Date of Award

5-2017

Document Type

Thesis

Degree Name

Master of Science (MS)

Legacy Department

Mechanical Engineering

Committee Member

Dr. Cameron Turner, Committee Chair

Committee Member

Dr. Georges Fadel, Committee Member

Committee Member

Dr. Gang Li, Committee Member

Abstract

Additive manufacturing (AM) is increasingly used in new product development: from prototyping to functional part testing, tooling and manufacturing. The flexibility of AM results in the ability to develop a geometrically complex part with reduced effort by moderating some manufacturing constraints while imposing other constraints. However, additively manufactured parts entail a certain amount of ambiguity in terms of material properties, microstructures effects and defects. Due to the intensive energy, rapid cooling and phase changes, parts made by Fused Deposition Modelling (FDM "“ a branch of AM) and other layer-manufacturing processes may deviate from the designed geometry resulting in inaccuracies such as discontinuities, curling, and delamination, all of which are attributed to the residual stress accumulations during geometry fabrication. Therefore, the FDM part can strongly differ from its design model, in terms of strength and stiffness. In performance critical applications, analyzing and simulating the component is necessary. Identifying appropriate methodologies to simulate and analyze additively manufactured parts accurately, enables better modelling and design of components. The Finite Element Method (FEM) is a widely used analysis tool for various linear and nonlinear engineering problems (structural, vibrational, thermal etc.). Therefore, it is necessary to determine the accuracy of FEA while analyzing the non-continuous, non-linear FDM parts. The goal of this study is to compare Finite Element Analysis (FEA) simulations of the as-built geometry with the experimental tests of actual FDM parts. A dogbone geometry is used as a test specimen for the study, with a set of different infill patterns. A displacement controlled tensile test is conducted using these specimens to obtain the experimental stress-strain results. Further, as built 3D CAD models of these specimens are developed and a displacement controlled tensile test is simulated using different material models in two FEA solvers. The stress-strain results of the analyses are compared and discussed with the experimental results. The metrics of the comparison are the precision and the accuracy of the results. This study found that FEA results are not always an accurate or reliable means of predicting FDM part behaviors, even when advance experimentally derived material models and as-built geometries are incorporated.

Share

COinS