About this Abstract |
Meeting |
MS&T24: Materials Science & Technology
|
Symposium
|
Additive Manufacturing Modeling, Simulation, and Machine Learning: Microstructure, Mechanics, and Process
|
Presentation Title |
MALAMUTE Directed Energy Deposition Process Modeling and Experimental Validation through Investigation of Laser and Powder Efficiency |
Author(s) |
Luis Nuņez, Wen Jiang, Dewen Yushu, Isabella van Rooyen, Michael Maughan |
On-Site Speaker (Planned) |
Luis Nuņez |
Abstract Scope |
Directed energy deposition (DED) process modeling can assist in understanding the physical phenomena related to the process fluctuations seen during DED fabrication. This study focuses on development and validation of a DED melt pool model using finite element method (FEM) using the Idaho National Laboratory Multiphysics Object Oriented Simulation Environment (MOOSE) framework-based Application Library for Advanced Manufacturing UTililiEs (MALAMUTE). Geometric measurements from parametric experimental study of laser power and powder feed rate of single-track clads fabricated on an Optomec LENS machine with 316L stainless steel and Inconel 718 are used to validate a 2D cross-sectional model. The developed MALAMUTE model is calibrated by investigating fluctuation in heat input with laser efficiency and mass input with powder efficiency to match the experimental measurements. Through a stochastic sampling of input parameters and efficiencies, machine learning surrogate models are trained to investigate and predict associated laser and powder efficiencies for reduced geometric error. |