Procedia Engineering 53 ( 2013 ) 156 – 166

1877-7058 © 2013 The Authors. Published by Elsevier Ltd. Open access under CC BY-NC-ND license.

Selection and peer-review under responsibility of the Research Management & Innovation Centre, Universiti Malaysia Perlisdoi: 10.1016/j.proeng.2013.02.021

Malaysian Technical Universities Conference on Engineering & Technology 2012, MUCET 2012Part 3 - Civil and Chemical Engineering

Innovative Laboratory Assessment of the Resilient Behaviour ofMaterials (Rigid, Elastic and Particulates)

Alvian John Lim Meng Siangª*, Devapriya Chitral Wijeyesekeraa, Lim Sin Meia ,Adnan Zainorabidina

aFaculty of Civil and Environmental Engineering, Universiti Tun Hussein Onn Malaysiaa

Abstract

This paper discusses the resilient behaviour of various materials of high rigidity, high elasticity and of particulate materials. Theresilient behaviour is measured by the resilient modulus (MR) test which is the mechanical response of soils to a dynamic cyclic loadwhich simulates traffic loading conditions and it is an important parameter in the design of both flexible and rigid pavements. Thematerials used are of a concrete cylinder, a rubber sample and particulate materials which are sands of different particle size and shape classification. The design resilient modulus follows a constitutive model and it is used to compare the resilient behaviour of each material.The constitutive model shows that the design resilient modulus is affected by the particulate characteristics and how it can be related to a stiff and flexible material.

© 2013 The Authors. Published by Elsevier Ltd.Selection and/or peer-review under responsibility of the Research Management & Innovation Centre, Universiti

Malaysia Perlis.Keywords: Resilient modulus; stiff; flexible; particulate materials.

Nomenclature

Cu Uniformity coefficientD10 Diameter which 10% of the total soil mass is passingD60 Diameter which 10% of the total soil mass is passingDr Relative Densityemax Maximum void ratioemin Minimum void ratio

d Deviator stressMr Resilient ModulusR RoundnessS Sphericity

r Resilient strain

* Corresponding author. E-EEE mail address: alvinjlms@gmail.com

Available online at www.sciencedirect.com

© 2013 The Authors. Published by Elsevier Ltd. Open access under CC BY-NC-ND license.

Selection and peer-review under responsibility of the Research Management & Innovation Centre, Universiti Malaysia Perlis

157 Alvian John Lim Meng Siang et al. / Procedia Engineering 53 ( 2013 ) 156 – 166

1. Introduction

Queen Elizabeth II in her 2012 New Year speech which she mentioned of her admiration for the human beings resilience to adapt to various natural catastrophes they meet. In the world of materials, its resilience is a function of the make up and behaviour. Hard as rock or concrete produces high resilience and that it can take large dynamic variations of loading.

form on removal of load. The understanding of resilience important in the application of pavement design as it simulates traffic loading conditions. The resilient modulus is not a constant stiffness property and its values are generally affected by factors such as the deviator stress, confining pressure, water content, dry density, method of compaction and freeze-thaw cycles [1]. Resilient modulus on cohesive soil is mainly a function of the applied deviator stress, whereas granular materials have an increasing resilient modulus with increasing confining pressure [2]. The axial permanent strain is directly related to deviator stress and inversely related to confining pressure. Thompson [3]; Hicks and Monismith [4] and Lekarp et. al. [1] has discussed the effects of the index properties and gradation of the soil on its resilient response. It was found that soil with increasing maximum particle size show high resilient modulus values than soils with increasing fines content. However, there is still much to be discussed on the effects of particle shapes characteristics. The resilient modulus value also increases with increasing density due to the increased number of particle contact per particle in the particulate system. It is also observe that dry samples have larger resilient modulus than wet samples due to capillary suction [5]. The presence of moisture content in a particulate assembly has some lubricating effect on particles and increases the deformation with a consequence in the reduction of the resilient modulus. In well graded materials with high proportion of fines, water is more readily available in the pores whereas in uniformly graded materials all the water will drain freely and rapidly [1].

