- Email: [email protected]
An investigation on the geotechnical properties of sand–EPS mixture using large oedometer apparatus
Construction and Building Materials 113 (2016) 773–782
Contents lists available at ScienceDirect
Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat
An investigation on the geotechnical properties of sand–EPS mixture using large oedometer apparatus Reza Jamshidi Chenari a,⇑, Mehran Karimpour Fard b, Sabina Pourghaffar Maghfarati c, Faranak Pishgar d, Sandro Lemos Machado e a
Department of Civil Engineering, Faculty of Engineering, The University of Guilan, P.O. 3756, Rasht, Guilan, Iran Department of Civil Engineering, Faculty of Engineering, The University of Guilan, Iran Azad University, Iran d Islamic Azad University, Fouman and Shaft Branch, Fouman, Guilan, Iran e Universidade Federal da Bahia, Brazil b c
h i g h l i g h t s Deformation and constrained modulus of EPS–sand mixture was investigated. Permeability and hydraulic conductivity of EPS–sand mixture was obtained. A new fabricated large size oedometer apparatus was designed and tested. A K0 coefficient measurement was designed through a ring pressure cell inside the large oedometer apparatus. Unconstrained deformation modulus to be back-calculated from oedometeric gauge readings was for the first time calculated.
a r t i c l e
i n f o
Article history: Received 16 July 2015 Received in revised form 1 March 2016 Accepted 17 March 2016 Available online 25 March 2016 Keywords: Expanded poly styrene (EPS) Permeability Deformation and strength characteristics Large scale oedometer
a b s t r a c t The applicability of Expanded PolyStyrene (EPS) beads mixed with sand in five different contents was investigated for use in geotechnical engineering applications utilizing a newly designed and fabricated large scale oedometer apparatus. Permeability, coefficient of earth pressure ‘‘at rest” and the volume compressibility coefficient were measured and calculated for different EPS beads contents. Consolidation and permeability tests were conducted under different overburden pressures. The main findings of major recent studies were compared with current study and verification of new results was undertaken. Permeability, coefficient of earth pressure ‘‘at rest” and the volume compression coefficient along with some important deformation and strength characteristics parameters namely, the internal friction angle, constraint modulus and the drained 3-D elasticity modulus were investigated. The Results revealed that permeability, internal friction angle, constraint modulus and 3-D Young modulus decreases with inclusion content. However, the volume compressibility coefficient and the K0 coefficient showed opposite trends. Predictive models were submitted in forms of Multi Linear Regression, MLR simulations and their performances were evaluated. Ó 2016 Elsevier Ltd. All rights reserved.
1. Introduction There have been growing interest in the use of nonconventional as lightweight geo-materials, and their introduction has presented both opportunities and challenges to researchers and engineers worldwide. Attention is specifically needed to be ⇑ Corresponding author. E-mail addresses: [email protected] (R. Jamshidi Chenari), [email protected] (M. Karimpour Fard), [email protected] (S. Pourghaffar Maghfarati), [email protected] (F. Pishgar), smachado@ ufba.br (S. Lemos Machado). http://dx.doi.org/10.1016/j.conbuildmat.2016.03.083 0950-0618/Ó 2016 Elsevier Ltd. All rights reserved.
drawn to the consideration of both the cost and environmental implications when any new material is introduced into construction. In recent decades, a successful conversion from academic excellence to commercial viability has occurred. Lightweight fill materials have a wide range of civil engineering applications around the world. They may be used as fill over soft clay sites to prevent excessive settlement; as backfill for retaining and basement walls to reduce the horizontal driving forces; as fill material to increase factor of safety for slopes by reducing driving forces; as seismic buffers to alleviate seismic forces, and so forth. Various types of lightweight fills, such as Expanded PolyStyrene (EPS)-block geofoam, EPS beads, tire waste products like shredded
774
R. Jamshidi Chenari et al. / Construction and Building Materials 113 (2016) 773–782
tires, tire chips and tire crumbs have been reported. Using such materials not only provides lightweight fill solution for civil engineering projects, but also helps to save the environment by recycling these materials instead of stockpiling them. EPS-block geofoam gained popularity due to its wide application areas such as compressible inclusions [14,15,22], reduction of swelling pressure caused by expansive sub-soils ([6,16], fill materials in highway embankments [28] and remediation of sandy slopes [1,24]. In order to overcome shortcomings of EPS-block geofoam namely transportation issue, lack of formability to fill irregular volumes and compatibility with on-site soil, lightweight fill material composed of a mixture of EPS beads with soil and alternative additives is recommended [29–31,20,32,9,10,21,26,11]. Miao et al. [21] mixed dredged sand with EPS-beads and Portland cement. They conducted different geotechnical tests to investigate their potential use for mitigation of settlement problems associated with bridge approach embankments over soft soil. Padade and Mandal [25] proposed a geomaterial by blending fly ash instead of soil with EPS-beads and cement. Using compression tests they showed that the compressive strength of EPGM (Expanded Polystyrene-bead Geo-Material) increases considerably if cement-to-fly ash ratios of 10%, 15% and 20% are used. Some researchers mixed EPS particulates with sand to create a lightweight fill and measured the stress–strain characteristics of the modified soils in the laboratory using direct shear and triaxial compression tests [20,32,9,10,21,11]. Deng and Xiao [9,10] studied the stress–strain behavior of EPS–sand for a single type of EPS bead– sand mixture. They showed a systematic decrease in drained strength with increasing EPS content. So far previous studies for lightweight geo-materials are mainly about the mechanical behaviors. When used in practical engineering, lightweight soil may undergo various hydraulic and seepage loads; so it is necessary to investigate its hydraulic properties along with the mechanical and stress–strain behavior. The main aim of this study is to conduct some large Oedometer tests on fine grained sand mixed with EPS beads with different weight contents to investigate its effect on mechanical and hydraulic properties of the proposed lightweight geo-material.
2. Experimental study 2.1. Materials The experiments are carried out on ‘‘Chamkhaleh Sand”, supplied from Chamkhaleh Beach adjacent to Chamkhaleh River, located on SW of Caspian Sea. The particles are quartz–based with gray color and uniform particle distribution as shown in Fig. 1. The index properties of the used host sand are given in Table 1. The specific gravity was determined according to ASTM D 854. Maximum and minimum dry unit weights were determined based on ASTM D 4253 [4] and ASTM D 854 [5], respectively. It had a specific gravity of 2.63, a maximum dry unit weight of 16.1 kN/m3 (i.e., minimum void ratio of 0.63) and a minimum dry unit weight of 14.2 kN/m3 (i.e., maximum void ratio of 0.85). The particle size distribution of the sand is given in Fig. 2. The sand had a coefficient of uniformity of 1.54, a coefficient of curvature of 0.95, and was classified under the Unified Soil Classification System as SP (poorly graded sand) and under the AASHTO Soil Classification System as A-3. The friction angle of the sand was 38° when its relative density Dr was 60%. The EPS beads used in this study is super light polymer foam, pre-puffed from polystyrene resin, provided by a local EPS block moulding company which had been manufacturing EPS geofoam blocks. The beads were white, even, and spherical, sizing between 2–7 mm (Fig. 3). The relevant index properties of EPS bead are presented in Table 1. Determination of the unit weight and specific gravity of the EPS beads were conducted employing a procedure modified from comparable standard test method for fine aggregates (i.e., ASTM C128 [3]). Beads were placed into a 1-L hydrometer until the volume of the hydrometer was apparently occupied. Beads were placed into the hydrometer without noticeable compaction effort so as to reach a moderate compaction state. Unit weight of the beads is then easily calculated by scaling the net weight of beads, used to fill the bottle. The unit weight obtained for EPS beads was 0.08 kN/m3. Specific gravity (Gs) of beads was also calculated by filling the voids between EPS beads with distilled water and then to calculate the net volume of beads and determine the specific gravity of beads, which was 0.013.
Fig. 1. Magnified photos of Chamkhaleh Sand [12]. 2.2. Specimen preparation EPS–sand specimens (Fig. 4) were formed by mixing EPS beads with sands at a dry mass ratio g of EPS beads over sands, which was thought as the most significant factor controlling the unit weight, mechanical and hydraulic behavior of the mixtures. Investigated ratios were 0.1%, 0.2%, 0.3%, 0.5% and 1% for EPS beads by weight. Sufficient water was added to the EPS–sand samples. The added water provided bonding between specifically the EPS beads and sand, which made it possible to mix without segregation. Initial water content was measured precisely to control the target dry density of prepared samples for test. For each designated mixing ratio, the mass-based proportions of sand, EPS bead were determined beforehand. The proportioned materials were mixed thoroughly until the mixtures were homogenous enough. Material was molded into the chamber of oedometer apparatus at a designated weight and dry density of 14.6 kN/m3. To reach the target dry density, a handy tamper was employed. It should be noted that the dry density was defined based on total solid constituents, namely EPS beads and sand particles. Some researchers try to maintain a constant skeletal or matrix dry density for sand portion [2], however authors believe that considering a constant skeletal relative density is a tricky situation which is strictly affected by compressibility of inclusions and needs precise estimate of sand portion volume. In real world applications, practitioners would rather to adopt a simple procedure based on either maintaining a constant total dry sand portion weight assumption [18] or adopting a constant overall bulk dry density which is the case in this study. Specifically, the target dry density of EPS–sand mixtures was achieved by quantifying the weight of proportioned EPS and sand to be placed within a volume (i.e., 57,000 cm3). Compaction was completed in seven layers. A list of specimens and corresponding weight and volume ratios is provided in Table 2 for EPS beads mixed with sand. Totally 6 large Oedeometer tests were performed in this investigation. The first test was conducted on employed host sand. Each mixture was tested at three different overburden pressures, 160, 260 and 375 kPa. The volumetric ratio of EPS over the combination of EPS and sand in the mixture, v, was calculated based on the mixing ratios and specific gravities of particles for each mixture, as presented in Eq. (1). If the EPS content increased above 1%, the EPS volumetric ratio exceeded 70%, which substantially led to reduced sand inter-particle interactions. g
v ¼ 100GSEPS g Gss
þ GsEPS
ð1Þ
775
R. Jamshidi Chenari et al. / Construction and Building Materials 113 (2016) 773–782 Table 1 Physical properties of tested materials. Material
Specific gravity, Gs Dry unit weight (kN/m3) Effective size D10 (mm) Mean grain size D50 (mm) Uniformity coefficient, Cu Coefficient of curvature, Cc
Sand 2.63 EPS bead 0.013
14.2 (min) 16.1 (max) 0.08
0.17 2.60
0.21 4.00
1.54 1.70
0.95 0.90
Fig. 2. Grain size distribution curves of EPS bead and sand used for mixtures.
