Probability Model Of Resistance Attenuation For Existing RC Bridge Member Based On Information Updating

There are about 658100 highway bridges of China, which concrete bridge account for above 90%. Corrosion due to the environment and vehicle loads has become a serious problem in these concrete bridges, which is a key factor to affect the resistance attenuation and service security. It is difficult to gain the law of life-cycle based on experiment and statistical regression method because great uncertainty lie in evaluation of existing bridge. Therefore, it is an important task to evaluate the performance of degraded for RC bridge by fully using inspection data.Aim at the above problems, the performance and resistance attenuation of RC bridge due to corrosion in chloride ions environment are studied. This work is supported by National Natural Science Fund of China under the Project NO.50908023"Research on Time-variant Reliability of Existing RC Bridge under Fuzzy and Random Information"and NO.50878031"The Degradation Model and Reliability of Shear of Existing RC Bridge". The main research works include as follows:(1) The prior probability model of corrosion rate is established. On the basis of Fick diffusion law, the distribution form and the probability density function (PDF) of corrosion initiation time, propagation time and the sensitivity of input parameters are analyzed. Monte Carlo simulation method is also applied. The effect of concrete cover and concrete strength on the probability distribution of the corrosion depth are analyzed.(2) The posterior probability model of corrosion rate is established based on information updating. The inspection data of bridge are incorporated and the prior probability model of corrosion rate is updated using bayes method, in addition, the probability density function is obtained. The results show that the updated model is more reliable and closer to the actual situation. This will provide scientific basis for durability assessment of RC bridge.(3) The tensile test on 452 corroded reinforcements from the electrochemical accelerated corrosion members and the field bridge members are conducted. The effect of corrosion rate on failure model are also discussed. The relation between steel corrosion rate and yield strength is established. The pullout tests on the bonding specimens with the deformed reinforcing bar installed by the interior strain gauge are conducted, and the associate expression of bond between steel and concrete is introduced. (4) The dynamic probability model of resistance attenuation of flexural member for RC bridge is established. In this model, the reduction of steel cross-sectional, and the decrease of corroded steel strength, and the degradation of bond properties between corroded steel and concrete are considered. For structural members with no inspection data, the effect of uncertainty information on resistance is also studied. The results show that the Bayes Model can efficiently combine the prior information and the specific information of RC bridge so that can reduce knowledge-based epistemic uncertainties, thus making resistance analysis more accurate.