Probabilistic Calibration of Stress-Strain Models for Confined Normal-Strength Concrete

• Published in: Journal of structural engineering (New York, N.Y.) N.Y.), 2021-08, Vol.147 (8)
• Main Authors: Yu, Bo, Qin, Chenghui, Tao, Boxiong, Li, Bing
• Format: Article
• Language: English
• Subjects: Calibration; Concrete; Concrete columns; Confidence intervals; Markov chains; Model accuracy; Monte Carlo simulation; Probabilistic models; Statistical analysis; Stress-strain curves; Stress-strain relationships; Structural engineering; Technical Papers; Vector space
• ISSN: 0733-9445; 1943-541X
• DOI: 10.1061/(ASCE)ST.1943-541X.0003092

AbstractA probabilistic calibration for traditional deterministic stress-strain models of square confined concrete columns was conducted based on the proposed probabilistic stress-strain model and a large number of experimental data. The probabilistic models for both peak stress and peak strain (strain corresponding to peak stress) of confined normal-strength concrete (NSC) were established first based on the Bayesian theory and the Markov chain Monte Carlo method. Then, a probabilistic stress-strain model of confined NSC was established based on the proposed probabilistic models for peak stress and peak strain. Finally, the confidence level and computational accuracy of four typical deterministic stress-train models of confined NSC were calibrated based on the proposed probabilistic models and a large amount of experimental data. The proposed probabilistic models not only describe the probabilistic characteristics of peak stress, peak strain, and the stress-strain curve, but also provide an efficient approach to calibrate the confidence level and computational accuracy of traditional deterministic models.