In a laboratory test, a desirable series of distinct load pulses can best simulate vehicle loading conditions on pavements materials. Cylindrical specimens of soil are subjected to a series of load pulses applied with a distinct period, simulating the stresses caused by multiple wheels moving over the pavement. A constant all around confining pressure is applied on the specimen to simulate the lateral stress caused by the overburden stress and applied wheel load. The resilient modulus (MR) values are obtained to a measure the degree to which a soil can recover from stress levels commonly placed upon soils by traffic [6]. The significance of this is to provide a relationship between stress and deformation of a soil material for the structural analysis of layered pavement systems. Resilient modulus is defined as:

= (1)

where Mr d r is the recoverable axial strain. As indicated in Equation 1, the larger the recoverable deformation, the smaller the resilient modulus resulting in a flexible subgrade soils. The equation is somewhat similar to the stiffness modulus equation which is:

E = ( ) ( ) (2)

where the difference is in the strain parameter which the stiffness modulus accounts for permanent strain. Fig. 1 shows typical strain behaviour of a resilient modulus test. At the initial stage of load applications, there can be considerable permanent deformations as indicated by the plastic strain in the figure. As compared to a static load test, the general stress versus strain graph as shown in Fig. 2 shows that the elastic behaviour occurs until a point where it changes to a plastic state and permanent strain occurs after that. Normally for granular soils, it shows that recoverable strains occurs after numerous loading repetitions as long as the deviator stress applied is not in excess of the shear strength of the soils.

158 Alvian John Lim Meng Siang et al. / Procedia Engineering 53 ( 2013 ) 156 – 166

Fig. 1. Strain behaviour for a resilient modulus test [7]

Fig. 2. Typical stress versus strain relationship of a static load test [8]

The AASHTO 1986 Guide for Design of Pavements Structures [9] recommended American highway agencies to use resilient modulus test for the design of subgrades and has led to the efforts of a large number of researchers to obtain more accurate, straightforward and reasonable Mr values which are representative of the field conditions. Resilient modulus has been incorporated by American highway agencies for many years [5], and the current standard test method now in the determination of resilient modulus is according to AASHTO Designation: T 307-99 [9]. Laboratory tests are performed over a range of vertical stresses and confining pressure to evaluate the nonlinear elastic behaviour of soils. Thus, the resilient modulus test does not necessary result in a single modulus value, but defines the modulus at different stress levels. Hence, the modulus is dependent on the stress regime/levels used. Various types of relationship have been suggested to represent the repeated load resilient modulus test results of coarse-grained soils according to the standard test method [10],[3],[11],[12]. A constitutive model that has been widely used for fined grain soils is of the form given in equation 3 [12].

Mr = K1 d)K2 (3)

d = Bulk stress K1 and K2 are regression constants

For granular materials, the most significant parameter that influences MR is the 3) and MR is modelled

in terms of the following equation [12]:

Mr = K1 3)K2 (4)

3 = Confining stress K1 and K2 are regression constants Pezo [10] suggested a stress dependent model that incorporates both deviatoric and confining stresses which can be used

for both cohesive and cohesionless soils.

Mr = K1 d)K2 3)K

2 (5)

159 Alvian John Lim Meng Siang et al. / Procedia Engineering 53 ( 2013 ) 156 – 166

K1 and K2 and K3 are regression constants This type of model was found to suit the resilient modulus (MR) value very well for all types of soils and stress

environment as demonstrated by Al-Suhaibani et. al. [13]. In this research, relationships can be established between the constitutive models of resilient modulus and the particulate properties of the soil.

2. Materials and Testing Methodology

Three types of materials with different characteristics were used for the resilient modulus test. The materials consist of a rigid, elastic, viscoelastic and particulate materials. A very stiff material of a concrete cylinder is used to simulate a very rigid material. A rubber and polystyrene sample is also used as flexible materials as shown in Fig. 4 respectively. The particulate material consists of river sands from Kahang Johor and also Leighton Buzzard sand from UK. Kahang sands are angular in shape which comprised of well graded sand (SW), uniformly graded sand (SPuKahang) and gap graded sand (SPg). Leighton B uzzard sand however has rounder particle shapes and is documented as being uniformly graded sand (SPuL.