Fig. 4. Sample preparation for EPS–sand mixture.
Table 2 Summary of Oedometer testing program on EPS–sand mixtures. Designation Content by weight g(%)
Content by volume v(%)
Dry density (kN/m3)
Overburden pressure (kPa)
Sand EPS–Sand-1 EPS–Sand-2 EPS–Sand-3 EPS–Sand-4 EPS–Sand-5
0 16.8 28.8 37.8 50.3 67
14.6 14.6 14.6 14.6 14.6 14.6
160, 160, 160, 160, 160, 160,
0 0.1 0.2 0.3 0.5 1
260, 260, 260, 260, 260, 260,
375 375 375 375 375 375
2.3. Large oedometer testing setup
Fig. 3. Optical microscope photos of EPS beads. where v denotes the EPS volumetric percentage in the mixture and GsEPS and Gss denote the specific gravities of EPS bead and sand, respectively. After sample preparation completion, a perforated plastic pad was placed over the specimen and loading plunger along with the top cap was fixed to the position and saturation process was commenced. Two O-rings were placed around the cylindrical plunger to seal the chamber. The specimen was saturated using a very small upward hydraulic gradient, from bottom of samples to expel air in voids among particles. The saturation was thought to be completed when no air bubble was observed to be escaped from the top drainage line.
Most commercial geotechnical laboratories have been utilizing some kinds of Casagrande-type oedometers [7] for routine measurements of consolidation parameters of fine-grained soils for decades. A large specimen may be more representative of field conditions in some civil engineering applications. Very few apparatuses are available for the measurement of consolidation parameters of large specimens as recovering large undisturbed specimens is difficult, expensive and time-consuming. The hydraulic oedometer developed by Rowe and Barden [27] is available for specimen diameters ranging typically from 75 to 254 mm (BSI 1990). Although the size of the specimen is increased significantly, it may still be too small for many civil engineering applications. A large cylindrical Oedometer with a diameter of 492 mm and height of 550 mm to measure the load-deformation parameters of geo-materials under a vertical applied stress up to 1.5 MPa was designed, fabricated and calibrated at the engineering school of the University of Guilan. This oedometer was employed to evaluate the hydraulic and mechanical properties of non-conventional materials like EPS Geofoam, tire shreds, tire chips and TC mixed with soils. The fabricated large Oedometer is not conceptually different from the large Casagrande-type Oedometer designed and fabricated by Ng et al. [23]. The apparatus consists of a loading frame, a computer-controlled hydraulic loading plunger, an Oedometer cell, displacement, pore pressure, normal and lateral pressure transducers, and a computerized data acquisition system. In Fig. 5, the complete setup of oedometer and its major components are illustrated. The primary function of the loading frame is to provide a robust support for the hydraulic loading actuator and the Oedometer cell, and a reaction for the hydraulic loading actuator to apply the load to the specimen. A 50 mm thick steel plate is used to support the Oedometer cell to minimize any deflection of the cell base and to distribute the load to the supporting members uniformly. The steel plate is supported by four wheels that transmit the load to the supporting frame members. The base plate is equipped with five outlet ports connected to a drainage line to facilitate bot-
776
R. Jamshidi Chenari et al. / Construction and Building Materials 113 (2016) 773–782
tom drainage and excess pore water pressure measurements. Stainless steel Oedeometer cell of 5 mm thick, 500 mm outer diameter and 500 mm high was conceived strong enough to assume that lateral deformation of the new Oedometer cell is negligible even when a very high vertical stress is imposed on the specimen. It should be noted that the thickness of the specimen is 200 mm shorter than the height of the cell to allow for the storage of free water atop the specimen. This means that the target specimen thickness is 300 mm. A loading piston of 80 mm diameter is connected to a plunger of 488 mm diameter with 500 mm stroke to convert and apply hydraulic pressure to the specimen top pad. A schematic of the oedometer cell is depicted in Fig. 6. A high pressure line is provided atop the specimen for maintenance of a constant prescribed pore water pressure at the top boundary of the specimen to permeability measurement purposes. The new large Oedometer has some major features as listed: Ability to apply 1500 kPa reciprocating pressure with 500 mm stroke for highly compressible materials. Measurement of the coefficient of at-rest pressure, K0 by bottom and lateral hydraulic earth pressure transducers. Measurement of the coefficient of permeability in a constant head permeability test scheme provision. A few compressibility tests were performed on sand material considered rapiddraining soil to evaluate the applicability of the new apparatus in comparison to the standard size Casagrande-type oedometer. The e-log r curves of the same material obtained from the newly designed large oedometer and a standard size Casagrandetype oedometer are presented in Fig. 7. Two measurements performed on each of the apparatuses are presented. The repeatability of results are clear, indicating the precise performance of newly designed apparatus in measurement of deformation characteristics of soil.
3. Results and discussion 3.1. Permeability Permeability is an important factor that significantly influences the behavior of fill under saturated conditions, and often dictates the suitability of fill for specific applications. This is because a well-drained fill prevents the development of excess pore water pressure during loading and also accelerates consolidation of
underlying low permeability foundation soils by providing a drainage path [8]. According to Fell et al. [13], a material that has permeability greater than 1 106 m/s is considered as a good drainage material. In this study, a series of modified constant head permeability tests adapted to the large oedometer setup are conducted on the sand–EPS beads mixtures and the results are shown in Fig. 8. The permeability tests were conducted after applying the overburden stress in each stage of compressibility test and stabilizing related deformations. A constant hydraulic head was applied through a water reservoir pressurized by air as shown in Fig. 6. Applying flow through sample was continued after a constant discharge rate was recorded. Coefficient of permeability is then calculated under a constant head assumption. It can be seen that there is a general trend of decreasing permeability with the increase of overburden pressure and also the admixture content. This decrease in permeability may be attributed to the joint effects of the skeletal relative density of the sand and the impermeable inclusions. As the amount of EPS beads increases, the skeletal relative density of sand increases due to the constant overall dry density of the mixture and the EPS beads clog the sand matrix and introduces blockage of the water flow lines inside the mixture and thus permeability decreases. This effect is more highlighted when overburden pressure is increased because high overburden pressure converts the EPS beads into impermeable membranes looked as cushions for sand particles. It can also be seen that the permeability of all mixtures used in this study are greater than 1 106 m/s, which indicates good drainage materials. Another observation from Fig. 8 is that an optimum EPS content is expected when there is negligible variation between the 0.5% and 1% EPS beads-sand mixtures. In order to isolate the effects of EPS beads and overburden pressure, a series of standard constant head permeability tests was conducted as superimposed on Fig. 8, implying zero overburden pressure. Two observations are made, first it is observed that the EPS beads inclusion effect is more highlighted in zero overburden pressure. In high overburden pressure when the EPS beads are squeezed, the skeletal relative density is the controlling factor. Second observation is that overburden pressure higher than 160 kPa has no extra effect on permeability of the mixture. 3.2. Lateral Earth pressure coefficient Lateral earth pressure is the pressure that soil exerts in the horizontal direction. The lateral earth pressure is important because it affects the consolidation behavior and strength of the soil and because it is considered in the design of geotechnical engineering structures such as retaining walls, basements, tunnels, deep foundations and braced excavations. The coefficient of lateral earth pressure, K is defined as the ratio of the horizontal effective stress, r’h, to the vertical effective stress, r’v. For a level ground deposit with zero lateral strain in the soil, the ‘‘at-rest” coefficient of lateral earth pressure, K0 is obtained. At rest lateral earth pressure, represented as K0, is the in situ lateral pressure. It can be measured directly by a dilatometer test (DMT) or a borehole pressuremeter test (PMT). As these are rather expensive tests, laboratory scale oedometer test can be employed to have an estimate of the K0 if virgin and undisturbed samples are taken from soil deposit. Empirical relations have also been created in order to predict ‘‘at-rest” pressure with less involved soil testing, and relate to the angle of shearing resistance. Jaky [17] proposed a famous equation for normally consolidated soils:
K0 ¼ 1 Sinu0
Fig. 5. Large Oedometer testing apparatus.
ð2Þ
In the fabricated oedometer apparatus, the bottom pressure cell is employed to measure vertical pressure. The cell consists of two 30 cm circular stainless steel plates welded together around their
R. Jamshidi Chenari et al. / Construction and Building Materials 113 (2016) 773–782
777
Fig. 6. Details of the large Oedometer cell.
Fig. 7. Evaluation and repeatability check of the new large Oedometer apparatus.
