SIMULATION OF A LARGE-SCALE SHAKE TABLE TEST USING PM4SAND PLASTICITY MODEL TO STUDY LIQUEFACTION-INDUCED RESPONSE-Paper

Paper

Pdf Summary

The study presents a two-dimensional FLAC simulation of a large-scale UC San Diego shake-table test examining how liquefaction affects a shallow foundation. The model used the PM4Sand plasticity model and represented a layered Ottawa sand deposit, including a 1.3 m-thick loose, liquefiable layer beneath a shallow foundation. Its response was compared with measurements from the baseline test, which applied a peak acceleration of 0.398g.<br /><br />The simulation predicted an average foundation settlement of 21.94 cm, compared with 26.05 cm measured. However, it substantially overpredicted free-field heave: 12.36 cm versus 1.57 cm measured. It also underestimated excess pore-pressure generation in the middle of the liquefiable layer, overestimated it in the dense layer, and dissipated pore pressure more quickly after shaking than observed in the experiment.<br /><br />Strain results showed that shear strain was the main contributor to settlement during strong shaking. As shaking subsided, volumetric strain—associated with post-liquefaction compression and reconsolidation—became more influential. Shear strain concentrated around the foundation–soil interface and extended through the loose sand layer.<br /><br />Differences between simulation and experiment were attributed to the model’s limited mesh size, potentially inaccurate representation of boundary conditions, and insufficient refinement of soil parameters. Despite these limitations, the analysis offers insight into liquefaction-induced soil–foundation response and highlights the importance of accurately modeling boundaries, drainage, and soil behavior.

Keywords

liquefaction

shallow foundation

FLAC simulation

PM4Sand

settlement

pore pressure

Ottawa sand

shear strain

volumetric strain

shake-table test