Detecting Process Anomalies in the GMAW Process by Acoustic Sensing with a Convolutional Neural Network (CNN) for Classification

Abstract

Today, the quality of welded seams is often examined off-line with either destructive or non-destructive testing. These test procedures are time-consuming and therefore costly. This is especially true if the welds are not welded accurately due to process anomalies. In manual welding, experienced welders are able to detect process anomalies by listening to the sound of the welding process. In this paper, an approach to transfer the “hearing” of an experienced welder into an automated testing process is presented. An acoustic measuring device for recording audible sound is installed for this purpose on a fully automated welding fixture. The processing of the sound information by means of machine learning methods enables in-line process control. Existing research results until now show that the arc is the main sound source. However, both the outflow of the shielding gas and the wire feed emit sound information. Other investigations describe welding irregularities by evaluating and assessing existing sound recordings. Descriptive analysis was performed to find a connection between certain sound patterns and welding irregularities. Recent contributions have used machine learning to identify the degree of welding penetration. The basic assumption of the presented investigations is that process anomalies are the cause of welding irregularities. The focus was on detecting deviating shielding gas flow rates based on audio recordings, processed by a convolutional neural network (CNN). After adjusting the hyperparameters of the CNN it was capable of distinguishing between different flow rates of shielding gas.

Influence of Process Parameters in Gas-Metal Arc Welding (GMAW) of Carbon Steels

“…To develop mechanisms for overcoming the disturbances and uncertainties observed during GMAW process, several researchers have developed mathematical (statistical and numerical) models that describe the physical GMAW process by employing fundamental principles of science, statistical and/or experimental techniques [6,12,13,32,40,41], and artificial intelligence models [14,15,59]. The major aim of these models is to provide a framework for describing and understanding the behavior of the GMAW system in order to provide an effective feedback control [6].…”

“…Studies on GMAW welding have developed a considerable understanding on how to address the challenges experienced in the GMAW process in many applications. Some studies have used statistical quality control tools, such as statistical models [12,13], numerical models [14][15][16][17][18], and artificial intelligent models [19,20], to predict the desired output parameters by refining the input process parameters [19,20]. This has led to process advancements in terms of equipment and process control for improvement in the weld quality and productivity, even though some challenges still exist [11].…”

D7.2 - Arc Welding Process Monitoring Using Neural Networks and Audio Signal Analysis

“…Additive manufacturing techniques such as arc welding gain importance in the producing industry and quality monitoring plays a vital role in the welding process to ensure the quality of the outcome. Fluctuations in the process parameters such as speed, power, shielding gas rate, and oil contamination can lead to pores in the arc welding seams and thus to poor quality [1]. To generate an appropriate dataset for our analysis, additive-manufactured Aluminium walls with 50 layers were produced.…”