Masters Theses

Date of Award

5-2025

Degree Type

Thesis

Degree Name

Master of Science

Major

Mechanical Engineering

Major Professor

Bradley H Jared

Committee Members

Chad Duty, Bradley Jared, Tony Schmitz

Abstract

Wire-Arc Additive Manufacturing (WAAM) is an emerging metal manufacturing technology that utilizes welding processes to build large-scale components in significantly reduced timeframes at lower costs with less material waste. However, maintaining consistent part quality remains a challenge due to variations in weld bead size, shape, and overall deposition characteristics. The enclosed research explores the development of a low-cost melt pool monitoring solution using readily available and inexpensive hardware and opensource software to analyze the deposition process and identify observable weld characteristics from an optical camera feed. The system’s effectiveness was evaluated by testing various weld parameters, introducing controlled defects, and using different materials to assess the correlation between melt pool behavior and process conditions. Post-processing of the recorded frame data was conducted using a combination of traditional image analysis techniques and machine learning algorithms to extract data on weld features and establish relationships between weld parameters and visual melt pool characteristics. The final camera system utilized a neutral density filter and low exposure times to mitigate the arc flash from the deposition process. The melting of the welding wire into the melt pool, the solidification front, and the arcflash itself were visible in the video frames. These components were detected by a machine learning algorithm and the data about their size and location within the video feed were used to compare the differences in the feed when weld parameters were changed.

Comments

1st revision after approval by committee.

Download
Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Included in

Manufacturing Commons

Share

COinS