Statistical tool wear evaluation for cutting processes

The storage of large volumes of high speed data in many manufacturing processes can be a problem. This paper analyses an experimental methodology for sensor data information obtained from several sensors. The use of the most accurate and reliable sensors is in order to measure and control some important variables and because all the information obtained by these sensors may not be necessary and some of them may be redundant. Sensor data interpretation can be done by several statistical techniques as an alternative. A statistical evaluation of a multisensor system in cutting processes to obtain variables in critical tool wear stages is made simultaneously. Some parameters are monitored (force, temperature, electric current and acoustic emission), and tool wear estimated by some computed characteristic parameters.