Laboratory testing and model calibration of a crushed stone backfill in a geosynthetic reinforced soil wall

Auburn University, Department of Civil and Environmental Engineering, 205 W. Magnolia Ave, , AL 36849, Auburn, USA

^* Corresponding author: cco0013@auburn.edu

Abstract

The objective of this study was to understand the effect of the choice of backfill constitutive models on the numerical simulation of a geosynthetic reinforced soil-integrated bridge system (GRS-IBS). The #89 material, a crushed angular granite stone, characterized by maximum particle sizes of 9 mm, was used for the reinforced backfill for the retaining wall. Consolidated drained triaxial tests were conducted at confining stresses of 48 kPa, 83 kPa, and 117 kPa. The average friction angle was 42 degrees. Two soil models (the linear-elastic perfectly-plastic and the Hardening Soil) were calibrated using the laboratory test results. The calibrated soil models were then used to model the reinforced backfill material in finite element simulations of the abutment. Results of the calculated vertical stresses at the instrument position showed reasonable agreement with the field measurement. The non-linear model predicted about twice the deformation (lateral displacement and bridge seat settlement) compared to the linear-elastic perfectly plastic model. Settlement results showed maximum values corresponding to 0.22% and 0.45% of the abutment height for linear-elastic perfectly-plastic model and Hardening-Soil model, respectively, which are less than 0.5% limit recommended by US FHWA, while the lateral displacements were 0.22% and 0.44% of the abutment height for linear-elastic perfectly-plastic model and Hardening Soil model, respectively, both less than the 1% limit recommended by the US FHWA.