Ali Kahirdeh - Mechanical Engineering Department Iran university of Science and Technology
Mir Saeed Safizadeh - Mechanical Engineering Department Iran university of Science and Technology

In this paper, a method for detection and classification of cyclic bearing faults is introduced. The method uses hierarchical clustering to classify severity of flaws. Also it is examined in classification of flaw’s location. Features are extracted from acceleration signals in three different domains which are time, frequency and timefrequency representations of the signal. Encouraging results indicate the potential of the demonstrated method in condition monitoring and fault diagnosis of rolling element bearings.

Hierarchical clustering, Rolling element bearing condition monitoring, Severity classification, Feature extraction