Real-Time Defect Detection of Die Cast Rotor in EV Induction Motor Based on Circular Flux Sensing Coils

Abstract

The complex casting process of die-cast rotors are easily leads to porosity or other defects in production. Accurate inspection for die-cast rotors is the key to guiding the improvement of the casting process as well as increasing the yield. However, many sensors cannot accurately evaluate the severity of porosity or broken bar. This study develops a novel sensor to inspect and analyze the rotor quality according to the electromagnetic model of squirrel cage induction motor (SCIM). The sensor consists of the electromagnetic flux sensing coils directly from a SCIM stator. By injecting a DC voltage at phases A and B of the sensor and dragging the rotor around the sensor, the induced voltage is generated in phase C. A quantitative fault indicator (QFI) is constructed on the basis of the signal analysis algorithm to analyze the acquired signal, and the trend of the QFI with respect to fault severity is studied. The different faults are made on the experiment rotors to simulate the defects that are likely to occur in rotor production. By building the experimental platform to detect different fault rotors, the effectiveness of the sensor is validated.