Stiffness and dimensional accuracy experimental validation of metallic 3d printed lattice structures

S. SARASOLA; A. CISILINO; L. STRAUSS; J. MONTERO

Mar del Plata

Congreso; XII Latin-American Congress of Artificial Organs and Biomaterials; 2023

Univ. Mar del Plata, Univ. Rosario, CONICET

Introduction and Goal: Metallic laser-based powder bed fusion enables sub-millimeter scale 3D printing, allowing the design of biocompatible multi-scale cellular structures applicable in the medical sector for customized bone substitutes and prostheses. The tetrakaidecahedron base cell is a well-known geometry that has been studied in prosthetics due to its similarity to low-density foams and space-filling capacity. This work aims to experimentally validate the dimensional accuracy of printed parts and its effect on the elastic behavior of a tetrakaidecahedron- based microstructure against theoretical designs.Methodology: Three batches of specimens were designed. Each batch contained five specimens with 5x5 tetrakaidecahedrons in the cross-section and eight tetrakaidecahedrons in height. The batches differ in the orientation of the tetrakaidecahedrons, i.e., they comprised three orientations according to the three orthogonal spatial directions. The batches were manufactured in a laser-based powder bed fusion machine with Ti6Al4V material and 30 μm of layer height. The stiffness of the base material was characterized via nanoindentation testing. Computer Tomography (CT) scans were used to quantify the solid fraction and trabecular thickness distribution. The specimens were tested in compression. FEM models of the test were made to study the effect of dimensional distortions and model boundary conditions.Results and discussion: The comparison between the CT of printed specimens and the CAD designs shows a solid fraction 15% higher in the physical parts. The measured strut thickness distribution, i.e., trabecular thickness, also differed from the digital design. While a good correlation was found in the first peak at 800 μm, there was a disparity at the second peak. There, the CT scan measured 1000 μm, yet in the digital design, the second peak was 875 μm. These differences are caused by excess material around the struts, particularly on the vertices of the structure, which causes a pseudo-fillet. Based on the CT results, the digital model of the structure was adjusted and analyzed via FEM and compared with a model without geometry deviations. The comparison of both models, considering buckling due to the lack of lubrication in compression, results in a stiffness increase of 10%.Conclusions: The comparison of results against an asymptotic homogenization proves that the presented specimen design contains enough cells to represent the periodic arrangement of cells. Engineers should note that the dimensional deviations of the printed parts should be considered when designing the microstructure. The induced filleting effect in sharp geometric transitions, such as the vertices of the structure, may impact the stiffness by up to 25% with a marginal effect on the solid fraction. Still, the excess material around the struts increases the solid fraction by up to 15% without considerable impact on the stiffness.