Ultrasonic phased array imaging and data post processing with applications to resistance spot welding

Resistance spot welding is a widely used metal joining method in the automotive industry due to its low cost, ease of automation and high throughput. Resistance spot welding is a process that joins two or more metal sheets through the application of force and high current in a localized region, or spot. With the average car containing thousands of such spot welds, some of which compose structural and safety components, the evaluation of the quality metrics related to this process is of key interest to the automotive sector. With this in mind, the evaluation of spot weld parameters including, size, penetration depth and yield strength have been evaluated using both destructive and non-destructive techniques. One of the more recent advancements involves the use of real time monitoring of the resistance spot weld using a single element transducer placed inside the welding electrode. This transducer allows for a determination of the penetration depth based on ultrasonic reflection at a single point within the weld stack. From this, other properties are estimated using the geometry and materials that are being welded. Such an approach works well in many cases but falls short as the industry shifts to ultra-high strength steels and aluminum materials, which are more prone to off-centre defects such as cracks and porosity. This work aims to overcome many of these challenges by replacing the current probe with a linear phased array version. Such a change allows for the monitoring of not only the weld penetration depth but also the lateral size of the weld alongside the detection of defects such as pores. This system would offer a significant advantage over current methods from a quality assessment standpoint. In order to realize these benefits, this work aims to bring such a system into a production-ready capacity. This includes designing a phased array system that is low cost and easy to maintain in a production environment. In addition to this, the methods of processing ultrasound data acquired with such a system require changes to existing methods, as the break in symmetry results in complications from an analysis standpoint. This work covers the methods and approaches used to reach these goals and bring the system closer to a production ready deployment.

Name thesis:degree_name

M.A.Sc.

Level thesis:degree_level

Masters

Discipline thesis:degree_discipline

Electrical and Computer Engineering

Grantor

University of Windsor

Year dc:date.issued

2024

Author dc:creator

• Draskovic, Milos

Advisor dc:contributor.advisor

• Roman Maev

Contributors dc:contributor

• scholarship@uwindsor.ca

Licence

• http://creativecommons.org/licenses/by-nc-nd/4.0/

Language dc:language.iso

en_CA

Handle dc:identifier.uri

https://hdl.handle.net/20.500.14776/8351

OAI identifier oai:identifier

oai:uwindsor.scholaris.ca:20.500.14776/8351