PITA Fiscal Year 2013 Projects

Multi-scale and Anisotropic Computational Modeling and Simulation of Additive Manufacturing Process

Lead University: Carnegie Mellon University
PI: Kenji Shimada, Mechanical Engineering
Co-PI: Soji Yamakawa, Mechanical Engineering
PA Industry: Ciespace Corp, Autodesk Inc, The ExOne Company

The goal of the proposed PITA project is to develop a computational modeling and simulation framework for additive manufacturing, or 3D printing, to predict accurately the geometric and structural properties of 3D-printed parts. Such a computational framework will assist the users of a 3D printing device in optimizing the manufacturing process variables to achieve the target geometric tolerance and structural strength without wasteful and time-consuming trial and error.

The key challenge for accurate simulation and prediction of geometric and structural properties of a 3D-printed part is to model multi-scale geometric structures and derive anisotropic material properties in the subsequent numerical analysis, finite element method (FEM) and/or discrete element method (DEM). In the proposed work, we will represent geometry by two scales, meso and global. After the local anisotropic material property is calculated using a meso-level model, its aggregate anisotropic material property will be mapped to a global-level model.

The accuracy of the computer simulation and prediction of geometry and structural properties will be measured and validated by physical experiments using two types of 3D printing device: (1) a fuse deposition modeling (FDM) device, and (2) a metal/ceramic printing (MCP) device. The FDM experiments will be conducted at Carnegie Mellon University using its in-house FDM printer, and the MCP experiments will be conducted at the facilities of an industry partner, The Ex One Company LLP, a Pittsburgh-based manufacturer of various MCP machines.

If successful, in addition to reducing the cost of additive manufacturing by eliminating the need for a for trial-and-error-based search for an optimal set of manufacturing-process variables, the proposed computational modeling and simulation framework will help the developers of additive manufacturing devices improve their device design by offering new insights on how local meso-level structures will affect the global geometric and structural properties of a final 3D-printed part.