Mechanical properties of carbon fiber-reinforced polymers (CFRP) are strongly dependent on the fiber volume fraction. Recent advances in additive manufacturing (AM) enable the use of endless fiber-reinforced filament within fused filament fabrication (FFF) processes. Therein, the fiber volume fraction depends on the path geometry, i.e., path width and layer height. In order to deposit the filament on non-planar printing layers for an improved load-orientation of the reinforcement fibers, the layer height has to be varied within one layer during slicing [1]. In consequence, the fiber volume fraction and therefore the mechanical properties such as stiffness and strength of the manufactured structure vary locally. The possibilities of non-planar FFF are exploited by anisotropic topology optimization, where in addition to the material distribution the local orientation of material is optimized for every element within the continuum structure [2]. The material properties are commonly considered in a deterministic way without spatial variation. Slices as well as printing paths for AM are computed from the optimized density and orientation fields. Only after the slicing procedure a local fiber volume fraction can be estimated and projected onto the numerically optimized simulation model. The authors present a method to relate properties of the vector field representing material orientation such as curvature or divergence to the fiber volume fraction of the manufactured composite structure and therefore the mechanical properties, so that the numerical optimization model can consider them as design-dependent material properties [3]. REFERENCES [1] J. Kipping and T. Schüppstuhl, “Load-Oriented Nonplanar Additive Manufacturing Method for Optimized Continuous Carbon Fiber Parts,” Materials, vol. 16, no. 3, Art. no. 3, Jan. 2023, doi: 10.3390/ma16030998. [2] K. Steltner, J. Kipping, B. Kriegesmann, and T. Schüppstuhl, “A workflow for designing stiffness-optimized structures in the context of additive manufacturing of endless fiber-reinforced composites,” Journal of Thermoplastic Composite Materials, 2025, Accepted for publication. [3] B. Kriegesmann, “Robust design optimization with design-dependent random input variables,” Struct Multidisc Optim, vol. 61, no. 2, pp. 661–674, 2020, doi: 10.1007/s00158-019-02388-3.