Statistical analysis of electrical breakdown behaviour in rolling bearings under varying operating conditions

Edward Klemm

Summary

In the context of energy transition and e-mobility, drive systems with modern converters are subject to bearing damage caused by electric discharges. This results in a current flow in tribological contact, the damage potential and frequency of which depend on various boundary conditions. The breakdown behaviour of the tribological system, consisting of the bearing and lubricant, is of particular importance in this regard, as a breakdown within an impedance measurement for example can falsify the results and thus has to be avoided.

In a complex system such as a bearing, the breakdown voltage cannot be described with one value, but rather a probability function. In this context measuring voltage has to stay a certain breakdown probability.

This presentation statistically describes the breakdown behaviour in the tribological contact of a thrust bearing. A methodology for determining the probability distribution of the first breakdown and the dependence of this probability on the operating conditions of the bearing, the kind of lubricant as well as its condition and the properties applied voltage signal is described. Hereby the voltage is constantly increased with 10 V/s up to 1000 V/s. The statistical reliability of these measurements was ensured by taking up to 100 measurements per operating point with high temporal resolution. In addition, the breakdown behaviour in voltages above the first breakdown and the energy transferred as a function of the boundary conditions are presented. Furthermore the transferability of the acquired probability distribution to other voltage signals, such as a sinus voltage is shown.