On the derivation of the pre-lockup feature based condition monitoring method for automatic transmission clutches

• The pre-lockup features useful for clutch condition monitoring are theoretically derived.
• The relationships of the features with the mean coefficient of friction (COF) of a clutch are revealed.
• The features sensitivity with respect to the operational variables is addressed.
• The theoretical predictions are experimentally verified.

This paper discusses how a qualitative understanding on the physics of failure can lead to a theoretical derivation of effective features that are useful for condition monitoring of wet friction clutches. The physical relationships between the features and the mean coefficient of friction (COF) which can be seen as the representation of the degradation level of a wet friction clutch are theoretically derived. In order to assess the accuracy of the theoretical relationships, Pearson's correlation coefficient is applied to experimental data obtained from accelerated life tests on some commercial paper-based wet friction clutches using a fully instrumented SAE#2 setup. The analyses on the experimental data reveal that the theoretical predictions are plausible.