Buzzard). Classifications of the samples are in accordance with BS 1377-1: 1990. Fig. 3 shows the particle size distribution curves of all the samples illustrating the different grading curves. The minimum and maximum void ratio (emin and emax) is obtained by the procedure specified by Head [14] and BS 1377-1: 1990. XRD tests show similar mineral composition where quartz is the dominant mineral for all the samples, thus minimizing the variables in the test in terms of mineralogy. The shape of the particles ware quantified by techniques used by Powers [15], Krumbein and Sloss [16] and Cho et. al. [17], Tsomokos and Georgiannou [18], Lim et. al. [19]. The mean sphericity and roundness of the particles are obtained to differentiate the shape parameters. Table 1 is a factual summary of all the main properties of the samples in this study. Higher mean sphericity and roundness values show rounder particle shapes. Fig. 6 shows microscopic images of the sand particles which show the different grading and shapes.

All the sand samples were homogenized with moisture content of 10%, thoroughly mixing it and storing in a closed container for 24 hours to enable a uniformly distribution of the water within the soil. The sand samples were formed in a 50mm diameter and 100mm height spilt mould prepared using the under-compaction method [20] where the samples are compacted with a tamping rod at five parts of equal weight. All the sand samples were prepared with two different relative densities. A cylindrical concrete and rubber sample of similar dimensions were also tested to compare it resilient behavioural patterns. The samples are tested on a Geocomp fully automated resilient modulus machine shown in Fig 5 which it uses air pressure as the confining stress. Fig. 6 shows the schematic diagram of the resilient modulus apparatus.

Fig. 3. Particle size distribution curve

0102030405060708090

100110

0.01 0.1 1

Perc

ent F

iner

Particle Diameter

Fine Medium Coarse Fine

Sand Gravel

Legend: SW SPu(Kahang) SPu(L.Buzzard) x

160 Alvian John Lim Meng Siang et al. / Procedia Engineering 53 ( 2013 ) 156 – 166

Table 1: properties of the test samples

Fig. 4: Cylindrical concrete, rubber sample and polystyrene

Fig. 5. (a) GEOCOMP fully automated resilient modulus apparatus (b) Schematic diagram of resilient modulus apparatus

Properties

Type of Sand SW SPu

Kahang SPg SPu

L.Buzzard D60 1.5 0.65 1.5 0.71 D10 0.38 0.35 0.13 0.48 Cu 3.95 1.86 11.53 1.48 emax 0.91 0.95 0.78 0.73 emiin 0.39 0.57 0.4 0.57 Mean Sphericity (S) 0.616 0.715 Mean Roundness (R) 0.358 0.591

Material Concrete Rubber Polystyrene

Density (kg/m3) 2208 1191 26.60

Concrete Rubber Polystyrene

(a) (b)

161 Alvian John Lim Meng Siang et al. / Procedia Engineering 53 ( 2013 ) 156 – 166

(a) (b) (c) (d)

Figure 6: Magnified pictures of the sand particles (a) SW sand (b) SPuKahang sand (c) SPuL.Buzzard sand (d) GPg sand. (Magnification of x10)

3. Results and Discussions

As stated, the set of loading conditions are based on the AASHTO Designation: T 307-99 [9] standard. The third constitutive model suggested by Pezo [10] was used to determine the design resilient modulus (MR) value. The MR values are calculated with a deviator stress of 68.9 kPa and a confining pressure of 41.4 kPa fitted in the equation. Table 2 shows the summary of the test results from the resilient modulus test together with its regression constants (K1, K2, and K3) which were automatically generated by the GEOCOMP resilient modulus software. The main finding is described as below: It can be seen that concrete, due to its high compressive strength and high stiffness has the lowest resilient strain as

r) respectively. It also can be seen that SPu(L.Buzzard) has a larger value MR as compared to the rest of the sands. Based on the AASHTO

soil classification system, granular soils are said to perform well as good road subgrades, however uniformly graded r) on the

SPu(L.Buzzard) sand. Regardless of that, SPu(Kahang) which is also uniform sands however has a lower Mr value than SPu(L.Buzzard). This can only

r) of the sample as angular sands can be compacted much more than rounder sands.

