METHOD AND DEVICE FOR DETECTING TOOL ABNORMALITY IN MACHINE TOOLS

Method and device for detecting tool abnormality in machine tools in which a threshold value is determined based on the condition of a tool in machining and tool abnormality is detected by comparing data successively obtained from the tool during machining with the threshold value. The threshold value is determined by computing the tool condition data in machining for a period when an externally provided switch is on and stored in a memory of the device when the switch turns off. The application of this invention to numerical controlled machine tools is also disclosed in which a plurality of threshold values are determined according to machining conditions in each machining process and tool abnormality is detected by comparing tool condition data during machining with one of the threshold values corresponding to the condition of the machining.