A Method to Decompose the Streamed Acoustic Emission Signals for Detecting Embedded Fatigue Crack Signals
AbstractThe data collection of Acoustic Emission (AE) method is typically based on threshold-dependent approach, where the AE system acquires data when the output of AE sensor is above the pre-defined threshold. However, this approach fails to detect flaws in noisy environment, as the signal level of noise may overcome the signal level of AE from flaws, and saturate the AE system. Time-dependent approach is based on streaming waveforms and extracting features at every pre-defined time interval. It is hypothesized that the relevant AE signals representing active flaws are embedded into the streamed signals. In this study, a decomposition method of the streamed AE signals to separate noise signal and crack signal is demonstrated. The AE signals representing fatigue crack growth in steel are obtained from the laboratory scale testing. The streamed AE signals in a noisy operational condition are obtained from the gearbox testing at the Naval Air Systems Command (NAVAIR) facility. The signal addition and decomposition is achieved to determine the minimum detectable signal to noise ratio that is embedded into the streamed AE signals. The developed decomposition approach is demonstrated on detecting burst signals embedded into the streamed signals recorded in the spline testing of the helicopter gearbox test rig located at the NAVAIR facility. View Full-Text
Share & Cite This Article
Zhang, L.; Ozevin, D.; He, D.; Hardman, W.; Timmons, A. A Method to Decompose the Streamed Acoustic Emission Signals for Detecting Embedded Fatigue Crack Signals. Appl. Sci. 2018, 8, 7.
Zhang L, Ozevin D, He D, Hardman W, Timmons A. A Method to Decompose the Streamed Acoustic Emission Signals for Detecting Embedded Fatigue Crack Signals. Applied Sciences. 2018; 8(1):7.Chicago/Turabian Style
Zhang, Lu; Ozevin, Didem; He, David; Hardman, William; Timmons, Alan. 2018. "A Method to Decompose the Streamed Acoustic Emission Signals for Detecting Embedded Fatigue Crack Signals." Appl. Sci. 8, no. 1: 7.