Parametric evaluation of sampling schemes for LRFD of top deflections in laterally loaded piles

Load and Resistance Factor Design (LRFD) provides a reliability-based framework for improving pile foundation design under uncertainty. This study develops an LRFD framework for laterally loaded piles that incorporates random finite element methods to assess pile top deflection. Random field sampling is used to represent borehole investigation data, and different sampling schemes, characterised by distance and depth, are systematically evaluated. The results indicate that sampling distance critically affects the worst-case scales of fluctuation associated with high failure probabilities. For a fixed distance from the pile, increasing the sampling depth enhances the understanding of site and reduces unnecessary conservatism. The pile installation location is found to be a particularly effective sampling point among the cases examined, providing reliable subsurface information without requiring excessive depth. These findings highlight the trade-offs between investigation effort and reliability-based design outcomes. The proposed framework provides practical guidance for evaluating sampling strategies and their implications for reliable deflection design of laterally loaded piles.