Downdrag in pile design: testing instrumented h-piles to advance engineering practice in Manitoba

Abstract

Design of piles subject to ground settlement requires evaluation of three limit states including the geotechnical ultimate limit state (ULS), the structural ULS, and serviceability limit state (SLS). The structural ULS and SLS can be evaluated by performing a neutral plane analysis to calculate pile settlement and the maximum force in the pile under serviceability loads. The resistance distribution and tip stiffness of the pile are required input for a neutral plane analysis. The theoretical approach for evaluating the geotechnical ULS can differ for various design codes. The Canadian Highway Bridge Design Code (CHBDC) neglects negative skin friction when assessing the geotechnical ULS. The AASHTO LRFD Bridge Design Specifications includes drag force as an additional load on the pile and neglects shaft resistance where negative skin friction exists. Both codes are specified in Manitoba, and it is unclear how this discrepancy in codes affects bridge foundation design. A test pile program was developed to test the underlying theory behind evaluating the geotechnical ULS for the two bridge design codes and to quantify impacts on foundation design. The study was also designed to improve estimates of resistance distribution and tip stiffness of piles using dynamic testing with CAPWAP analysis and piezocone testing. Instrumented test piles were installed through a compressible clay layer to a hard end bearing stratum, then subjected to ground settlement by construction of a surrounding embankment. The capacity of the piles was measured during installation with dynamic testing and after experiencing ground settlement with static load testing. A load transfer model was calibrated to observations from the static load tests to model the pile capacity, settlement, and drag force for varying serviceability loads and distributions of negative skin friction. The results show that negative skin friction did not detrimentally affect the pile capacity. It was found that estimates of resistance distribution and pile tip stiffness using CAPWAP analysis were improved by considering residual stresses and setup. It is concluded that the geotechnical ULS calculations according to the AASHTO LRFD Bridge Design Specifications can result in more conservative foundation designs, requiring more piles, than compared to the CHBDC.