Table of contents Search Author index
Geosynthetics - 7 th ICG - Delmas, Gourc & Girard (eds) © 2002 Swets & Zeitlinger, Lisse ISBN 90 5809 523 1
Testing and application of coir based geotextiles
G.V. RAO & R.K.DUTTA,
Department of Civil Engineering, Indian Institute of Technology, Delhi, New Delhi, India
ABSTRACT: This paper presents some of the recent developments made in developing coir based geotextiles along with their physi-
cal characteristics and tensile strength behavior for their application in rural roads. Both monotonic and cyclic load tests carried out in
this connection are very encouraging. The paper highlights the significant behavioral changes obtained by using the coir products.
1.0 INTRODUCTION of 2 kPa after one minute of application of pressure. The tensile
strength of natural geotextiles can be taken as that corresponding
India is well endowed with natural fibers like jute and coir avail- to a wide width specimen (200 mm wide * 100 mm long) at a
able in a processed form, inexpensively. It is the first largest deformation rate of 10 mm/min determined in a constant rate of
country (66% of world production), producing coir fiber from extension machine, and the overall life of jute and coir can have
the husk of coconut fruit. The coir fibers (50 to 150 mm long and a life of more than one and two/three years respectively. It is also
0.2 to 0.6 mm diameter) till recently were being spun into coir pertinent to point out that the tensile strength testing has been
yarn and then woven to obtain woven nettings. The fibers are conducted on specimens of varying length of 25 mm to 200 mm
also now a days being air laid, needle punched or adhesive keeping the width as 200 mm and also by varying the width from
bonded to obtain non-woven products or blankets. Like their 25 mm to 200 mm and keeping a length of 100 mm constant.
polymeric counterparts geotextiles can be synthesized for spe- The deformation rate used varied from 2 to 300 mm/min. Five
cific applications in civil engineering like erosion control, different varieties of jute and coir have been used in this study.
ground improvement etc. Most of the present day products are Hence the above parameters have been used in the present
being developed with an eye on erosion control applications (for study for the evaluation of the different products.
vegetative growth), particularly, because among naturals they
have much longer life. Their biodegradability has not encoun-
tered users for more permanent applications. In fact they are yet 3.0 EXPERIMENTAL WORK
to be standardized , for their tensile behaviour and their biode-
gradability characteristics. Studies in this direction were initiated
3.1 Coir geotextiles used
at Indian Institute of technology, Delhi (Venkatappa Rao and
Balan (1994), Balan (1995) and Venkatappa Rao (1997)). These
3.1.1 Physical properties
studies have broadly indicated that their biodegradability can be
Four different varieties of non-woven coir geotextiles designated
used to advantage and the coir based geotextiles have potential
as A, B, C, D and four woven coir geotextiles designated as
of being used for many geotechnical applications. Keeping this
E,F,G,H were used in the present study.
in view , over the years many varieties of woven and non-woven
The non-woven coir geotextile Type A is 100% de-curled coir
products have been developed in India and are now commer-
fiber web of 650 g/m2 encased over top and bottom with stable
cially available.
woven heavy jute netting. The matrix is stitched together on 5
The paper presents a brief review of the earlier work carried
cm centres with 2-ply jute yarn. The mass per unit area of top
out on characterizing the tensile strength behaviour and biode-
and bottom jute netting is 100 g/m2 each.
gradability of coir geotextiles. It then presents tensile strength
The Type B is composed of 100% de-curled coir fiber web of
characteristics of four woven products and four non-woven
400 g/m2 encased over top and bottom with brown PP netting.
products. Results of monotonic and cyclic load tests carried in a
The matrix is stitched together on 5 cm centres with white PP
model test tank simulating unpaved roads with geotextiles are
thread dipped in black natural glue. The mass per unit area of top
presented. The results are encouraging, for use in developing
and bottom netting is 7.1 g/m2 and 4.8 g/m2. Type C is similar to
countries (like India) in rural roads (village roads) that are yet to
Type B except that the coir fiber web is 750 g/m2
be developed to connect as many as 0.2 million villages (with a
The Type D is 100% de-curled coir web with 390 g/m2 en-
population less than 1,000). Most of these roads happen to be on
cased over top heavy duty woven coir netting of 700 g/m2 and
soft soils.
bottom brown UV stabilized PP netting of 4.8 g/m2. The matrix
is stitch bonded together on 5.0 cm centres with heavy 2 ply jute
thread.