Fig. 8. Effect of EPS content on the permeability of EPS–sand mixture.
periphery and spaced apart by a narrow cavity filled with de-aired oil. Changing overburden pressure squeezes the two plates together causing a corresponding increase of fluid pressure inside the cell. An electrical pressure transducer converts this pressure into an electrical signal which is transmitted to the data logger. In the case of lateral earth pressure the same procedure was employed. In this case, 10 cm belt-like cell pressure has been attached to the inner perimeter of oedometer cell. Test procedure in large oedometer apparatus involves spontaneous application of overburden pressure of specified magnitudes to the specimen in undrained condition. Lateral and vertical earth pressure transducers render continuous reading of the earth pressure in horizontal and vertical directions respectively. A dynamic variation of the coefficient of earth pressure ‘‘at-rest” is provided in Fig. 9 for sand mixed with EPS beads under three different overburden pressures. It indicates that the K0 starts from 1 at the beginning and stops at a final value when the induced pore pressure dissipates. In fact, it moves from a full undrained condition at the beginning when the overburden pressure is applied, to a full drained condition at end when all of the applied pressure is fully transferred to the skeleton. The ultimate values of K0 are then drawn for different inclusion contents and overburden pressures as illustrated in Fig. 10. It is observed that both overburden pressure and inclusion content are clearly affecting the K0 coefficient. According to Deng and Xiao [10], depending on the EPS content there are three major interaction mechanisms. Sand–sand, sand– EPS and EPS–EPS mechanisms play dominating effect due to the rigidity difference between sand and EPS beads. EPS beads behave like balloons filled with air. In low overburden pressure when there is no remarkable deformation in beads, especially when the inclusion content is high, the EPS–EPS mechanism controls and the mixture behaves like fluid and consequently a hydrostatic fluid condition is dominant and thus the K0 value is not far from the incompressible fluid state. In Fig. 9(f), at lower overburden stress of 160 kPa, the value of K0 is considerably higher for the g = 1% in comparison to lower EPS contents. Overburden pressure, however is affecting the coefficient of lateral earth pressure in a different way. Increasing the overburden pressure causes compression of EPS beads and thus mobilization
778
R. Jamshidi Chenari et al. / Construction and Building Materials 113 (2016) 773–782
Fig. 9. Time variation of the coefficient of earth pressure ‘‘at-rest” for EPS–sand mixture; a) g = 0%; b) g = 0.1%; c) g = 0.2%; d) g = 0.3%; e) g = 0.5% and f) g = 1%.
of shear strength is expected by inducing an apparent cohesion to the mixture due to the mobilized reinforcing effect. This is equivalent to virtually applying more confining pressure and thus less horizontal pressure is activated due to the virtual confinement induced. The implication is that the K0 decrease with an increase in overburden pressure. Another justification is that the sand–sand mechanism becomes more highlighted when the overburden pressure is remarkable. Due to the rigidity difference and higher level of overburden pres-
sure, the EPS components deform and experience substantial volumetric compression, causing densification of the mixtures remarkably more than a low consolidation pressure did. Accordingly, a more compact and rigid structure led to a decreased K0 value. Internal friction angle of the sand mixed with EPS is backcalculated from Jacky’s equation as described earlier and depicted in Fig. 11 for different inclusion contents and overburden pressures. Results show that the internal friction angle decreases as
R. Jamshidi Chenari et al. / Construction and Building Materials 113 (2016) 773–782
Fig. 10. Effect of EPS content on the K0 coefficient of sand mixtures.
the inclusion content increases. Results of current study are qualitatively comparable with findings of others, although direct comparison is not possible due to different shear failure mechanisms. Deng and Xiao [10] reported three principal shear mechanisms in terms of sand–sand, sand–EPS and EPS–EPS granular interaction. Particle rigidity difference between sands and EPS beads as well as the packing difference among different compositions leads to higher shear strength mobilization for sand–sand mechanism in comparison to sand–EPS and EPS–EPS mechanisms. This means that increasing EPS content might decrease the friction angle. Karimpour et al. [19] investigated shear strength behavior of sand–EPS beads mixture in low weight contents. The study was conducted employing a large direct shear test apparatus. Different inclusion contents were tested under a wide range of overburden pressures. Results clearly demonstrated that the internal friction angle decreases with the increase of EPS content, again confirming the findings of current study.
are conducted to investigate the compressibility of specimens of sand mixed with EPS or TC. The specimens are subject to overburden pressures of, 160, 260 and 375 kPa, and the data logger keeps interrogating the pore pressure transducer until a zero pore pressure is reached. Effective stress is calculated while maintaining constant overburden pressure during each test. A potentiometer reads and measures the vertical displacement induced in specimen when the sample is laterally constrained. The strain–stress behavior observed on the sand–EPS beads mixtures are shown in Fig. 12 for different overburden pressures. In order to evaluate the behavior of the EPS–sand mixtures, a control experiment using pure sand was conducted. The results of this control experiment can be used to normalize the volume compressibility coefficient (mv) and other measured and calculated parameters obtained from EPS–sand mixture samples for subsequent regression and modeling analyses. Illustrations provided in Fig. 12 clearly show that the strain– stress behavior of sand–EPS mixtures is strongly affected by the inclusion content. EPS inclusion leads to the increased volume compressibility of the mixture. This means that the rate at which the oedomteric sample deforms, increases due to the inclusion of less stiffer aggregates. However, the overburden pressure has a different effect when looking into Fig. 12. Under high overburden pressure, the EPS beads are subject to significant volumetric compression. This resulted in densification of EPS beads within the mixture, which is the cause of the strain hardening behavior, and the stress–strain behavior of the mixture is mainly dominated by EPS beads [11]. The relationship between g and mv is presented in Fig. 13 for EPS–sand mixture. Whereas g increases with all other variables remaining constant (overburden pressure in oedometeric specimens), the mv increases linearly (Fig. 13). The mv is a function of EPS content in the mixture and overburden pressure. The volume compressibility coefficient of EPS–sand mixture was decreased with an increase in overburden pressure (Fig. 13). Constraint modulus was calculated from the volume compressibility coefficients depicted in Fig. 13 by inversing mv values. Furthermore, adopting an oedometric condition will lead to the derivation of Poisson’s ratios from K0 values (Eq. (3)). Drained elasticity modulus for 3-D condition was then calculated from Eq. (4) and illustrated in Fig. 14.
3.3. Composite compressibility
K0 ¼ The stability of any structures made by geo-materials, depends mainly on the strength and deformation characteristics of the materials used. A series of large oedometer stress controlled tests
Fig. 11. Effect of EPS content on the internal friction angle of sand mixtures.
779
E0 ¼
m0
1 m0
ð1 þ m0 Þð1 2m0 Þ 0 EOed ð1 m0 Þ
ð3Þ ð4Þ
Fig. 12. Effect of EPS content on the stress-strain behavior of sand mixtures.
780
R. Jamshidi Chenari et al. / Construction and Building Materials 113 (2016) 773–782
Fig. 13. Effect of EPS content on the volume compressibility coefficient of sand mixtures.
where m’ is the Posisson’s ratio, E’ is the drained 3-D elasticity modulus obtained conventionally from triaxial CD or even CU tests and E0Oed ¼ Mv ¼ m1v is the drained constrained modulus computed from the volume compressibility coefficient obtained from the oedometric tests. Superimposed on the same graph are the findings of Edinçliler and Özer [11] for undrained condition obtained from CU triaxial tests demonstrated for comparison. Results show that both drained and undrained elasticity moduli decrease as the EPS content increases. This is as expected as the rigid sand particles are replaced with compressible EPS beads and EPS–EPS interaction mechanism dominates when the EPS beads content rises. Another observation from Fig. 14 is that confining pressure and overburden pressure have the same effect as explained by Edinçliler and Özer [11] earlier. Lower overburden pressure or confining pressure in triaxial sense leads to formation of sand–EPS or EPS– EPS mechanisms which pose more deformation potential and less composite rigidity is then exhibited. 4. Skeleton compressibility In the case of an ordinary soil sample, compressive volume change is caused by a reduction in pore phase owing to the rear-
Fig. 14. Effect of EPS content on the drained and un-drained elasticity modulus of sand mixtures.
rangement of the soil skeletal structure. This is why a soil particle is assumed to be almost rigid. However, in the case of a mixed soil sample containing deformable particles such as EPS beads, larger compression occurs caused by additional volume change owing to the compression of the deformable particles. In such a situation, the additional volume change was thought to consist of: a volumetric compression of the deformable particles themselves and further pore phase reduction by the rearrangement of the particle skeleton caused by the deformation of deformable particles. EPS beads are soft inclusions which are subject to compression and volume change during overburden application in large oedometer experiments. Both in drained and undrained loading schemes, EPS beads undergo compression. However, differentiation between the mixture compressibility and the skeleton volume change is difficult and impossible within a single loading scheme. For this reason, in addition to the drained loading cases, a few large oedometer experiments were conducted in undrained condition to isolate different effects as discussed before. Fig. 15 shows the variation of oedometric strain with the overburden stress for both drained and undrained experiments. Two observations are deduced from this figure. The influence of the skeleton compression exhibited as volume compression in undrained loading condition is noticeable and becomes more remarkable when the inclusion content increases. Another observation is that a nontrivial volume compression was observed for pure sand model which should not be considered as skeleton compression because sand particles are rigid elements and do not undergo volume change during undrained loading condition. This volume change is attributed to the compression of air bubbles trapped inside the mixture.
5. Predictive models 5.1. MLR regression analysis Mechanical and hydraulic properties of sand mixed with EPS are not only a function of g but also depend on the overburden pressure applied to oedometric specimens. In addition, test results indicate that some of these parameters linearly vary with increasing input parameters. By using the linear behavior obtained from the test results, multiple linear regression (MLR) models were run to investigate the predictive performance and reasonableness of the developed regression equations. Ultimately, the independent variables chosen for the final MLR models to predict these param-
Fig. 15. Effect of inclusion compressibility on overall volume change of the mixture.
R. Jamshidi Chenari et al. / Construction and Building Materials 113 (2016) 773–782 Table 3 Data variables sets and MLR models for oedometer parameters. Data set Independent variables g K (EPS) r 100 , Pa g K0 (EPS) r 100 , Pa mv (EPS) g , r 100 Pa
R2 (%) Equationa g 0:7832 r 0:9434 ¼ 0:1796 100 Pa 0.8312 K 0EPSSand g 0:1292 r 0:3160 ¼ 1:7435 K 0Sand 100 Pa 0.8302 mv EPSSand g 0:5080 r 0:5831 ¼ 4:9669 mv Sand 100 Pa 0.9866
K EPSSand K Sand
a From the MLR model given in Eq. (5), and regression output by using Matlab MathWorks.
