A new Time-domain Feature called the Really Keen Metric (RKM) for Acoustic Emission Transient Signal Processing

摘要

A series of dynamic tests were performed where certain equipment was monitored with AcousticEmission (AE). Resonant AE sensors were used and coupled to MISTRAS AE instrumentation.The recorded data comprises data files with 8,000 to 10,000 Hits recorded over a very short periodof time <15 seconds). Of interest during these tests were AE signals that are truly indicative oftransient signals that have a clear beginning and end and an exponential decay after the peakamplitude is observed. Unfortunately, the data sets were contaminated with considerableamounts of extraneous emission that could only be filtered by manually examining the waveformcollected for each Hit. The use of correlation plots to separate these AE signals was helpful to acertain extent but through the process of manual data filtering, it became clear that a well-definedtime-domain feature could be very efficient in separating the transient signals from extraneousemissions. This paper presents a concept for a new time-domain feature that is capable of betterdefining transient-like AE signals as defined above. This feature is called the Really Keen Metric(RKM) and will be presented with actual examples recorded during the above testing.