Continuous fiber-reinforced additive manufacturing (CFRAM) holds significant promise for fabricating lightweight, load-bearing cellular structures. However, conventional fiber alignment strategies, which typically based on principal stress vector fields, are primarily optimized for fixed loading conditions, which rarely reflect real-world applications. To address this, we propose a global optimization framework guided by Rotational Symmetric (RoSy) frame fields to accommodate variable loading conditions and ensure geometric continuity. By employing a representation vector field together with a neural solver to resolve the inherent mathematical challenges of direction ambiguity and singularities, our method enables the smooth alignment between multi-directional fiber paths and integrated principal stress fields. The resulting structures are decomposed into simultaneously optimized matrix and fiber layers, achieving a critical balance between stress-oriented reinforcement and structural integrity. Both simulations and physical experiments demonstrate that this frame-field-guided approach significantly enhances the mechanical performance and versatility of cellular structures under complex and varying loads. • A frame field guided design method for fiber-reinforced cellular structures. • Stress-aligned and smooth fiber paths generated to enhance load-bearing strength. • Stable load-bearing performance achieved under variable loading conditions.
Xu et al. (Fri,) studied this question.