Dimensioning Tips

[lecuyero]

I have a question regarding proper dimensioning of a small part. This miniature part has 6 "sharp" tips that must lie within a plane parallel to datum A. The overall location of the plane from the datum is not a big concern. However, the 6 tips must be within .02 mm of each other relative to datum A. I do not believe I am constraining these tips correctly in the drawing. I am calling out a FCF for a parallel plane but how am I to show that all 6 tips (not just the 3 needed to establish the plane) need to be within these tolerances?

I am thinking I need to remove the "6X " from the length but I know more must be done after that...

 

[KENAT]

Normally for similar situations where you have a number of surfaces that need to be coplanar you'd use profile of a surface as shown in ASME Y14.5M-1994 section 6.5.6.

Not quite so sure on the 'sharp points' but I'm guessing that would be the fundamental approach.

What drawing standards do you work to out of interest and to help others trying to answer the question?

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[lecuyero]

Our title blocks only says ISO-2768-1989 and that would be the only thing a manufacturer would be able to go by from looking at our drawings.

Officially our handbook says we use ISO 129, 406, 1101, and ANSI Y14.5M-1982 but everybody here uses the 1994 standards.

 

[looslib]

ISO-2768 requires modifiers when used as a spec reference.


"Wildfires are dangerous, hard to control, and economically catastrophic."

Ben Loosli

 

[SteveMartin]

I can't think of a good way to make this constraint symbolically (without tightly constraining other things you don't appear to need tight control over). I would write out the requirement in a note. GTOLs can be great for making consistent and interpretable drawings, but they cannot cover everything.

 

[KENAT]

lecuyero , do you have Y14.5M-1994 to hand. If so look at 6.5.6.

I think you could probably use the same approach but rather than '6X surfaces' say '6X PEAKS' or similar. As Steve says you may need a bit more than just an FCF but I think the approach may be OK.

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[ptruitt]

'Profile of a surface' can control almost anything, (and probably this part) but more views of the part would be needed to suggest specific profile call-outs.

Peter Truitt
Minnesota

 

[prdave00]

Maybe this is too far out of the box, but what it you switch datum A to target datums where each of the 6 points is a point target and then check the parallelism of the flat face (formerly datum A). Then use a surface profile tolerance without any datums in the FCF to control the form of the surfaces that come to points.

Maybe I'm creating the cardinal sin of choosing datums based on inspection methodology and not function, but you thought I'd put it out there anyway.

 

[KENAT]

If I understand you, would six planar points to create a datum be appropriate/correct?

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[prdave00]

Probably not since you can never gaurantee all 6 points are coplaner. Maybe I should have thought about my response for another 10 minutes. Now I'm a bit embarrassed.

 

[SteveMartin]

prdave,

Six datum targets (one/point) won't assure the six points are coplanar/nearly so. While a primary datum is defined by 'at least three target points or areas' and can therefore I presume legally use more than three, A physical datum simulator will only end up picking up the three highest points - the other three could be well off the plane established by the first three. Defining a plane by four or more points is overdefining it.

kenat,

I'd be hesitant to extrapolate 6.5.6 (which deals with nominally flat surfaces) to this example (which doesn't). In this example there's no clear demarcation of what wouldn't have to stay within the profile zone. If the tips were flattenned, sure; but I don't know about coplanarity of points or of surfaces that aren't planar to themselves (depending on whether or not you discount the .03 MAX radius).

I don't like profile of a surface unless the OP is interested in(or at least accepting of) tight control over more than just the tip of the points - ie some portion of the surfaces that define the points. My interpretation of the original post is that the OP is only interested in controlling the six individual high points. Points by definition are not surfaces.

 

[KENAT]

I still think using the idea/concept of the coplanarity application of surface profile makes the most sense. However, I already conceded you may need some additional explanation, perhaps even more than just "6 PEAKS".

Perhaps all you do is try and copy the terminology associated with surface profile in this application rather than trying to use a surface profile FCF at all. Or maybe you have the FCF, with '6 PEAKS' and a flag to a note giving an explanation that the tips of the peaks are to be coplaner within the limits indicated, or some such.

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[DeanD3W]

As has been mentioned already, those points are a problem. GD&T tolerances apply to surfaces, not edges or points between surfaces. Adding "6X PEAKS" or "6X POINTS" should overcome that problem well enough. The parallelism you show doesn't work, since profile is needed to control coplanarity.

You mention that all 6 points must be within 0.02 wrt datum A, so you also need more than just coplanarity. If you reference datum feature A in order to get the needed orientation constraint on the profile call-out you will need then you normally also get the location control wrt A, which is more than you need. So, I think you're left with either a note applied to the profile call-out to release the translational degree of freedom that A constrains or you need to use the "Customized datum reference frame" approach from section 4.22 of ASME Y14.5-2009... If you set up datum reference frame coordinate axes with Z normal to datum A, for instance, then the datum feature reference for the profile call-out would be [A(u,v)]. Then you get the rotational constraints you need from datum feature A with the power to control coplanarity that the profile of a surface call-out will provide.

To measure those points, maybe a vision system could be used with datum feature A against a nice angle block and the part indexed so all six points can be seen.

I hope this helps.

Dean

http://www.d3w.engineering.com