GD&T with general tolerances between datums

[PetkovStoyan]

Hi all,

I am trying to compose tolerance callout for the part attached. The idea is that overall dimension can vary more than position of holes relative to datums B and C, and I wonder how to properly annotate it. I am wondering should dimension 1600+-3 be basic? Should I use composite callout for datums B and C? Maybe I have other mistakes, please let me know if you find any. UF stands for "united feature", I think this is the equivalent for CF in ASME

https://files.engineering.com/getfi...af58-2a72088c9283&file=GD_T_Example062023.pdf

 

[greenimi]

Just tossing it out there, maybe another option is to use CZ inside the tolerance indicator.
Can you tell us how this part functions, as datum features B and C seems a little strange.

 

[Burunduk]

Datum features B and C are not referenced in the same tolerance indicator (tolerance frame) so each datum - B and C, will be established normal to datum A, but independently of each other or of the length between those features. The length dimension doesn't have to be specified as TED. It can be ±3 without issues.

 

[PetkovStoyan]

greenimi,

this is part of drag link conveyor, see updated sketch attached. Datum A slides on guiding plates, and the holes locate and fasten the chain link. Initially it made more sense to me to use the holes as datums, but then I had to change the dimension scheme and include the distance 1560 between the holes. And it was confusing, as this dimension is part of the holes location. So I couldn't realize how to callout tolerance Ø1mm on holes location, but in the same time to allow the distance 1560 to vary more than that, i.e. to have larger tolerance. So I used the sides for datums instead to avoid this.

https://files.engineering.com/getfi...0&file=2023-06-12_GD_T_Skrepka_TSH2-Model.pdf

 

[3DDave]

That is double dimensioning and is not a good practice. It is identified by having multiple dimension chains locating the same feature(s).

I would make the holes a pattern, locations described with basic and use them as a datum feature to locate the profile of the part. The primary feature should be the face of the part; that is the face one would set onto a mill or drill press to orient the part before drilling the holes.

 

[PetkovStoyan]

3DDave,

can you explain exactly what it means to "make the holes a pattern"? As far as I can understand, this means to make dimension 1560 basic, right? And this is exactly my problem, I don't want this dimension to be constrained with the same tolerances as the other hole-to-hole and hole-to-datum dimensions

 

[3DDave]

If you don't want the control remove the dimension and its tolerance.

can you explain exactly what it means to "make the holes a pattern"?

That is explained in the standard.

 

[PetkovStoyan]

It is not that I don't want it, but rather I want it to have different tolerance compared to the other hole locating dimensions

 

[3DDave]

You shouldn't locate them twice. Perhaps bidirectional positional tolerancing? Nothing in the standard about having overlapping tolerance zones; some experts recommend them.

 

[jassco]

Hi, PetkovStoyan:

As 3DDave indicated, you double dimensioned. That is no-no.

If I were you, I would make the large face (far side) as datum A. Label 4 holes as a pattern.

Then you can add profile tolerances to edges of the part.

If you want to use sides of the plate as a datum, you will need to use centerline of feature 1600 (BASIC) as datum B. After all, your part is symmetric. Datum C will be your current datum A. Then you can use bi-directional positional tolerances.

Best regards,

Alex

 

[PetkovStoyan]

Sure, this is not how the part will be toleranced. I put the dimension on the drawing for the others to know which dimension I am reffering to when I write "dimension 1560", just for the sake of the discussion.

3DDave,

Thank you for the tip about bidirectional location tolerances, I wasn't aware of that. I think this answers my question exactly to the point.
But just of curiosity, will it be considered mistake if I make all hole locating dimensions basic, except dimension 1560, giving it tolerance +-3? I suppose most people will find this annotation more meaningful than the bidirectional tolerance, even though it might be invalid.

 

[3DDave]

It would be considered a mistake, but I am not your boss, so I won't fire you.

 

[Burunduk]

Your initial version made sense.
Each pair of holes was related to a separate datum system, each datum system was based on one end of the part. If the threaded holes on the mating part are toleranced similarly, it would be functionally correct.
The 1560±3 between holes totally ruined it.
What do you need it for?

 

[PetkovStoyan]

I just added the dimension for reference in order to discuss different dimension schemes, just for the discussion. I will not dimension this way the part, it doesn't make sense, I am aware of that

 

[Wuzhee]

Don't add dimensions to discuss differences or clarify something that is already clear. You'll reach the exact opposite. visible confusion