QA Info on Drawings 1

[pdybeck]

What is everyone's stance on putting QA information on a drawing. We have some individuals here that would like to see some QA info on a drawing, and others that think the drawing should not contain information like that. The QA check for this particular drawing is a test of a circuit to verify proper resistances at certain points. This gives a quick check of whether there are proper connnections among several components. I take the stance that information like this should not be on a drawing, but maybe I'm wrong. What are your thoughts?

 

[ctopher]

Any info that does not have anything to do with creating the part should always be a separate doc.

Chris
Sr. Mechanical Designer, CAD
SolidWorks 05 SP3.1 / PDMWorks 05
ctopher's home site (updated 06-21-05)
FAQ559-1100
FAQ559-716

 

[Hydroformer]

In automotive we identify 'key and/or critical characteristics' on the print. These can include safety items, fit or functional items, etc. that must be met (and monitored) for compliance. Additional checks (risk management) are applied to these elements. This is the design team's way of speaking to QA.

The program manager (ei: me) will bring together the design and quality teams to address how we intend to treat each characteristic. In your case, apply a resistivity check at 'X' frequency across points A&B within a range of 'tolerance'.

Identifying these characteristics on the print throws the flag up for all involved and this is good communication. A proper DFMEA is the tool to identify whether a particular characteristic is key, critical or merely an attribute.

*Without data, you're just another person with an opinion.*

Hydroformer

 

[MadMango]

In the medical device field, I have seen many drawings that had Quality Requirement as notes on the drawing, separate from the General Notes. I have also seen full-blown Inspection or Quality drawings made from the fabrication drawings to cover such things.

According to ASME Y14.100-2000, Section 4.26.6 (j)
Notes shall not include contractual requirements, such as statements of costs; time and place of delivery; methods of payment; and requirements for submission, approval, or distribution of data, reports, or plans.

[green]"I think there is a world market for maybe five computers."[/green]
Thomas Watson, chairman of IBM, 1943.
Have you read faq731-376 to make the best use of Eng-Tips Forums?

 

[ctopher]

Hydroformer is correct. Key/critical characteristics are OK to show on dwgs. Sometimes it is required.

Chris
Sr. Mechanical Designer, CAD
SolidWorks 05 SP3.1 / PDMWorks 05
ctopher's home site (updated 06-21-05)
FAQ559-1100
FAQ559-716

 

[ajack1]

On formed “sheet” parts in the automotive trade it is also common to show the areas where a part will be located for inspection, it can make a big difference.

 

[MintJulep]

I never understood the "key/critical" feature concept.

If a dimension needs to be on the drawing to allow the part to be made, then it is necessary - period. Let the QA guys determine what sampling level (both number of parts and features measured on individual parts) they need to ensure the desired level of scrap is consistently produced.

In pdybeck's case, it sound like a test procedure is needed. In this case I would say that the drawing should contain a note referencing the test procedure.

 

[pdybeck]

MintJulep,

To play devil's advocate... Why not just put the test procedure on the drawing if your are going to put a note about it anyway. What I am getting at is what is the metric you use to decide when to put information on another document outside the drawing? To everyone else, the same question applies. Why locate inspection reference points on a drawing if as Ctopher says "Any info that does not have anything to do with creating the part should always be a separate doc." Is it merely for not having to create another document outside the drawing that references the part? Ctopher, do you take your stance b/c now there is extra information that might end up revising the print that has nothing to do with the creation of the part?

Pete

 

[MintJulep]

pdybeck,

In the current case you have determined that measuring the resistance at a few point is a sufficient check. A few simple steps.

If your next project is more complicated you might have a dozen steps in your test procedure.

More complicated yet, and you could have a few hundred steps.

At some point is become impractical to do this on the drawing. Where is that point? Any answer other than "the test is a separate document, referenced on the drawing" will result in product-to-product inconsistency in your production methodology.

Also, where do you record that the test has been performed, or for more complex tests, record the actual results and values measured? Where do you record what instrument was used to make the measurement?

 

[pdybeck]

MintJulep,

I agree with you, I was just digging for the reason behind the answer. Thanks for the reply. Maybe Hydroformer or ajack1 have a different metric....

 

[Heckler]

One instance I can think of why you wouldn't want a test procedure on a print is revision control. If the test engineer wants to revise the test procedure then the drawing has to be revised. If the test procedure number is called out on the print then the current rev is used.

We call out the document numbers to Test procedures (ATPs), and Procurement specifications. It just makes of better document control.

 

[ajack1]

In the automotive trade well at least with panels it is now almost unheard of to have a drawing in the old sense of the word.

