Uniaxial Compression of 3D Printed Samples With Voids: Laboratory Measurements Compared With Predictions From Effective Medium Theory

3D printing technology offers the possibility of producing synthetic samples with accurately defined microstructures. As indicated by effective medium theory (EMT), the shapes, orientations, and sizes of voids significantly affect the overall elastic response of a solid body. By performing uniaxial compression tests on 20 types of 3D-printed samples containing voids of different geometries, we examine whether the measured effective elasticities are accurately predicted by EMT. To manufacture the sample, we selected printers that use different technologies; fused deposition modelling (FDM), and stereolithography (SLA). We show how printer settings (FDM case) or sample cure time (SLA case) affect the measured properties. We also examine the reproducibility of elasticity tests on identically designed samples. To obtain the range of theoretical predictions, we assume either uniform strain or uniform stress. Our study of over two hundred samples shows that measured effective elastic moduli can fit EMT predictions with an error of less than 5% using both FDM and SLA methods if certain printing specifications and sample design considerations are taken into account. Notably, we find that the pore volume fraction of the designed samples should be above $\approx 1\%$ to induce a measurable softening effect, but below $\approx 5\%$ to produce accurate EMT estimations that fit the measured elastic properties of the samples. Our results highlight both the strengths of EMT for predicting the effective properties of solids with low pore fraction volume microstructural configurations, and the limitations for high porosity microstructures, particularly, those with interactive pores geometries.