An evolutionary optimization method for modular structures made of predefined building blocks

In recent years, modular structures have been extensively used in engineering applications due to their streamlined manufacturing, reduced assembly time, and enhanced sustainability. Although many topology optimization techniques have been developed, the design optimization of structures using a small set of predefined modules remains underexplored. This study addresses this gap by proposing a bi-directional evolutionary structural optimization (BESO)-based approach that optimizes both the global material distribution and the spatial arrangement of the predefined modules. The proposed approach begins by initializing the design problem through a tailored initial configuration constructed from the given modules. An update scheme for the structure is then employed to iteratively remove inefficient element patches (i.e., groups of elements marked for removal) and add efficient blocks to improve the current design. Moreover, we propose two block-type adjustment schemes that refine the optimization process and strengthen connections between adjacent blocks. Several numerical examples demonstrate the effectiveness of the proposed approach, highlighting its capability to generate efficient modular structures from various specified block sets. This work holds significant potential in the field of prefabricated construction, particularly for projects requiring rapid construction, as it enables the direct use of existing modules to design efficient modular buildings. • An evolutionary optimization method for modular structures using prescribed blocks. • A tailored initial design constructed directly from prescribed building modules. • Structure updated iteratively by removing patches of inefficient elements. • Two block-swapping methods enhance connectivity and optimize module arrangement. • Different block sets generate diverse and structurally efficient modular designs.