Session: 02-11-02: Session #2: Laser-Based Advanced Manufacturing and Materials Processing

Paper Number: 95055

95055 - An Adaptive Thermal Finite Element Simulation of Direct Energy Deposition With Reinforcement Learning: A Conceptual Framework 

During last decades, metal additive manufacturing (AM) technology has transitioned from rapid prototyping application to industrial adoption owing to their flexibility in product design, tooling and process planning. Thus, understanding the behavior, interaction and influence of the involved processing parameters on the overall AM production system in order to obtain high-quality parts and stabilized manufacturing process is crucial. Despite many advantages of the AM technologies, the difficulties arise due to modelling the complex nature of the process–structure–property relations, which prevent its wide utilization in various industrial sectors. It is known that many of the most important defects in direct energy deposition (DED) are associated to the volume and timescales of the evolving melt pool. Thus, the development of methodologies for monitoring, controlling the melt pool is critical. In this study, an adaptive numerical transient solution is developed, which is fed from the set of experiments for single-track scanning of super alloy Inconel 625 on the hot-tempered steel type 42CrMo4. An established exponential formula based on the response surface methodology (RSM) that quantifies the influence of process parameters and geometries of deposited layers from experiments are considered to activate the volume fraction of passive elements in the finite element discretization. The melt pool dynamics and stability are driven by the temperature field in the melt pool and the deposition volume varies increasingly by accumulating temperature in the process. By resorting to the FORTRAN language framework capabilities, commercial finite element method software ABAQUS has been steered in order to control unfavorable defects induced by localized and rapid heating and cooling and unstable volume of the melt pool. A thermodynamic consistent phase-field model is coupled to the transient thermal simulation to track the material history. A Lagrangian description for the spatial and time discretization is used. The goal is to present a closed-loop approach to track the melt pool morphology to a reference deposition volume profile which is established based on deep reinforcement learning (RL) architecture to avoid instabilities, defects and anomalies by controlling the laser power density adaptability. Therefore, standardization and process control are being enhanced, towards the consistency of the materials, equipment, process parameters, the final properties, and their characterization methods. On the production of AM parts, RL is able to assist practitioners in pre-manufacturing planning, and product quality assessment and controlling stage. The derived results from the implemented model enriched with a reinforcement learning architecture are compared with the non-enriched model to shed light on the capability of the approach.

Presenting Author: Joao Sousa FEUP

Presenting Author Biography: Jo&#227;o Sousa is a Ph.D. student and researcher at Faculty of Engineering of University of Porto (FEUP, Portugal) in the field of Mechanical Engineering. He has been involved in several research projects where laser-based processes such as laser cutting, welding and additive manufacturing are covered, in partnership with Institute of Science and Innovation in Mechanical and Industrial Engineering (INEGI), FEUP and Portuguese original equipment manufacturers (OEMs). <br/>His research activities have resulted in publications in the international journals, as well as attendance and presentation in conferences. His main research areas are Advanced Manufacturing Processes, Artificial Intelligence and IIoT. He has a particular interest in Control and Monitoring of laser-based processes with the application of novel developments in artificial intelligence and machine learning models in order to minimize waste, increase quality and reliability. <br/>ORCID: https://orcid.org/0000-0003- 3879-6908

Authors:

Joao Sousa FEUP
Roya Darabi INEGI
Ana Reis INEGI
Marco Parente FEUP
Luís Paulo Reis FEUP
Jose Cesar De Sa FEUP