Finite Element Modeling of 3D Printed Parts With Defects
thesisposted on 2021-05-24, 10:54 authored by Geethanjali Chandramouli
To manufacture a component, one first needs to assess its structural design performance, damage tolerance, and service experience, and validate them with pertinent test results. Finite Element (FE) modeling can predict mechanical performance, save time and cost by limiting required structural testing. 3D printing is a layer-by-layer manufacturing technology that has been widely used for rapid prototyping applications in product design and development. Recently, there has been a move towards manufacturing functional products using 3D printing, which requires materials mechanical characterization and simulation. Mechanical characterization testing results are available for 3D printed ASTM D638 tensile coupons without defects, i.e. tension along (0°) and transverse (90°) to the printing direction, and a quasiisotropic stacking sequence. In addition, tensile test results of a quasi-isotropic coupon with intentional defects are also available. In this project, FE models of the coupons are created to obtain their tensile strength, modulus, and failure strain. First ply, last ply failure and stiffness reduction iterative approach have been implemented on a 2D shell model. MSC Software is used to simulate the analyses due to its ease of use for composites using 2d shell elements. This simulation is then extended to predict strength and stiffness of a quasi-isotropic coupon with defects. The analysis is also extended to implement progressive failure analysis to predict the ultimate strength of the laminate. For coupons without defects, FE models estimated test results of stiffness and strength within 1% error, while the error for estimating failure strain is higher. For coupons with defects, the error in calculating stiffness and strength is below 8%, while it is higher for failure strain. Although the stress-strain curve from FE simulation looks similar to experimental result, it is found that progressive failure analysis is necessary for obtaining failure strain values with acceptable error percentage.