GD&T on a Hex Bar 2

[HightechHillbilly]

Howdy,

I am really at a loss as to how to GD&T a hex bar. In my research, I've found that it's classified as an "Irregular Feature of Size" and that most ppl tend to favor a profile tolerance.

Attached is what I've started, but I've hit a roadblock because I'm not even sure if I'm on the right track.

The functionality of the bar is that it also has 2 "dimples" on opposing flats, which act as friction retention in a press fit application when pressed into a blind hex hole. The surface that I have tentatively labeled Datum "A" will contact the surface at the end of the blind hex hole. Obviously, the shape of the hex itself will keep it from rotating, and the "dimples" will deform when the hex bar is pressed in, causing a tight friction press fit such that it cannot slide out of the hole.

So, not only is my issue in GD&T'ing the hex bar itself, but also in relating the 2 dimples - I'm not sure if what I'm calling Datum "A" is correct, and/or if I need any other datums, and how to relate everything.

Any assistance will be greatly appreciated.

Thanks,
Rob

 

[drawoh]

HightechHillbilly,

On your end view, don't bother with the point to point dimension. Show the width with 3X, and show an angle, again, 3X.

The datums depend a lot on your requirements. Your datum[ ]A looks legal, interpretable and reasonable to me. Your datum[ ]B could be one or both faces the hole goes through. Your datum[ ]C could be one or both of the adjacent faces. Your hole requires all three datums for the positional tolerance.

At some point, you are better off if your end face is a tertiary datum rather than the primary one.

--
JHG

 

[HightechHillbilly]

OK, so I've done some more work on this, and came up w/ 2 other paths (see attached).

I was wondering if someone could please let me know which of the 3 paths (my 1st post being the 1st path, and this post being paths 2 & 3) is the correct one? If none are correct, could someone please set me on the proper path?

Thank you,
Rob

 

[HightechHillbilly]

drawoh,

Looks like we must've posted at around the same time!! heh...

Anyway, the "hole" you speak of is not a through hole, but is really 2 "dimples" - and what that means is that a punch is used to deform the material twice (one on one side, and the 2nd on the opposing side). What the "dimples" do is create a small cavity (indentation), on each face, but the material that is deformed to form the cavity spreads out & up to make sort of a "volcano" of material jutting out on each side. You can see the "jutting" out of material on each side via the end view.

All those dimples are used for is to provide a friction fit when the hex bar is pressed into a hex hole, as the dimples deform to keep the hex bar from falling out of the hole.

So based on your comments, it looks like what I need to do is sort of a combination of my Figure 2 & Figure 3. So, I've attached yet a 4th path (figure), based on your comments... hopefully this is the correct one.

However, I still have some questions:

1.) I don't think I have Datum "B" correct (or perhaps even Datum "A") - because doesn't the definition as I have it listed mean the center-axis, as opposed to an actual face? And if that's the case, then how would I define the Datum A & B - by a Continuous Feature (meaning, I want 2 opposing faces to be 1 datum)?

2.) Would I just be able to set the profile of the hex itself as Datum "A"?

Thanks,
Rob

 

[Tunalover]

Is the hex bar fabricated from scratch? If not then STOCK dimensions and tolerances apply.


Tunalover

 

[HightechHillbilly]

Tunalover,

Yes, it's made from scratch, as it has a specific tolerance needed to fit the hex hole on the other piece.

That being said, is what I've done in Figure 4 (my last attachment) on the right track? Also, any thoughts on the 2 questions I asked above?

Thank you,
Rob

 

[Tunalover]

Rob,
It looks complicated but I think what you have is sufficient. I don't think you need the EQ SP notes. Also I think the multiplier on the angle dimensions should be 6. Otherwise, when you call out the feature size (what is it, a hole?) of course you want to attach the positional tolerance feature control frame to that dimension.


Tunalover

 

[HightechHillbilly]

Tunalover,

Thanks for the feedback, but what do you mean it "looks complicated?" What looks complicated? My GD&T? Is it overly complicated?

My main issue is the placement of the Datums - I just arbitrarily chose Datum "B" - I could've very easily chosen the other 2 opposing faces.

Why would I not choose the "points" for the 2nd datum (what I chose in my original attachment)?

And the "holes" you are referring to are the "dimples" I described earlier. And yes, when I officially call it out on the actual drawing, I'll attach the Feature Control Frame (FCF) properly - for these attachments in this thread, I was just using Powerpoint and making my own FCF.

Thank you for your assistance,
Rob

 

[Tunalover]

Rob,
Your approach is sufficient. Sometimes sufficient means simple, sometimes sufficient means complicated. Sometimes when the GD&T is sufficient it looks complicated. Your choice of datums is just fine too. I think you've got it right. Sorry I didn't realize the feature was a dimple as I did not look at your original attachment.
Regards,
TL



Tunalover

 

[HightechHillbilly]

OK, thanks for the clarification re: "complicated"

However, I'm still curious about Datum "B" - as I said, I just arbitrarily chose those 2 faces, I could've just as easily chosen the opposing 2 faces.

Also, I thought that the definition of the "3 Plane System" used in GD&T was 3 mutually perpendicular planes (ie, at 90 degrees to one another). If that's the case, then Datum "B" fails that requirement.

Which brings me back to my previous question - why can I not use the "points" that I had chosen before as Datum "B"?

Sorry that I keep beating this horse, it's just that I want to make sure this is correct & I understand it.

Thank you,
Rob

 

[Tunalover]

Just imagine the part being clamped across datum A and clamped across datum B. Is the part then completely constrained against motion in that "x-y" plane? The answer is yes so those two datums together are sufficient.
Regards,
TL


Tunalover

 

[Tunalover]

Rob,
Actually I should have worded that "with both datum A and B constrained against normal translation the part is fully constrained against motion in the plane of the page." The final constraint, as you know, is datum C constraining the part against motion normal to the plane of the page. If Mr. Crawford or Mr. Foster are following this maybe they can word it better.



Tunalover

 

[drawoh]

HightechHillbilly,

Your third example is close to what I was suggesting before.

Your datum[ ]B across the points is not very good. Orthogonal datums are desirable, but not absolutely necessary. On the real part, those points will be rounded a bit. It is hard to predict what the actual measurement will be. This is an issue when you use a feature of size as a datum. A datum[ ]B across a second set of flats will be more reliable. The first requirement of your datums is that they must immobilize the part.

Your second example is not good for me. I think you need three datums for this particular part.

--
JHG