A new approach to eliminating enclosed voids in topology optimization for additive manufacturing

•A new approach to controlling the structural connectivity is proposed.•Graph theory is introduced in combination with the BESO technique to eliminate enclosed voids.•Highly efficient structural designs without enclosed voids can be achieved.•The new approach solves a problem of practical importance in additive manufacturing. Topology optimization is increasingly used in lightweight designs for additive manufacturing (AM). However, conventional optimization techniques do not fully consider manufacturing constraints. One important requirement of powder-based AM processes is that enclosed voids in the designs must be avoided in order to remove and reuse the unmelted powder. In this work, we propose a new approach to realizing the structural connectivity control based on the bi-directional evolutionary structural optimization technique. This approach eliminates enclosed voids by selectively generating tunnels that connect the voids with the structural boundary during the optimization process. The developed methodology is capable of producing highly efficient structural designs which have no enclosed voids. Furthermore, by changing the radius and the number of tunnels, competitive and diverse designs can be achieved. The effectiveness of the approach is demonstrated by two examples of three-dimensional structures. Prototypes of the obtained designs without enclosed voids have been fabricated using AM.