Revision Levels and Traceability Of Specifications and Controlled Documents

[KP1SP]

Hello All,


Wondering if anyone know if there is a standard for controlling revision levels for specifications and documents? We have to many personal opinions flying back and forth at work and I need something concrete as evidence for a change, simply to just get something standardized here. We have an audit coming up and we feel there are some holes in our current protocols.

Thanks,

-Kyle Paolino

 

[swertel]

Officially, specifications are always the most current version available, therefore you "NEVER" reference the revision of the specification on your drawings or documents. The only exception is ASME Y14.5 because the interpretation of some of the symbols changes over the years but the drawings don't, so the revision has to be listed.

Now that doesn't always sit well with 1) auditors, 2) failure investigations, 3) design groups who spend more time verifying revisions than designing, 4) production groups who always have to order another spec when only a typo is different, or 5) quality groups who don't know how to read specs anyway.

My solution is to put the revision in the listing of referenced documents (often "Section 2" of a report) but NOT to put the revision when actually calling out the specification as part of a requirement. In the list of reference documents, I have a sentence that states "revisions shown here were the current revision at the time of this writing" or something similar.

--Scott
www.wertel.pro

 

[KENAT]

KP1SP - if you can base it upon drawing revision control then see ASME Y14.100 & Y14.35m

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[KP1SP]

swertel/KENAT,

We are an ISO13485 certified company so we need to have very detailed SCN. Every change that affects form, fit or function needs to be documented. There has been debate here whether or not we need a revision level when the product specification is still going back and forth to the customer pending approval. What I am ultimately trying to find out is if we really need to have a revision level change every time I send the spec to the customer, even if it is still in the R&D stage. Currently we just change the date on the drawing and leave it in "DRAFT" until the spec is agreed to by the customer. Then it would become Rev 000 in our EBR system. This was mainly to avoid entering in part numbers that are at Rev 011 before we even introduce them to our system.

And we may be seeking ISO9001 certification later on, which I believe requires far more extensive document controls than current.

Thanks,

-Kyle Paolino

 

[KP1SP]

Also,

When a customer places a PO we need the most current part number and revision listed for both companies. So I do need to have a revision level on the part number. But I was hoping there might be a standard for this so that I may just shut down the opinionated suggestions going back and forth in meetings lately. No one can agree and no one can compromise, so I figure if there is a standard protocol then that's an easy argument for change.

-Kyle Paolino

 

[KENAT]

So the ASME specs I mentioned only deal with 'released documents' - of course one can debate when the drawing is released.

Where I work, prior to formal release revision control of drawings is essentially left to the individual drawing creator using revision numbers and it's not well controlled but in principle:

- Any time draft document gets changed & shared (either with vendor, customer or large group internally) it will get a numeric revision.​

- Draft documents should have a reference only not released drawing or similar 'stamp'.​

- When document is formally released it gets a revision letter and is then under full ECO control in a secure part of our server.​

So, revs 1, 2, 3... for initial drafts & even prototypes. Revs A, B, C... for properly released drawings.

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[KP1SP]

KENAT,


That is what the QA manager and myself are pushing for as the new protocol; to use a letter or number for draft and opposite for official release. Others are pushing for us to have an internal release and new revision for every little change along the R&D path. I don't think they realize how much more work that will add for QA and myself in approvals. The main problem is that I have non engineers trying to develop my protocols and it's just not working.


Thanks,

-Kyle Paolino

 

[ewh]

Try reading Frank Watts book on engineering documentation. It may be of help.

http://www.amazon.com/Engineering-Documentation-Control-Handbook-Edition/dp/0815515952

“Know the rules well, so you can break them effectively.”
-Dalai Lama XIV

 

[Duwe6]

KP1SP, I feel your pain. You might consider 'judo' - just relax and let the auditors deliver the drubbing that these folks have earned and deserve. Then you and the QA Manager can roll out your [correct and industry standard] revision control & documentation procedure as a response to that audit finding.

BTW, the standard seems to be letters - Draft-A, Draft-B for unreleased editions, and numbers Revision-0, Rev/1, Rev/2 for the Released-for Fabrication drawings. And for ISO documents and procedures, it is best to complete an MOC - Management-of-Change for each revision after Rev/0.

 

[KENAT]

Duwe6 - for what it's worth ASME Y14 series only really talk about letter revisions, and that's for 'released' drawings. (They allow - for initial release as I recall.)

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[Duwe6]

This is one of the places ANSI Y14 diverges from common industry usage. Never seen A, B, C used for released drawings, except maybe medical device fabrication. Refineries, chemical plants, powerhouses, nukes that I've been at have all used 0, 1, 2 for released drawings and spec's. YMMV