Multi-sensor solutions for predicting and monitoring tool wear

Recent research has shown that the use of several sensors is the best way to measure the amount of wear and tear during machining. The combination of different physical parameters seems to say more than a single signal.

In machining, the amount of tool wear is one of the most important parameters for controlling the process. Tools with excessive wear often lead to inferior quality and can fail suddenly. So far, the only way to measure wear has been by visual inspection. For a long time now academics and researchers have been trying to find connections between data that can be recorded during the process and the amount of wear. Attempts have been made to predict wear by using a single indicator, such as force, power or an acoustic measurement. In practice, however, these turned out not be accurate enough.

Correlations between measurements

Recent studies have shown the need to move to a multi-sensor-solution. That is, combining signals from different sensors. The combination of physical parameters says more than a single signal. In the European research report Cornet Dynatool, Sirris and KU Leuven looked for relationships between the amount of wear on the one hand and force measurements and acoustic signals on the other. The measurements showed correlations with a classic tool life curve.

The amount of tool wear rose quickly to start with, leading to a stable zone, followed by a rapid rise towards the end of the service life. Forces give a coarser indication of the wear curve, while acoustic emissions give more insight into the detailed aspects of the process. You can imagine that adding more data can lead to better control and prediction algorithms.

Sirris and KU Leuven intend to continue working on this. Are you interested in these themes? Then get in touch with us