Topology optimization of 3D-printed material architectures: Testing toolpath consideration in design

Topology Optimization (TO) methods applied to the design of material architectures allow for a wider exploration of the possible design space when compared to common geometry parameter controlled design methods. These optimal designs are often realized using Direct Ink Writing methods which exhibit characteristic features of discrete bead sizes and weak bead bonding. The resultant lack of design fidelity and toolpath dependent anisotropy has been found to negatively impact structural performance if not accounted for in the design. This paper addresses both characteristics in the design process of cellular material architectures by expanding upon the Nozzle Constrained Topology Optimization algorithm and experimentally validating the results against a typical baseline. An experimental method of deriving bond region material properties is detailed. A direct toolpath generation method from topology optimized results is proposed. Comparisons are made with conventional topology optimization design methods and performance is measured both experimentally and numerically against theoretical bounds. At relative densities $\le 70\text{\%}$, designs with nozzle constraints were able to more closely align numerical and experimental results for both performance and design fidelity (measured by relative density). In contrast, conventional topology optimized designs had higher overall performance, but little alignment between intended design and resultant experimental result. Typical designs consistently overdeposited material and inconsistently predicted performance.