Wong, J., Sanders, E. D., & Rosen, D. W. (2026). Toolpath-integrated topology optimization for design of additively manufactured fiber-reinforced structures considering limits on fiber curvature. Composite Structures, 378, 119897. https://doi.org/10.1016/j.compstruct.2025.119897
Abstract:
Topology optimization together with additive manufacturing is a promising approach to design and manufacture fiber-reinforced polymer composite structures with high mechanical efficiency attributed to geometrically complex fiber layouts. Nevertheless, structural performance is often degraded in transitioning from design to manufacturing when fiber toolpaths are generated in post-processing with properties that do not match those used during design. We propose a density-based topology optimization formulation for volume-constrained compliance minimization that not only enforces manufacturing-specific local fiber curvature constraints to reduce deviations between as-designed and as-manufactured fiber layouts, but that also integrates a toolpath generation step into each optimization iteration such that mechanical properties associated with as-manufactured fiber–matrix distributions are considered during design. Fiber orientation design variables are defined at support points of radial basis functions so that the curl field associated with the fiber orientations can be expressed as differentiable functions and used to impose the local fiber curvature constraints. To define the fiber–matrix layout concurrently with the design iterations, a two-term truncated Fourier series representation of the fiber orientation design variable field is projected onto a fiber–matrix composite microstructure with finite length scale, and mechanical properties of fiber and matrix are considered directly in the analysis. Several design examples are provided to illustrate how curl constraints can be controlled to meet process-specific fiber curvature limits, how parameters of the Fourier series and projection operation can be tuned to achieve process-specific fiber–matrix distributions, and how the optimized topology, geometry, and fiber layout change when such constraints and parameters are varied.
License type:
Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0)
Funding Info:
This research was supported by David Rosen's A*STAR start-up fund.