Fig. 16. Comparison of measured and predicted performance for different parameters from oedometric tests.
eters were: g and the overburden pressure (r). These variables were used to predict mechanical and hydraulic parameters, by dividing them into six different models, as presented in Table 3. From an application standpoint, the regression models should not be dependent on units. Consequently the variables were normalized with those of pure sand in the mixture and r was normalized with atmospheric pressure (Pa) to make the regression variables dimensionless. All regression analyses shown in Table 3 were performed using MATLAB MathWorks. These models have the general form:
y ¼ n0 xn11 xn22
ð5Þ
This equation is re-written in the following form:
log y ¼ logn0 þ n1 logx1 þ n2 logx2
ð6Þ
A comparison of predicted parameters in Table 3 and laboratory measured values from large oedometer tests in normalized forms can be seen in Fig. 16. As shown in Fig. 16 models from Table 3 provide reasonably close prediction of the laboratory results for EPS– sand mixture. 6. Conclusion The applicability of EPS–sand mixtures was investigated for use in geotechnical engineering applications utilizing a newly designed and fabricated large size oedometer apparatus. The new designed large size oedometer apparatus was first calibrated and tested for repeatability and accuracy. For this aim some standard Casagrande type oedometer tests were conducted as well and results were compared with those of new fabricated large size oedometer. Permeability, coefficient of earth pressure ‘‘at rest” and the volume compressibility coefficient were measured and calculated for dif-
781
ferent EPS contents. A comprehensive discussion and comparison with similar studies found in literature primarily in forms of triaxial drained and un-drained tests was undertaken. The main findings of major recent studies were superimposed on graphs of current study and verification of new results was accomplished. Predictive models were also worked out in order to provide an efficient regression based prediction of the parameters under study. For this aim, multi linear regression (MLR) procedure was employed to render predictions comparable with measured values. The results obtained in this study led to the following conclusions: 1- The new designed and fabricated large size oedometer apparatus proved an efficient tool to reflect hydraulic and mechanical behavior of non-conventional materials. Results obtained from the evaluation program indicate that the new apparatus is comparable to the typical Casagrande-type oedometer. 2- Permeability of EPS–sand mixtures exhibited a decrease with inclusion content due to replacement of sand particles with impermeable EPS beads. 3- Overburden pressure has a decreasing effect on permeability of sand mixed with EPS. The reason is that increasing effective stress leads to higher inter-granular stresses which in effect induce more compression in EPS beads and less permeability is thus expected. 4- EPS beads are compressible materials transferring overburden pressure to different directions and behave more or less similar to incompressible fluids as long as the overburden pressure is low. The increase in inclusion content leads to an increase in K0 values. However, this behavior vanishes when the effective stress dramatically rises. Indeed the sand–sand or sand–EPS mechanisms are controlling effects in such case and rigidity of mixture increases, implying smaller Poisson’s ratio and thus smaller K0 values are rendered. 5- Internal friction angle for sand mixed with EPS was backcalculated from K0 values adopting Jacky’s equation. Results showed that the internal friction angle decreases as the inclusion content increases due to the sand–sand or sand– EPS mechanisms to be less effective in higher inclusion content. Overburden pressure; however increases interaction between sand particles and the internal friction angle is thus increased as expected. 6- Volume compressibility coefficient of the EPS–sand mixtures is a function of the equivalent rigidity of the mixture which is a function of both inclusion content and the overburden pressure. EPS inclusion induces less rigidity to the mixture and thus more deformability and compressibility is expected. This effect decreases with overburden pressure due to the densification induced by overburden pressure. 7- Drained elasticity modulus of the EPS–sand mixture was also extracted indirectly from constraint modulus values and the Poisson’s ratios back-calculated from K0 values. Results showed that inclusion content leads to dramatic decrease in elasticity modulus when the EPS content is less than 0.5% by weight. However, the rate of decrease ceases for higher contents implying an optimum EPS content of 0.5%. 8- Predictive models for permeability, coefficient of earth pressure ‘‘at rest” and the volume compressibility coefficient adopting MLR procedure proved comparatively acceptable and efficient in simulation of these parameters. Although a decrease in permeability, internal friction angle and constraint modulus is not desirable in most geotechnical
782
R. Jamshidi Chenari et al. / Construction and Building Materials 113 (2016) 773–782
applications, the lightweight EPS beads might have other benefits due to the low unit weight and specific gravities in comparison to quartz sand. Very low inertial force in seismic applications provokes one to consider such material as backfill material for retaining walls, trench filling material, etc. References [1] Onur Akay, A. Tolga Özer, Garey A. Fox, Steven F. Bartlett, David Arellano, Behavior of sandy slopes remediated by EPS-block geofoam under seepage flow, J. Geotext. Geomembr. 37 (2013) 81–98. [2] M. Amel-Sakhi, Influence of optimized tire shreds on strength of sandreinforced with tire shreds, in: In Partial Fulfillment of the Requirements the Degree of Master of Science, Faculty of Engineering, Isfahan University of Technology, Isfahan, Iran, 2001. [3] ASTM C128, in: Standard Test Method for Density, Relative Density (Specific Gravity), and Absorption of Fine Aggregate, 2012. [4] ASTM D 4253, in: Standard Test Methods for Maximum Index Density and Unit Weight of Soils Using a Vibratory Table, 2006. [5] ASTM D 854, in: Standard Test Methods for Specific Gravity of Soil Solids by Water Pycnometer, 2014. [6] Mustafa Aytekin, Numerical modeling of EPS geofoam used with swelling soil, J. Geotext. Geomembr. 15 (1) (1997) 133–146. [7] A. Casagrande, The Determination of Preconsolidation Load and its Practical Significance, Proc., 1st International Conference on Soil Mechanics Foundation Engineering, Cambridge, 1936. [8] H.R. Cedergren, Seepage, Drainage, and Flow Nets, third ed., John Wiley & Sons, New York, 1989, pp. 19–72. [9] An Deng, Yang Xiao, Modeling stress-strain behavior of sand-EPS beads lightweight fills based on cam-clay models, in: In Recent Advancement in Soil Behavior, in Situ Test Methods, Pile Foundations, and Tunneling@ Selected Papers From the 2009 GeoHunan International Conference, 2009, pp. 55–61. [10] An. Deng, Yang Xiao, Measuring and modeling proportion-dependent stressstrain behavior of EPS-sand mixture 1, Int. J. Geomech. 10 (6) (2010) 214–222. [11] A. Edinçliler, A.T. Özer, Effects of EPS bead inclusions on stress–strain behaviour of sand, J. Geosynthetics Int. 21 (2) (2014) 89–102. [12] K. Fakharian, F. Shabani, Monotonic and cyclic experimental results of Chamkhaleh SAND in southern Caspian Sea, in: Proceedings of the International Offshore and Polar Engineering Conference, 2011, pp. 373–380. [13] R. Fell, P. McGregor, D. Stapledon, Geotechnical Engineering of Embankment Dams, Balkema, Rotterdam, Brookfield, 1992. [14] J.S. Horvath, Expanded polystyrene (EPS) geofoam: an introduction to material behavior, J. Geotext. Geomembr. 13 (4) (1994) 263–280. [15] John.S. Horvath, The compressible inclusion function of EPS geofoam, J. Geotext. Geomembr. 15 (1) (1997) 77–120.
[16] S. Banu Ikizler, Mustafa Aytekin, Evin Nas, Laboratory study of expanded polystyrene (EPS) geofoam used with expansive soils, J. Geotext. Geomembr. 26 (2) (2008) 189–195. [17] J. Jaky, Pressure in silos, 2nd ICSMFE, London, 1, 1948, pp. 103–107. [18] C, R. Jamshidi, Dynamic Behavior of Sand Reinforced with Randomly Distributed Carpet Waste Strips behind Sheet-pile Retaining Walls in Shaking Table Tests (Ph.D. Dissertation), Iran University of Science and Technology, Tehran, Iran, 2008. [19] M. Karimpour Fard, R. Jamshidi Chenari, F. Soheili, Shear strength characteristics of sand mixed with EPS beads using large direct shear apparatus, Electron. J. Geotech. Eng. 20 (8) (2015). [20] Han-long Liu, An Deng, Jian Chu, Effect of different mixing ratios of polystyrene pre-puff beads and cement on the mechanical behaviour of lightweight fill, J. Geotext. Geomembr. 24 (6) (2006) 331–338. [21] Linchang Miao, Fei Wang, Jie Han, Lv Weihua, Li Jing, Properties and applications of cement-treated sand-expanded polystyrene bead lightweight fill, J. Mater. Civ. Eng. 25 (1) (2012) 86–93. [22] George P. Murphy, The influence of geofoam creep on the performance of a compressible inclusion, J. Geotext. Geomembr. 15 (1) (1997) 121–131. [23] A.M.Y. Ng, A.T. Yeung, P.K.K. Lee, L.G. Tham, Design, fabrication, and assembly of a large oedometer, J. Geotech. Test. 29 (4) (2006) 298–305. [24] A. Tolga Özer, Onur Akay, Garey.A. Fox, Steven.F. Bartlett, A new method for remediation of sandy slopes susceptible to seepage flow using EPS-block geofoam, J. Geotext. Geomembr. 42 (2) (2014) 166–180. [25] A. Padade, J. Mandal, Expanded polystyrene-based geomaterial with fly ash, Int. J. Geomech. 14 (6) (2014) 06014013. [26] Nicholas T. Rocco, Ronaldo Luna, Mixtures of clay/EPS particulates and undrained shear strength, Geo-Congress 2013 Stability and Performance of Slopes and Embankments III (2013) 2059–2068. [27] P.W. Rowe, L. Barden, A new consolidation cell, J. Geotech. 16 (2) (1996) 162– 170. [28] T.D. Stark, D. Arellano, J.S. Horvath, D. Leshchinsky, Guideline and recommended standard for geofoam applications in highway embankments, in: NCHRP Report 529, Transportation Research Board, Washington, DC, USA, 2004. [29] T. Tsuchida, Super geo-material project in coastal zone-development of high quality geo-material by recycling bulky urban wastes, in: Proc. Int. Sym. Coastal Ocean Space Utilization, 1995, pp. 391–401. [30] Takashi Tsuchida, Ali Porbaha, Nobuyuki Yamane, Development of a geomaterial from dredged bay mud, J. Mater. Civ. Eng. 13 (2) (2001) 152–160. [31] Gil.-Lim. Yoonz, Sang-Soo Jeon, Byung-Tak Kim, Mechanical characteristics of light-weighted soils using dredged materials, J. Mar. Georesour. Geotechnol. 22 (4) (2004) 215–229. [32] Wei Zhu, Li Mingdong, Zhang Chunlei, Zhao Gan, Density and strength properties of sand-expanded polystyrene beads mixture, in: GeoCongress 2008, Characterization, Monitoring, and Modeling of GeoSystems, 2008, pp. 36–43.
