Time domain method for gear fault diagnosis

In the process of condition monitoring and fault diagnosis, we often use the vibration time domain signal directly to analyze and give the results, which is the simplest and most direct method, especially when the signal obviously contains simple harmonic component, periodic component or instantaneous pulse component. Of course, this method requires the analyst to have rich practical experience. The vibration time-domain waveform is a wave curve of time history. According to different types of sensors, the amplitude of the curve can represent displacement, velocity or acceleration. In waveform analysis, the following characteristic quantities, also known as indicator, are mainly used:

(1) The vibration amplitude includes peak value, effective value (RMS) and average amplitude, and the peak value is divided into zero peak value and peak peak value.

(2) Vibration period and frequency, different fault sources usually produce different frequencies of mechanical vibration, so frequency analysis plays a very important role in fault diagnosis.

(3) Phase, in practical application, phase is mainly used to compare the relationship between different vibration movements, or to determine the vibration status of one component relative to another. Generally, the vibration produced by different sources has different phases.

(4) In order to effectively describe complex vibration, other indicators are often used in practical applications, such as skewness and kurtosis. Sometimes, dimensionless indicators are needed to complete diagnosis or trend analysis, such as: peak state factor, waveform factor, pulse factor, peak factor, margin factor and other dimensionless indicators. Their diagnosis ability is in the order of peak state factor – margin factor – pulse factor – peak factor – waveform factor.