Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations SSS148 on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

General Tolerances: Where do you datum from? Please help 1

Status
Not open for further replies.

kahlju

Mechanical
May 12, 2009
18
Hello,

So my questions is for both ASME and JIS. When you have features that are complex or considered not so functionally important and are not dimensioned (dimension line), we use General Tolerances. What is the rule for General Tolerances? Where is the datum for these dimensions (dimension line)? What standard and section of that standard is General Tolerancing rules specified?

Please include the JIS equivalents as my conversation goes to Japan. I am having trouble explaining that general tolerances have legality but they asked me where the datum was and I responded with, "they are the datums already used on the drawing." I need to back this up with evidence. JIS 0405 says, "General Tolerances only apply to dimensions without tolerance." I don't know if this is poor wording or if they are using the second definition of dimension (space).

Thank you,
~Justin
 
Replies continue below

Recommended for you

I don't understand the question in light of your second sentence. General tolerances apply to dimensions that are not directly given a tolerance. But your second sentence asks about things that "are not dimensioned." Are you referring to implied coaxial features? Or a print that appeals to the CAD model for the dimensions?

Maybe that's what your very last sentence is asking -- if so, then we should read "dimensions" in that statement from JIS 0405 as real numbers such as length, width, height, diameter.

John-Paul Belanger
Certified Sr. GD&T Professional
Geometric Learning Systems
 
My question is how to appropriately handle general tolerances for dimensions not on the drawing. For instance, "3D is master" but the drawing has a general tolerance, in this case 3D determines nominal. Where would the general tolerances be measured from? Any point to any point, from the datums, etc.?

Lets do an example where something is not functionally important to check with a gauge so you don't need to put GD & T on the drawing to request quality checking. Say it is a body exterior part that takes many turns and twist and has a lot of curvature. You decide to call out the areas around a hole that you want to control with GD & T for contact. The rest you leave to "3D is master" and put General Tolerancing. Where do you measure the tolerancing from?

Maybe I'm misunderstanding. Does General tolerance only apply to dimensions on the drawing that are just void of tolerances or is it a general tolerance to the nominal condition of all points in respect to one another on the part?
 
JIS 0405 is Japanese equivalent of ISO 2768.

ISO 2768 is dealing with dimensions (on the drawing) without tolerances, not dimensions you query from CAD model.

"For every expert there is an equal and opposite expert"
Arthur C. Clarke Profiles of the future

 
CheckerHater's reply should help -- it's the dimensions on the drawing that are void of tolerances. But as to an implied datum, that may indeed be open to debate. (ISO/JIS doesn't have the Rule #1 that ASME does, so it might not be as big of an issue?)

John-Paul Belanger
Certified Sr. GD&T Professional
Geometric Learning Systems
 
Take a step back.

If all you had was a conventional drawing and a part in front of you and you were asked to inspect the part for conformance to the drawing what do you do?

You measure the features that are given dimensions on the drawing. You compare the measurement against the allowed tolerance (either explicit for that dimension, or the general).

You make judgement on the part based on those measurements and those measurements only.

What you do not do is make measurements between randomly selected features and try to assess if they are "in tolerance".

Why would you change just because there is a 3D model?
 
In ISO 2768 Part 2 dealing with geometrical tolerances, for conditions like parallelism or perpendicularity the longer feature is taken as a datum.

That's the only mention of a datum you can find there.

"For every expert there is an equal and opposite expert"
Arthur C. Clarke Profiles of the future

 
So I guess now it's split into two questions. We've established that JIS 0405/ISO 2768 is only applied to dimensions not specified to a tolerance on the drawing and GD & T is typically used to control the surfaces.

The first Question I have from this applies to objects we don't have a good way to dimension or aren't very super critical but still complex such as a thick dip in your part/stamping. It's possible but not necessary to make a check fixture for it. In terms of the amount this part can vary from the 3D, how do we control it? Is there a rule?

Second question, what does "3D is master" actually mean if there is no rule for general toleranceing 3D? It's on almost all OEM's and supplier's drawings, so what does it actually mean? Is there a rule?

Thank you,
~Justin
 
At Mintjulup, you use the 3D to make the part not the drawing in mass manufacturing. In stamping you use 3D models to make your die not the drawing. Afterwards, you tweak it, cam it, etc. to get in the tolerance zone you need depending on the shape. So what does "3D is master actually mean?"

