Calibration of Resistance Factor Based on Pile Load Test Conducted to Failure

In Load and Resistance Factor Design (LRFD), the resistance factor for pile foundation is used to account for uncertainties in pile capacities. With pile load tests conducted, the uncertainties associated with pile capacity prediction can be greatly reduced, thus, the resistance factor can be updated based on the load test results. In this paper, a probabilistic approach based on Bayes' theorem and the Monte Carlo Simulation (MCS) is proposed to calibrate the resistance factor based on pile load tests conducted to failure. To illustrate the proposed approach, the pile capacity predicted by the standard penetration test (SPT) is adopted. The measured pile capacity is normalized by the predicted pile capacity to obtain the resistance bias factor. Parametric studies are performed to evaluate the effect of the number of load tests, mean and minimum resistance bias factors on the resistance factor. Results show that the resistance factor increases when the number of load tests increases if all resistance bias factors are larger than 0.75. The resistance factor increases when the mean resistance bias factor increases. For the same mean resistance bias factor, the resistance factor decreases as the minimum resistance bias factor decreases. The results may explain why the specification in Eurocode 7 recommends the equivalent resistance factors depending on the mean and minimum measured pile capacities in load tests.

Keywords: LRFD, resistance factor, pile load test, Bayes' theorem, MCS.