Compressibility characteristics of TDA from OTR (off-the-road) tires: A numerical approach

Abstract

This paper aims to use the discrete element method (DEM) to study the performance of discarded off-the-road (OTR) tire chips under compression. The results of small-scale laboratory tests were used to calibrate a model built in PFC3D. This model was validated using the results of a large-scale laboratory experiment. The results show that the discrepancy between the calculated strain and the strain measured in the laboratory was 1.94%. Furthermore, the calculated elastic modulus was 0.11 MPa lower than the elastic modulus determined in laboratory measurements. A sensitivity analysis done on the parameters included in the simulation indicated that Poisson’s ratio and the local damping ratio had a significant impact on the simulation results, whereas the loading rate, particle size and scale effect of the loading cylinder had a limited effect on the simulation results. The compressibility characteristics of OTR tire chips in the field embankment were simulated using a cuboid filled with the same size clumps as the TDA chips used in laboratory tests, and the simulation results were compared with available information. The results indicated that the parameters calibrated based on the small-scale laboratory tests can be used to predict the deformation of OTR tire chips in the field. The discrepancy between the calculation and field measurements for strain was 1%. Due to the extremely low winter temperatures in Edmonton (as low as −40 °C), the impact of temperature on the mechanical properties of OTR tire chips was also considered by varying the values of elastic modulus used in the simulation. The results show that the tire chip embankment had a larger strain in summer than in winter as a result of the high elastic modulus in warmer temperatures; however, the difference between the summer and winter strain values is only 0.4%. This paper presents valuable results on the deformation characteristics of OTR tire chips, using a combination of field measurements and laboratory measurements, and exploring the impact of temperature on the deformation of OTR chips using a numerical approach.

Introduction

The disposal of spent tires has significant environmental impacts worldwide due to the sheer volume of used tires discarded every year. For instance, in Alberta, more than 5 million tires are discarded every year [19], [20]. Thus, using tire-derived aggregate (TDA) as a fill material in civil engineering projects has become an attractive option over the past two decades [2], [12], [31]. Passenger and light truck tires are a very common source of TDA [27]. In industrial regions such as northern Alberta, Canada, discarded off-the-road (OTR) tires are readily available, making OTR tires an attractive source material for TDA. However, before using TDA chips derived from OTR tires as a fill material, the compressibility characteristics and potential settlement should be thoroughly understood.

There are two general approaches to investigating the compressibility of materials such OTR-derived TDA: numerical simulations and physical experiments. Physical experiments include both laboratory experiments (small-scale and large-scale tests) and field tests. Yi et al. [31] summarized the results of physical compression tests for TDA. The diameters of the molds used ranged from 63.5 to 900 mm, with heights from 25.4 to 1000 mm [30], [25]. Different methods have been reported for sample preparation; e.g., standard compaction (with one layer or five layers), or no compaction [9], [23]. The maximum TDA particle size used in the different tests ranges from 10 to 300 mm [30], [23]. Since particle size, porosity, density and specific gravity were not same among different tests, the maximum stress applied varied from 110 kPa to 700 kPa [23].

In field testing for various projects, a layer of TDA (1.3–3.0 m) was placed above different soils, such as marine silty clay and plastic clay [25]. It was observed that TDA compacted in the field has a higher initial unit weight than samples compacted in the laboratory, with densities ranging from 6.0 to 8.8 kN/m³ for particle sizes of 300 mm and 75 mm, respectively. The maximum strain observed in these cases was 5.3% and 19.1%, respectively. It was found that laboratory measurements of strain-stress relationships overestimate the strain measured in the field [25]. This difference may be due to the involvement of surcharge and pavement in the field, which can result in compacted tire chips with a low void ratio.

