Technological simulation of the resulting bead geometry in the WAAM process using a machine learning model

2025/09/15 by Denkena, B., Wichmann, M., Böß, V. +1
#600 #Additive manufacturing #Konferenzschrift #Planning #Simulation #Welding

paper · doi:10.15488/19630

Abstract

In contrast to most subtractive processes where a specific tool geometry is available, process planning in the CAD/CAM chain of additive manufacturing is not as accurate unless the deposited geometry is known. Therefore, a dexel-based process simulation for wire and arc additive manufacturing is implemented to predict the resulting geometry of the deposited material depending on the process parameters. In order to make accurate predictions and consider the effects of the process parameters on the geometry, a multi-stage model is developed for three different materials. The results of this prediction pipeline show an R of 0.82 for the width and 0.76 for the height. Finally, the simulation method is evaluated in terms of computational effort, and the ratio of simulation time to process time is found to be reasonable for simulation-based process planning.

Related

• A comprehensive review and future perspectives of simulation approaches in wire arc additive manufacturing (WAAM)
• Enhancing Experimental Efficiency in Materials Design: A Comparative Study of Taguchi and Machine Learning Methods
• Physics-informed machine learning surrogate for scalable simulation of thermal histories during wire-arc directed energy deposition
• Comprehensive Quality Investigations of Wire-feed Additive Manufacturing by Learning of Experimental Data
• A hybrid machine learning framework for clad characteristics prediction in metal additive manufacturing