A grammar L-system-based generative design process to improve the seismic performance of structures with steel dampers

This study aims to significantly advance the practical application of steel dampers by introducing a novel approach that integrates building standards with the optimization of problems involving multiple conflicting objectives. Departing from conventional approaches that rely solely on linear calculations or simplified optimization, this research strategically combines linear and nonlinear analyses within a robust optimization framework. A key contribution is the explicit consideration of separate requirements for the structural system without energy dissipators. This aspect, often overlooked in prior works, leads to more realistic and code-compliant designs. Furthermore, to address the limitations of single-objective optimization, which fail to capture the inherent trade-offs between critical design objectives such as drift, element performance, and cost, this study employs a multi-objective optimization algorithm. To enhance the practicality and efficiency of the design process, a fractal is implemented using L-systems to parametrically define damper properties, substantially reducing the number of design variables compared to optimizing individual damper characteristics. Numerical results for two different building types demonstrate that the proposed methodology successfully produces optimized designs that meet seismic performance objectives. Crucially, this is achieved with significantly reduced computational time, from days to mere hours, compared to traditional iterative approaches. This innovative approach offers a more robust and implementable methodology for designing effective steel-damped structures, ultimately contributing to safer and more cost-efficient earthquake-resistant buildings.