Contents lists available at ScienceDirect
Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat
An investigation on the geotechnical properties of sand–EPS mixture using large oedometer apparatus Reza Jamshidi Chenari a,⇑, Mehran Karimpour Fard b, Sabina Pourghaffar Maghfarati c, Faranak Pishgar d, Sandro Lemos Machado e a
Department of Civil Engineering, Faculty of Engineering, The University of Guilan, P.O. 3756, Rasht, Guilan, Iran Department of Civil Engineering, Faculty of Engineering, The University of Guilan, Iran Azad University, Iran d Islamic Azad University, Fouman and Shaft Branch, Fouman, Guilan, Iran e Universidade Federal da Bahia, Brazil b c
h i g h l i g h t s Deformation and constrained modulus of EPS–sand mixture was investigated. Permeability and hydraulic conductivity of EPS–sand mixture was obtained. A new fabricated large size oedometer apparatus was designed and tested. A K0 coefficient measurement was designed through a ring pressure cell inside the large oedometer apparatus. Unconstrained deformation modulus to be back-calculated from oedometeric gauge readings was for the first time calculated.
a r t i c l e
i n f o
Article history: Received 16 July 2015 Received in revised form 1 March 2016 Accepted 17 March 2016 Available online 25 March 2016 Keywords: Expanded poly styrene (EPS) Permeability Deformation and strength characteristics Large scale oedometer
a b s t r a c t The applicability of Expanded PolyStyrene (EPS) beads mixed with sand in five different contents was investigated for use in geotechnical engineering applications utilizing a newly designed and fabricated large scale oedometer apparatus. Permeability, coefficient of earth pressure ‘‘at rest” and the volume compressibility coefficient were measured and calculated for different EPS beads contents. Consolidation and permeability tests were conducted under different overburden pressures. The main findings of major recent studies were compared with current study and verification of new results was undertaken. Permeability, coefficient of earth pressure ‘‘at rest” and the volume compression coefficient along with some important deformation and strength characteristics parameters namely, the internal friction angle, constraint modulus and the drained 3-D elasticity modulus were investigated. The Results revealed that permeability, internal friction angle, constraint modulus and 3-D Young modulus decreases with inclusion content. However, the volume compressibility coefficient and the K0 coefficient showed opposite trends. Predictive models were submitted in forms of Multi Linear Regression, MLR simulations and their performances were evaluated. Ó 2016 Elsevier Ltd. All rights reserved.
1. Introduction There have been growing interest in the use of nonconventional as lightweight geo-materials, and their introduction has presented both opportunities and challenges to researchers and engineers worldwide. Attention is specifically needed to be ⇑ Corresponding author. E-mail addresses: [email protected] (R. Jamshidi Chenari), [email protected] (M. Karimpour Fard), [email protected] (S. Pourghaffar Maghfarati), [email protected] (F. Pishgar), smachado@ ufba.br (S. Lemos Machado). http://dx.doi.org/10.1016/j.conbuildmat.2016.03.083 0950-0618/Ó 2016 Elsevier Ltd. All rights reserved.
drawn to the consideration of both the cost and environmental implications when any new material is introduced into construction. In recent decades, a successful conversion from academic excellence to commercial viability has occurred. Lightweight fill materials have a wide range of civil engineering applications around the world. They may be used as fill over soft clay sites to prevent excessive settlement; as backfill for retaining and basement walls to reduce the horizontal driving forces; as fill material to increase factor of safety for slopes by reducing driving forces; as seismic buffers to alleviate seismic forces, and so forth. Various types of lightweight fills, such as Expanded PolyStyrene (EPS)-block geofoam, EPS beads, tire waste products like shredded
774
R. Jamshidi Chenari et al. / Construction and Building Materials 113 (2016) 773–782
tires, tire chips and tire crumbs have been reported. Using such materials not only provides lightweight fill solution for civil engineering projects, but also helps to save the environment by recycling these materials instead of stockpiling them. EPS-block geofoam gained popularity due to its wide application areas such as compressible inclusions [14,15,22], reduction of swelling pressure caused by expansive sub-soils ([6,16], fill materials in highway embankments [28] and remediation of sandy slopes [1,24]. In order to overcome shortcomings of EPS-block geofoam namely transportation issue, lack of formability to fill irregular volumes and compatibility with on-site soil, lightweight fill material composed of a mixture of EPS beads with soil and alternative additives is recommended [29–31,20,32,9,10,21,26,11]. Miao et al. [21] mixed dredged sand with EPS-beads and Portland cement. They conducted different geotechnical tests to investigate their potential use for mitigation of settlement problems associated with bridge approach embankments over soft soil. Padade and Mandal [25] proposed a geomaterial by blending fly ash instead of soil with EPS-beads and cement. Using compression tests they showed that the compressive strength of EPGM (Expanded Polystyrene-bead Geo-Material) increases considerably if cement-to-fly ash ratios of 10%, 15% and 20% are used. Some researchers mixed EPS particulates with sand to create a lightweight fill and measured the stress–strain characteristics of the modified soils in the laboratory using direct shear and triaxial compression tests [20,32,9,10,21,11]. Deng and Xiao [9,10] studied the stress–strain behavior of EPS–sand for a single type of EPS bead– sand mixture. They showed a systematic decrease in drained strength with increasing EPS content. So far previous studies for lightweight geo-materials are mainly about the mechanical behaviors. When used in practical engineering, lightweight soil may undergo various hydraulic and seepage loads; so it is necessary to investigate its hydraulic properties along with the mechanical and stress–strain behavior. The main aim of this study is to conduct some large Oedometer tests on fine grained sand mixed with EPS beads with different weight contents to investigate its effect on mechanical and hydraulic properties of the proposed lightweight geo-material.
2. Experimental study 2.1. Materials The experiments are carried out on ‘‘Chamkhaleh Sand”, supplied from Chamkhaleh Beach adjacent to Chamkhaleh River, located on SW of Caspian Sea. The particles are quartz–based with gray color and uniform particle distribution as shown in Fig. 1. The index properties of the used host sand are given in Table 1. The specific gravity was determined according to ASTM D 854. Maximum and minimum dry unit weights were determined based on ASTM D 4253 [4] and ASTM D 854 [5], respectively. It had a specific gravity of 2.63, a maximum dry unit weight of 16.1 kN/m3 (i.e., minimum void ratio of 0.63) and a minimum dry unit weight of 14.2 kN/m3 (i.e., maximum void ratio of 0.85). The particle size distribution of the sand is given in Fig. 2. The sand had a coefficient of uniformity of 1.54, a coefficient of curvature of 0.95, and was classified under the Unified Soil Classification System as SP (poorly graded sand) and under the AASHTO Soil Classification System as A-3. The friction angle of the sand was 38° when its relative density Dr was 60%. The EPS beads used in this study is super light polymer foam, pre-puffed from polystyrene resin, provided by a local EPS block moulding company which had been manufacturing EPS geofoam blocks. The beads were white, even, and spherical, sizing between 2–7 mm (Fig. 3). The relevant index properties of EPS bead are presented in Table 1. Determination of the unit weight and specific gravity of the EPS beads were conducted employing a procedure modified from comparable standard test method for fine aggregates (i.e., ASTM C128 [3]). Beads were placed into a 1-L hydrometer until the volume of the hydrometer was apparently occupied. Beads were placed into the hydrometer without noticeable compaction effort so as to reach a moderate compaction state. Unit weight of the beads is then easily calculated by scaling the net weight of beads, used to fill the bottle. The unit weight obtained for EPS beads was 0.08 kN/m3. Specific gravity (Gs) of beads was also calculated by filling the voids between EPS beads with distilled water and then to calculate the net volume of beads and determine the specific gravity of beads, which was 0.013.
Fig. 1. Magnified photos of Chamkhaleh Sand [12]. 2.2. Specimen preparation EPS–sand specimens (Fig. 4) were formed by mixing EPS beads with sands at a dry mass ratio g of EPS beads over sands, which was thought as the most significant factor controlling the unit weight, mechanical and hydraulic behavior of the mixtures. Investigated ratios were 0.1%, 0.2%, 0.3%, 0.5% and 1% for EPS beads by weight. Sufficient water was added to the EPS–sand samples. The added water provided bonding between specifically the EPS beads and sand, which made it possible to mix without segregation. Initial water content was measured precisely to control the target dry density of prepared samples for test. For each designated mixing ratio, the mass-based proportions of sand, EPS bead were determined beforehand. The proportioned materials were mixed thoroughly until the mixtures were homogenous enough. Material was molded into the chamber of oedometer apparatus at a designated weight and dry density of 14.6 kN/m3. To reach the target dry density, a handy tamper was employed. It should be noted that the dry density was defined based on total solid constituents, namely EPS beads and sand particles. Some researchers try to maintain a constant skeletal or matrix dry density for sand portion [2], however authors believe that considering a constant skeletal relative density is a tricky situation which is strictly affected by compressibility of inclusions and needs precise estimate of sand portion volume. In real world applications, practitioners would rather to adopt a simple procedure based on either maintaining a constant total dry sand portion weight assumption [18] or adopting a constant overall bulk dry density which is the case in this study. Specifically, the target dry density of EPS–sand mixtures was achieved by quantifying the weight of proportioned EPS and sand to be placed within a volume (i.e., 57,000 cm3). Compaction was completed in seven layers. A list of specimens and corresponding weight and volume ratios is provided in Table 2 for EPS beads mixed with sand. Totally 6 large Oedeometer tests were performed in this investigation. The first test was conducted on employed host sand. Each mixture was tested at three different overburden pressures, 160, 260 and 375 kPa. The volumetric ratio of EPS over the combination of EPS and sand in the mixture, v, was calculated based on the mixing ratios and specific gravities of particles for each mixture, as presented in Eq. (1). If the EPS content increased above 1%, the EPS volumetric ratio exceeded 70%, which substantially led to reduced sand inter-particle interactions. g
v ¼ 100GSEPS g Gss
þ GsEPS
ð1Þ
775
R. Jamshidi Chenari et al. / Construction and Building Materials 113 (2016) 773–782 Table 1 Physical properties of tested materials. Material
Specific gravity, Gs Dry unit weight (kN/m3) Effective size D10 (mm) Mean grain size D50 (mm) Uniformity coefficient, Cu Coefficient of curvature, Cc
Sand 2.63 EPS bead 0.013
14.2 (min) 16.1 (max) 0.08
0.17 2.60
0.21 4.00
1.54 1.70
0.95 0.90
Fig. 2. Grain size distribution curves of EPS bead and sand used for mixtures.
