Experimental investigation of vibrational signal in a fault induced Francis runner

Hydro turbines are prone to failure and the detection of fault in turbine is essential to ensure the reliability of power plant. This study investigates vibrational signals in a fault-induced Francis turbine using an experimental test setup to identify the trends that could be helpful in diagnosis of turbine faults. By analyzing the vibrational signal, the study aims to correlate the turbine's dynamic behavior. Faults in the turbine have been introduced by adding masses to the blades, and the experimental tests are conducted under two different conditions: dry and wet testing conditions for both normal and faulty turbine blades. The turbine's operating condition is determined with the help of pressure, flow, and RPM sensors. The turbine's speed is varied using a variable frequency drive. For the acquisition of vibration signals, the NI-LabVIEW system is employed along with a uniaxial vibration sensor located at the turbine bearing. The obtained vibration data are analyzed using the Fast Fourier Transform (FFT) algorithm and wavelet transform algorithm to identify frequency-domain characteristics. While studying and comparing the fundamental frequency of the turbine shaft, it is found that turbine faults can either increase or decrease the amplitude of the resonant peak frequency of the system, but the amplitude at other frequencies remains almost unaffected.