Impact of Proof Load Test Programs on the Reliability of Foundations

Abstract

Traditionally, proof-load tests have been utilized to validate design methods and construction procedures. There is currently an inconsistency in the recommendations that are available in pile design codes and practices regarding the required number of proof-load tests and the level of the proof loads. Recently, and in the framework of reliability-based design, researchers have shown that information from pile load tests may have a considerable impact on reducing the probability of failure of piles at a site, thus allowing for the use of lower factors of safety for the piles. This paper presents the results of a thorough investigation that is conducted to study the effect of choosing different proof-load test programs on the reliability of piles. This is achieved by utilizing a Bayesian approach to update the capacity distributions of piles at a site given the results of the proof-load test program. In the updating exercise, an effort is made to update both the mean and the lower-bound capacity to maximize the benefit of the collected proof load data. The significance of the results presented lies in the fact that these results constitute necessary input to any practical decision framework for choosing the number and the magnitude of the proof load test that would maximize the value of information of the test program.