Metal forming defect detection method based on recurrence quantification analysis of time-series load signal measured by real-time monitoring system with bolt-type piezoelectric sensor

Abstract

In recent years, research related to smart manufacturing processes, which aim to advance production technology through productivity improvement and manufacturing technology innovation, has been actively pursued. In the field of sheet metal press forming, intelligent process management is required to improve productivity and reduce costs. In this field, a method of indirectly monitoring signals generated in the forming process without the need to modify existing tools is advantageous, as it can be easily applied to the mass-production process. In this study, a monitoring system for a laboratory-scale cup drawing process was established that uses a bolt-type piezoelectric sensor. First, a numerical analysis was conducted to select the location where the sensor was to be installed in the tool, and the feasibility of load monitoring was confirmed. Then, the forming load of the cup drawing process using a servo press was measured according to process variables such as holding force, specimen size, and initial specimen position. Forming defects such as wrinkles and fractures occurred according to process variables, and it was confirmed that the measured load had a specific pattern. To quantitatively analyze the time series data of the measured load, recurrence quantification analysis was used, and a system for real-time monitoring of formability was developed based on the structural shapes formed by the resulting recurrence plots.