Thorough Understanding and Experimental Validation of Current Sideband Components in Induction Machines Rotor Monitoring

The MCSA diagnostic procedures for the rotordiagnosis of induction machines are mostly based on thesideband lines near the supply frequency line that appear inthe input current spectrum. The left component (1-2s)f is theeffect of rotor backward rotating field caused by the cagerotor asymmetry and it varies with the asymmetry degree andwith the rotor current, i.e. with the machine load.The right component (1+2s)f is caused by the speed ripplecreated by the left component, being the combined machineloadinertia the main parameter that spreads the asymmetryeffect into the two sideband components.Theoretically it can be shown that the sum of the twocomponents is not affected by the speed ripple effect andtherefore this sum may be used as an effective diagnosticindex to state rotor conditions.The above results can be experimentally validated only once asuitable test set-up is realized that allows changing the inertiavalue. To this aim a test bed was designed so that the virtualinertia of the mechanical system can be chosen by the user.This allows to prove the theoretical claims and to obtain athorough understanding about the effect of inertia onsideband components, improving the performance of rotorcondition monitoring techniques.

The MCSA diagnostic procedures for the rotor diagnosis of induction machines are mostly based on the sideband lines near the supply frequency line that appear in the input current spectrum. The left component (1-2s)f is the effect of rotor backward rotating field caused by the cage rotor asymmetry and it varies with the asymmetry degree and with the rotor current, i.e. with the machine load. The right component (1+2s)f is caused by the speed ripple created by the left component, being the combined machine-load inertia the main parameter that spreads the asymmetry effect into the two sideband components. Theoretically it can be shown that the sum of the two components is not affected by the speed ripple effect and therefore this sum may be used as an effective diagnostic index to state rotor conditions. The above results can be experimentally validated only once a suitable test set-up is realized that allows changing the inertia value. To this aim a test bed was designed so that the virtual inertia of the mechanical system can be chosen by the user. This allows to prove the theoretical claims and to obtain a thorough understanding about the effect of inertia on sideband components, improving the performance of rotor condition monitoring techniques. ©2006 IEEE.