Fig. 4. Sample preparation for EPS–sand mixture.
Table 2 Summary of Oedometer testing program on EPS–sand mixtures. Designation Content by weight g(%)
Content by volume v(%)
Dry density (kN/m3)
Overburden pressure (kPa)
Sand EPS–Sand-1 EPS–Sand-2 EPS–Sand-3 EPS–Sand-4 EPS–Sand-5
0 16.8 28.8 37.8 50.3 67
14.6 14.6 14.6 14.6 14.6 14.6
160, 160, 160, 160, 160, 160,
0 0.1 0.2 0.3 0.5 1
260, 260, 260, 260, 260, 260,
375 375 375 375 375 375
2.3. Large oedometer testing setup
Fig. 3. Optical microscope photos of EPS beads. where v denotes the EPS volumetric percentage in the mixture and GsEPS and Gss denote the specific gravities of EPS bead and sand, respectively. After sample preparation completion, a perforated plastic pad was placed over the specimen and loading plunger along with the top cap was fixed to the position and saturation process was commenced. Two O-rings were placed around the cylindrical plunger to seal the chamber. The specimen was saturated using a very small upward hydraulic gradient, from bottom of samples to expel air in voids among particles. The saturation was thought to be completed when no air bubble was observed to be escaped from the top drainage line.
Most commercial geotechnical laboratories have been utilizing some kinds of Casagrande-type oedometers [7] for routine measurements of consolidation parameters of fine-grained soils for decades. A large specimen may be more representative of field conditions in some civil engineering applications. Very few apparatuses are available for the measurement of consolidation parameters of large specimens as recovering large undisturbed specimens is difficult, expensive and time-consuming. The hydraulic oedometer developed by Rowe and Barden [27] is available for specimen diameters ranging typically from 75 to 254 mm (BSI 1990). Although the size of the specimen is increased significantly, it may still be too small for many civil engineering applications. A large cylindrical Oedometer with a diameter of 492 mm and height of 550 mm to measure the load-deformation parameters of geo-materials under a vertical applied stress up to 1.5 MPa was designed, fabricated and calibrated at the engineering school of the University of Guilan. This oedometer was employed to evaluate the hydraulic and mechanical properties of non-conventional materials like EPS Geofoam, tire shreds, tire chips and TC mixed with soils. The fabricated large Oedometer is not conceptually different from the large Casagrande-type Oedometer designed and fabricated by Ng et al. [23]. The apparatus consists of a loading frame, a computer-controlled hydraulic loading plunger, an Oedometer cell, displacement, pore pressure, normal and lateral pressure transducers, and a computerized data acquisition system. In Fig. 5, the complete setup of oedometer and its major components are illustrated. The primary function of the loading frame is to provide a robust support for the hydraulic loading actuator and the Oedometer cell, and a reaction for the hydraulic loading actuator to apply the load to the specimen. A 50 mm thick steel plate is used to support the Oedometer cell to minimize any deflection of the cell base and to distribute the load to the supporting members uniformly. The steel plate is supported by four wheels that transmit the load to the supporting frame members. The base plate is equipped with five outlet ports connected to a drainage line to facilitate bot-
776
R. Jamshidi Chenari et al. / Construction and Building Materials 113 (2016) 773–782
tom drainage and excess pore water pressure measurements. Stainless steel Oedeometer cell of 5 mm thick, 500 mm outer diameter and 500 mm high was conceived strong enough to assume that lateral deformation of the new Oedometer cell is negligible even when a very high vertical stress is imposed on the specimen. It should be noted that the thickness of the specimen is 200 mm shorter than the height of the cell to allow for the storage of free water atop the specimen. This means that the target specimen thickness is 300 mm. A loading piston of 80 mm diameter is connected to a plunger of 488 mm diameter with 500 mm stroke to convert and apply hydraulic pressure to the specimen top pad. A schematic of the oedometer cell is depicted in Fig. 6. A high pressure line is provided atop the specimen for maintenance of a constant prescribed pore water pressure at the top boundary of the specimen to permeability measurement purposes. The new large Oedometer has some major features as listed: Ability to apply 1500 kPa reciprocating pressure with 500 mm stroke for highly compressible materials. Measurement of the coefficient of at-rest pressure, K0 by bottom and lateral hydraulic earth pressure transducers. Measurement of the coefficient of permeability in a constant head permeability test scheme provision. A few compressibility tests were performed on sand material considered rapiddraining soil to evaluate the applicability of the new apparatus in comparison to the standard size Casagrande-type oedometer. The e-log r curves of the same material obtained from the newly designed large oedometer and a standard size Casagrandetype oedometer are presented in Fig. 7. Two measurements performed on each of the apparatuses are presented. The repeatability of results are clear, indicating the precise performance of newly designed apparatus in measurement of deformation characteristics of soil.
3. Results and discussion 3.1. Permeability Permeability is an important factor that significantly influences the behavior of fill under saturated conditions, and often dictates the suitability of fill for specific applications. This is because a well-drained fill prevents the development of excess pore water pressure during loading and also accelerates consolidation of
underlying low permeability foundation soils by providing a drainage path [8]. According to Fell et al. [13], a material that has permeability greater than 1 106 m/s is considered as a good drainage material. In this study, a series of modified constant head permeability tests adapted to the large oedometer setup are conducted on the sand–EPS beads mixtures and the results are shown in Fig. 8. The permeability tests were conducted after applying the overburden stress in each stage of compressibility test and stabilizing related deformations. A constant hydraulic head was applied through a water reservoir pressurized by air as shown in Fig. 6. Applying flow through sample was continued after a constant discharge rate was recorded. Coefficient of permeability is then calculated under a constant head assumption. It can be seen that there is a general trend of decreasing permeability with the increase of overburden pressure and also the admixture content. This decrease in permeability may be attributed to the joint effects of the skeletal relative density of the sand and the impermeable inclusions. As the amount of EPS beads increases, the skeletal relative density of sand increases due to the constant overall dry density of the mixture and the EPS beads clog the sand matrix and introduces blockage of the water flow lines inside the mixture and thus permeability decreases. This effect is more highlighted when overburden pressure is increased because high overburden pressure converts the EPS beads into impermeable membranes looked as cushions for sand particles. It can also be seen that the permeability of all mixtures used in this study are greater than 1 106 m/s, which indicates good drainage materials. Another observation from Fig. 8 is that an optimum EPS content is expected when there is negligible variation between the 0.5% and 1% EPS beads-sand mixtures. In order to isolate the effects of EPS beads and overburden pressure, a series of standard constant head permeability tests was conducted as superimposed on Fig. 8, implying zero overburden pressure. Two observations are made, first it is observed that the EPS beads inclusion effect is more highlighted in zero overburden pressure. In high overburden pressure when the EPS beads are squeezed, the skeletal relative density is the controlling factor. Second observation is that overburden pressure higher than 160 kPa has no extra effect on permeability of the mixture. 3.2. Lateral Earth pressure coefficient Lateral earth pressure is the pressure that soil exerts in the horizontal direction. The lateral earth pressure is important because it affects the consolidation behavior and strength of the soil and because it is considered in the design of geotechnical engineering structures such as retaining walls, basements, tunnels, deep foundations and braced excavations. The coefficient of lateral earth pressure, K is defined as the ratio of the horizontal effective stress, r’h, to the vertical effective stress, r’v. For a level ground deposit with zero lateral strain in the soil, the ‘‘at-rest” coefficient of lateral earth pressure, K0 is obtained. At rest lateral earth pressure, represented as K0, is the in situ lateral pressure. It can be measured directly by a dilatometer test (DMT) or a borehole pressuremeter test (PMT). As these are rather expensive tests, laboratory scale oedometer test can be employed to have an estimate of the K0 if virgin and undisturbed samples are taken from soil deposit. Empirical relations have also been created in order to predict ‘‘at-rest” pressure with less involved soil testing, and relate to the angle of shearing resistance. Jaky [17] proposed a famous equation for normally consolidated soils:
K0 ¼ 1 Sinu0
Fig. 5. Large Oedometer testing apparatus.
ð2Þ
In the fabricated oedometer apparatus, the bottom pressure cell is employed to measure vertical pressure. The cell consists of two 30 cm circular stainless steel plates welded together around their
R. Jamshidi Chenari et al. / Construction and Building Materials 113 (2016) 773–782
777
Fig. 6. Details of the large Oedometer cell.
Fig. 7. Evaluation and repeatability check of the new large Oedometer apparatus.
