Curved surface as datums 1

[Madhu454]

Hi All,

Recently I saw a drawing where the curved surface is defined as datum. But there is no datum targets used in this case, datum feature symbol is directly attached to the curved surface.

Please refer to the attached simplified drawing. The slotted holes in the bracket is located w.r.t datum’s A, B & C with Positional Tol.

I referred to the ASME 1994 Std, the section 4.5.10.1 allows the use of curved surface as datum under " mathematically defined surface concept. It says the theoretically true geometric counterpart of the shape is used to establish the datum.

I am not clear about the concept, I have the following doubts.
1) The inner radius is defined with R875+/- 5 mm and the thickness of the sheet is 3 +/- 0.2 mm. my question is to make use of the mathematically defined surfaces as datum’s do we need to define the surface with the basic dimension?? here in this drawing it is defined with +/- tolerance. is this valid?

Suppose if I defined the surface with Basic dimension, then do I need to tolerance it using profile of a line or profile of a surface?

there is not much explanation about mathematically defined surfaces in the 1994 std., (I have only ASME 1994 std with me , I don’t have 2009 std.)

Can anyone help me in understanding these concepts and to work with this kind of drawings.

Thanks for your time.

 

[Belanger]

The 2009 standard gives a little more info on this idea (paragraph 4.13) and even gives a picture in Fig. 4-28. But notice that it still uses the datum target symbol to clarify datum feature A.

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

 

[fsincox]

Very interesting application, I believe it is obvious that the feature pictured is a correct datum based on what it appears (in my mind) the function would most likely be. I also think the statement was made to cover things just like this. If the mating part is say, another rounded like form I feel you are correct, whether it is shown in the book or not.
Frank

 

[SeasonLee]

Just talked about this topic two months ago.

http://www.eng-tips.com/viewthread.cfm?qid=303763

SeasonLee

 

[drawoh]

Madhu454,

Your curve is a radius, so it is mathematically defined. If it were some free-form thing, I would strongly recommend datum targets.

How accurate is it? It is hard to use sloppy surfaces as datums.

How flexible is it? You may have to read ASME Y14.5M-1994, Section[ ]6.8.1 on free state variation.

JHG

 

[pmarc]

Madhu454,

First of all I would reconsider the usage of directly toleranced radius for R875 because for arcs less than 90 degrees it is really hard to get repeatable measurements of their center and therefore their radii values.

That is why I would recommend basic radius approach, however, if you strongly insist on staying with directly toleranced radius, I think you can still get to mathematical definition of datum feature simulator. In that case if:
- A was specified on RFS (RMB) basis anywhere on the drawing the radius of datum feature simulator would have to be adjustable depending on value of actual radius of bigger arc;
- A was specified on MMC (MMB) basis anywhere on the drawing, the radius of datum feature simulator would have to equal to maximum smaller arc plus maximum thickness of the part (880+3.2=883.2).

Everything would be much easier when the bigger radius was basic, for instance 878, and a profile of surface or a line was applied to it, let's say equal bilateral 2. Then per Y14.5-2009 you could indirectly define size of datum feature simulator A in any feature control frame which uses A as a reference. This could be for example:
- |pos|x.x|A[BSC]|B|C| - means that radius of simulator A equals to 878;
- |pos|x.x|A(M)||B|C| - means that radius of simulator A equals to 879;
- |pos|x.x|A(L)||B|C| - means that radius of simulator A equals to 877;

Or you could do it directly:
- |pos|x.x|A[R878]||B|C|;
- |pos|x.x|A[R879]||B|C|;
- |pos|x.x|A[R877]||B|C|;

Y14.5-2009 is really doing a great job in the area of datum feature simulator size definition.

 

[Madhu454]

Thank you all for explaining in detail and thanks for your time.

 

[MechNorth]

Good response, PMARC.

Jim Sykes, P.Eng, GDTP-S
Profile Services

http://www.profileservices.ca

TecEase, Inc.

http://www.tec-ease.com