Evaluation of drained residual interface strength between silty/dry sand & various surfaces at low effective stresses

The motivation of this study is to investigate how a sand-surface interface behaves under low effective stresses. The main objectives of this study are to- (1) record the peak and residual interface friction between sandy soils and different interfaces using tilt-table apparatus; (2) observe whether different sample thickness changes the shear strength significantly in sandy soils; (3) study the effect of interface coating on the failure mechanism; (4) study the influence of normal stresses on residual failure angle or friction coefficient; (5) compare frictional coefficient of polypropylene-sand interface from model flowline axial resistance tests & tilt-table tests; (6) study the drained response of silty sand. Tilt-table tests have been used to find the interface friction for mostly clay and solid interfaces, so there is a need to study the behavior of sand-solid interface. A series of 76 tests have been conducted. A total of 62 tilt-table tests have been conducted (47 on silty sand & 15 on dry sand) on five different interfaces and a total of 14 model flowline axial resistance tests have been conducted to compare with the data obtained from tilt-table tests. These tests provide a better idea of the interaction between the sandy soil and several surfaces under low effective stresses. The following conclusions were drawn from this study- (1) Changing the effective stresses may lead to change in failure type. The failure angle measured when the internal failure starts may not be a correct estimation of the interface strength as the soil fails internally in some cases; (2) At very low effective stresses, the failure angle on smooth surfaces is governed by the soil-interface interaction while on rough surfaces, it can be governed by internal soil particle interaction; (3) The friction coefficient obtained from tilt-table test and model flowline axial resistance test on dry sand were similar. It is recommended for future tests that the model pipes are loaded as closely as possible to the vertical center of mass on the pipe.