Inspectors failing parts due to Reference Dimensions 2

[Sparweb]

Have I forgotten something important about dimensioning? Am I not reading ASME Y14.5 correctly?

I have an inspector who won't accept a part because it doesn't conform to a reference dimension on a drawing.
This has happened many times to my colleagues and they have been revising drawings to suit.
Now it's one of my drawings, and it seems someone has already removed some reference dimensions but apparently not enough, yet.
FYI it is a machined part with a sequence of holes to be drilled/csk that are centered on pre-existing holes in the part.
The new holes are fully dimensioned linearly, but I included reference data about the existing holes in case the fabricator wants to just pick up on them, or to check that the new holes line up.
The part would fit & function just fine no matter how the machining job is set up.

1.3.24 Dimension, Reference
dimension, reference: a dimension, usually without a tolerance, that
is used for informational purposes only.

NOTE: A reference dimension is a repeat of a dimension or is
derived from other values shown on the drawing or on related
drawings. It is considered auxiliary information and does not govern
production or inspection operations. See Figs. 1-19 and 1-20.
Where a basic dimension is repeated on a drawing, it need not be
identified as reference. For information on how to indicate a reference
dimension, see para 1.7.6.

It's not the first time I've tried to get this inspector to stop inspecting references. Before I go squeal to his boss I just want to be sure I'm on solid ground.
(Yeah, I guess someone is going to say: "don't put reference dimensions on drawings..." But I use them when I sincerely believe they will help a fabricator set up or locate a tool.)


STF

 

[Sparweb]

Hi CWB1,
Some of the things you're asking about are not always so explicitly defined that one can rub a guy's nose in the letter of the law.
Our company standard doesn't define reference dimensions because it refers to the ASME Y14 instead - our designers should be familiar with that stuff FIRST before getting into the peculiarities of our company drawings (styles and preferred practices based on the things we typically work on and put on drawings).

My last comments may have sounded a bit prejudiced but rest assured the inspector and I had a respectful and collegial discussion about how to interpret this drawing. I purposefully kept the tone light while still driving home my argument because I need to have his ears open the NEXT time the issue comes up. I can tell that it will come up again, and that's what was getting my goat when I posted on Thursday.

...The shop should have a purchase-level print and already know the info you are adding as a reference on your modification-level. If they decide to combine the info via a shop markup or manufacturing print that is their responsibility, design should NOT be involved in this effort nor should design prints ever devolve to include manufacturing details. As a general rule of thumb, keep design engineers out of the shop and manufacturing engineers out of the design office as very few have the experience necessary to properly fill the other's role...

Here, I can't support what you're saying at all. Many of the assumptions built into your statement are false in this case and/or are poor as a general rule. Except perhaps "...design prints ever devolve to include manufacturing details..." and I see where you're going with that, but I rather insist that designers get into the shop and understand the tools that are used. That must be understood before designing any part. We can't all design like we have a 3D printer on our desk.

STF

 

[gabimo]

I agree also with "As a general rule of thumb, keep design engineers out of the shop and manufacturing engineers out of the design office as very few have the experience necessary to properly fill the other's role. "

Make a clear difference and distinction between PRODUCT drawing and a PROCESS drawing.

Copy-paste from a different thread

They are simply two different drawings: One is a PRODUCT drawing (goals to achieve and not instructions on how to make the part) and the other is a PROCESS drawing (the PROCESS requirements are the instructions *for one particular supplier's chosen method* to achieve those goals that were stated on the PRODUCT drawing.)

The PRODUCT drawing is the legal requirement and obligation of the supplier. The PROCESS drawing is the chosen method for a given supplier, but another supplier could chose a different process to achieve the same PRODUCT requirements

Product drawing is controlled by Design.
Process drawing is controlled by Manufacturing.

 

[drawoh]

Nescius,

I agree with almost everything what you are saying, however, remember that Y14.5 is a voluntary standard.

...

I am not sure I understand "voluntary" in the context of ASME[ ]Y14.5. The standard explains what all the symbols and notes on your drawing mean. You can prepare your drawing any way you damn well please.

I believe there is an old joke about somebody going to a French restaurant and ordering in French to impress his lady friend.

--
JHG

 

[Nescius]

Gabimot,

Regarding standards, I agree. However, there has to be some unspoken common ground. This goes far beyond line styles, or colors, extensions of principles, or any number of adaptive quirks that an entity might institute to make their lives easier. Company policy or not, the things that CWB1 described cannot be reconciled. It boils down to the inspectors demanding the part satisfy more than a fully-defining set of the dimensions: the ones shown AND the ones they are inventing.

Edited: reworded for clarity.

 

[gabimo]

Re: Quote drawoh: "The standard explains what all the symbols and notes on your drawing mean"

But the meaning can be driven by your own company standard and not by ISO or ASME.

It is not a building code/standard or an electrical code....................It is not a MANDATORY standard.

 

[mkcski]

I have been following this thread - please let me add my two cents.

To avoid any confusion with being "voluntary" or not, our drawings state: Interpret per Y14.5-2009. We take great effort to "stick with standard". We consider it a "dictionary' of how to "read" the drawing symbology. Now you can put anything on a drawing, and you can make up symbols all you want, but if there is no common understanding from a "dictionary" you are fooling yourself.