2.0 PREVIOUS WORK
The woven geotextiles Type E, F, G and H are netting com-
posed of 100% coir fiber spun into yarn 240 m/kg (warp and
weft) and woven in conventional flat bed looms in widths of 1,2
2.1 Evaluation of natural fiber geotextile or 4m.
Extensive testing has been carried out at IIT Delhi (Balan (1995)
3.1.2 Tensile strength
and Venkatappa Rao (1997)) on jute and coir geotextiles in order
The tensile strength testing has been carried out as per Balan
to arrive at a rational procedure for evaluating the physical and
(1995) and Venkatappa Rao (1997). The physical and mechani-
strength characteristics as well as their biodegradability. Based
cal properties are presented in Tables 1 and 2 for all varieties of
on these studies, it was recommended that the thickness of natu-
non-woven and woven products respectively.
ral geotextiles can be taken corresponding to a normal pressure
955
3.1.3 Model testing 1.50 kN/m to 1.87 kN/m and 5.2 kN/m to 13.75 kN/m in m/c di-
rection with a mean of 1.70 kN/m and 10.63 kN/m and variance
Model tests were carried out in a tank made of 10 mm thick per- of 0.02 and 9.91 for five specimens tested. The corresponding
spex side plates fitted inside a rigid aluminum frame. The inter- failure strain were found to be 35 % and 31.67 % respectively.
Similarly the tensile strength in x-m/c direction for Type B and
nal dimensions as shown in Figure 1 are 350 mm * 350 mm in Type G ranged between 0.66 kN/m to 0.99 kN/m and 5.83 kN/m
plan and 420 mm in depth. A typical test model consisted of to 8.92 kN/m with a mean of0.84 kN/m and 7.5 kN/m and a
saturated variance of 0.02 and 0.91 respectively. The corresponding failure
soft subgrade (270 mm thick) overlain by 75 mm thick sand strain were found to be 15.83 % and 26.25 % respectively. De-
layer representing a granular course. At the outset, a thin sheet of spite the variation that could be expected in a natural material,
polythene was fixed with cello tape over the internal surfaces of there is a fair level of uniformity in the specimens tested. The
the tank in an attempt to minimize the friction. At the bottom of average tensile strength curves for geotextile Type B and G both
the tank a thin layer of grease was applied. in m/c and x-m/c directions are depicted in Figure 2&3 respec-
tively.
Table 1 Physical and mechanical properties of non-woven coir
geotextiles
Property Non-woven coir geotextiles
A B C D
Roll width (m) 2.2 2.2 2.2 2.0
Roll length (m) 25 25 42 25
Roll weight (kg) 33 22 69 50
Thickness (mm) 12 8.9 9 13.6
Mass per unit area (g/m2) 865 420 750 1175
Tensile strength m/c
7.6 1.7 2.3 15.4
(kN/m)
% Elongation m/c 11.3 35 27.5 27
Tensile strength x-m/c
3 0.8 0.7 8.3
(kN/m)
% Elongation x-m/c 9.7 15.8 19.2 24
Initial tangent
Modulus at 5 mm defor-
68 10 13.8 49
mation (kN/m)
Figure 1 Model test tank.
Table 2 Physical and mechanical properties of woven coir geotextiles
Property Coir geotextile
E F G H
Roll width (m) 1.2 1,2,4 1,2,4 1.2
Roll length(m) 25 50 50 24
1/20,2/40 1/35,2/70,4/
Roll weight (kg) 10 42
,4/80 140
Runnage of yarn 240 240
Aperture size
45*30 25*25 6.25*6.25 7*4
(mm*mm)
Thickness (mm) 7.3 6.7 8 9.5
Mass per unit
335 360 610 1335
area (g/m2)
Tensile strength
3.86 6.34 10.63 31.50
m/c (kN/m)
% Elongation
20 19.17 31.67 42
m/c Figure 2 Load versus elongation curves for non-woven coir geotex-
Tensile strength tile Type B in m/c and x-m/c direction
2.50 4.38 7.50 12.73
x-m/c (kN/m)
% Elongation
27.5 31.67 26.25 18
x-m/c
Initial tangent
Modulus at 5
mm deformation 30 48 46 100
(kN/m)
Nextly, kaolinite clay ot a moisture content 36 % is placed into
the tank by hand kneading. The resulting dry density obtained is
12.35 kN/m3. After the preparation of subgrade, a sand layer of
75 mm thickness was placed over this in two layers 40 mm and
35 mm each at a density 14.95 kN/m3. Where required, a geo-
textile could be conveniently placed over the clay before placing
the sand layer. The model thus prepared was placed on the
Hounsfield Universal Testing Machine. The models were tested Figure 3 Load versus elongation curves for woven coir geotextiles
under monotonic and cyclic loading through a square steel plate Type G in m/c and x-m/c direction
of 75 mm * 75 mm at a deformation rate of 4 mm/min and 75
