Development of Bayesian-based Solution to Update the Resistance Factors of Piles Using Combined Static and Dynamic Pile Load Test Results

The Load and Resistance Factor Design (LRFD) for deep foundations provides limited guidance on how to update Resistance Factors (ϕ) that incorporate the site-specific Static Pile Load Tests (SPLTs) and/or Dynamic Pile Load Tests (DPLTs) into design. This study develops a comprehensive Bayesian interface analytical solution for updating resistance factors of driven Precast Prestressed Concrete (PPC) piles using combined SPLT-DPLT information. Three calibration components are presented: update ϕ using SPLT alone, update ϕ using DLTs alone, and (3) apply a new analytical Bayesian updating solution to combine priors with multi-based type likelihoods (SPLTs + DPLTs) through corrected population variance and site-specific variability. The solution is applied to Laboratoire Central des Ponts et Chaussées (LCPC) CPT-based design method for driven PPC piles evaluated using a database of 80 pile load tests in Louisiana as prior information, and site variability categorized into low, medium, and high Coefficient of Variation (COV) ranges. Monte Carlo simulation is used to calibrate ϕ for varying testing scenarios and likelihood bias levels. The results show that most of the improvement from both SPLT and DPLT testing is realized after three SPLT tests and about 30–35% of the piles tested using DPLT. Incorporating the combined SPLT-DPLT likelihood information reduces the uncertainty more quickly, leading to higher and more stable ϕ than using results of STLTs or DPLTs alone.