Master of Science in Mechanical Engineering (MSME)
Controlling the quality of welds is an essential part of high quality fabrications and ensuring structural integrity of welded joints and welded structures. Industry is in an endless pursuit to find more cost effective methods of testing welds for defects and eliminating them entirely. This Study contributes to this goal by investigating a new system with the ability to monitor welding conditions during the forging process. Monitoring the condition of the weld, while the metals are joined together is faster and more cost- effective than conventional methods of post-weld defect detection. In this study, the feedback signals are first collected from friction stir (FS) welded specimens of Aluminum Alloy 2219-T87, next signals were segmented into discrete windows to imitate data availability in an on-line monitoring system, and finally features were extracted using discrete wavelet decomposition (DWD). Features were then selected using Ant Colony Optimization (ACO) and loaded into two machine learning algorithms for testing. The quality of the weld is considered to be a classification problem that is evaluated using the wrapper method with the Decision-tree (DT) and K-Nearest-Neighbor (KNN) algorithms. The test results show that feature extraction and selection on signals obtained in friction-stir-welding (FSW) can be a powerful method of defect detection when used with a machine learning algorithm for classification. Predictive power of models trained with selected features were able to yield 98.8487% classification accuracy. The work completed in this study provides a novel foundation for the creation of a new type of on-line sensing system for classification of the quality of FS- welded specimens, which is believed to be the first of its kind since limited work is available in the open-literature at this time.
Document Availability at the Time of Submission
Secure the entire work for patent and/or proprietary purposes for a period of one year. Student has submitted appropriate documentation which states: During this period the copyright owner also agrees not to exercise her/his ownership rights, including public use in works, without prior authorization from LSU. At the end of the one year period, either we or LSU may request an automatic extension for one additional year. At the end of the one year secure period (or its extension, if such is requested), the work will be released for access worldwide.
Roberts, James Devecmon, "Weld Quality Classification from Sensory Signatures in Friction-Stir-Welding (FSW) Using Discrete Wavelet Transform and Advanced Metaheuristic Techniques" (2016). LSU Master's Theses. 4559.
Available for download on Friday, February 21, 2025