Centrifuge modelling of models of seismic effects on saturated earth structures

Centrifuge modelling of seismic effects on saturated, liquefiable, cohesionless soil structures is discussed in this paper. A methylcellulose solution (metolose) was used as a substitute pore fluid to alleviate problems arising from the conflict in time scaling in the modelling of dynamic and dissipative phenomena. A modelling of models type of study was performed on retaining wall models with saturated cohesionless backfill using both metolose and water as the pore fluid. These experiments were analysed using the finite element program DIANA-SWANDYNE II, and parametric studies were also performed. Based on the experimental and numerical results, it was confirmed that the conflict between dynamic and consolidation time scales exists and a substitute pore fluid is necessary for modelling prototype behaviours. It was shown that metolose is a suitable replacement pore fluid. It was also concluded that seismic modelling of models can be achieved only if a suitable substitute pore fluid is used. However, depending on the type of model, type of earthquake motion, and drainage boundary conditions, a false indication of achievement of seismic modelling of models can be obtained with water as the pore fluid which may not necessarily represent real prototype behaviour.