结合贝叶斯-MCMC更新的RC梁抗剪承载力概率模型的可靠度分析

Reliability Analysis of the Updated Probabilistic Model of RC Beams Shear Capacity combined with Bayesian-MCMC

为分析钢筋混凝土(RC)梁抗剪承载力,首先,基于国内外100组RC梁抗剪承载力试验数据作为先验信息,结合我国GB 50010-2010规范的抗剪承载力,综合考虑混凝土与箍筋的材料强度、截面尺寸、剪跨比、配箍率等因素的影响,通过贝叶斯-马尔科夫链蒙特卡洛(MCMC)方法对我国规范模型进行更新与修正;随后,结合Monte Carlo模拟,对所建立的RC梁概率抗剪承载力模型进行可靠度分析,验证了该模型具有较好的计算精度与可靠性。结果表明:概率模型均值与试验值比值K的均值与标准差分别为1.013与0.171,试验值落在概率模型的95%置信区间范围内,且可靠指标分布在4.0左右,说明建立的概率模型具有较好的预测效果与可靠性。

To analyze shear capacity of reinforced concrete(RC)beams,this study utilizes Bayesian Markov Chain Monte Carlo(MCMC)method to update and revise the Chinese GB 50010-2010 code model.The paper involves considering factors such as material strength of concrete and stirrups,section dimensions,shear-span ratio,and reinforcement ratio,based on prior information from 100 sets of RC beam shear capacity test data from references.Subsequently,a reliability analysis of the probabilistic shear capacity model for RC beams is conducted using Monte Carlo simulation.The results validate the model′s favorable computational accuracy and reliability.The findings indicate that the mean and standard deviation of the ratio K,representing the ratio between the mean value of the probabilistic model and the experimental value,are 1.013 and 0.171,respectively.The experimental values fall within the 95%confidence interval of the probabilistic model,with reliability indices distributed around 4.0.This suggests that the probabilistic model established in this study exhibits good predictive nature and reliability.