The log-normal distribution of strength modeling which is widely used in concretecompressive strength statistics is not match the practice in terms of its domain ofdefinition Furthermore, for the complex structures, there is a lack of an effectivemethod in the qualitative and quantitative analysis of safety. To resolve the aboveproblems, based on the data collected from a large amount of practice and projects inreinforced concrete frame building, this paper aims to:(1) Develop the bounded normal distribution for concrete compressive strength.23,037values of concrete compressive strength from construction sites in Beijingwere obtained from the year of2009to2012with standard cubes testing method. Themean values, standard deviations, coefficients of variations, maxima and minimavalue were derived from the original strength values and then were analyzed andcompared with the data of1989. The results were improved in terms of its deficiencyin C25/C30, and the maxima/minima of every strength class of concrete in the formerdata. Consequently, a bounded normal distribution for concrete compressive strengthwas developed. At the same time, a goodness-of-fit of both log-normal distributionand bounded normal distribution were tested with Geary’s test.(2) Improve PSO-Based JC Method. The inertia weight was modified in this paperto improve the PSO-based JC method. The test example was separately analyzed withJC method, BP-based JC method, PSO-based JC method and improved PSO-based JCmethod, and the results were compared with MC method which was regarded as anapproximate exact solution.(3) Propose Statistics-Based MC Simulation. In this paper the statistics-based MCsimulation was presented based on probability-statistics theory and was investigatedin computing time and its accuracy by comparing with basic MC simulation andBP-based MC Simulation.(4) Conduct the comprehensive assessment of structure safety for reinforcedconcrete frame buildings. A reinforced concrete frame structure was analyzed withstochastic model building and its safety was evaluated based on the area method. Thesignificant failure modes of the building were identified with β-unzipping method andthe reinforcement scheme for failed members was introduced in brief. |