Probabilistic assessment of torsion in concrete beams externally strengthened with CFRP composites

This paper illustrates an analytical probabilistic study of concrete beams subjected to torsion that are strengthened with carbon fiber reinforced polymer (CFRP). Hii and Al-Mahaidi’s method is one of the best analytical models for evaluating the torsional capacity of CFRP strengthened reinforced concrete beams. The first-order reliability method is carried out to probabilistically assess the capacity of CFRP-strengthened beams. For this aim, the statistical characteristics of design variables and model errors have been considered, followed by the determination of the average reliability indexes of the strengthened beams. The effect of each design variable on the average reliability is also determined. The assessment shows that the Hii and Al-Mahaidi’s analytical model is unconservative. In order to correct the situation, a more relaxed set of resistance factors for use in a load and resistance factor design format are needed. These are then determined for two target reliability levels of $ \beta_{\text T} $  = 3.0 and $ \beta_{\text T} $  = 3.5. It is found that factors of 0.9200 and 0.8225 are needed for target reliability levels, $ \beta_{\text T} $  = 3.0 and $ \beta_{\text T} $  = 3.5, respectively. Values of 0.9 and 0.8 are suggested for use in real practice depending on the target reliability sought.