Mixed and Competing Failure Modes - What's the difference?

[Steve]

In Weibull++, what is the fundamental difference between "Mixed" and "Competing Failure Modes"? I know that with CFM you can label the failures in order to separate their effects, but is that the only difference?

[David]

In Weibull++, what is the fundamental difference between "Mixed" and "Competing Failure Modes"? I know that with CFM you can label the failures in order to separate their effects, but is that the only difference?

Hi Steve,

One main difference is that mixed Weibull is a distribution while competing failure modes is a type of analysis. While using mixed Weibull, you are simply trying to use this distribution to model your data. In general, the mixed Weibull distribution is used model data that does not follow a straight line on a Weibull probability plot. This can be indicative of multiple failure modes within the data. Additional information on the Mixed Weibull distribution can also be found at http://www.weibull.com/LifeDataWeb/the_mixed_weibull_distribution.htm.

For competing failure modes, there are also multiple failure modes. But in this case the failure modes are "competing" to cause a failure. Keep in mind that this does not apply to a repairable system. Since the failure modes are in a sense competing to generate a failure, this can be viewed as a series system. A separate analysis must be performed for each failure mode. The failure times for the modes not being considered are suspensions since it is known that the unit would have failed due to the mode under consideration at some time in the future. The analysis for each of the individual failure modes are then combined to provide an overall model. The overall reliability is simply then the product of the reliabilities of each on the individual failure modes. A more detailed explanation of competing failure modes can be found at http://www.weibull.com/LifeDataWeb/competing_failure_modes.htm.

I hope this helps.

David