The SKS02 centrifuge model test examined liquefaction-induced downdrag in a predominantly uniform liquefiable soil deposit. The soil profile consisted of a 1 m of coarse sand, underlain by 4 m of clay, 9 m of liquefiable loose soil, and a deeper dense soil (prototype scale). The test included two 635 mm diameter (D) piles, with their tip embedded to the depth of 0D and 5D in the dense sand layer. The model was shaken with a suite of scaled Santa Cruz earthquake motions with peak horizontal accelerations ranging from 0.025 g to 0.4 g. The model was heavily instrumented to track the development of pore-pressures, axial loads, and soil and pile settlements during and post shaking. With multiple shakings, the drag load on the piles increased and saturated at the load curve estimated from the mobilization of full negative skin friction (assuming it equal to the interface shear strength). A higher drag load was observed along the deep-embedded 5D pile as compared to the shallow-embedded 0D pile. While soil settlements were large, the piles underwent considerably smaller settlements. Most of the pile settlement (>90%) occurred during shaking and very minimal settlement (<5mm) occurred during reconsolidation.

The SKS03 centrifuge model test examined liquefaction-induced downdrag in a layered liquefiable soil deposit with a surface clay crust. The soil profile consisted of 1 m of coarse sand, 2 m of clay crust, and 4.7 m of loose sandy soil (prototype scale). Below the loose sand was about 1.3 m of a clayey silt layer underlain by 5 m of medium dense sand and dense sand beneath it. The test included three medium 635 mm diameter (D) piles, with their tips embedded to a depth of 3D in the dense sand layer. The loads applied on the piles were 500 kN, 1500 kN, and 2400 kN, respectively. The model was shaken with a suite of scaled Santa Cruz earthquake motions with peak horizontal accelerations ranging from 0.08g to 0.61g. A pile load test was performed at a constant penetration rate to obtain the pile capacity (including the shaft and tip resistance). Similar to SKS02, drag loads were observed to increase along the piles with multiple shakings. Most of the pile settlement occurred during shaking with very minimal settlement (< 10 mm) post shaking. Among all piles, the heavily loaded ones suffered the most settlement. Higher drag loads were observed on lightly loaded piles as compared to the heavily loaded piles. The neutral plane was deeper for the lightly loaded pile and correspondingly much shallower for the heavily loaded pile.