Validation of a three‐dimensional constitutive model for nonlinear site response and soil‐structure interaction analyses using centrifuge test data

Abstract

Summary The capability of a bounding surface plasticity model with a vanishing elastic region to capture the multiaxial dynamic hysteretic responses of soil deposits under broadband (eg, earthquake) excitations is explored by using data from centrifuge tests. The said model was proposed by Borja and Amies in 1994 (J. Geotech. Eng., 120, 6, 1051‐1070), which is theoretically capable of representing nonlinear soil behavior in a multiaxial setting. This is an important capability that is required for exploring and quantifying site topography, soil stratigraphy, and kinematic effects in ground motion and soil‐structure interaction analyses. Results obtained herein indicate that the model can accurately predict key response data recorded during centrifuge tests on embedded specimens—including soil pressures and bending strains for structural walls, structures' racking displacements, and surface settlements—under both low‐ and high‐amplitude seismic input motions, which was achieved after performing only a basic material parameter calibration procedure. Comparisons are also made with results obtained using equivalent linear models and a well‐known pressure‐dependent multisurface plasticity model, which suggested that the present model is generally more accurate. The numerical convergence behavior of the model in nonlinear equilibrium iterations is also explored for a variety of numerical implementation and model parameter options. To facilitate broader use by researchers and practicing engineers alike, the model is implemented as a “user material” in ABAQUS Standard for implicit time stepping.

Deterministic ground motion simulations with shallow crust nonlinearity at Garner Valley in Southern California

“…This has motivated the development of more practical, yet still effective constitutive models that can reproduce the main features of dynamic soil behavior using only a small number of parameters. 31,[35][36][37] Among them, the bounding surface multiaxial plasticity model by Borja and Amies 31 (BA model) is one of the extensively used models, which has consistently shown good performance in small-scale applications studied to date [38][39][40] and therefore is adopted in this study to be used for large-scale ground motion simulation problems.…”

Effectiveness of particle tuned mass damper devices for pile‐supported multi‐story frames under seismic excitations

“…A calibrated and validated nonlinear soil model with a viscous damping‐bounding surface plasticity model (NLV) 47 is also adopted in the nonlinear analyses. The parameters of the NLV model are set according to the soil properties—namely, the radius of the bounding surface is set as R ≈ 0.0235 G max , where G max is the initial shear modulus; parameters for the intensity of the hardening are set as h = 0.11 G max , m = 0.93; and the viscous damping coefficient is set as ξ 0 = 0.014.…”

Effect of Flexibility Ratio on Seismic Response of Cut‐and‐Cover Box Tunnel

“…Slip and normal behaviour of the soil-structure interface was simulated using the surface-to-node interface. An interface slip coefficient was defined as 0.7 tan(φ soil ), as used by Deng et al [34] and Zhang et al [35] to validate the numerical model. Tsinidis [7] reported that simulating only the slip and restraining the potential separation between the soil and tunnel produces an unrealistic transmission of the tensile stress to the soil in contact.…”