Fig. 8. Effect of EPS content on the permeability of EPS–sand mixture.
periphery and spaced apart by a narrow cavity filled with de-aired oil. Changing overburden pressure squeezes the two plates together causing a corresponding increase of fluid pressure inside the cell. An electrical pressure transducer converts this pressure into an electrical signal which is transmitted to the data logger. In the case of lateral earth pressure the same procedure was employed. In this case, 10 cm belt-like cell pressure has been attached to the inner perimeter of oedometer cell. Test procedure in large oedometer apparatus involves spontaneous application of overburden pressure of specified magnitudes to the specimen in undrained condition. Lateral and vertical earth pressure transducers render continuous reading of the earth pressure in horizontal and vertical directions respectively. A dynamic variation of the coefficient of earth pressure ‘‘at-rest” is provided in Fig. 9 for sand mixed with EPS beads under three different overburden pressures. It indicates that the K0 starts from 1 at the beginning and stops at a final value when the induced pore pressure dissipates. In fact, it moves from a full undrained condition at the beginning when the overburden pressure is applied, to a full drained condition at end when all of the applied pressure is fully transferred to the skeleton. The ultimate values of K0 are then drawn for different inclusion contents and overburden pressures as illustrated in Fig. 10. It is observed that both overburden pressure and inclusion content are clearly affecting the K0 coefficient. According to Deng and Xiao [10], depending on the EPS content there are three major interaction mechanisms. Sand–sand, sand– EPS and EPS–EPS mechanisms play dominating effect due to the rigidity difference between sand and EPS beads. EPS beads behave like balloons filled with air. In low overburden pressure when there is no remarkable deformation in beads, especially when the inclusion content is high, the EPS–EPS mechanism controls and the mixture behaves like fluid and consequently a hydrostatic fluid condition is dominant and thus the K0 value is not far from the incompressible fluid state. In Fig. 9(f), at lower overburden stress of 160 kPa, the value of K0 is considerably higher for the g = 1% in comparison to lower EPS contents. Overburden pressure, however is affecting the coefficient of lateral earth pressure in a different way. Increasing the overburden pressure causes compression of EPS beads and thus mobilization
778
R. Jamshidi Chenari et al. / Construction and Building Materials 113 (2016) 773–782
Fig. 9. Time variation of the coefficient of earth pressure ‘‘at-rest” for EPS–sand mixture; a) g = 0%; b) g = 0.1%; c) g = 0.2%; d) g = 0.3%; e) g = 0.5% and f) g = 1%.
of shear strength is expected by inducing an apparent cohesion to the mixture due to the mobilized reinforcing effect. This is equivalent to virtually applying more confining pressure and thus less horizontal pressure is activated due to the virtual confinement induced. The implication is that the K0 decrease with an increase in overburden pressure. Another justification is that the sand–sand mechanism becomes more highlighted when the overburden pressure is remarkable. Due to the rigidity difference and higher level of overburden pres-
sure, the EPS components deform and experience substantial volumetric compression, causing densification of the mixtures remarkably more than a low consolidation pressure did. Accordingly, a more compact and rigid structure led to a decreased K0 value. Internal friction angle of the sand mixed with EPS is backcalculated from Jacky’s equation as described earlier and depicted in Fig. 11 for different inclusion contents and overburden pressures. Results show that the internal friction angle decreases as
R. Jamshidi Chenari et al. / Construction and Building Materials 113 (2016) 773–782
Fig. 10. Effect of EPS content on the K0 coefficient of sand mixtures.
the inclusion content increases. Results of current study are qualitatively comparable with findings of others, although direct comparison is not possible due to different shear failure mechanisms. Deng and Xiao [10] reported three principal shear mechanisms in terms of sand–sand, sand–EPS and EPS–EPS granular interaction. Particle rigidity difference between sands and EPS beads as well as the packing difference among different compositions leads to higher shear strength mobilization for sand–sand mechanism in comparison to sand–EPS and EPS–EPS mechanisms. This means that increasing EPS content might decrease the friction angle. Karimpour et al. [19] investigated shear strength behavior of sand–EPS beads mixture in low weight contents. The study was conducted employing a large direct shear test apparatus. Different inclusion contents were tested under a wide range of overburden pressures. Results clearly demonstrated that the internal friction angle decreases with the increase of EPS content, again confirming the findings of current study.
are conducted to investigate the compressibility of specimens of sand mixed with EPS or TC. The specimens are subject to overburden pressures of, 160, 260 and 375 kPa, and the data logger keeps interrogating the pore pressure transducer until a zero pore pressure is reached. Effective stress is calculated while maintaining constant overburden pressure during each test. A potentiometer reads and measures the vertical displacement induced in specimen when the sample is laterally constrained. The strain–stress behavior observed on the sand–EPS beads mixtures are shown in Fig. 12 for different overburden pressures. In order to evaluate the behavior of the EPS–sand mixtures, a control experiment using pure sand was conducted. The results of this control experiment can be used to normalize the volume compressibility coefficient (mv) and other measured and calculated parameters obtained from EPS–sand mixture samples for subsequent regression and modeling analyses. Illustrations provided in Fig. 12 clearly show that the strain– stress behavior of sand–EPS mixtures is strongly affected by the inclusion content. EPS inclusion leads to the increased volume compressibility of the mixture. This means that the rate at which the oedomteric sample deforms, increases due to the inclusion of less stiffer aggregates. However, the overburden pressure has a different effect when looking into Fig. 12. Under high overburden pressure, the EPS beads are subject to significant volumetric compression. This resulted in densification of EPS beads within the mixture, which is the cause of the strain hardening behavior, and the stress–strain behavior of the mixture is mainly dominated by EPS beads [11]. The relationship between g and mv is presented in Fig. 13 for EPS–sand mixture. Whereas g increases with all other variables remaining constant (overburden pressure in oedometeric specimens), the mv increases linearly (Fig. 13). The mv is a function of EPS content in the mixture and overburden pressure. The volume compressibility coefficient of EPS–sand mixture was decreased with an increase in overburden pressure (Fig. 13). Constraint modulus was calculated from the volume compressibility coefficients depicted in Fig. 13 by inversing mv values. Furthermore, adopting an oedometric condition will lead to the derivation of Poisson’s ratios from K0 values (Eq. (3)). Drained elasticity modulus for 3-D condition was then calculated from Eq. (4) and illustrated in Fig. 14.
3.3. Composite compressibility
K0 ¼ The stability of any structures made by geo-materials, depends mainly on the strength and deformation characteristics of the materials used. A series of large oedometer stress controlled tests
Fig. 11. Effect of EPS content on the internal friction angle of sand mixtures.
779
E0 ¼
m0
1 m0
ð1 þ m0 Þð1 2m0 Þ 0 EOed ð1 m0 Þ
ð3Þ ð4Þ
Fig. 12. Effect of EPS content on the stress-strain behavior of sand mixtures.
780
R. Jamshidi Chenari et al. / Construction and Building Materials 113 (2016) 773–782
Fig. 13. Effect of EPS content on the volume compressibility coefficient of sand mixtures.
where m’ is the Posisson’s ratio, E’ is the drained 3-D elasticity modulus obtained conventionally from triaxial CD or even CU tests and E0Oed ¼ Mv ¼ m1v is the drained constrained modulus computed from the volume compressibility coefficient obtained from the oedometric tests. Superimposed on the same graph are the findings of Edinçliler and Özer [11] for undrained condition obtained from CU triaxial tests demonstrated for comparison. Results show that both drained and undrained elasticity moduli decrease as the EPS content increases. This is as expected as the rigid sand particles are replaced with compressible EPS beads and EPS–EPS interaction mechanism dominates when the EPS beads content rises. Another observation from Fig. 14 is that confining pressure and overburden pressure have the same effect as explained by Edinçliler and Özer [11] earlier. Lower overburden pressure or confining pressure in triaxial sense leads to formation of sand–EPS or EPS– EPS mechanisms which pose more deformation potential and less composite rigidity is then exhibited. 4. Skeleton compressibility In the case of an ordinary soil sample, compressive volume change is caused by a reduction in pore phase owing to the rear-
Fig. 14. Effect of EPS content on the drained and un-drained elasticity modulus of sand mixtures.
rangement of the soil skeletal structure. This is why a soil particle is assumed to be almost rigid. However, in the case of a mixed soil sample containing deformable particles such as EPS beads, larger compression occurs caused by additional volume change owing to the compression of the deformable particles. In such a situation, the additional volume change was thought to consist of: a volumetric compression of the deformable particles themselves and further pore phase reduction by the rearrangement of the particle skeleton caused by the deformation of deformable particles. EPS beads are soft inclusions which are subject to compression and volume change during overburden application in large oedometer experiments. Both in drained and undrained loading schemes, EPS beads undergo compression. However, differentiation between the mixture compressibility and the skeleton volume change is difficult and impossible within a single loading scheme. For this reason, in addition to the drained loading cases, a few large oedometer experiments were conducted in undrained condition to isolate different effects as discussed before. Fig. 15 shows the variation of oedometric strain with the overburden stress for both drained and undrained experiments. Two observations are deduced from this figure. The influence of the skeleton compression exhibited as volume compression in undrained loading condition is noticeable and becomes more remarkable when the inclusion content increases. Another observation is that a nontrivial volume compression was observed for pure sand model which should not be considered as skeleton compression because sand particles are rigid elements and do not undergo volume change during undrained loading condition. This volume change is attributed to the compression of air bubbles trapped inside the mixture.
5. Predictive models 5.1. MLR regression analysis Mechanical and hydraulic properties of sand mixed with EPS are not only a function of g but also depend on the overburden pressure applied to oedometric specimens. In addition, test results indicate that some of these parameters linearly vary with increasing input parameters. By using the linear behavior obtained from the test results, multiple linear regression (MLR) models were run to investigate the predictive performance and reasonableness of the developed regression equations. Ultimately, the independent variables chosen for the final MLR models to predict these param-
Fig. 15. Effect of inclusion compressibility on overall volume change of the mixture.
R. Jamshidi Chenari et al. / Construction and Building Materials 113 (2016) 773–782 Table 3 Data variables sets and MLR models for oedometer parameters. Data set Independent variables g K (EPS) r 100 , Pa g K0 (EPS) r 100 , Pa mv (EPS) g , r 100 Pa
R2 (%) Equationa g 0:7832 r 0:9434 ¼ 0:1796 100 Pa 0.8312 K 0EPSSand g 0:1292 r 0:3160 ¼ 1:7435 K 0Sand 100 Pa 0.8302 mv EPSSand g 0:5080 r 0:5831 ¼ 4:9669 mv Sand 100 Pa 0.9866
K EPSSand K Sand
a From the MLR model given in Eq. (5), and regression output by using Matlab MathWorks.
