An initial investigation into the performance of acoustic condition monitoring in the detection of structural faults in turbine blades has been carried out. The focus is to design a non-contact condition monitoring method which might allow the detection of incipient faults in the turbine blades therefore preventing major breakdown and potentially reducing maintenance downtime. Such systems, may be deployed remotely from the turbine and provide a non-fixed and thus more economic solution for operation and maintenance of wind turbines.
An initial investigation into the performance of an acoustic monitoring system has been carried out both in laboratory as in-situ. A number of signal analysis methods are evaluated against structural faults in turbine blades. A lab measurement was obtained using a 3 blade small fan of constant speed. The in-situ measurements were carried outdoor on a 5 blade micro wind turbine with speed dependent on wind speed.
Faults were planted on a single blade for each test system, in the form of added weights and delaminations near the tip of the blade. Acoustic data acquisition was obtained by placing a single microphone close to the blades. Time synchronous analysis (TSA), ensemble empirical mode decomposition (EEMD) and spectrum of individual intrinsic mode functions were investigated as feature extraction and fault finding methods.
Results show that the effect of faults is to shift energy in the IMFs, particularly around the har-monic frequencies of the system. Faults can also be detected in terms of increased harmonic energy, observable through magnitude spectra determined from individual valid IMFs.
[1]
Harvey H. Hubbard,et al.
Aeroacoustics of large wind turbines
,
1991
.
[2]
Babak Hossein Khalaj,et al.
Grey Prediction Based Handoff Algorithm
,
2007
.
[3]
Hong Fan,et al.
Rotating machine fault diagnosis using empirical mode decomposition
,
2008
.
[4]
Q. Du,et al.
Application of the EMD method in the vibration analysis of ball bearings
,
2007
.
[5]
Norden E. Huang,et al.
Ensemble Empirical Mode Decomposition: a Noise-Assisted Data Analysis Method
,
2009,
Adv. Data Sci. Adapt. Anal..
[6]
N. Huang,et al.
The empirical mode decomposition and the Hilbert spectrum for nonlinear and non-stationary time series analysis
,
1998,
Proceedings of the Royal Society of London. Series A: Mathematical, Physical and Engineering Sciences.
[7]
Jonathan A. Hargreaves,et al.
Perception of noise from large wind turbines(EFP-06 Project)
,
2010
.
[8]
Bruno Fazenda,et al.
Gearbox fault diagnosis under different operating conditions based on time synchronous average and ensemble empirical mode decomposition
,
2009,
2009 ICCAS-SICE.