This paper presents a methodology for determining criticality of failure modes in mechanical system at conceptual design stage on the basis of functional and structural knowledge of the system, using statistical method ‘K-Clustering’. First of all component failure matrix (CF) is derived then its transpose is obtained. Then similarity matrix (SM) is obtained by pre-multiplying component failure matrix (CF) by its transpose. By adding each row corresponding to the given function and arranging the sums in ascending or descending order and selecting a range to determine critical failure modes. Prediction of critical failure modes during conceptual design stage using the component failure matrix (CF) and function component matrix (EC) matrices, the function failure mode matrix can be computed as EF = EC × CF.

This content is only available via PDF.
You do not currently have access to this content.