|About this Abstract
||2022 TMS Annual Meeting & Exhibition
||Additive Manufacturing: Advanced Characterization with Synchrotron, Neutron, and In Situ Laboratory-scale Techniques II
||Predicting WAAM Material Properties via Machine Learning
||Pinelopi Kyvelou, Leroy Gardner, Lei Zou, Stasha Lauria, Carlos Gonzales, Filippo Gilardi, Odysseas Krystalakos, Amine Ammar, Victor Champaney, Mustafa Megahed
|On-Site Speaker (Planned)
Predicting material quality in metallic additive manufacturing (AM) is challenging due to the large number of parameters controlling the final output. Plates of different steel grades, thicknesses, toolpath trajectories and dwell times are manufactured using Wire-arc additive manufacturing (WAAM). From these plates, coupons are extracted at different orientations and subjected to tensile tests to determine their mechanical properties. The voltage and current signals employed during AM are monitored and analyzed to identify anomalies both manually and via machine learning. The trained neural network predicts the input signals reliably. The specimens are characterized to obtain mechanical properties as functions of the parameters studied. Data analytics identifies the most influential input parameters on mechanical properties and a regression model is developed to predict mechanical properties as a function of input signals. This presentation will discuss the experiments and the predicted mechanical properties arising from WAAM processes using the developed digital platform.
||Additive Manufacturing, ICME, Mechanical Properties