The table also shows that lower relative density of the sample which is around Dr = 0.5 give lower MR values. It is also appears very interestingly that although SPu(L.Buzzard) has very large MR values, but it also shows that it has the highest maximum permanent strain. This shows that rounder sand particles tend to have low resilient strain but are more easily to be compacted.

Fig. 7 shows the resilient modulus values versus its deviator stress according to the confining pressure. It can be seen that the rubber and polystyrene samples respond very differently from the others as the MR values tend to decrease

he deviator stress.

The cylindrical concrete sample also behaves differently than the sand samples. It can be seen that, the MR values do not tend to be affected by the different confining stress levels. It is however that the concrete sample shows drastic increases in MR values as the deviator stress increases due to its very low resilient strain.

As for the sand samples, the confining stress plays an important factor in the resilient behaviour. The confining stress shows a direct relationship with the MR values where high confining stress shows high MR values. The increasing deviator stress also tend to decrease the resilient strain and resulting in higher MR values.

Fig 8 shows the Mr increases with an increase in uniformity coefficient (Cu) at high relative density. At lower Dr, the Mr only increases slightly, thus it can be said that the Mr is dependent on the relative density (Dr) of the samples. This also shows that the grading characteristics shows an important factor in the resilient behaviour of the soil as gap graded sand tend to show high Mr values due to the increasing fines content in the soil.

Fig. 9 also shows that the increase in the shape parameters of sphericity and roundness will shows a trend of an increase in the resilient modulus values.

The effect of the void ratio (e) is shown in Fig. 10, where as the void ratio increases, the Mr value however decreases. Fig. 11 shows the overall maximum resilient modulus (MR) for all the samples based on the concrete, rubber,

polystyrene and sand materials on its density. It can be seen that the concrete samples has the highest density with very

162 Alvian John Lim Meng Siang et al. / Procedia Engineering 53 ( 2013 ) 156 – 166

high stiffness, resulting in very high MR r). The higher the density of the material, thelower will be the resilient r) and the higher would be the MR. The relationship between the Resilient Modulus and the Stiffness Modulus is shown in Figure 12. It can be seen thatResilient Modulus increases with the Stiffness Modulus.

4. Related Applications

Based on the findings, it is certain that particulates of granular materials show very interesting resilient behaviour. TheResilient modulus is an important parameter in the application of pavement design and it can be explained in detail by themechanistic-empirical method which shows the loading reactions on each of the pavement layers to predict types of distress[21]. Fig. 13 shows a typical diagram of a pavement showing each of the soil layers and its characteristic. The top layer of the pavement is made of hot mix asphalt and it can best described as a flexible material. The materials below it areconsidered to be stiffer materials. Fig. 14 shows the mechanistic response on the pavement which shows how the loadinginputs are transferred. It can be seen that tensile stress occurs at the hot mix asphalt and compressive stress occurs at the subbase and subgrade materials. Hence, the resilient modulus can give a good indication of the reaction of each of the layers.Obviously, the higher the design resilient modulus of a subgrade, the higher will be the tensile stress on the asphalt.

Table 2: Resilient modulus results

Densitykg/m3

Resilient modulus constantsMR

Value(MPa)

Max resilientMax

permanentStiffnessModulus

Material K1 K2 K3 Strain % Strain % (MPa)Rubber 1191 234010 -1.62 0.457 1.35 2.09 10 3.297Polystyrene 26.60 4833 -0.11 0.055 3.72 1.42 0.5 -Concrete 2208 12122 0.711 0.232 583.0 0.01 0.1 30SW, Dr = 0.76 1729 23958 0.0772 0.354 124.1 0.06 1.6 222.22SW, Dr = 0.50 1651 17316 0.00869 0.333 62.1 0.14 5.2 -SPu(Kahang), Dr = 0.71 1573 36882 0.0928 0.172 103.6 0.09 5 208.33SPu(Kahang), Dr = 0.50 1500 18376 -0.0339 0.413 74.1 0.13 16 -SPu(L.Buzzard), Dr=0.70 1620 25270 0.276 0.321 268.5 0.03 36 250SPu(L.Buzzard), Dr=0.48 1588 14261 0.457 0.236 237.6 0.05 40 -SPg, Dr = 0.70 1758 16966 0.289 0.259 151.2 0.06 0.76 285.71SPg, Dr = 0.50 1662 10417 -0.018 0.559 77.4 0.13 22.5 -

