PFMEA Severity - What is Process FMEA Severity estimation based on?

Re: Severity estimation in a PFMEA is a bit confusing

I've been away for a long period of time, but the "unspoken" question push me to wrote me opinion
The problem with DFMEA's and PFMEA's starts from their purposes.
To those in charge of design we told them that their responsibility is only for the product -they don't know the manufacturing process (even if they know that some product failure modes are cause by processes), and to the process engineers we told them to take care of the processes presuming that the product is well designed (but they still need the DFMEA as input because some causes from DMEA's becomes failure modes in PFMEA's).
I don't think I've answer to your question, but I presented an approach used very often in automotive industry.
The negative side of this is that the 2 parties rarerly comunicate properly.

Re: Severity estimation in a PFMEA is a bit confusing

Sanda said:

I've been away for a long period of time, but the "unspoken" question push me to wrote me opinion
The problem with DFMEA's and PFMEA's starts from their purposes.
To those in charge of design we told them that their responsibility is only for the product -they don't know the manufacturing process (even if they know that some product failure modes are cause by processes), and to the process engineers we told them to take care of the processes presuming that the product is well designed (but they still need the DFMEA as input because some causes from DMEA's becomes failure modes in PFMEA's).
I don't think I've answer to your question, but I presented an approach used very often in automotive industry.
The negative side of this is that the 2 parties rarerly comunicate properly.

Click to expand...

Welcome back... I agree with your comment, but remember, there is no hard line separating the DFMEA and the PFMEA. The PFMEA still deals with product failure modes, but focuses on those caused by the Process. The focus is still product failure modes - and their effect on the assembly line and the end user.

Re: Severity estimation in a PFMEA is a bit confusing

Helmut Jilling said:

Welcome back... I agree with your comment, but remember, there is no hard line separating the DFMEA and the PFMEA. The PFMEA still deals with product failure modes, but focuses on those caused by the Process. The focus is still product failure modes - and their effect on the assembly line and the end user.

Click to expand...

That's true only if the AIAG guidelines are adopted. There is benefit to be had (as I've said before) in concentrating the PFMEA on risks not to the customer or end user, but on "the organization's" own processes. We should be able to assume that risks to customers and end users are addressed in the specifications.

Re: Severity estimation in a PFMEA is a bit confusing

Jim Wynne said:

That's true only if the AIAG guidelines are adopted. There is benefit to be had (as I've said before) in concentrating the PFMEA on risks not to the customer or end user, but on "the organization's" own processes. We should be able to assume that risks to customers and end users are addressed in the specifications.

Click to expand...

True, I took the practice from there since Sandra mentioned the automotive reference. However, in any case, I advocate a broad approach. There is nothiing to be lost in identifying all the relevant failure modes. If there is an issue that will affect the user, let's say part rusting due to insufficient paint thickness, there is no reason that should be left out. That is the purpose of a PFMEA. Identify risks to the internal processes, the customers, end users, etc.

Re: Severity estimation in a PFMEA is a bit confusing

Helmut Jilling said:

True, I took the practice from there since Sandra mentioned the automotive reference. However, in any case, I advocate a broad approach. There is nothiing to be lost in identifying all the relevant failure modes. If there is an issue that will affect the user, let's say part rusting due to insufficient paint thickness, there is no reason that should be left out. That is the purpose of a PFMEA. Identify risks to the internal processes, the customers, end users, etc.

Click to expand...

Corrosion is the result of a process failure, and that (or those) process failures are best characterized as process failure modes. Corrosion is the effect of a process failure mode, and any potential process failure modes should be considered fair game. This does not mean, however, that effects on end users per se should be considered. One of the effects of failure in end use (failure that's a result of process failure, not design failure) is something bad happening to the provider. That's the relevant risk.

Re: Severity estimation in a PFMEA is a bit confusing

Jim Wynne said:

Corrosion is the result of a process failure, and that (or those) process failures are best characterized as process failure modes. Corrosion is the effect of a process failure mode, and any potential process failure modes should be considered fair game. This does not mean, however, that effects on end users per se should be considered. One of the effects of failure in end use (failure that's a result of process failure, not design failure) is something bad happening to the provider. That's the relevant risk.

Click to expand...

Jim, I don't think you answered my question.
If there is an issue that will affect the user, we have to consider it somewhere. Why shouldn't we use the FMEA that we are doing anyway ofr everything?
I used the example of a part that recived insufficient paint thickness. That is clearly a process failure. The part looks perfectly good when it was shipped, buut it will rust prematurely.

There is no reason that should be left out. That is the purpose of a PFMEA. Use one document to identify risks to the internal processes, the customers, end users, etc. That is what the darn thing was made for...

