UC Davis Institute of Transportation Studies

A quality assurance (QA) process is proposed to prevent the possibility of legal issues for a multibridge retrofit project recently conducted by the California Department of Transportation (Caltrans) to increase the horizontal shear resistance of decks. Providing a solution requires determining an appropriate number of samples with the corresponding sampling scheme to ensure 95% compliance with the specification requirements. The determination of the appropriate sample size must recognize the practical considerations of cost, time, statistical simulation results, and the binomial distribution theory. It is not uncommon for many agencies to base quality assurance on just three samples. Using a binomial distribution, a discussion is also presented why it is not appropriate to take this limited number of samples for quality assurance. With the selection of sample size for each bridge, a representative sampling scheme that is random and unbiased was developed using the concept of uniform design as sampling strategy. The recommended acceptance criteria are specified based on the hypothesis testing results with the normal approximation of a binomial distribution.