Windchill Risk and Reliability Desktop Help > Windchill FRACAS > FRACAS Theory > Key FRACAS Terms > Failure Classification Definitions
Failure Classification Definitions
According to the Failure Reporting, Analysis and Corrective Action System (FRACAS) Application Guidelines, the classification of failure events helps to determine:
• The level of analysis that should be performed on each failure.
• The corrective action that should be taken (and when) to eliminate the failure or at least to minimize its recurrence.
The following table lists the generic categories that this document uses to classify failures. Although the FRACAS module uses these same categories by default, you can customize them to use whatever terminology you want.
Classification
Definitions
Failure, Relevant
A failure that has been verified and can be expected to occur in normal operational use. Relevancy indicates whether a specific failure should “count” in the calculation of reliability. (See Failure, Chargeable below.)
Failure, Non-Relevant
A failure that has been verified as having been caused by a condition not defined for normal operational use.
Failure, Chargeable
A relevant primary failure under test, and any secondary failures resulting from a single failure incident. This definition of failure is typically limited to formal, contractually required reliability tests (performed in-house or in the field).
Failure, Non-Chargeable
A non-relevant failure, or a relevant failure caused by a previously agreed to set of conditions that eliminates the assignment of failure responsibility to a specific functional group. This definition of failure is typically limited to formal, contractually required reliability tests (performed in-house or in the field).
Failure, Pattern
The occurrence of two or more failures of the same part or function in identical or equivalent applications, where the failures are caused by the same basic failure mechanism, and the failures occur at a rate inconsistent with the expected failure rate.
Failure, Multiple
Simultaneous occurrence of two or more verified independent failures. When two or more failed parts are found during troubleshooting, and assignable causes cannot be verified as dependent, multiple failures are presumed to have occurred.