Initial PFMEA

[Sebastian]

Hello,

I had occasionally joined team during PFMEA.
It was during process development stage, post claim cases, trainings.
I had never opportunity to see how initial PFMEA, before tooling launch was done.
Now, we want to do it and there are some unclear points.
Maybe is it somewhere described, but I did not find it in AIAG FMEA-4 or AIAG & VDA FMEA.

Could you share your experiences in this matter?

For example I have first question.
We know, that prevention against improper orientation of part before processing is to use shaped mold (jaws).
During PFMEA for new product, we don't have it in our hands, but we know we need them and later we will have them.
Should we put it in advance into "Current process" section or "Action results" is much proper one, as first it needs to be done?

Next one is related.
Potential effect of other failure is different when you assume, that process could be done without shaped mold, e.g. failure flow outs to customer.
With shaped mold implemented, it results in part damage.
So proper location of "know how" is right or left side of RPN column?
What is your approach?

 

[John C. Abnet]

Hello,

I had occasionally joined team during PFMEA.
It was during process development stage, post claim cases, trainings.
I had never opportunity to see how initial PFMEA, before tooling launch was done.
Now, we want to do it and there are some unclear points.
Maybe is it somewhere described, but I did not find it in AIAG FMEA-4 or AIAG & VDA FMEA.

Could you share your experiences in this matter?

For example I have first question.
We know, that prevention against improper orientation of part before processing is to use shaped mold (jaws).
During PFMEA for new product, we don't have it in our hands, but we know we need them and later we will have them.
Should we put it in advance into "Current process" section or "Action results" is much proper one, as first it needs to be done?

Next one is related.
Potential effect of other failure is different when you assume, that process could be done without shaped mold, e.g. failure flow outs to customer.
With shaped mold implemented, it results in part damage.
So proper location of "know how" is right or left side of RPN column?
What is your approach?

Good day @Sebastian ;

1= "Action Results/"Recommended actions" . (Once implemented and confirmed, it can be moved to "action taken and any SOD/RPN adjustments made).

2= Not sure if I understand your question. I infer from your post that the "shaped mold jaws", are NOT yet implemented. Therefore, the potential for damage is/should be part of your risk assessment. So, if I understand correctly...
a) NOT yet implemented
b) WILL be implemented
c) WILL eliminate problems with part orientation (SOD and RPN adjusted accordingly once confirmed), but will ALSO potentially cause part damage (SOD and RPN adjusted accordingly once confirmed).
If none of this part of the current process, it should only be placed in RECOMMENDED actions at this time. Then, once tested and/or implemented, the SOD needs to be reconfirmed and adjusted if/as necessary.

Hope this helps.

Be well.

 

[Sebastian]

Thank you very much John.
Interesting why there is only one comment. Maybe performing initial PFMEA analysis when tooling is available, is a common practice.

It's about how to record current status of process development and "know how" we have.
For other product, already in production we use "shaped mold jaws". Btw. is it proper wording?
For product during implementation, we need to design, manufacture and use this tool, because product shape is different.
So, we agree, that it should be placed in "Recommended actions" section.

In following part of my post I wanted to confirm, that defect consequences can transform in line with manufacturing process development progress. Initially without individual tooling, failure will result in defected part flow out to customer. Later when tooling is available, same failure will result in damage made by tooling on following process step, because input part was out of specification.

 

[John C. Abnet]

"shaped mold jaws"

Very intuitive term...("i.e. shaped mold jaws"). It was very clear what you were speaking of.

Be well.

 

[Tidge]

For the "left side" Process FMEA I start with two things:

1. What is the flow of the process I am implementing?
2. Have I been directed by the customer to add specific process steps?

An external customer (or separate design team) won't necessarily know what steps the factory will take in a manufacturing process, so I always start with that. For example, a customer buying molded plastic parts from a supplier may be unaware that mold release is used. A PFMEA can certainly explain why you need mold release.

The second item is important because some customers will require a manufacturer to add specific controls to certain process steps; the PFMEA will contain the rationalization for those controls. Some process steps won't have much effect on if non-conforming parts will be produced, yet I still included them in Process FMEA. Detailed analysis of outcomes of failures in some steps may be better analyzed in other lines of the PFMEA.

As far as "ratings" (detectability, severity, occurrence) on the "left side" of PFMEA, I try not to overthink them. In the absence of any process controls, I generally treat severity as worst case, that is "I can probably expect to get non-conforming products when I have no controls". I allow the process experts to sway my opinion on left-side detectability (e.g. how obvious the failure mode would be to the uninitiated?) and occurrence (e.g. just how likely is this failure mode even likely to occur?). The right side (after implementation of controls, analysis of the controls, establishment of control charts, etc.) is where the process engineers get to demonstrate their expertise.

 

[Sebastian]

Tidge, thank you very much for sharing your experience in this matter.

It's a pity, FMEA manual does not go into details, how analysis progress may looks like.

2 views on subject only?
Maybe a lot of tooling is made with outsourcing services or through remote supporting functions.
I have doubts, PFMEA is done and audited properly in these cases.

Focus is on latest pre-serial revision and correspondence with control plan and then post-claim reviews and/or revisions.
None cares when initial PFMEA was made and how it looks like.

 

[Tidge]

Initial ("left-side") work on Process Failure Modes Effects and Analysis does have (market) advantages, even when a manufacturer has plenty of experience with both their processes and controls. Many different manufacturers have adopted APQP processes steps as a part of 'supplier controls'; my experience has been that smaller (than automotive) manufacturers who adopt APQP elements find themselves stumbling over (some of) the steps because of "waterfall or iterative?" or "checklist or checkbox?" types of issues in APQP... suppliers who know how to punch out a "left-side only" PFMEA can provide some value to their customers. Sometimes the customer only wants a 'complete' Process FMEA, but sometimes the customer wants to get some "skin in the game" too. Mileage may vary.