Nylon lattice design parameter effects on additively manufactured structural performance.

Lattice structures are used in a multitude of applications from medical to aerospace, and their adoption in these applications has been further enabled by additive manufacturing. Lattice performance is governed by a multitude of variables and estimating this performance may be needed during various phases of the design and validation process. Numerical modeling and constitutive relationships are common methodologies to assess performance, address risks, lower costs, and accelerate time to market for innovative and potentially life altering products. These methods are usually accompanied by engineering rationales to justify the methods appropriateness. However, engineering analyses and numerical models should be validated using experimental data when possible to quantify the accuracy of their predictions under conditions relevant to their planned use. In this work, a set of lattice design parameters are evaluated using numerical modeling and experimental methods under quasi-static tensile, compressive, and shear modalities. Regular body centered cubic (BCC) and stochastic Voronoi Tessellation Method (VTM) lattices are constructed with three different cell lengths (2.5 mm, 4.0 mm, 5.0 mm) and various strut diameter thicknesses (ranging from 0.536 mm-1.3506 mm) while maintaining the lattice's relative density (0.2 and 0.3). Some strut diameters were selected to challenge the AM process limits. Specimens were fabricated in nylon 12 on a laser powder bed fusion system. Optical microscopy showed up to a 28.6% difference between as-designed and fabricated strut diameters. Simulated reaction loads revealed up to a 4.6% difference in BCC lattices within a constant relative density at a 1.4 mm displacement boundary condition while the VTM samples had up to a 19.5% difference. Errors between the experimental and simulated lattice reaction loads were as high as 97.0%. This error magnitude appears to strongly correlate with lattice strut diameter. These results showcase that a computational estimation, even one with reasonable assumptions, may erroneously characterize the performance of these lattice structures, and that these assumptions should be challenged by experimentally evaluating and validating critical quantities of interest.