Reliability analysis of studded shear connectors in composite slabs subjected to shear loading

Abstract

This study investigated the reliability of stud connectors in composite slabs subjected to shear loading, focusing on the adequacy and efficiency of design provisions outlined in four major international codes: SANS 10162-4, EN 1994, AISC 360, and AS/NZS 2327:2017. The analysis evaluated the safety margins and conservatism inherent in these codes, using both the First Order Reliability Method (FORM) and Monte Carlo Simulation (MCS) to assess the probability of failure of stud connectors under various loading and material conditions. The study began by analysing model uncertainty, revealing significant differences in bias and variability among the design codes. The SANS 10162-4 model showed no bias and minimal variability, while AISC 360 demonstrated the largest bias and variability. EN 1994 and AS/NZS 2327:2017 displayed moderate levels of conservatism. Reliability indices were computed for each design code, with results indicating that all models provided sufficient safety margins, though EN 1994 and AISC 360 exhibited excessive conservatism that may lead to overdesign. A sensitivity analysis based on FORM identified variable actions as the most critical factor affecting reliability, followed by stud diameter. The implications of these findings were used to propose adjustments to partial safety and reduction factors to optimize design efficiency without compromising safety. For instance, a reduction in the partial safety factor from 1.25 to 1.1 for EN 1994:2005, and an increase in the partial reduction factor for AISC 360 from 0.65 to 0.80, were recommended. The study concludes that while the current design standards provide adequate safety, there is significant potential for optimization, particularly in reducing conservatism in certain models. Future research is recommended to refine design models, account for more complex loading conditions, and explore probabilistic methods to further enhance the reliability and efficiency of stud connector designs. Limitations of the study include the use of simplified load conditions and assumptions about material properties, as well as the exclusion of long-term effects such as creep and shrinkage. In summary, this study evaluates the reliability performance of existing stud-connector design models and develops calibrated resistance factors that enhance their consistency with target reliability levels, providing evidence that can inform future improvements to design provisions while maintaining adequate safety margins.