Imaging Acoustic Signature Analyses for Plant Turbo-machinery

The proposed acoustic evaluation technology is a major improvement over past art because, in addition to monitoring naturally occurring operational noise, it uses artificial stimulation to determine component "ring-out" frequency response. This independently establishes excitation forces separate from key component (e.g. bladed disk) natural frequencies, shifts in either of which imply different kinds of mechanical degradation. Artificial stimulation is produced by a speaker near the machine casing. Phased array microphones will allow acoustic intensity vector phase-correlated pink-noise frequency response data to be gathered over key surfaces or volumes. The acoustic intensity space-time-frequency maps would be interpreted by AI software to determine the nature of developing problems, and the remaining life of specific components, as predicted by fatigue, fracture mechanics, and tribological analytical methods. Phase I will build and test a proof-of-principle prototype using COTS hardware and software. Phase II will develop a dedicated acoustic-based diagnosis/ prognosis system which will be proof-tested in actual plant turbomachinery installations.