Revision History... 1

[mustangod]

OK, I am having a debaucle with my Doc. Control person, and wanted to see if any experienced drafters or those familiar with ANSI could help me out. In our doc control system, a part number was released without a print under rev A. I created the print for rev B, and since the print didn't exist at rev A, the first rev shown in the block is B. Since the part number is tracked in our doc control system (we use Master Control software), there is history for that part number at the A rev level. The doc control person wants me to put the rev level A on the print's rev block, but since the PRINT wasn't released on rev A, I think it would be misleading to look at the rev block and see that initial release was A (of course) as the rev block states, even though the print that shows it was initial released at rev A didn't exist at that time.

Any ideas here? Thanks!

Mike

 

[ewh]

Mike,

I have to agree with you, as long as the drawing is not the only record of the part revision history. Back in the days of board drawings, when a drawing was redrawn, the revision history did not go onto the new drawing, only the statement that it was redrawn per ECN XXXX (for example). It doesn't make sense to show a part (as opposed to drawing) revision or release on a drawing that didn't exist, and any previous part revisions (or releases) should be documented elsewhere. It is also misleading, making it appear that the part and drawing were released at the same time. How do you document the approval signatures for the initial release?

Good luck in finding a standard to back you up!

Eric

 

[mustangod]

Eric, thank you for your reply! All of our ECO approval records are in the form of 'electronic signatures' in our Doc Control system. So, say you wish to find all revision history for p/n xyz1... You do a query on that part in our software, and it returns all of the different revisions, including document info such as prints for that p/n... I think ASME has standards for this type of thing, I will post back with what I find (if anything!)

Mike

 

[leanne]

Check out:

Mil-Std-973
ANSI/EIA-649
ISO10007

I have worked on programs where drawings were released skipping a rev letter - but it was only when there was an error on an engineering change or when a developmental drawing had changes with hardware in existance (hardware was bult with redlined developmental drawings).

Up until hardware is created a drawing should be released with no revision and upon the first change, becomes Revision A (or Revision 1). Each subsequent CCB (Configuration Control Board) approved change whether incorporated on the drawing or not, should bump the rev letter (or number).

 

[mustangod]

Thanks for the reply Leanne, but I think you are looking at this from a different angle than intended. The post was regarding revision block history of a print vs. a part, whereas I believe you are addressing a subject of having a drawing to begin with, which is not the case here. Either way, your input is appreciated!

Mike

 

[leanne]

mustangod,

Color me confused then 'cause I thought I addressed the original question. I was a Configuration Manager for several years, now I'm back in Quality (I always return to QE).

What I read was:
[ul][li]part number introduced into CM system at Rev A, no drawing released (revision dash or no rev condition skipped)[/li] [li]drawing released in CM system at Rev B [/li][/ul]

Is that not correct? I read your question not as a drafting question, but as one related to configuration management. If I'm not understanding your issue, can you frame it differently so that I "get" it?

[ul][li]Do parts exists against any iteration of your drawing? [/li][li]Is it your company's standard practice to release drawings with a revision letter? [/li][li]What do your contracts (customer requirements) call out with regard to CM and design changes?[/li][/ul]

Where I work, all drawings except those as listed above (i.e., developmental drawings with hardware built against red-lined interations) are released as a baseline with no revision status. Any change after baseline release is accompanied by a revision bump starting with A (or 1 if a numeric revision system is used.) This system is used for both hardware & software.

 

[mustangod]

The issue was a drafting question, with misinterpretation from Doc. Control (Config. Mgmt.). The question is related to how revision blocks of a drawing are affected by rules of doc. control, which should be a non-issue, since doc. control is supposed to follow standards set forth by ANSI for drafting practices.
Per ANSI standards, the revision block of a print is there to detail the history of the PRINT, not the part number. The doc control system is there to keep track of the part number and its respective revision history. The only purpose of adding a part number to a print is to catalog the part in the order and doc control system as to when their existance came to be. If a print of the part didn't exist at rev A, then one should begin the rev. block of the print at whatever revision the print for the part was released. Example: I issue a part number in doc control that references an OEM part, basically a std. part spec, no print required. I release this part number in my doc control system at rev A. I then decide that the OEM part does not fit my requirements in some certain aspect, and I can find no other OEM part to fit the bill, so I create my own. Assuming the part is backwards-forwards compatible, I am encouraged to use the same part number per ANSI Y14.7 drafting stds. So, I create the print, at rev B. Why on earth would I note in the rev block of the print anything about rev A when the PRINT I created had nothing to do with rev A??? That's what the doc control system is there for.

Does this help to explain the situation? I hope I haven't offended you in any way, I surely don't mean to. I just want to make sure we're on the same page )

FYI, I won the battle with doc control after providing this explanation ;o)

Mike

 

[BML]

Just my opinion ...

Part revisions on a drawing should never be greater than the revision of the drawing itself. So if a part was released at A with no print, then the next revision would be B for both the part and the drawing. The drawing revision could mention something like "changed supplier" or "revised to match doc control system."

By the same token, if there are multiple parts on one drawing, it would be okay to revise only some of the parts to rev C, change the drawing rev to C, and leave the remaining parts at rev B.

 

[swertel]

BML,

That assumes that the drawing is the controlling document. In some places, the model is the controlling document and the drawing is just for reference.

i.e. A model is used to develop a CNC program which builds the parts (or an STL file for rapid prototyping or molding). When the part is inspected to the drawing, it is wrong. But, when inspected to the model geometry, it is correct. This is a because the model is the controlling document and a change was made to the model, but the reference documents didn't have the budget to get updated.

--Scott

For some pleasure reading, try FAQ731-376

 

[BML]

swertel, in that case why even have a drawing? Put the revision history on the model and leave it at that. Otherwise the drawing is just one other extraneous piece of information that serves (almost) no purpose.

 

[mustangod]

BML, A drawing in that case serves the purpose of having a document for a QA inspector to reference to gauge part accuracy. Material thicknesses, for example, are not usually checked during production on injection molded parts, just first articles. So the print would stay the same if you changed from .080" thick to .100" thick walls. The print serves the purpose of clarifying and pointing out "critical dimensions", which basically mean if they are out of spec, the part will not work with the rest of the parts in an assembly.

As far as the statement of the part being inspected to the drawing and being wrong, that is a statement often coveted by designers. Designers live in the model world and can't give any less of a crap about the doc control world. They know how mfg. actually works, and the rest is just a paper trail. The drawings SHOULD match the conceptual model, no exceptions.

 

[BML]

mustangod, in my job, I would hate for someone to think that wall thickness isn't a critical feature. It's directly tied to the structural stability of the part, and is most definitely controlled by the drawing (or model).

It seems to me that there is a distinction being made between a drawing used to convey design intent vs a drawing used for inspection. We have this breakdown at our company, and we handle it in the following way.

Manufacturing is required by our documentation processes to provide shop-floor information, such as equipment setup and run procedures, maintenance instructions, and instructions on which features to check and how to check them. There may even be some further detail, such as "if the part is in tolerance but more than .XXX from the nominal, do procedure YYY."

However, the design intent drawing/model always supercedes the shop-floor documentation. If the shop documentation references an old dimension, then it must be revised. In this case, each document (design vs shop) has its revision increased any time any item on the document changes.

 

[mustangod]

BML, wall thickness isn't always directly related to structural integrity, in my line of business, it is for aesthetics as often as structure, and sometimes we only change it at a certain point for the part to release easier from the mold.

As far as drawing differences, there is only one drawing, and it should be made to satisfy every criteria possibly required by all parties.