Certified Sr. GD&T Professional

 

[CWB1]

Nescius,

Nonsense. Nobody gets to ignore fundamental rules about drawing conventions because they are "not governed by ASME".

Fundamental rules? That's a rather comical notion given the ambiguities present in many society "standards." In this case there's a decent example - do reference dims have a tolerance or not? Per ASME either is allowable. Beyond that, tradesmen typically arent given access to society standards and are often trained on the job with minimal excess, so freely distributable company standards are typically created to resolve any ambiguities and confusion in these matters. Admittedly, often they are little more than a detailed rewrite of a society standard with clarification added, however they help drive standard process and avoid confusion. Not bashing anyone in the least but had the OP had a company standard that covered interpretation of reference dims he could've resolved this in ~1 minute if he was even called.

What is a "suggested inspection area"? That is another arbitrary interpretation of a reference dimension.

Its one which designers feel inspection should focus on due to non-critical quality concerns, often the most likely failed non-critical dim. Like anything else production-related on a design print, its purely a suggestion.

Please tell me the inspectors aren't calculating a dimension not shown on a print, then measuring it and applying the default tolerance to it.

Please tell me you don't believe parts are only inspected via dimensions explicitly called out on a print? In reality inspectors calculate many of the infinite number of inspection dims and points using a combination of the given dims and GD&T.


Sparweb,

Here, I can't support what you're saying at all. Many of the assumptions built into your statement are false in this case and/or are poor as a general rule. Except perhaps "...design prints ever devolve to include manufacturing details..." and I see where you're going with that, but I rather insist that designers get into the shop and understand the tools that are used. That must be understood before designing any part. We can't all design like we have a 3D printer on our desk.

Glad to hear you resolved the issue but out of curiosity what assumptions of mine are wrong in that paragraph? Do shop supervisors and manufacturing engineers not have access to a concise structure of prints starting with a purchase level and ending with the highest locally modified level? If not then I would highly encourage improving standards, print structure, and process to eliminate future confusion. Arm the shop with proper tools and everyone becomes more efficient. As someone that grew up on and spent the first decade of his working adult life on the shop floor before attending college I can attest that you can never communicate too clearly nor be too organized. While I also wholly support putting design engineers to work in the shop (esp when assembling prototypes), when it comes to setting up production in the skilled trades there is no replacement for experience. Unless they've attended trade school or spent years of their life in the shop, a design engineer will never be remotely knowledgable enough to spec many manufacturing details so there really is need to keep them, their process, and their prints separate. Many will disagree but IMHO that is well beyond arrogant and usually drives production costs unnecessarily high.

 

[Nescius]

CWB1,

You can defend the position with a shield of "internal policy", but that's not much of a defense when you're dealing with the outside world. The policy didn't specify a special symbol or similar device to communicate your requirement; you hijacked an existing concept.

If a reference dimension has a tolerance, and the part is fully defined with non-reference dimensions, the part is over-defined. The part could satisfy all non-reference dimension tolerances (as it should) but still fail to meet the bogus tolerance on a reference dimension.

Perhaps your drawings don't fully define the part with non-reference dimensions...the references are "needed". In that case, you've simply applied an existing symbology to some dimensions and deemed them "reference" dimensions.

Please tell me the inspectors aren't calculating a dimension not shown on a print, then measuring it and applying the default tolerance to it.

Please tell me you don't believe parts are only inspected via dimensions explicitly called out on a print? In reality inspectors calculate many of the infinite number of inspection dims and points using a combination of the given dims and GD&T.

You didn't answer my question. Of course, inspectors may calculate any number of dimensions to, let's say, build a path to get where they need to go. The point is, NONE of these calculated dimensions are judged against a tolerance.

 

[drawoh]

Nescius,

The problem is that everyone needs to know how to interpret stuff on the drawing. When you specify that the drawing is to be interpreted as per ASME Y14.5-2009, you provide a reference that everybody can acquire and look at. There is nothing to stop Dominion Consolidated Widgets Incorporated from creating their own dimensioning and tolerancing standard, and calling that up. They are doing a great deal of work that, on the whole, will confuse everyone they work with.

--
JHG

 

[Nescius]

Drawoh,

I agree 100%. To be clear, when I say:

If a reference dimension has a tolerance, and the part is fully defined with non-reference dimensions, the part is over-defined. The part could satisfy all non-reference dimension tolerances (as it should) but still fail to meet the bogus tolerance on a reference dimension.

Perhaps your drawings don't fully define the part with non-reference dimensions...the references are "needed". In that case, you've simply applied an existing symbology to some dimensions and deemed them "reference" dimensions.

I'm highlighting the insanity of that policy, not trying to rationalize it.

 

[Sparweb]

Hi CWB1

Glad to hear you resolved the issue but out of curiosity what assumptions of mine are wrong in that paragraph?

This one:

...keep design engineers out of the shop and manufacturing engineers out of the design office...

