Vibroacoustic method of diagnosing rotating machines operating under variable load

Diagnosing machines operating under variable load using vibration signals requires the application of advanced computational methods. The variable load of rotating machinery affects their rotational speed and the amplitude of generated vibration signals. Synchronous methods are commonly used to negate the effect of speed variations, such as Synchronous Sampling, Order Analysis or Time-Synchronous Averaging. However, the problem remains with the impact of load on the amplitude values of diagnostic signals. This paper proposes new order spectra insensitive to varying load. The components of the spectra are determined from the dependence of order amplitudes as a load function. The proposed spectra are a reference to a reference spectrum determined for the correct operating condition of the machine. In order to test the effectiveness of the new method, an experiment was carried out on a laboratory bench to diagnose a planetary gearbox. The method gave unambiguous results in identifying the alignment of the driveline under test.