Reliability-Based Optimization of Steel Beam Designs for Elevated Temperature Applications

In this article, a novel algorithm is presented for reliability-based topology optimization of steel beams under elevated temperatures.The proposed framework integrates elastoplastic limit analysis by employing the concept of a plastic ultimate load multiplier.Additionally, the algorithm incorporates the location of the applied load as a random variable to enhance reliability-based design.This technique facilitates the analysis of other critical characteristics, such as geometrical imperfections, volume fraction, and material properties, all assumed to follow a normal distribution to address uncertainties.The optimal layouts are generated using these improvements within the bi-directional evolutionary structural optimization method to minimize structural weight while maintaining high performance.Furthermore, different reliability indices result in varied topologies, indicating the sensitivity of the optimization process to these values.By incorporating the reliability index as a constraint, the algorithm effectively regulates the optimization procedure.This method offers an efficient design strategy that considers probabilistic conditions, thereby enhancing the safety and durability of structures exposed to high temperatures.