Modeling Robotic Wire Arc Additive Manufacturing Process Using Machine Learning

WAAM is a promising additive manufacturing process that makes use of current technologies and materials that are widespread and much more easily accessible compared to other metal AM processes. However, a significant challenge with current WAAM processes is the high surface roughness and variance of the height and width of the produced parts. These parts often require additional machining to achieve the desired dimensions and tolerances. By applying Machine Learning, this proposal aims to predict the surface roughness and dimensions of WAAM-produced parts by adjusting the welding parameters (voltage, current, travel speed, feed speed, amplitude, and wavelength), potentially minimizing or eliminating the need for post-process machining. These modeled parameters will then be used to generate the toolpath to fill any shape, regular or irregular. Current literature has focused on single passes of the torch, much like conventional 3D printers. This proposal plans to improve the process by using a weaving toolpath. For characterization, a Cognex DS 1300R laser scanner will generate a 3D point cloud of the welds produced by the WAAM process with a resolution in the micrometer range.

Name thesis:degree_name

Master of Science

Level thesis:degree_level

Masters

Discipline thesis:degree_discipline

Mechanical and Manufacturing Engineering

Grantor

Texas State University

Year dc:date.issued

2024

Author dc:creator

• Gonzalez, Aaron

Advisor dc:contributor.advisor

• Chen, Heping

Committee members dc:contributor.committeemember

• Valles Molina, Damian
• Asiabanpour, Bahram

• machine learning
• WAAM
• additive manufacturing

Language dc:language.iso

en

Handle dc:identifier.uri

https://hdl.handle.net/10877/20434

OAI identifier oai:identifier

oai:digital.library.txst.edu:10877/20434