Reliability analysis of bearing capacity for piles based on random weighting method

In pile foundation engineering, due to high cost and time limit for a project, the data of test piles are often scarce and the dispersion degree is very large, which makes the error of calculated results of reliability for piles very large. T he theory of the random weighting method is employed to improve the bias factor (the ration of measured value of calculated value) of bearing capacity for piles into big samples meeting the engineering requirements, and the formulation of reliability is presented by using the first-order second moment when the values of the bias factor are improved. The results from case study indicate that the variance of the bias factors become small when the times of weighing become large, which shows that the values are more concentrated. In addition, with the times of weighing increasing, the reliability index becomes large, which means that the pile foundation is much safer, therefore it is unreasonable to calculate the reliability only using the original values, thus the danger maybe exists in the projects.