Probabilistic Moment Bearing Capacity Model and Fragility of Beam-Column Joints with Cast Steel Stiffeners

Abstract

Beam-column joint with cast steel stiffeners (CSS) is a new type of joint with a large degree of design freedom. The joint stress distribution can be improved by designing a reasonable cross-sectional shape of the CSS with high rigidity, high integrity, and good seismic performance. Due to the construction specificity, the exact theoretical formula for the moment bearing capacity of the CSS joint is hard to deduce. Some researchers have proposed empirical or simplified theoretical formulas for the prediction of moment bearing capacity. However, the formulas are biased and cannot capture uncertainties in the data measurement and modeling process. In addition, current formulas cannot be updated efficiently over time, and no work has been conducted regarding the reliability of the CSS joints subject to different loading conditions. In this paper, a new approach to address the above issues is proposed. A probabilistic model for the joint capacity is established to capture the uncertainties and correct the bias. A Bayesian method is proposed for model training, which allows the model to be updated efficiently whenever new experiment or simulation data are available. A fragility analysis is conducted using the proposed capacity model to quantify the failure probability of joints under different loading conditions. The advantages of the proposed approach are validated by analyzing joints in a database obtained from experiments and numerical simulations. Results show that the proposed capacity model provides unbiased and more accurate estimates of the bending moment than the currently available ones. New factors such as column thickness and concrete filling are found to significantly impact the moment capacity. The bending fragility of CSS joints can be lowered at different degrees by increasing concrete strength, steel strength, column thickness, etc. Guidance on CSS joint design and retrofitting based on the capacity model and fragility analysis is also presented at the end of this paper.