Revising a drawing to eliminate a nonconformance

This is my first post (in any forum not related to cars/trucks). I'm a quality engineer by day, and an issue has come up at work that I'd sure be interested in hearing outside opinions on.

When a nonconformance exists on a production part, we all know a nonconformance record should be created to control the nonconforming material. If the nonconformance is considered to be acceptable as is, a Use-As-Is disposition is proposed. In some instances, however, at my place of employment, instead of proposing a use-as-is disposition, the nonconforming part waits while a drawing change is processed to make the engineering definition match (or at least tolerate) the nonconforming characteristic. Then the nonconformance record is closed, cancelled, or dispositioned as "rework."

I am opposed to this method, and I think it is not in accordance with the intent of the QMS; however, I cannot find anything that specifically states it can't be done. My take is that the part should be held to the released engineering at the time of its manufacture. Thoughts?

 

Thanks for the response. I would be in favor of a Use-As-Is disposition, as a Use-As-Is disposition basically translates to "the part is nonconforming, but the proposal is to use it in its current configuration anyway." My gripe is that after the engineering definition is revised, the part in question is then considered conforming and the disposition becomes "rework," rework means actions were performed to the part to make it conform to the engineering definition (which is not exactly what happened). I see no provisions for doing this in our QMS procedures for processing nonconforming material - the procedure has options and potential paths for how to disposition nonconforming material, not one of which is to wait and revise the engineering definition so that it tolerates the nonconformance.

 

I appreciate your opinion/input on the matter, and I understand what you're saying as far as the paper trail telling an accurate configuration tale. It still seems to me that drawing change effectivity should only apply to parts not yet built. Applying drawing changes to parts already built seems crooked (it alters defect metrics, for one thing). But, the fact that I can't find anything that specifically forbids it should be an indicator to me that it isn't as offensive as I think.

 

Another follow on thought to this... the organization has certain obligations related to nonconforming product, such as notifying the customer of delivered nonconforming product, or if not yet delivered, to seek customer approval. The organization may be side-stepping some of these requirements/obligations by revising the engineering definition instead of treating the product as nonconforming.

 

Good day RPNYVB;
A few questions if I may..
1- is your organization design responsible (or does your organization simply mfg to the customer provided drawings?)
2- what is the contract with the customer ? (Is there an agreed upon print? ; Is the specification in question on a control plan or other method sent to/“contracted” with the customer? )

 

1 - The design is the organization's, not provided by the customer. The organization is the prime contractor and does have design authority. There are policies and procedures related to drawing changes, although I don't believe the procedure is so detailed as to specify whether the change can affect parts already built or not. Sorry for the vague terms, I have some (likely irrational) concern about being too specific on a public forum.
2 - The contract calls out the engineering drawing #s, but there are provisions for drawing revisions during production. There is a process for processing drawing changes, and (eventhough I'm not directly involved in this part), the customer is somehow involved in concurring with those changes.

 

To give an example, let's say 100 parts are made with 10 holes slightly oversized. Everyone agrees that the holes can still be used in their current condition (no technical concern). The quality org, however, doesn't want the 100 x 10 defects to hurt their metrics, so they let the parts sit on a shelf while the engineering dept. processes a design change to open the tolerance to tolerate these 100 parts. The nonconformance record is then changed to show the parts meet design, and some additional processing steps/approvals are avoided. The engineering dept. owes the quality manager a favor now, but in the meantime it seems the fidelity of the process has been jeopardized.

And as Andy mentioned, it does seem like there is an issue with Product Design Verification (I am not as familiar with that section of ISO 9001). I'll have to revisit that section before commenting much more on it, but I recall reading the product verification section and thinking that it didn't jive with the backwards design change trick I've been describing.

 

Good day @RPNYVB ;
Based on what I understand, it sounds as if resident expert @Andy Nichols 's council is spot on specific to your current situation. I do, however, agree with your concerns. Specifically, as stated by @Andy Nichols , the seemingly poor job done during NPD (new product development) in ensuring a capable process. This is "quality 101" and can/will surely lead to real or potential problems in the future (as you clearly seem to be aware).

Beyond the inherent risks to your organization, there is a risk (if an astute auditor) of a QMS non-conformance specific to 8.3.5-b , whereas, "the organization shall ensure that design and development outputs are adequate for the subsequent processes..."
In other-words, the product design and development "...shall ensure..." = DFM- Design For Manufacture. (i.e. do product design engineers have the necessary information as to process capability, and are they considering that when the design?)

If your organization was NOT design responsible, then the responsibility would be part of 8.2.3.1 (Review of requirements (i.e 'feasibility" review).
Since your organization IS design responsible, then the aforementioned 8.3.5-b applies.

Hope this helps.

Be well.

 

Ah, yes, I see what you're saying. The ding would be against Manufacturing, not the Quality dept. I'm clearly not involved in that part of the process!

 

You’re way overthinking it and over complicating it. If the parts can be used as is, then are they really “defective?” Bring them in with a pencil and work on real issues.

 

Yes, they are defective if they do not match the engineering definition.

 

So you're saying making quick work of "use as is" parts by revising the print was costing them $8m? Obviously there's a lot more to that story. If it doesn't affect the part, it doesn't affect the part, and some drawing isn't going to change that.

 

"Defect" is not a term I commonly use, and when I do, I translate it as "nonconformance to engineering requirement." I revisited ISO9000 today, and was surprised at how ISO9000 defines "defect." I wish their definition were more clear, as they describe "defect" generically as a nonconformance that impacts "use." In my world, a part either meets the requirements or it doesn't. Some group of people somewhere decided what was important enough to be a requirement and what wasn't. Whose responsibility is it now to determine what nonconformances are defects and what aren't? And why does it matter? So, if we're using the ISO9000 definition of "defect," I'm not sure I stand by my earlier statement. I'll stick with using the term "nonconformance" in the future, as the term "defect" seems obscure to me.

 

I don't really get your point. If you have a "defective" part (defective because it doesn't match the print) but the "defect" will have no adverse affect on the part/final assembly, and it is "fixed" by modifying the print so this and subsequent parts will be in conformance, do you not reduce future costs by not constantly dealing with the issue.