Condition Monitoring and Fault Detection for Electrical Machines Using Advanced Sensing Techniques Based on Fibre Bragg Gratings

Abstract

Emerging techniques are being researched to expand the suite of condition monitoring solutions available for electric machines to adapt to a world of net zero carbon emissions. This research investigates the use of fibre bragg gratings (FBG) for condition monitoring and fault detection in three 2.2kW induction motors (IMs) using stray flux in a non-invasive manner. Optical fibre is immune to electromagnetic interference (EMI) which is an advantage but limits its direct use for magnetic field sensing. A magnetostrictive transducer, terfenol-D was bonded to FBG to form a composite sensor - FBG-T. The FBG-T was inserted into an acrylic tube - which is unaffected by magnetic field - and then positioned both axially and transversely relative to the machine’s rotor shaft at the drive end (DE). The transverse position showed better repeatability and sensitivity over different operating frequencies. Temperature and magnetic flux calibrations of the FBG-T sensor gave sensitivities of 20.77 picometre per degree Celsius (pm/°C) and 19.38 picometre per micro-tesla (pm/μT) respectively. Various investigations were carried out at different operating frequencies and under three motor conditions viz: healthy, broken rotor and inter-turn short circuit conditions. Experimental results confirm that the FBG-T sensor reliably distinguished each of the three machine conditions using different orders of magnitudes of braggshifts. The FBG-T sensor accurately detected faults with the short circuit condition reaching braggshifts of hundreds of pm. Healthy and broken rotor conditions reached braggshifts in the low–to-mid-hundred and high-hundred pm range respectively. Fast Fourier Transform (FFT) analysis performed on the measured stray flux showed that not only its amplitude but also the harmonic component of its spectrum, affected the magnetostrictive behaviour of the magnetic dipoles of the terfenol-D transducer. This effect was translated into strain on the FBG. The investigation proved that FBG technology can reliably and accurately monitor the condition of the motors as well as detect faults in a non-intrusive manner.