mm/min respectively. 4.2 Monotonic load tests
4.0 TEST RESULTS AND DISCUSSION 4.2.1 Woven coir geotextiles
4.1 Tensile strength of coir geotextiles The woven coir geotextiles Type E, F, G, H were used at the in-
terface of base course and sub grade.The bearing pressure versus
The tensile strength of Type B and G geotextiles ranged from deformation plots are shown in Figure 4. It can be seen from this
956
figure that there is an improvement in the bearing pressure of the The non-woven geotextile Type A reinforced model shows an
reinforced models over the unreinforced model. It is observed overall better performance than the other non-wovens. The bear-
that Type H reinforced model shows an overall better perform- ing pressures of model reinforced with Type A, B, C, Df, Db and
ance than the other geotextiles. The bearing pressures of model unreinforced models are 72.58 kPa, 63.71 kPa, 74.01 kPa, 71.84
reinforced with Type E, F, G, H geotextiles and unreinforced
models are around 57 kPa, 60.04 kPa, 62.69 kPa, 75.81 kPa and kPa, 63.13 kPa and 54.03 kPa corresponding to a deformation of
54.03 kPa corresponding to a deformation of 20 mm. The Type 20 mm. The Type A, B, C, Df and Db reinforced models show
E, F, G, H reinforced models show 1.05, 1.11, 1.16, 1.40 times 1.34, 1.18, 1.37, 1.33 and 1.17 times higher bearing pressure
higher bearing pressure than the unreinforced model. It is thus than the unreinforced model. Comparing the results of Type B
evident that Type H reinforced model behaves better than Type and Type C it is clear that model with Type C behaves better
G, F and Type E reinforced model which could be attributed to than with Type B, as expected, keeping in view that Type C has
difference in their tensile strength, initial tangent modulus and higher mass per unit area, tensile strength and initial tangent
aperture size. modulus compared to Type B. Further models reinforced with
Type A, D f and Db are performing better than the model rein-
forced with Type B. This is attributed to higher tensile strength
and higher initial tangent modulus of these products. Also Type
A reinforced model behaves better than Type Df and Db rein-
forced model. It may be attributed to its higher initial tangent
modulus.
4.2.3 Comparison
A comparison of the behaviour of models reinforced with non-
woven and woven geotextiles are shown in Figure 6. It is ob-
served that model reinforced with Type H show an overall better
performance than all others. This is attributed to high modulus
and high tensile strength of this product. Further it is also ob-
served that models reinforced with non-wovens show better per-
formance than models reinforced with Type G, F and Type E.
This could be be attributed to better surface friction of these
products. The variation in bearing capacity ratios of these mod-
Figure 4 Static load versus deformation behaviour of models els with respect to deformations are shown in Figure 7. It may be
reinforced with different types of woven geotextiles seen that BCR values for model using non-wovens and wovens
have an increasing trend upto 60 mm deformation.
4.2.1.1 Effect of aperture size
Herein, the influence of the aperture size of the woven geotex-
tiles on the behaviour of the model pavement is studied. From
Figure 4 it can be inferred that as the aperture size decreases, the
reinforced model can bear more bearing pressure. But with the
increase in aperture size, the trend is reverse. However, the rein-
forced model still behaves better than the unreinforced model.
This behaviour of the reinforced model is due to better friction
mobilization and more number of yarns per unit area in lower
aperture size coir geotextile.
4.2.2 Non-woven coir geotextiles
The bearing pressure versus deformation plots obtained in the
model tests with the non-woven geotextiles Type A, B, C, D are
shown in Figure 5. It may be noted Df represents the case where
the woven coir scrim is in touch with the subgrade and Db is the
reverse case. It is evident that there is an improvement in the
bearing pressure of the reinforced models over the unreinforced
model.
Figure 6 Static load versus deformation behaviour of models
reinforced with different Non-woven and woven coir
geotextiles.
