A strain sensing bolt for lathe cutting tool monitoring

The status of cutting tools is critical in metal cutting processes, as tool wear or breakage can result in increased production costs as well as unscheduled downtime. This study proposes a novel lathe cutting tool monitoring scheme using a specially designed bolt that can not only be applied for fasten the tool holder, but also capable of perceiving the strain of it. In conjunction with a front-end amplifier, a RC filter, and an analog-to-digital converter (ADC), the strain perceived by the bolt can be transformed into digital signals. Afterwards features relating to signal morphology and statistical characteristics are extracted, and the fuzzy neural network (FNN) classifier is employed for tool breakage detection. Using bearing outer ring machining as a practical case, a hit rate of 99.41% can be achieved, which demonstrates the validity of the proposed strain sensing bolt and analyzing method.