Fig. 16. Comparison of measured and predicted performance for different parameters from oedometric tests.
eters were: g and the overburden pressure (r). These variables were used to predict mechanical and hydraulic parameters, by dividing them into six different models, as presented in Table 3. From an application standpoint, the regression models should not be dependent on units. Consequently the variables were normalized with those of pure sand in the mixture and r was normalized with atmospheric pressure (Pa) to make the regression variables dimensionless. All regression analyses shown in Table 3 were performed using MATLAB MathWorks. These models have the general form:
y ¼ n0 xn11 xn22
ð5Þ
This equation is re-written in the following form:
log y ¼ logn0 þ n1 logx1 þ n2 logx2
ð6Þ
A comparison of predicted parameters in Table 3 and laboratory measured values from large oedometer tests in normalized forms can be seen in Fig. 16. As shown in Fig. 16 models from Table 3 provide reasonably close prediction of the laboratory results for EPS– sand mixture. 6. Conclusion The applicability of EPS–sand mixtures was investigated for use in geotechnical engineering applications utilizing a newly designed and fabricated large size oedometer apparatus. The new designed large size oedometer apparatus was first calibrated and tested for repeatability and accuracy. For this aim some standard Casagrande type oedometer tests were conducted as well and results were compared with those of new fabricated large size oedometer. Permeability, coefficient of earth pressure ‘‘at rest” and the volume compressibility coefficient were measured and calculated for dif-
781
ferent EPS contents. A comprehensive discussion and comparison with similar studies found in literature primarily in forms of triaxial drained and un-drained tests was undertaken. The main findings of major recent studies were superimposed on graphs of current study and verification of new results was accomplished. Predictive models were also worked out in order to provide an efficient regression based prediction of the parameters under study. For this aim, multi linear regression (MLR) procedure was employed to render predictions comparable with measured values. The results obtained in this study led to the following conclusions: 1- The new designed and fabricated large size oedometer apparatus proved an efficient tool to reflect hydraulic and mechanical behavior of non-conventional materials. Results obtained from the evaluation program indicate that the new apparatus is comparable to the typical Casagrande-type oedometer. 2- Permeability of EPS–sand mixtures exhibited a decrease with inclusion content due to replacement of sand particles with impermeable EPS beads. 3- Overburden pressure has a decreasing effect on permeability of sand mixed with EPS. The reason is that increasing effective stress leads to higher inter-granular stresses which in effect induce more compression in EPS beads and less permeability is thus expected. 4- EPS beads are compressible materials transferring overburden pressure to different directions and behave more or less similar to incompressible fluids as long as the overburden pressure is low. The increase in inclusion content leads to an increase in K0 values. However, this behavior vanishes when the effective stress dramatically rises. Indeed the sand–sand or sand–EPS mechanisms are controlling effects in such case and rigidity of mixture increases, implying smaller Poisson’s ratio and thus smaller K0 values are rendered. 5- Internal friction angle for sand mixed with EPS was backcalculated from K0 values adopting Jacky’s equation. Results showed that the internal friction angle decreases as the inclusion content increases due to the sand–sand or sand– EPS mechanisms to be less effective in higher inclusion content. Overburden pressure; however increases interaction between sand particles and the internal friction angle is thus increased as expected. 6- Volume compressibility coefficient of the EPS–sand mixtures is a function of the equivalent rigidity of the mixture which is a function of both inclusion content and the overburden pressure. EPS inclusion induces less rigidity to the mixture and thus more deformability and compressibility is expected. This effect decreases with overburden pressure due to the densification induced by overburden pressure. 7- Drained elasticity modulus of the EPS–sand mixture was also extracted indirectly from constraint modulus values and the Poisson’s ratios back-calculated from K0 values. Results showed that inclusion content leads to dramatic decrease in elasticity modulus when the EPS content is less than 0.5% by weight. However, the rate of decrease ceases for higher contents implying an optimum EPS content of 0.5%. 8- Predictive models for permeability, coefficient of earth pressure ‘‘at rest” and the volume compressibility coefficient adopting MLR procedure proved comparatively acceptable and efficient in simulation of these parameters. Although a decrease in permeability, internal friction angle and constraint modulus is not desirable in most geotechnical
782
R. Jamshidi Chenari et al. / Construction and Building Materials 113 (2016) 773–782
applications, the lightweight EPS beads might have other benefits due to the low unit weight and specific gravities in comparison to quartz sand. Very low inertial force in seismic applications provokes one to consider such material as backfill material for retaining walls, trench filling material, etc. References [1] Onur Akay, A. Tolga Özer, Garey A. Fox, Steven F. Bartlett, David Arellano, Behavior of sandy slopes remediated by EPS-block geofoam under seepage flow, J. Geotext. Geomembr. 37 (2013) 81–98. [2] M. Amel-Sakhi, Influence of optimized tire shreds on strength of sandreinforced with tire shreds, in: In Partial Fulfillment of the Requirements the Degree of Master of Science, Faculty of Engineering, Isfahan University of Technology, Isfahan, Iran, 2001. [3] ASTM C128, in: Standard Test Method for Density, Relative Density (Specific Gravity), and Absorption of Fine Aggregate, 2012. [4] ASTM D 4253, in: Standard Test Methods for Maximum Index Density and Unit Weight of Soils Using a Vibratory Table, 2006. [5] ASTM D 854, in: Standard Test Methods for Specific Gravity of Soil Solids by Water Pycnometer, 2014. [6] Mustafa Aytekin, Numerical modeling of EPS geofoam used with swelling soil, J. Geotext. Geomembr. 15 (1) (1997) 133–146. [7] A. Casagrande, The Determination of Preconsolidation Load and its Practical Significance, Proc., 1st International Conference on Soil Mechanics Foundation Engineering, Cambridge, 1936. [8] H.R. Cedergren, Seepage, Drainage, and Flow Nets, third ed., John Wiley & Sons, New York, 1989, pp. 19–72. [9] An Deng, Yang Xiao, Modeling stress-strain behavior of sand-EPS beads lightweight fills based on cam-clay models, in: In Recent Advancement in Soil Behavior, in Situ Test Methods, Pile Foundations, and Tunneling@ Selected Papers From the 2009 GeoHunan International Conference, 2009, pp. 55–61. [10] An. Deng, Yang Xiao, Measuring and modeling proportion-dependent stressstrain behavior of EPS-sand mixture 1, Int. J. Geomech. 10 (6) (2010) 214–222. [11] A. Edinçliler, A.T. Özer, Effects of EPS bead inclusions on stress–strain behaviour of sand, J. Geosynthetics Int. 21 (2) (2014) 89–102. [12] K. Fakharian, F. Shabani, Monotonic and cyclic experimental results of Chamkhaleh SAND in southern Caspian Sea, in: Proceedings of the International Offshore and Polar Engineering Conference, 2011, pp. 373–380. [13] R. Fell, P. McGregor, D. Stapledon, Geotechnical Engineering of Embankment Dams, Balkema, Rotterdam, Brookfield, 1992. [14] J.S. Horvath, Expanded polystyrene (EPS) geofoam: an introduction to material behavior, J. Geotext. Geomembr. 13 (4) (1994) 263–280. [15] John.S. Horvath, The compressible inclusion function of EPS geofoam, J. Geotext. Geomembr. 15 (1) (1997) 77–120.
[16] S. Banu Ikizler, Mustafa Aytekin, Evin Nas, Laboratory study of expanded polystyrene (EPS) geofoam used with expansive soils, J. Geotext. Geomembr. 26 (2) (2008) 189–195. [17] J. Jaky, Pressure in silos, 2nd ICSMFE, London, 1, 1948, pp. 103–107. [18] C, R. Jamshidi, Dynamic Behavior of Sand Reinforced with Randomly Distributed Carpet Waste Strips behind Sheet-pile Retaining Walls in Shaking Table Tests (Ph.D. Dissertation), Iran University of Science and Technology, Tehran, Iran, 2008. [19] M. Karimpour Fard, R. Jamshidi Chenari, F. Soheili, Shear strength characteristics of sand mixed with EPS beads using large direct shear apparatus, Electron. J. Geotech. Eng. 20 (8) (2015). [20] Han-long Liu, An Deng, Jian Chu, Effect of different mixing ratios of polystyrene pre-puff beads and cement on the mechanical behaviour of lightweight fill, J. Geotext. Geomembr. 24 (6) (2006) 331–338. [21] Linchang Miao, Fei Wang, Jie Han, Lv Weihua, Li Jing, Properties and applications of cement-treated sand-expanded polystyrene bead lightweight fill, J. Mater. Civ. Eng. 25 (1) (2012) 86–93. [22] George P. Murphy, The influence of geofoam creep on the performance of a compressible inclusion, J. Geotext. Geomembr. 15 (1) (1997) 121–131. [23] A.M.Y. Ng, A.T. Yeung, P.K.K. Lee, L.G. Tham, Design, fabrication, and assembly of a large oedometer, J. Geotech. Test. 29 (4) (2006) 298–305. [24] A. Tolga Özer, Onur Akay, Garey.A. Fox, Steven.F. Bartlett, A new method for remediation of sandy slopes susceptible to seepage flow using EPS-block geofoam, J. Geotext. Geomembr. 42 (2) (2014) 166–180. [25] A. Padade, J. Mandal, Expanded polystyrene-based geomaterial with fly ash, Int. J. Geomech. 14 (6) (2014) 06014013. [26] Nicholas T. Rocco, Ronaldo Luna, Mixtures of clay/EPS particulates and undrained shear strength, Geo-Congress 2013 Stability and Performance of Slopes and Embankments III (2013) 2059–2068. [27] P.W. Rowe, L. Barden, A new consolidation cell, J. Geotech. 16 (2) (1996) 162– 170. [28] T.D. Stark, D. Arellano, J.S. Horvath, D. Leshchinsky, Guideline and recommended standard for geofoam applications in highway embankments, in: NCHRP Report 529, Transportation Research Board, Washington, DC, USA, 2004. [29] T. Tsuchida, Super geo-material project in coastal zone-development of high quality geo-material by recycling bulky urban wastes, in: Proc. Int. Sym. Coastal Ocean Space Utilization, 1995, pp. 391–401. [30] Takashi Tsuchida, Ali Porbaha, Nobuyuki Yamane, Development of a geomaterial from dredged bay mud, J. Mater. Civ. Eng. 13 (2) (2001) 152–160. [31] Gil.-Lim. Yoonz, Sang-Soo Jeon, Byung-Tak Kim, Mechanical characteristics of light-weighted soils using dredged materials, J. Mar. Georesour. Geotechnol. 22 (4) (2004) 215–229. [32] Wei Zhu, Li Mingdong, Zhang Chunlei, Zhao Gan, Density and strength properties of sand-expanded polystyrene beads mixture, in: GeoCongress 2008, Characterization, Monitoring, and Modeling of GeoSystems, 2008, pp. 36–43.