Connectivity-Constrained Topology Optimization for Additive Manufacturing

Topology optimization (TO), while powerful for generating high-performance structural layouts, often yields designs with enclosed voids that hinder manufacturability in powder-based additive manufacturing (AM). To address this, this paper proposes an Adaptive Virtual Temperature Field (AVTF) method that enforces the connectivity constraint. The approach integrates a projection-based density filtering and flood fill algorithm to detect enclosed voids, combined with an adaptive penalty scheme that autonomously adjusts the virtual temperature penalty factor to eliminate disconnected regions. AVTF operates via a low-cost geometric feedback mechanism. Numerical examples demonstrate that the method effectively eliminates enclosed voids with only a marginal increase in compliance while significantly reducing the maximum virtual temperature. The resulting designs exhibit fully connected material layouts, ensuring powder removability. The method provides a practical and robust pathway toward AM-ready topology optimization, bridging the gap between structural performance and manufacturability.