Thank you,
~Justin

 
Mintjulup,

In fact (not the case for my example above to be clear), 3D data at some OEMs is being replaced like in the link below by removing 3D so there is no transference issues from 3D to 2D. I've seen this for Honda this way before. Since most design is 3D nowadays to begin, it's very advantageous to make the drawing 3D to avoid confusions and actually have capability to show the envelope. They explain the advantages in the last paragraph.

Thank you,
~Justin
 
Thanks for the link to the article Justin, interesting.

However I don't fully agree with their statement:
Machine Design said:
The term "product-definition data" is meant to encapsulate all the 3D annotation needed to document a design in 3D. It is mostly the same information found on a drawing, but it can now include more.

There is already a definition for "Product Definition Data" in Y14.41:
Y14.41 said:
product definition data: denotes the totality of data elements required to completely define a product. Product definition data includes geometry, topology, relationships, tolerance, attributes, and features necessary to completely define a component part or and assembly of parts for the purpose of design, analysis, manufacture, test, and inspection.

Note that inspection is included.

Note further than this definition is is completely independent of the existence or non-existence of a 3D model. Exactly the same information is needed for a drawing done by pencil as by CATIA. The important difference is that a fully annotated model can contain all of this entirely within the model, whereas the pencil drawing may need to be supplemented by other documents.

The point is, if there is a feature that is important enough to warrant inspection then the product definition data must say so, establish the features to measure and the tolerance.
 
I hope I didn't send this off in the wrong direction with my initial reply, but the OP's question was different from all this CAD stuff. I think he was simply asking what the datum should be if you're measuring a non-GD&T dimension. My thinking is that it's up for grabs; that's part of the problem with not employing GD&T.

John-Paul Belanger
Certified Sr. GD&T Professional
Geometric Learning Systems
 
Belanger,

This is true initially but after finding out that only dimensions on the drawing on covered by general tolerancing rules, I wanted to know if there is any rule for generally tolerancing dimensions that were left out of the drawing but relating it to the 3D. For example, complex contours that would be financially unnecessary to gauge with a check fixture (GD&T). This also made me wonder what it meant to write "3D is master" on a 2D drawing if general tolerancing is not the purpose. Is this a pointless note?

Thank you,
~Justin
 
Not a direct answer to your question but...

On the odd ocasion I use partial/hybrid MBD I have a note referencing the model for 'basic geometry' and then for 'general tolerance' put in a surface profile feature control frame that I usually put datum in (and obviously indicated the datum features on the drawing views).

Posting guidelines faq731-376 (probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484
 
"3D is master" means that the model is the source of theoretically exact dimensions. Dimensions - not tolerances.

It is your job to specify dimensions and geometric tolerances to your drawing / model. One of the possible ways was described by KENAT - use profile control.

The only "rule" you can find is "Definitive drawing" from ISO 8015:

5.3 Definitive drawing principle
The drawing is definitive. All specifications shall be indicated on the drawing using GPS symbology (with or without specification modifiers), associated default rules or special rules and references to related documentation, e.g. regional, national or company standards. Consequently, requirements not specified on the drawing cannot be enforced.


It's very simple - there are no hidden rules - only the rules you specify on the drawing.


"For every expert there is an equal and opposite expert"
Arthur C. Clarke Profiles of the future

 
I've gotten design packages from customers who give a general tolerance in terms of GD&T but without a datum reference. They'll have a note in their phrasing which basically means "3D is master" and general tolerances will include something to the effect of "subject to a profile tolerance zone of .060 with respect to the CAD data" which is essentially useless unless you /assume/ datums. Which may be ok. For the type of work we do, it's low quantity, skilled trade work that requires some ability to make judgment calls in fabrication so... it's 'safe'... though maybe not all that defensible in cases of disagreement.

But I always have been curious to see how others handle such things.
 
@ JNieman:
It's not the worst thing. As long as you have mathematically defined surface, you can apply profile tolerance to it.

If you look into any GD&T book, profile can be defined with or without datum(s), so not illegal either.

"For every expert there is an equal and opposite expert"
Arthur C. Clarke Profiles of the future

 
Like CH implys, specifying the datums makes it easier for subsequent inspection and probably captures function better if you're doing it right. However, having a general surface profile tolerance without datums probably isn't actually 'wrong' just harder to work with down stream.

Posting guidelines faq731-376 (probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484
 
It's essentially the profile "all over" idea.

John-Paul Belanger
Certified Sr. GD&T Professional
Geometric Learning Systems
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor