Bézier-based biased random-key genetic algorithm to address printability constraints in the topology optimization of concrete structures

The advancements of additive manufacturing (AM) technologies are typically coupled with research addressing topology optimization, whose aim is to use optimization methods to achieve effective expressions of free-form design. While many studies emphasize the breakthroughs that topology optimization could bring into structural engineering, there are just a few scientific contributions that address design feasibility, accounting for the technological constraints that characterize the different AM techniques. By formulating a stress-constrained topology optimization problem with a more technologically oriented approach, this study aims to optimize concrete structures while enforcing the cross-section width and path-traceability restrictions that affect the feasibility and performance of geometries obtained through the layered extrusion technique. In particular, this paper proposes a curve-based Biased Random-Key Genetic Algorithm that optimizes stress-constrained structures and generates topologies that can be implemented without post-processing operations. The proposed algorithm, when tested on a diverse set of concrete beam configurations, effectively achieved optimized solutions that used between 81 % and 75 % less material than the full beam configuration. Additionally, each one of the designed topologies adequately met the stress requirements and process-specific constraints. Lastly, two experimental cases also highlighted the printability effectiveness of the proposed approach in conjunction with design of optimized solutions.