from carpet industrial waste.
Journal of Material Science, 29, 4191–4199.

Waste Management & Research 114

Downloaded from wmr.sagepub.com at PENNSYLVANIA STATE UNIV on March 5, 2016