Potential Failure Mode & Potential Effect of Failure Definition

When I read the AIAG FMEA 4th edition for Process FMEA, I understand as follows:
- "Potential Failure Mode" is the process failure that make the process cannot meet particular process requirement.
- "Potential Effect of Failure" is the product failure that make the product does not meet the product spec.

However, some customer told the "Potential Failure Mode" must be product failure to meet product spec.

Please advise which one is correct, I think I have already understood but I am confused now.

 

Good day @Pongsakorn ;
Consider this....

e.g.
The name of the requirement/step is "properly install wheel nut"
During the manufacturing process, you are installing a wheel nut on a car. The controls include torque and torque angle. Something goes wrong in the process resulting in the "Failure Mode" : Improperly installed wheel nut.

In this example (appropriate in most examples) the "Failure Mode" is the opposite of the intent "requirement/step".

Now while the car is being driven, the wheel falls off. That is one potential "effect of failure" (i.e. the result that was allowed to occur due to the "Failure Mode").

Don't overthink this. The PFMEA is simply a tool to help us see what might occur if our process fails, and then apply risk based thinking to determine what actions, if any, are necessary to prevent that process failure from happening .

Specific to your question... , ...
a) the failure mode is a nonconforming output [i.e. "...fails to meet process requirements..."] (I would assume in context of your question that output is PRODUCT), which is potentially allowed to occur if the process fails.
b) the effect of failure ["effect of the failure mode as perceived by the customer..." ] is a potential result of that nonconforming output (NG product).

Hope this helps.

Be well.

 

@Andy Nichols is spot on. The failure mode must be properly identified. Keep in mind, it might be different depending on who the "customer" is. A cracked valve at the manufacturer may simply be "cracked valve body", but at the end user it may be "leaky valve". Both correct, but simply depend on how far out you are aware of the use and implications.

And of course, yes, there are commonly more than "one". For example, in the example I gave above, the "improperly installed wheel nut", may have "effect of failure" ranging from RATTLE, to WOBBLY WHEEL, to WHEEL FALLS OFF to LOSS OF CONTROL/CRASH. This is why the (under the current RPN system) the "highest" ("worst" ) result is used to calculate the RPN.

Hope this all makes sense @Pongsakorn

Be well.

 

Thank you so much John.
Your reply is helpful.

 

Andy, thanks for your reply.

 

Anyone have any examples of the new required format?

https://fpdl.vimeocdn.com/vimeo-pro...11-0xf86a6bd7d5f0b63d6a59083f86cdeba2e69a60fe