Tool condition monitoring of diamond-coated burrs with acoustic emission

Jessel, Thomas 2025. Tool condition monitoring of diamond-coated burrs with acoustic emission. PhD Thesis, Cardiff University. Item availability restricted.

Preview	PDF (Thesis) - Accepted Post-Print Version Download (48MB) | Preview
	PDF (Cardiff University Electronic Publication Form) - Supplemental Material Restricted to Repository staff only Download (338kB)

Abstract

Within manufacturing there is a growing need for autonomous Tool Condition Mon itoring (TCM) systems, with the ability to predict tool wear and failure. This need is increased, when using specialised tools such as Diamond-Coated Burrs (DCBs), in which the random nature of the tool and inconsistent manufacturing methods, create large vari ance in tool life. This unpredictable nature leads to a large fraction of a DCB’s life being underutilised, due to premature replacement, a significant concern considering the mass-manufacturing scale of DCB usage. To combat this widespread wastage, this thesis presents Acoustic Emission (AE) as an on-machine and indirect sensing technology, which through a range of processing and Machine Learning (ML) methods can monitor the wear state of DCBs. A developed DCB wear test methodology, allowed the systematic wearing of a 1.3mm #1000 DCB, whilst acquiring continuous AE and direct tool wear measure ments. Over this thesis, 22 wear tests were conducted with constant machining paramet ers. Therefore, enabling the identification of DCB wear mechanisms and phases, through monitoring DCB radial wear, and in turn quantifying the inherent variability of DCBs. Additionally, investigation into the effect of varying initial DCB runout levels, between 1–77µm, identified a clear correlation between increased runout and variability in total life. These wear tests also allowed extracted AE features, from both the time and fre quency domain, to be validated as useful indicators for both a DCB’s Remaining Useful Life (RUL) and its grinding effectiveness. Frequency domain partial powers within the AE Power Spectral Density (PSD) were seen to align well with overall DCB wear, whilst AERMS and AEkurt indicated a DCB’s contact and runout. With this knowledge, three TCM approaches were developed, each framing the prob lem differently. A threshold-based criterion method, using a transformation of selected AE features, to indicate a DCB’s transition into its final phase of wear was developed. Resulting in a computationally inexpensive and indirect warning system to indicate the decline of a DCB. Additionally, a regression and classification Artificial Neural Network (ANN), trained on the obtained TCM dataset, are able to predict a DCB’s mean ra dius with a Root Mean Square Error (RMSE) = 1.407µm, and a balanced accuracy of 0.938 when predicting the geometric tolerance of the ground workpiece. Both of which, providing valuable information to the operator about the state of grinding; enabling tool compensation or changes to occur with no measurement downtime. The completed work therefore, demonstrates the potential for developing an indirect TCM system to simultan eously reduce tool wastage, improve cycle times, and prevent unexpected tool failure.

Item Type:	Thesis (PhD)
Date Type:	Completion
Status:	Unpublished
Schools:	Schools > Engineering
Uncontrolled Keywords:	1. Tool Condition Monitoring 2. Acoustic Emission 3. Diamond Coated Burrs 4. Grinding 5. Machine Learning 6. Wear
Date of First Compliant Deposit:	20 March 2026
Last Modified:	23 Mar 2026 09:18
URI:	https://orca.cardiff.ac.uk/id/eprint/185787

Actions (repository staff only)

Edit Item