Synchronous Phase Averaging Method for Machinery Diagnostics

Abstract : This paper discusses a new diagnostic signature analysis algorithm called Synchronous Phase Averaging (SPA) which performs signal enhancement on quasi-periodic waveforms common in machinery and/or gearbox diagnosis. Time domain averaging (TDA) is an effective technique for extracting periodic signals from noisy or complex waveforms associated with machinery vibration. Due to the quasi-periodic nature of machinery shaft rotational speed, TDA is usually initiated with a trigger-pulse synchronous signal in order to synchronize the averaging process with the shaft rotation. This process is called Synchronous Time Averaging (STA), and is a common diagnostic tool in mechanical signal analysis. The Synchronous Phase Averaging method discussed in this paper can provide such signal enhancement performance but does not require a tachometer, or once-per-revolution input, as the reference signal. This is achieved through a novel synchronization process called the Phase Synchronized Enhancement Method (PSEM) which transforms a quasi-periodic synchronous (Sync or RPM) component within a vibration signal into a pure-tone discrete component. Within this newly transformed PSEM signal, since all the Sync-related components become discrete, standard Time Domain Averaging can be directly applied to the new signal to achieve signal enhancement. The elimination of tachometer monitoring reduces both instrumentation and data acquisition requirements, but most importantly, allows the effective implementation of non-intrusive sensors (accelerometer, acoustic) in applications such as gear box diagnostics, etc. Application examples using the SPA method for gearbox diagnostics will be demonstrated in this paper.