Multi-parameter tool condition monitoring

Aim of project
Currently, both experimental and commercially available tool condition monitoring (TCM) systems, are based on the measurements of physical phenomena that are correlated with tool wear and can be exploited as tool wear symptoms. Cutting force components and acoustic emission (AE) are most often used. The possibility of reliable tool wear evaluation based on one signal feature (SF) has been questioned because the measured feature depends not only on tool wear but also on a variety of other process parameters and random disturbances. Attempts at rectifying these shortcomings have focused on pursuing a multi-sensor fusion strategy. In the approach applied at WUT, a TCM system consists of two modules. The fi rst module estimates tool wear from all available SFs separately. The results are then integrated into the fi nal system response in the second module.

The main purpose of the project is: Quest for new signal features useful for tool condition monitoring, based on advanced methods of time series analysis (e.g. wavelet transform);

Development of algorithms of data acquisition, calculation and automatic selection of SFs and selection of optimal monitoring strategy;

Development of general strategy of catastrophic tool failure detection suitable to diff erent machining methods like turning, milling, drilling etc.;

Development of general strategy of tool wear monitoring, based on many automatically selected signal features.

Expected results
The result of the project will be several virtual tool condition monitoring systems as LabVIEW programs executed on PCs equipped with DAQ devices, so prepared to be used in real time.