5.0 CYCLIC LOAD TESTS
The behaviour of model pavements having sand as base course
and kaolinite as subgrade have been studied under the repeated
loading of 17.94 kPa, 35.88 kPa and 71.76 kPa. Out of the six
tests which were repeated twice, three were conducted at re-
peated load of 35.88 kPa for comparison purposes. The geotex-
tiles used at the interface were of the Type G and B. The varia-
tion in permanent vertical deformation with number of load
repetitions of unreinforced and reinforced models is shown in
Figure 8. Under a repeated load of 35.88 kPa the permanent
Figure 5 Static load versus deformation behaviour of models deformation increases upto 27.46 mm at 200 number of repeti-
reinforced with different types of non-woven geotextiles tions in the unreinforced model while in the model reinforced
with Type B and G, a significant improvement in the behaviour
957
of the models by restricting its permanent deformation upto Table 3 Percentage reduction in permanent deformation
11.30 mm and 12.17 mm. Number of Percentage reduction in permanent deformation
cycles Geotextile Type
B G
10 21.19 21.19
200 58.85 55.68
700 76.68 74.62
5.1 Apparent resilient modulus
The variation of apparent resilient modulus of different test
with the number of load repetitions are shown in Figure 9.This
figure shows that for the Type B and Type G reinforced models,
there is a distinct decrease in the apparent resilient modulus with
the number of load repetitions. For the unreinforced models,
under a repeated load of 35.88 kPa, the apparent resilient
modulus decreases from 25 kPa/mm at N = 2 to 12.3 kPa/mm at
N = 100 at which unreinforced sample failed. Similarly for the
model reinforced with Type B, the apparent resilient modulus
decreased from 23 kPa/mm at N = 2 to 18 kPa/mm at N = 100.
Whereas, the variation in apparent resilient modulus for the
Figure 7 Bearing capacity ratio versus deformation behaviour model reinforced with Type G is some what similar and it varies
of models reinforced with non-woven and woven coir geotextiles. from 19.65 kPa/mm at N = 2 to 15.1 kPa/mm at N = 100. It can
also be seen from Figure 9 that as the intensity of repeated load-
ing is doubled, the apparent resilient modulus is marginally in-
creasing.
Figure 8 Permanent vertical deformation versus number of load
repetitions
Figure 9 Apparent resilient modulus versus number of load repeti-
tions.
Similarly at 700 number of repetitions, the permanent vertical
deformation for the unreinforced model is 63.39 mm whereas it
is 14.78 mm and 16.09 mm for the model reinforced with Type 6.0 CONCLUSION
B and G geotextiles. Thus the effect of geotextile inclusion can
be very well noted. It can also be seen from Figure 8 that at 50 On the basis of results and discussion presented in this paper the
number of cycles the behaviour shown by both the types is simi- following conclusions can be drawn.
lar with more or less improvement in permanent deformation. 1. Use of coir geotextiles in road appears to be of great
But after 50 number of cycles Type B is behaving better than the potential in rural road application particularly on soft
Type G. This may be due to better friction offered by the Type B soils. The optimum product selection is possible by
than Type G though the mass per unit area of Type B is less than further detailed analysis.
Type G. From a comparison curve, it is observed that the rein- 2. As the coir geotextiles have life at least upto 5 years in
forced model consistently performed better than the unreinforced submerged/saturated condition, there is a possibility
model. In models reinforced with Type B and Type G the per- that the subgrade soil will improve in course of time as
manent deformation continue to increase upto 1000 cycles to a it serves as a drainage and filteration layer apart from
value of 16.30 mm and 18.04 mm. The failure in reinforced being separator.
models tend to be general shear type, while in the unreinforced
model punching shear failure occurs. All models were tested 7.0 REFERENCES
upto 1000 number of repetitions.
The percentage reduction in the permanent deformation for 1. Balan, K 1995. Studies on engineering behavior and uses of geo-
models reinforced with Type B and Type G are presented in Ta- textiles with natural fibers, Ph.D Thesis submitted to Indian Insti-
ble 3. The ranges of percentage reduction for geotextile Type tute of Technology Delhi (India).
and Type G reinforced models are 21.19 to 76.68% and 21.19 to 2. Venkatappa Rao, G & Balan, K 1994. Design and development of
7.62%. From permanent deformation versus number of repetition a natural fiber strip drain, Proc. Int. Symp. on Biocomposites and
plot, it is evident that Type B reinforced model performs better Blends Based on Jute and Allied Fibers, pp. 87 – 95.
than Type G reinforced model. Further when the repeated load 3. Venkatappa Rao, G. 1997. Geosynthetic testing – recent develop-
ments, Proc. Geosynthetics Asia'97, 26-29 Nov, Bangalore, India,
was doubled, the reinforced model was still behaving better than
Vol. II., pp. A137 – A168.
the unreinforced model. But the better behaviour of reinforced
model was observed at greater permanent deformation.
958