163 Alvian John Lim Meng Siang et al. / Procedia Engineering 53 ( 2013 ) 156 – 166

Fig. 7.

Fig. 8. The relationship between the resilient modulus values and the uniformity coefficient of Kahang sands

Fig. 9. The relationship between the resilient modulus values and the shape parameters

020000400006000080000

100000120000140000160000

0 5 10 15

Resil

ient

Mod

ulus

(MPa

)

Uniformity Coefficient Cu

Dr = 0.7

= 0.5Dr

0.0

50.0

100.0

150.0

200.0

250.0

300.0

0.4 0.45 0.5 0.55 0.6 0.65 0.7 0.75

Resil

ient

Mod

ulus

(MPa

)

Shape parameter

Sphericity Roundness Dr = 0.7 - - - Dr = 0.5

164 Alvian John Lim Meng Siang et al. / Procedia Engineering 53 ( 2013 ) 156 – 166

Fig. 10. The effects of the void ratio on the resilient modulus values

Fig. 11. The maximum resilient strain versus the density of the sample

Fig. 12. Relationship of the Stiffness Modulus and the Resilient Modulus

Fig. 13. Schematic diagram of the pavement layers [22]

Fig. 13. Schematic diagram of the pavement layers [22]

0.050.0

100.0150.0200.0250.0300.0

0.45 0.55 0.65 0.75 0.85

Resil

ient

Mod

ulus

(MPa

)

Void ratio e

Kahangdssan

uzzardL.BundsSan

RoundedRoundedSands

0.0

100.0

200.0

300.0

400.0

500.0

600.0

700.0

0 500 1000 1500 2000 2500

Max

Res

ilien

t Stra

in %

Density (kg/m3)

RoundedSands

Concrete

RubberPolystyrene

165 Alvian John Lim Meng Siang et al. / Procedia Engineering 53 ( 2013 ) 156 – 166

Fig 14. Tensile and compressive strains in flexible pavments [21]

5. Conclusion

The assessment of stiff (rigid), elastic (flexible) and particulate materials of different shapes and gradation characteristics will lead to a better understanding of the resilient behaviour. The main conclusions from the study are:

The permanent strain is a function of the stiffness modulus whereas the resilient strain is a function of the resilient

modulus. Elastic materials such as the rubber sample shows the MR values tend to decrease significantly as the deviator

stress increases. much.

A very stiff sample shows that the MR values do not tend to be affected by the different confining stress levels but shows significant increase in the MR values as the deviator stress increases due to its very low resilient strain.

For the sand samples, the confining stress plays an important factor in the resilient behaviour. High confining stress shows high MR values and the increasing deviator stress also tend to decrease the resilient strain, resulting in higher MR values.

Concrete, due to its high compressive strength has the lowest resilient strain as compared to a rubber sample which r). This results to large value of the design resilient modulus from the constitutive

model. SPu(L.Buzzard) has a largest value of the design MR as compared to the rest of the sands. It is also appears very

interestingly that although SPu(L.Buzzard) has very large MR values, but it also shows that it has the highest maximum permanent strain. This shows that rounder sand particles tend to have low resilient strain but are more easily to be compacted. Particle shapes tend to be a significant factor in the resilient behaviour.

Mr increases with an increase in uniformity coefficient (Cu). The effect of the void ratio (e) shows that as void ratio increases, the Mr value decreases. It can be seen that the concrete samples have the highest density with very high stiffness, resulting in very low r.

r). Acknowledgment

The authors would like to acknowledge the Research Center for Soft Soil (RECESS) laboratory, the laboratory staffs and

the advices from fellow researchers.