I think you have grossly oversimplified it to the point of being useless advice. Perhaps the tasks done by design engineers should not be seen as equivalent to the tasks done by production engineers, but there should ALWAYS be a conversation between them. If a design engineer cannot draw a part that can be economically manufactured, when a different design would both work and be easy to fabricate, then NO AMOUNT OF PRODUCTION ENGINEERING CAN FIX IT. The design engineer must have a means of fabrication in his/her head at all times while designing, otherwise you end up with bags on the sides of boxes and fasteners inside cavities with 1" diameter access holes!

Do shop supervisors and manufacturing engineers not have access to a concise structure of prints starting with a purchase level and ending with the highest locally modified level?

This is one of the problem assumptions. No. This shop makes what appears on the engineering drawings, and engineering is not involved in preparing production drawings, if any - in the particular shop I am dealing with. My company has other facilities that DO make production drawings based on customer drawings, but that's not the case in the shop I am sending this current batch of drawings to. There is a "Production Planner" at work in this shop who does a lot of interpretation between engineering and the shop, but the department as a whole discourages making any subordinate drawings at all (and the opposite of the other facilities we have - it's quite the split-personality problem).

As someone that grew up on and spent the first decade of his working adult life on the shop floor before attending college I can attest that you can never communicate too clearly nor be too organized. While I also wholly support putting design engineers to work in the shop (esp when assembling prototypes), when it comes to setting up production in the skilled trades there is no replacement for experience.
Unless they've attended trade school or spent years of their life in the shop, a design engineer will never be remotely knowledgable enough to spec many manufacturing details so there really is need to keep them, their process, and their prints separate. Many will disagree but IMHO that is well beyond arrogant and usually drives production costs unnecessarily high.

Like you, I spent just enough time in the shop to be truly dangerous, then learned that my calling is the design office! Since it is an asset to my experience, why shouldn't I encourage the junior designers to learn in similar fashion? I can't hold their hand every step of the way - they have to learn enough to find their own solutions, but that knowledge is not just in the engineering office. It's also downstairs in the minds of the technicians. Furthermore, I think your second and third statements are contradictory. Anyway, this doesn't really pertain to the inspection problems, which is what I see as the root of this particular problem.

STF

 

[CWB1]

You can defend the position with a shield of "internal policy", but that's not much of a defense when you're dealing with the outside world. The policy didn't specify a special symbol or similar device to communicate your requirement; you hijacked an existing concept.

Reread the post. I'm not defending tolerancing reference dims as internal policy, I'm saying every company should have internal standards that clarify the various "ors" and "mays" within society standards, the "or" in the Y14.5 quote above being just one example. The use of tolerance on reference dims needs no defense as its not uncommon.

As for the outside world, I authorize POs exceeding $10M annually, grew up in the family job shop, worked in several others, and have spent the last decade in large corp design working internationally. Not sure where this other "outside world" you speak of is but I'm often called on to bridge the communication gap between design and manufacturing, often due to lack of process and standards.

If a reference dimension has a tolerance, and the part is fully defined with non-reference dimensions, the part is over-defined. The part could satisfy all non-reference dimension tolerances (as it should) but still fail to meet the bogus tolerance on a reference dimension.

No, if something is on a print it MUST be correct or the print isn't valid, every person who signed off on said print has failed, and the shop should kick it back for simple stupidity. I can duplicate or show excess reference dims ad-infinitum so long as any applied tolerances are correct. The fact that many people screw this up is likely why so many don't apply tolerances to reference dims, but given the past few decades of CAD availability there's really no excuse for not being able to do this work correctly.

You didn't answer my question. Of course, inspectors may calculate any number of dimensions to, let's say, build a path to get where they need to go. The point is, NONE of these calculated dimensions are judged against a tolerance.

It was a silly question, obviously no. Regardless, youre incorrect again. Every point and dimension measured is viewed within its acceptable tolerance, hence the main purpose of GD&T - to establish tolerance/variance for the infinite number of inspection points via MMC and LMC regardless of the surface's complexity. Not every inspection point is explicitly called out on a print nor does standard tolerance apply to calculated/extrapolated dims.

 

[Nescius]

I can duplicate or show excess reference dims ad-infinitum so long as any applied tolerances are correct. The fact that many people screw this up is likely why so many don't apply tolerances to reference dims

OK, good luck. I'm not doubting your experience or your value at your current position, or anything like that. What I am saying is that you have a really, really non-standard notion of what "reference" dimensions are. Take it or leave it; I'm just some dude on the internet.

 

[CWB1]

SparWeb, I agree with most of your last post and think we're on the same page without realizing it. My point isn't that design engineers should stay out of the shop entirely nor even that they shouldn't be involved in manufacturing related discussion when necessary, simply that they need to be very careful to give the manufacturing team reign to make and inspect the part their way while refusing to enable their laziness. On the design side we own the design, but everything beyond that needs to be the shop's responsibility. I'll gladly work with them occasionally on minor design/print tweaks and review deviations but if they want changes to accommodate their process or add non-design reference info its on them. Rather than your specific reference diameters, I'd have ensured the shop had access to the print hierarchy from purchase-level up to your print (without the references) and called it done.