Key points are not available for this paper at this time.
This paper presents an optimization framework for steel trusses. The authors implemented a penalty-based approach to optimise the size, shape, and topology based on a dynamic grouping strategy to address the constructability challenges. The main contribution of the paper is the use of damped exponential constructability penalties. This approach ensures optimal designs by balancing structural complexity, through standardization in design, and minimizing the total number of members and variety of sections, with the overall structural cost. The paper also presents a detailed analysis that underscores the sensitivity of the optimization convergence to the algorithmic hyperparameters, emphasizing the role of cross-section assignments and stabilization of truss piece counts. The optimization framework is validated on a trussed roof structure based on the findings from the single truss optimization. The best truss topology proved to be the Howe truss configuration, highlighting its efficiency in meeting the defined objective function.
Cucuzza et al. (Tue,) studied this question.