Experimental validation of a robust diagnostic index for induction motors stator faults

Motor current signature analysis (MCSA) is used for non-invasive diagnosis of incipient or actual faults of grid supplied induction machine stator defects, being the negative sequence component of current space vector the diagnostic index most widely used. Supply unbalance and load influence of the machine however affect the negative component and they may cause the failure of the diagnostic procedure. Recently, deceleration test that may remove some drawbacks of MCSA has been proposed for rotor and stator fault diagnosis. In this paper it is shown that the residual voltage signature analysis (RVSA) obtained by the deceleration test provides, once removed the intrinsic manufacturing asymmetry, a diagnostic index directly related to the percentage of shorted turns. The main advantage of RVSA is that any influence caused by the supply unbalance and load is masked. Therefore a new diagnostic index is presented that is the residual voltage negative sequence component. The novel index performances have been proofed thanks to a simplified model of the faulted machine and experimentally validated by a special test motor that allows the introduction of different short circuits. This novel index can be used to confirm the MCSA-based diagnostic procedure developed by the same authors.