Non-destructive testing of machines to reduce maintenance time and cost

This paper presents time domain methods for early fault detection and severity analysis in non-destructive testing (NDT) of electromechanical rotating machines. An adaptive band pass filtering technique has been used to detect faulty pulses present in the signal emanating from a gearbox. A pulse position analysis method is used to calculate rotational speed of faulty components in the gearbox. Faulty pulse shape is further analyzed visually to estimate fault severity. Visual inspection of the gearbox is also performed to verify the results obtained through proposed methods. The proposed method is a useful input to manage maintenance personnel in an effective way and reduce maintenance costs.