Everything works from the model and is measured as such. All holes or slots are dimensioned and toleranced. Datum holes or faces are denoted and the areas that the part will be restrained by in the checking fixture are shown both on the drawing and model.

Some companies show mating faces and have a different tolerance for them to the general surface tolerance or even a unilateral tolerance. Beyond that you get a surface tolerance and a trim tolerance.

No one tells you how to measure the part although some companies do supply CMM programs or checking points, but they do tell you how it will be held to be measured, this is critical with complex 3D shapes.

 

[markn82]

I would have to agree with Heckler on the point of not putting the actual procedure on the print. I'm used to seeing Key/Critical Characteristic callouts often refering the person reading the print to procedure. If the test procedure is changed for whatever reason, you would have to change the print.

From a document control stand point, keep as many forms as generic as possible so changing one form or procedure does not require changing 100 or more prints. If you were going to follow this same procedure on a family of parts that has a lot of prints tied to it, each print would need to be updated if the procedure changes. If you just reference the procedure by title, number, whatever you use, don't specify revision level, you'll possibly save yourself a lot of work later.

 

[btrueblood]

It is normal, and correct, for an assembly drawing to have test procedures called out on it, either directly or by reference to other documents.

Examples include circuit board trimming/trim pot settings as in the OP's example, as well as hydrostatic test pressures for stuff that Mango or I would draw.

Ctopher wrote: "Any info that does not have anything to do with creating the part should always be a separate doc."

I agree with this, but argue against what I think he meant. Ctopher probably thought QA requirements should not be included on the part; generally he is correct. But if a certain component is required by engineering to be trimmed to a certain value in order for the assembly to behave as intended, that requirement should be spelled out on the drawing. Similarly, if a contract requires hydro testing (or leak testing, or flow calibration, or performance test, etc....) of the final assembly before shipping, that should be called out. Those requirements are part of what defines the assembly.

 

[MadMango]

The requirement would be spelled out, but not the actual process.

You would state on the drawing "Must pass hydrostatic test IAW ASTM E1003-95." You wouldn't print the entire standard on the fabrication drawing.

[green]"I think there is a world market for maybe five computers."[/green]
Thomas Watson, chairman of IBM, 1943.
Have you read faq731-376 to make the best use of Eng-Tips Forums?

 

[ctopher]

I agree with MadMango. In a note, point to a spec or internal process spec, but do not spell out the process on the dwg.
For a PCB, trimming pots or similar is acceptable, it is part of the process of mfg the part.
Test processes change and should not be spelled out. Also, if too many processes are spelled out in a note, the machinist or whoever is making the part, will quote to do all processes in addition to making the part.
In my experience with aerospace/military industries, Key/Critical Characteristic callouts are required most of the time to verify the part will fit according to customer specs and interface dwgs. These callouts are referenced to/from QA docs that is configuration controlled with the customer.
As Markn82 and Heckler pointed out, do not list revisions with specs on dwgs either. The only dwgs I have ever the need to list revisions is military dwgs that are owned/controlled by the military or their main supplier (our customer).
Also, a lot of companies have proprietary dwgs that are not seen by their customers. But, during an audit, the customer may ask to see test procedures (or any procedure) to insure them they are getting a quality product. Usually these procedures are given to the customer to review. If listed on the dwg, they do not have the dwg to review.
Drawings should be controlled by Engineering, Drafting and/or Design depts; quality docs/test proc/etc controlled by Quality.

Chris
Sr. Mechanical Designer, CAD
SolidWorks 05 SP3.1 / PDMWorks 05
ctopher's home site (updated 06-21-05)
FAQ559-1100
FAQ559-716

 

[Jabberwocky]

I just wanted to thank everybody for this great thread. We have been going through some of this at work lately, trying to decide on the best way. Our French parent company appears to have three versions of a print:
the drawing itself - for manufacturers
the drawing with only (the most) "critical" dimensions - for QA
and a drawing with only two or three major dimensions - for marketing

I think the only issue with this system is the document control concern, but I have not come across any problems yet with the bookkeeping.

 

[pdybeck]

Thanks also to everyone who added their input. I had taken the stance that QA methods should not be placed on the drawing and read some good advice on this thread that solidfies that position. A refernce on the drawing to a separate document where that information is kept is the best method in my opinion, and it seems like most here. Having to revise the drawing for any slight changes in say an OP Sheet, QA document, etc... would drive the cost of the design up with the unnecessary drawing changes without the part/assembly ever really changing. Again, thanks to all involved.

Pete Yodis