Perpendicularity of an Imaginary Line

[Lee Saxon]

This seems like a pretty straightforward design case if you're trying to orient but not necessarily locate two parts, but it's not something I've ever found a great example for.

I've got two holes. An imaginary line drawn between their centers would need to be *oriented* perpendicular to a flat surface Datum A. But the holes aren't *located* to anything with any precision (except each other).

http://files.engineering.com/getfil...a23-4772-829b-f61927cb8d5c&file=Imaginary.pdf

 

[CheckerHater]

I am having hard time visualizing your problem.
Maybe little picture will help?

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

 

[Lee Saxon]

Good idea, I've added a PDF of a simple example.

 

[Belanger]

CH -- see attached sketch. I presume that the holes are not coaxial. (The OP says that need to be located to each other, but offset from each other or coaxial?)

One possible work-around is to make the holes the datum and then have the surface be toleranced for perpendicularity from the holes; see attached. This bypasses the problem of creating a location tolerance between the holes and the surface. The holes probably need a datum reference yet I hesitate to use the back face as a datum because that brings in another element that wasn't asked for in the question.

Another possibility is to make the surface the datum feature and then use profile of a surface on the two holes, referencing that datum. Any location dims of the hole would not be basic but just toleranced dimensions. (After all, you still need to give some relationship there.) The downside to this method is that profile would also control the form and size of the two holes, which you aren't interested in.

Anyhow, there's some stuff to get the discussion going.

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

 

[CheckerHater]

Thinking out loud:
If you apply perpendicularity without diameter symbol to both holes that will mean that axis of every one of them resides in plane, perpendicular to A.
As holes are controlled by the FCF, basic zero dimension between the two implies as well.
That brings us very close to the solution.
Ah, Friday

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

 

[mkcski]

Lee:

The

But the holes aren't *located* to anything with any precision (except each other).

I am having trouble with avoiding the use of position control due to the lack of "precision". I suggest specifying more datums and use composite position and refine the perpendicularity to A in the lower FCF. The upper FCF could be MUCH larger than the lower - .500 vs .010 inch.

Certified Sr. GD&T Professional

 

[axym]

I agree with mkcski's suggestion. Composite position would be the appropriate method here.

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[Belanger]

I agree too, but notice that the OP doesn't want to locate the holes. So it's sort of an academic discussion -- how to get that perp idea but not location.

Any comments on my post and sketch?

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

 

[mkcski]

J-P:

How can you make the part if you don't locate the holes? Location needs dimension and tolerance.

Certified Sr. GD&T Professional

 

[Belanger]

I get that, but let's not think of this as a real part. There are plenty of examples in the standard that illustrate some point without covering all the proper bases that you'd need in a real design.
Treat it only as a thought experiment for how to control perpendicularity of the imaginary center plane of two holes relative to the surface (or vice versa), yet not touching the idea of location. Dicey, but fun!

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

 

[mkcski]

J-P:

OK. If it's not a real part and just a case study, then I could accept your sketch - reversing the relationship between the holes and the surface.

Lee:

Just to be clear, GDT controls physical features and not theoretical construction (center) lines.

Certified Sr. GD&T Professional

 

[Nescius]

Would it be considered cheating to use a customized datum reference frame?

 

[greenimi]

Would it be considered cheating to use a customized datum reference frame?

Do you really need it?

Can you use a plier to drive the screws/ nuts? Yes. It is a good idea? No.
Can you use a caliper to hammer the nails? Yes. It is a good idea? No.


It is not a proper tool for the job. But we actually don’t know the job.

 

[Nescius]

Why would it not be a proper tool for the job? See sections 4.22 and 4.23 in the 2009 standard.

My asking if it would be considered cheating was tongue-in-cheek, and meant to subtly admit that it's a direct solution, but not an artful one.

 

[mkcski]

The OP needs to chime-in so we can get some clarity for more posts. Lee?

Certified Sr. GD&T Professional

 

[axym]

J-P,

I agree that it's a bit of an academic discussion. Y14.5 doesn't have a tool to control the orientation of the features in a pattern without controlling their location (only the lower segment of a composite FCF will do it).

Here are some comments on your sketch:

-The perpendicularity tolerance on the planar surface would require it to be flat within the same tolerance

-I'm not sure that the tolerance on the planar surface could be called perpendicularity. The planar surface is nominally parallel to the axes of the two holes, so I think it would be called parallelism.

-Having the holes as the datum feature, referenced MMB, introduces datum feature shift. I believe this would allow more angular variation between the planar surface and each hole axis, than there would be if the planar surface was the datum feature.

Nescius,

I hadn't thought of the customized datum reference frame. That would do the job as well. The FCF would be |POS|dia 0.1|A[u,v]|

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[Belanger]

Good point, Evan -- I forgot to include the MMB modifier. But yes, it sounds like the customized datum reference frame would be your answer, Lee!

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

 

[pylfrm]

Lee Saxon,

In addition to providing the hole pattern orientation control you desire, a composite position tolerance or a regular position tolerance with a customized datum reference frame will also control the distance between the holes to the same tolerance value. If this is undesired, then I would suggest using the bidirectional positional tolerancing method described in ASME Y14.5-2009 para. 7.4.4 and illustrated in Fig. 7-28. Since you are only concerned with control in one direction, you could skip the tolerance in the other direction. It might not technically be bidirectional anymore, but I think the principle should apply just the same.

Attached is a modified version of Fig. 7-28 illustrating what I propose.


pylfrm