Process FMEA - Need help to clarify O (Occurrence) and D (Detection) rankings

[qsmso]

Dear friends,

I have the question about FMEA rating as the followings:

1. The Occurrence (O) is the frequency of failure mode or the frequency of failure cause?

2. If the (O) rating is for the frequency of failure casue, how can we use the statistic data to count it because normally, we monitor ppm or % defect for only failure mode.

3. Also, for the Detection (D), this is for detection of failure mode or detection of failure cause?

Thank you for the help.

QSMSO

 

[Al Dyer]

Good Morning qsmso,

Perform a search using failure+mode and so much info on FMEA's will pop up that you will be busy for days if you want.

Short and sweet;

1-Failure Cause
2-You would have to start monitoring the causes. Possibly a Pareto Chart.
3-Failure Cause

Al...

 

[Bill Ryan - 2007]

Sorry Al but number 3 = Failure Mode (at least with every one of our automotive customers). It is based in the thinking that WHEN the Failure Mode happens, what is your ability to catch it.

 

[Al Dyer]

FMEA Detection

Hmmmm....... Let's take a look at the above link, I sure have been wrong before and I do have fat fingers.

Al...

 

[Jim Wynne]

1. The Occurrence (O) is the frequency of failure mode or the frequency of failure cause?

What you want to ask is, "How likely is it that this particular failure (either the mode or the cause) will occur?" Remember--it's possible to predict failure without necessarily knowing all of the possible causes.

2. If the (O) rating is for the frequency of failure casue, how can we use the statistic data to count it because normally, we monitor ppm or % defect for only failure mode.

You don't necessarily need statistical data per se in order to make a prediction but the prediction should be based on experience with similar processes and products. Make the best estimate you can with the data available.

3. Also, for the Detection (D), this is for detection of failure mode or detection of failure cause?

Detection of the failure mode as manifested in the process output. You should be looking for potential process failures, and providing for ways to detect nonconforming product before it's delivered to the customer, or before further value is added to it.

Thank you for the help.

 

[Jim Wynne]

FMEA Detection

Hmmmm....... Let's take a look at the above link, I sure have been wrong before and I do have fat fingers.

Al...

The "Detection" factor is based on the ability to detect defects before the product is shipped, or before further value is added. Here's what SAE J1739 (the "parent" document of the AIAG FMEA manual) says:

Assume the failure has occurred and then assess the capabilities of all "Current Process Controls" to prevent shipment of the part having this failure mode or defect (Emphasis added)

 

[Al Dyer]

Thanks js!!!


Al...

 

[bgwiehle]

I generally agree with the answers that you've received... when you rate occurrence and detection, you are essentially looking at a line item on your FMEA -- the combination of a particular failure mode and cause and the particular controls in place at that process.

Controls may be poka-yokes and/or characteristic controls [failure mode]. Many process controls are prevention controls [cause] and should not contribute to the detection rating, although their effectiveness affects occurrence.

If you are only tracking scrap rates of a defect (failure mode) without a lot of data on the contribution from each cause, use the total scrap rate for that defect as a starting point. The occurrences of each cause should total the occurrence of the defect. Use process or part history to assign proportions. The rating process may also point up missing causes when the occurrence numbers just won't add up logically.

B.G. Wiehle
PFMEA Co-ordinator

 

[Jim Wynne]

If you are only tracking scrap rates of a defect (failure mode)

In a PFMEA, the failure mode should reflect a failure in the process, not the manifestation in the product, regardless of what the SAE and AIAG guidelines say. Use of part defects as failure modes is a good way to miss prevention opportunities.

The occurrences of each cause should total the occurrence of the defect.

Not necessarily. It's possible for a potential cause to occur without resulting in defective output. For example, in a machining operation, failire to change tools at a prescribed interval might be a process failure mode, but it might not result in defects every time it happens. For every defect there's a cause, but there isn't necessarily a defect every time the cause happens.

 

[astudent]

I'm still confused! (
Maybe I need a summary.

As to (D) on the AIAG worksheet there is a row for each inspection, and there is an inspection for each cause.
So Do we inspect causes or modes?
I can't understand. ()
Is (D) referred to the secund column of actual controls?
Can someone paste what the AIAG standard defines the two kind of controls to discuss them?