Methods

The compaction mold for a large triaxial test specimen of tire shreds was fabricated from a 1.2 m inside-diameter polypropylene culvert (provided by CIWMB); 1.25 mm latex sheeting was used to fabricate the triaxial test membrane. The end platens were 25 mm (top platen) and 12.5 mm (bottom platen) thick steel plates. The load frame was an existing structural steel frame used for consolidating large soil samples at the National Geotechnical Centrifuge at UC Davis. Vertical load was provided by dual-acting, 76 mm-diameter, hydraulic rams with a 76 cm-stroke and a maximum operating pressure of about 20 MPa. Vacuum loading for confining pressure was provided using a high-volume Thomas Industries H50A-60 vacuum generator working off air pressure. Instrumentation for measurement of P- and S-waves consisted of 42 piezoelectric, ICP accelerometers by PCB Piezotronics.

After the latex membrane was secured to the bottom platen, the mold was secured in place and the combined apparatus was placed on a specially fabricated lifting frame to provide a solid stable base for the bottom platen to rest on during sample preparation. Two horizontal and two vertical rows of accelerometer arrays were established as shown schematically in Fig. 1.

Figure 1: Schematic of Accelerometer Arrays

The tire shreds were placed in approximately 150 mm lifts, about 100 kg per lift, and compacted with a walk-behind vibrating plate compactor (Wacker model) with a 172 kg static weight. As each lift was brought to the level of accelerometer placement, precise accelerometer locations were determined using a pre-marked rod and plumb bob. The completed triaxial test set up, placed in the load frame, is shown in Fig. 2.

Figure 2: Triaxial Test Set Up

As discussed earlier, the directional Young's moduli and shear moduli were obtained from three-dimensional P- and S- wave velocities. The P- and S- waves were generated by striking appropriately positioned metal or PVC striker blocks (either welded to the top and bottom platens or inserted during specimen preparation) with a 0.5 kg, double-headed 'dead blow' hammer. The P- and S- wave velocities were measured through the accelerometer arrays shown in Fig. 1 and recorded via a data acquisition system with a scan rate of 10,000 readings per second from each accelerometer. To model in situ conditions, the embankment profile shown in Table 1 was assumed. The tire shred specimen was tested under four triaxial load conditions representing the top and bottom of each tire shred layer as tabulated in Table 1. Table 2 presents the loading conditions modeled during our testing.

Table 1: Assumed Embankment Profile

Thickness [m]	Material	Density [$Mg/m^3$]
1.2	Paving and compacted soil	2.0
3	Tire shreds	0.8
1	Compacted soil	2.0
3	Tire shreds	0.8

Table 2: Selected Loading Conditions for Tests

load number	$\sigma_v$ [kPa]	$\sigma_h$ [kPa]	$K_0$
1	23.9	11.3	0.47
2	47.9	13.4	0.28
3	65.8	18.4	0.28
4	89.9	25.1	0.28