References [1] Lekarp, F., Isacsson, U. And Dawson, A. State of the Art I: Resilient Response of Unbound Aggregates. Journal of Transporting Engineering.

Jan/Feb 66-75, 2000. [2] Nazarian S., & Feliberti M., Methodology for resilient modulus testing of cohesionless subgrades, transportation research record 1406, National

Research Council, Washington, D.C. pp. 108-115., 1983 [3] Tompson M.R., Factors affecting the resilient moduli of soils and granular materials, workshop on resilient modulus testing, Oregon State

University, Corvallis, Oregon., 1989. [4] Hicks G.R., and Monismith C.L. Factors influencing the resilient response of granular materials, Highway Research Record 345, National Research

Council, Washington, D.C., pp. 15-31, 1981 [5] Kim, D., and N. Z. Siddiki. Simplification of Resilient Modulus Testing for Subgrades. Joint Transportation Research Program, Indiana Department

of Transportation and Purdue University, West Lafayette, Indiana, 2006.

166 Alvian John Lim Meng Siang et al. / Procedia Engineering 53 ( 2013 ) 156 – 166

[6] Davish P., Labuz J.F., Guzina B., Drescher A. Small strain and resilient modulus testing of granular soils (Final Report). Minesota Departmet of transportation research services section, 2004.

[7] Mathew, Significance of cyclic confining stress inrepeated load triaxial testing of granular material. In Transportation Research 113 Record 537, Transportation Research Council, Washington, D.C., pp 49-58, 2003.

[8] Whitlow,R. (2001), Basic Soil Mechanics, Prentice Hall, Harlow. [9] AASHTO Designation: T 307-99. Standard method of test for determining the resilitn modulus of soils and aggregate materials. 2003 [10] Pezo, R. A general method of reporting resilient modul nd Annual meeting of transportation

research board, Jan, 10 14, Washington, D.C., 1993. [11] Mohammad, C., Puppala, A. and Alavilli, P., Effect of strain measurement on resilient modulus of sands. Dynamic Geotechnical Testing II, ASTM

STP 1213, Ronald J. Ebelher, Vincent P. Drnevich, and Bruce L. Kutter, Eds, American Society of Testing and Materials, Philadelphia, 2004. [12] Moussazadeh, J. And Witczak, M. Prediction of subgrade moduli for soil that exhibits nonlinear behaviour, Transportation research board,

Transportation research record 810: 9-17, Washington, D.C., 1981. [13] Al-Suhaibani, A., Al-Refeai, T. and Noureldin, A. Characterization of subgrade soil in Saudi Arabia; A study of resilient behaviour, KACST Project

No. Ar-12-51, Final report, 1997 [14] Head, K.H. Manual of soil laboratory testing. Vol 1: Soil classification and compaction test. Pentech Press, London, 1992. [15] Powers, M. G. A new roundness scale for sedimentary particles. Journal ofSedimentary Petrology 23, No. 2, 117-119., 1953 [16] Krumbein, W. C. Measurement and geological significance of shape and roundness of sedimentary particles. Journal of Sedimentary Petrology 11,

No.2, 64-72. (1941). [17] Cho, G.C, Dodds, J. & Santamarina, J.C. Particle shape effects on packing density, stiffness and strength: Natural and crushed sands. Journal of

Geotechnical and Geoenvironmental Engineering, ASCE, Vol. 132, No 5, 2006. [18]

Vol. 47, Number 5.pg 539 to 551., 2010. [19] ngth and dilatancy

characteristics. International Journal of Integrated Engineering. (In Press), 2012 [20] Ladd R.S. Preparing test specimen using undercompaction. Geotechnical Testing Journal 1(1), pp. 16 23, 1978. .[21] Yang, H.,H. Pavement analysis and design. Second Edition. Presntice Hall, Upper Saddle River, 2004. [22] Harold N.,A. Highway Materials, Soils, and Concretes, 3rd Edition. Prentice Hall, Columbus, Ohio, USA