To date, there have been very few tests done on TDA derived from OTR tires. Meles et al. [18] carried out a small-scale test using a compression cell with an inner diameter of 570 mm and a height of 1120 mm. The sides of the cylinder was covered with a plastic sheet, effectively reducing the friction between the rubber chips and the steel cylinder. OTR tire chips were tested over a range of porosities by varying the amount of tire chips packed in the cell (from loose to dense packing). Regression equations between the applied stress and the observed strain were suggested to predict the immediate compression of OTR with different unit weights [18]. In other tests, Meles et al. [19] used a polyvinyl chloride (PVC) pipe with a diameter of 1230 mm to carry out compression testing for TDA derived from OTR tires. The authors calculated a bulk lateral earth pressure coefficient of 0.33 and Poisson’s ratio of 0.25. Meles et al. [20] also presented the results of field measurements of an embankment constructed using OTR tire chips. Two layers of OTR tire chips (3 m thick) were placed and compacted, with a 0.5 m layer of soil between them. Settlement plates were used to monitor the performance of the OTR embankment, and a maximum settlement of approximately 450 mm was observed at the top layer. Yi et al. [31] noticed that performance varies with differences in the initial void ratio, even when the same OTR tire chips are used.

Compressibility models for granular material are also applicable to OTR tire chips. Cundall [34] introduced a discrete element method (DEM) to study problems in rock mechanics, and this DEM was then applied to soils by Cundall and Strack [7]. A particle flow code (PFC) model developed based on this DEM has been used in the simulation of confined compression tests [5], [16], [26], [28], [33]. Ding et al. [8] studied the effect of model scale and particle size distribution on simulation results, and found that the particle size distribution, ratio of model diameter to average particle size, and ratio of maximum to minimum particle diameter all have an impact on the predicted mechanical properties of the corresponding material. However, the observed discrepancy in strain is not very significant. It was also noticed that the results of simulations using three-dimensional PFC (PFC3D) models have a lower coefficient of variation compared to two-dimensional PFC (PFC2D) simulations.

Valdes and Evans [29] simulated the performance of rubber chips under cycle loading conditions using PFC2D. In this work, rubber particles were mixed with sand in different volume fractions (V_rub/V_total, where V_rub is the volume of rubber, and V_total is the total volume) ranging from 0.4 to 1.0. In the simulation, the diameter of rubber particles ranged from 0.6 to 1 mm. The shear modulus and friction coefficient were used as key parameters to describe the mechanical properties of the rubber chips, which varied with changes in axial strain and contact stress. Results indicated that the type of loading (such as uniaxial or triaxial loading) affects the stress-strain relationship. Also, the selection of proper parameters, such as friction coefficient and shear modulus, was important in the simulation. Lopera Perez et al. [17] studied the impact of particle size and volumetric content of rubber particle on the mechanical properties of rubber/sand mixtures. The Hertz-Mindlin contact model was used to simulate the mechanical interaction between particles. The results indicated that different particle size ratios and rubber particle content have effects on strength and deformability. Kim and Santamarina [14] carried out compression tests on mixtures of sand and rubber particles. For pure rubber particles with an average size of 3.5 mm, the bulk modulus varied from less than 0.2 MPa to around 2 MPa when the applied stress increased from around 12 to 200 kPa. The shear wave velocity in the compacted rubber particles has a linear relationship with the applied confining stress. In another study, a finite element method (FEM) was used for simulation and a linear elastic model was adopted to simulate the mechanical properties of rubber [21].

Numerical simulations can provide information about the micro mechanical properties of a material, and these can then be verified and calibrated using the macro parameters, (such as Young’s modulus and Poisson’s ratio) measured from conventional experiments [10]. However, very few researchers have researched the properties of OTR tire chips (micro and macro) using DEM based on laboratory test results. Also, the effect of temperature on the mechanical properties of OTR tire chips has never been studied. In this research, the micro mechanical properties of OTR tire chips were determined by using stress-strain relationships obtained by small-scale laboratory tests. Next, the results from large-scale laboratory experiments were used to validate the model. A sensitivity analysis was carried out on the mechanical properties of the OTR tire chips, and the results were evaluated using the discrepancies of strain and elastic modulus between calculations and field measurements. The effects of loading rate, cylinder size and particle size on the stress-strain relationship were also studied. Finally, the effect of temperature was incorporated in the simulation to investigate the compressibility characteristics of tire chips under cold conditions. This paper combines a numerical simulation approach with laboratory measurements of micro mechanical properties and field measurements to investigate the compression behavior of OTR tire chips for a practical project.