Online tool condition monitoring of galling wear in sheet metal forming using acoustic emissions

Abstract

Sheet metal forming is a common manufacturing technique widely used in industry for high volume production. Galling wear involves the adhesive transfer of material from the sheet to tool and is a well-known problem in sheet metal forming. Current methods to evaluate galling require stopping production to assess the integrity of the tool surfaces. Acoustic emission is showing promise as a method for real time, tool condition monitoring by measuring elastic waveforms in forming tools during manufacture.
This project aims to further the use of AE as a condition monitoring system. The major focus of this research is the quantifiable early detection of the galling phenomena through extracted AE phenomena. Scratch tests at low loads and low sliding speeds were used to enable an in-depth and high-precision study of the initiation of galling wear damage mechanism and the resulting AE. A range of techniques were used to evaluate the sensitivity of various AE features to the initiation of galling wear and the optimal features were found. Subsequently these features were investigated on a full-scale, semi-industrial stamping set up.
The novelty in this work comes from the emphasis on the initiation of galling wear, the cross examination of an industrially relevant set-up and a focus on feature selection for predictive maintenance. When test conditions were constructed to allow for a focussed study on the initiation of galling wear, the observed damage mechanisms were different compared to what was expected from the literature and was later characterized in a novel way. The subsequent analysis of the AE employed t-tests, cluster, and discriminant analysis, as well as linear regression and multinomial logistic regression modelling to investigate the AE signal for novel features in the sheet metal forming field. These methods and features have been derived from other applications of the AE technology and may be used for detecting this onset of galling wear, by investigating predictive maintenance models.
The knowledge gained from this research provides key insights towards the development of online wear condition monitoring systems for production sheet metal forming. This has the potential to provide significant productivity, economic and environmental improvements for manufacturing industry, as well as adding knowledge to the Industry 4.0 sector by helping to establish the use of AE as a predictive wear measurement and allowing for the establishment of preventative maintenance strategies based on this technology

Date of Award	25 Mar 2024
Original language	English
Awarding Institution	Coventry University Deakin University
Supervisor	James Griffin (Supervisor), Ping Lu (Supervisor), James Berriman (Supervisor), Michael Pereira (Supervisor) & Bernard Rolfe (Supervisor)