Gas metal arc welding process monitoring with acoustic sensing

This project aims to optimise and, where possible, automate the gas metal arc welding process in order to deliver consistent high-quality welds.

Artificial intelligence

Construction & asset management

Data science

Industry 4.0

Infrastructure, utilities & transport

Robotics and automation

Wireless communications & Internet of Things

}

Duration

3 years

Academic team

Professor Xiaojun Qiu

Professor Ian Burnett

Dr Sipei Zhao

Lab

Engagement model

Funding

ARC Linkage

Sound Intuition

Future applications

Range of weld processes

Development of remote welding processes

Acoustic sensing for quality control, productivity & training

Challenge

Auditory cues are critical for expert welders to adjust the weld process and maintain quality. The challenge lies in using these sound emissions to develop an automated process that enables weld quality to be controlled remotely. As the mechanisms underpinning sound generation and control are not well understood, this project will develop fundamental knowledge of sound emissions from welding process.

 

Solution

The UTS Tech Lab team is developing a model that correlates the acoustic signal with other weld parameters for gas metal arc welding. By investigating the physical mechanisms of sound generation, the key acoustic features can be identified. This knowledge supports the development of a database of multimodal weld recordings, and a prototype monitoring system based on the key acoustic features identified.