Re: Severity estimation in a PFMEA is a bit confusing

Helmut Jilling said:

Jim, I don't think you answered my question.
If there is an issue that will affect the user, we have to consider it somewhere. Why shouldn't we use the FMEA that we are doing anyway ofr everything?
I used the example of a part that recived insufficient paint thickness. That is clearly a process failure. The part looks perfectly good when it was shipped, buut it will rust prematurely.

There is no reason that should be left out. That is the purpose of a PFMEA. Use one document to identify risks to the internal processes, the customers, end users, etc. That is what the darn thing was made for...

Click to expand...

The specifications for the product are supposed to derive from all of the requirements. That's why a job shop--which will never have access to an OEM DFMEA--should concentrate its efforts on meeting the specifications, and not what might happen if the specifications are met and something bad happens anyway. Meet the specifications and if something rusts or blows up or goes over the edge of a cliff, the design-responsible party will be responsible. It's a dangerous thing for a job shop to try to accomplish things that the designer is responsible for.

Re: Severity estimation in a PFMEA is a bit confusing

Jim Wynne said:

The specifications for the product are supposed to derive from all of the requirements. That's why a job shop--which will never have access to an OEM DFMEA--should concentrate its efforts on meeting the specifications, and not what might happen if the specifications are met and something bad happens anyway. Meet the specifications and if something rusts or blows up or goes over the edge of a cliff, the design-responsible party will be responsible. It's a dangerous thing for a job shop to try to accomplish things that the designer is responsible for.

Click to expand...

Jim, I agree with your premise, but I don't think you are answering my question. A painter MUST evaluate the consequences of underpainting a part. And, those consequences only occur later to the end user. The failure mode is premature rusting. Isn't that the very purpose of a PFMEA? The spec indicates the paint mill thickness, the PFMEA considers the risk of failures. Why do you have a problem with that? The purvey of a designer is to specify the paint, the rest is process management.

Re: Severity estimation in a PFMEA is a bit confusing

Helmut Jilling said:

Jim, I agree with your premise, but I don't think you are answering my question. A painter MUST evaluate the consequences of underpainting a part. And, those consequences only occur later to the end user. The failure mode is premature rusting. Isn't that the very purpose of a PFMEA? The spec indicates the paint mill thickness, the PFMEA considers the risk of failures. Why do you have a problem with that? The purvey of a designer is to specify the paint, the rest is process management.

Click to expand...

Having some experience with automotive paint processes, I can tell you from firsthand experience that the goal of the process is to meet the specifications and in particular to pass the performance testing that's almost always required in those cases. If I understand the specifications and can demonstrate, using the customer's criteria, that I can meet them given specific process controls, my goal is to maintain those process controls and continue to meet the specifications. What might happen in the field is not a part of the thinking that goes on; it's all about successful test results.

I experienced two instances where automotive parts (not painted metal) failed in the field, and in both instances, despite the customer's dogged attempts to blame the failures on us, we were able to demonstrate that the parts we made consistently passed the required testing and otherwise met all of the specifications. There were no safety issues involved; these were interior trim parts. When safety issues are involved and product liability is an issue, it's even more important that a job shop is able to demonstrate a pattern of consistently meeting the customer's specifications.

There are innumerable instances every day of job shops making parts without knowing exactly how and where they're used. They might know in a general way what a part is used for, but to think that contract shops do or should always understand the potential field failure modes is just not realistic.

Re: Severity estimation in a PFMEA is a bit confusing

Jim Wynne said:

Having some experience with automotive paint processes, I can tell you from firsthand experience that the goal of the process is to meet the specifications and in particular to pass the performance testing that's almost always required in those cases. If I understand the specifications and can demonstrate, using the customer's criteria, that I can meet them given specific process controls, my goal is to maintain those process controls and continue to meet the specifications. What might happen in the field is not a part of the thinking that goes on; it's all about successful test results.

I experienced two instances where automotive parts (not painted metal) failed in the field, and in both instances, despite the customer's dogged attempts to blame the failures on us, we were able to demonstrate that the parts we made consistently passed the required testing and otherwise met all of the specifications. There were no safety issues involved; these were interior trim parts. When safety issues are involved and product liability is an issue, it's even more important that a job shop is able to demonstrate a pattern of consistently meeting the customer's specifications.

There are innumerable instances every day of job shops making parts without knowing exactly how and where they're used. They might know in a general way what a part is used for, but to think that contract shops do or should always understand the potential field failure modes is just not realistic.

Click to expand...

Jim, this is going to be my last attempt to synchronize with you on this thread. Again, I agree with your premise, but again, you did not answer my question -

"What is the purpose of a PFMEA, automotive style particularly, if not to evaluate the failure mode and the risks involved?"

If applying a RPN number to a permaturely rusting part is not a failure mode that a painter must evaluate on a PFMEA, then we are in a parallel universe...