FMEA link to Control Plan frequency and sample size

[yoledie]

I had a recent finding from an auditor that our Control Plan is not linked to FMEA or other sources available to justify sample sizes or test frequencies?
Could someone help me to explain how is the requirement for linking of FMEA vs Control plan? And how FMEA can justify a sample sizes or test frequencies from control plan? Thanks in advance

 

[Alan_DB]

Does your FMEA rely on trending of incidents / testing of product to determine if risk mitigations are effective? If so, the post mitigation likelihood should drive your samples size / frequency in the control plan.

 

[yoledie]

Does your FMEA rely on trending of incidents / testing of product to determine if risk mitigations are effective? If so, the post mitigation likelihood should drive your samples size / frequency in the control plan.

Hello Allan, do you have any example? I can't still get it..
For example: Control Plan states the following
Control: Dimension measurement Sample Size: 100% Frequency: Per lot/material
How can I use the FMEA to justify the control with sample size/frequency in control plan?

 

[Sebastian]

I would change thread title, e.g. importance of frequency in PFMEA analysis, as current one is too general.

Detection table ranking (5), 6 - 10 per AIAG does not depend on sample size and frequencies.
What was your ranking?
Sample size depends on parts quantity achieved during one production cycle, e.g. when multiple cavity molds are used, parts coming from all cavities need to be verified. Other aspect, when failure source is setup related, so setup inspection size should be reliable, e.g. 5 pcs.
Frequency starts from level of 100%, but when capability analysis results are satisfactory, you can implement sampling, e.g. first off/ last off.
I can't tell auditor was right or wrong without reviewing both documents, but usually setting sampling is post-PFMEA activity.

 

[Miner]

I had a recent finding from an auditor that our Control Plan is not linked to FMEA or other sources available to justify sample sizes or test frequencies?
Could someone help me to explain how is the requirement for linking of FMEA vs Control plan? And how FMEA can justify a sample sizes or test frequencies from control plan? Thanks in advance

What type of audit was this (i.e., ISO 9001, IATF, etc.)? Were there Customer Specific Requirements (CSRs) involved?

 

[yoledie]

I would change thread title, e.g. importance of frequency in PFMEA analysis, as current one is too general.

Detection table ranking (5), 6 - 10 per AIAG does not depend on sample size and frequencies.
What was your ranking?
Sample size depends on parts quantity achieved during one production cycle, e.g. when multiple cavity molds are used, parts coming from all cavities need to be verified. Other aspect, when failure source is setup related, so setup inspection size should be reliable, e.g. 5 pcs.
Frequency starts from level of 100%, but when capability analysis results are satisfactory, you can implement sampling, e.g. first off/ last off.
I can't tell auditor was right or wrong without reviewing both documents, but usually setting sampling is post-PFMEA activity.

Thanks Sebastian, I think I am getting your point but I think the auditor wanted me to show how FMEA justified the Sample size and frequency from the control plan. Do you have any examples that could share how it was being done? thanks!

 

[yoledie]

What type of audit was this (i.e., ISO 9001, IATF, etc.)? Were there Customer Specific Requirements (CSRs) involved?

Hi Miner, it is a VDA audit. No customer specific requirements involved. I am not sure how to show or justify the sample size and frequency from the control plan using the FMEA? and how will it be linked.

 

[Sebastian]

Some time ago, I've read AIAG FMEA manual searching for this subject and I've presented result in previous post.
Detection table is the only one related thing I found.
Maybe someone else will find more.

Very simplified - PFMEA ends where control plan starts.
The only aspect where we could say, PFMEA goes a little bit on control plan territory, is evaluation/measurement technique.
During PFMEA you identify detection method and then per manual, ranking is different for attribute and variable gauging.
I understand intention - variable gauging is preferred, as through this kind of inspection you get much more valuable data. You can monitor trends, you can calculate capability while attribute inspection result is "OK" or "NOK".
I won't look for justification here, so I can't share examples.

 

[Miner]

I think that the auditor is over simplifying this. When I make a determination of sample size and frequency, I consider all of the following:

• Understanding the process (process flow diagram)
• Understanding the areas of risk (PFMEA) - This determines where I should add control points
• Understand the dominant source of variation in the areas of risk (i.e., setup, machine, operator, material, component, maintenance, tooling, fixture/pallet or environment)
• Identify how often changes occur in the dominant source of variation to establish a frequency (e.g., every batch change, every setup change, etc.)
• Sample size will depend on your control method (i.e., are you sampling, using SPC, type of control chart (i.e., IMR, Xbar/R), etc.)
• History of escapes on similar control points

 

[yoledie]

I think that the auditor is over simplifying this. When I make a determination of sample size and frequency, I consider all of the following:

• Understanding the process (process flow diagram)
• Understanding the areas of risk (PFMEA) - This determines where I should add control points
• Understand the dominant source of variation in the areas of risk (i.e., setup, machine, operator, material, component, maintenance, tooling, fixture/pallet or environment)
• Identify how often changes occur in the dominant source of variation to establish a frequency (e.g., every batch change, every setup change, etc.)
• Sample size will depend on your control method (i.e., are you sampling, using SPC, type of control chart (i.e., IMR, Xbar/R), etc.)
• History of escapes on similar control points

Thanks Miner for the response, this is helpful.. I've found a clause from control plan that states

"The organization shall have a control plan for pre-launch and production that shows linkage and
incorporates information from the design risk analysis (if provided by the customer), process flow diagram, and manufacturing process risk analysis outputs (such as FMEA)."

The organization shall, if required by the customer, provide measurement and conformity data
collected during execution of either the pre-launch or production control plans. The organization shall include in the control plan:
a)
b)
.
.
i) at a set frequency based on a risk analysis


This might be the reference of auditor. Although I am still not sure how to show it using the FMEA, do I need to explain on every fail mode/causes in FMEA how should be the frequency and sample size analyse? that's tough and hard to maintain.
Could someone help me.. what are your suggestions how to address this.