Development of a real-time failure detection system for stamping die.

Modern industrial products require high precision and material strength. These have a greater effect on a die and so there is accelerating die failure during mass production. Die failure damages product accuracy so it must be detected in real time. This study develops a system that detects die failure directly from the servo press machine itself. The servo motor torque data and punch time data and press slide position data are used to detect the failure status of the die and punch. SUS304 and SAE1065 steel sheets were used as test pieces. The failure mode and the effect of failure on feedback data are studied to determine the feasibility of the detection mechanism. The results show that the punch time and the torque vary with the failure mode and failure grades. For SUS304 sheets, after 1000 cycles of artificial wear, the maximum torque is 9% greater than that for a punch that is not worn. For SAE1065 sheets, the torque is 4.4% greater for the same degree of punch failure. The punch time respectively increases by 26 ms and 22 ms for SUS304 and SAE1065. Using these results, a LabVIEW DAQ module with a servo stamping machine is used to construct an online wear detection system. The LabVIEW program is tested. The test results on the LabVIEW front panel agree well with previous results. This study shows that punch/die failure can be detected accurately using this real-time failure detection system. [ABSTRACT FROM AUTHOR]