Methods for Dimensioning a Simple Plate with Holes

[MrGearhead]

Reference attached. This is what I’d believe to be a basic question, but I’ve not found any clear, cohesive answer that points back to ASME 14.5. Though I did find a reference to “dimensioning all features off a datum”, but I can’t find that post now.
See Scheme 1 of attached. Nothing crazy here, just a plate with 6 holes (could be 4, 2, etc…). Assume this is a “Baseplate” with no immediate function of the edges relative to anything else in the design.
The top left hole is the “keyhole”… Since I really have no need for the hole pattern to be held with respect to Datums B or C, I just use a 2 place rectangular dimension (assume +/-.01”) to hold the hole to the plate. Once it’s located to within .01”, then the dimensioning scheme implies a pattern that needs to be held “tighter” so that the pattern holes align to their mating part (with the same pattern)…with a new “implied” datum formed by the key hole, yes? And… That all the holes should be within .005” TP of each other, yes?
I’ve seen many plates dimensioned this way and I thought it was the “most correct”, and it matched my old way of thinking when using rectangular coordinates.

But… Isn’t there some things wrong here in this thinking which would make “Scheme 2” more correct?
Scheme 2: ALL holes are held to the A, B, C datums. Advantages on Scheme 2 appear to be:
1) The “keyhole” is held with respect to the same datums as the other holes (no perpendicularity issues with holes being different from one another).
2) The machinist has to block up the plate to sides A, B, and C anyway when making the part, so why not make this easy for him to set all the coordinate CLs for each hole based on the same datums?. This is especially an advantage for a CNC programmed part (less chance for error by simply putting in all X, Y coordinates for each hole with no “translation” made to offset that first hole from the edge since they are ALL for the same edge).
Can someone comment on these two methods with respect to each other and the merit of either?

I’ve seen some references to a 14.5 section that implies scheme 2 is more correct but as said at the beginning of the post, I can’t find it now…

 

[pmarc]

The attachment can't be opened.
Change the name of attached file. Do not use "&" symbol.

 

[MrGearhead]

Hmmm.... I did not find a way to EDIT my post so I'll attach the new attachment (with a new filename) here.

 

[PeterStock]

What I would do is set the holes as datum B with the true position relative to A and B at max material. I would control the edges with a surface profile to A and B at max material. See the example attached.

Peter Stockhausen
Senior Design Analyst (Checker)

http://www.linkedin.com/profile/view?id=30064526&trk=tab_pro

 

[powerhound]

Scheme 1 is incorrect because the use of positional tolerance requires that features be related back to their respective datums through basic dimensions. Scheme 2 is correct but does not do what you need it to do. One solution is composite position. You can specify a loose tolerance for the location of the pattern within the plate in the upper segment and then tighten up the tolerance of the location of the holes to themselves in the lower.

John Acosta, GDTP S-0731
Engineering Technician
Inventor 2013
Mastercam X6
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

 

[MrGearhead]

Why does solution 2 not do what I need it to do???

It gives a true position tolerance to each hole of .005" with respect to datums. I guess if you are saying that I'm holding the first hole "too tight", I could agree... But, in practice, the machinist is just going to set up the part and drill the clearance holes in one tool pass... The machine will stop at the *exact* XY coordinates given. I suppose in theory, the first hole could be "off" and the others could magically correctly themselves, but if the first hole is off, all the holes will probably be off anyway...

I never thought that such a simple plate would require a composite tolerance... Can anyone point to any spec 14.5 or otherwise that covers this basic dimensioing scheme that probably covers 95% of all plate design? I find it amazing that there isn't one spec/standard I've seen that covers this definitively. In fact, I've seen several "examples" that mix the schemes, etc...

 

[fsincox]

I would tend to do it the same as Peter, What you, and many others want, is the standard to say there is only "one way". My impression is the standard is intended to give multiple ways of solving a problem so people can find the one that is right for their particular application.
The way you propose is a very traditional "old school" manufacturing oriented method. As you have noticed the standard has shown it that way many times. Even the way you showed in the first method was used in the past (without the hole pattern referencing B & C). Basically, as methods evolve traditional manufacturing methods will become obsolete, are you familiar with the new rapid prototyping methods? I personally believe these will someday make "traditional" methods obsolete.
The standard has grown to advocate functional methods of dimensioning, parts like your misrepresent the importance of (2) outside surfaces solely due to manufacturing methods, while, this is acceptable it is not functional in many cases, yours may be different. I do not know the functional application, here.
Frank

 

[Belanger]

I agree with Powerhound. You ask why solution 2 doesn't do what you need ... it's because that scheme holds the hole-to-hole position to the same accuracy as the hole-to-edge position, where you stated that you don't want that.

Scheme 1 isn't correct, as already mentioned. Some people will try to pick one hole as a "pilot" hole, positioning it back to the edges, and then make that hole a new datum for the others to reference. That's not the best way; first, because all holes are of the same size, and second, when the other 5 holes are given a position tolerance, there would still be a need for an edge datum to control rotation.

The way to do this is composite position. It's not complicated, and it's actually ideal for the situation you describe. Granted, the Y14.5 standard might make it look imposing, but that's just because of the many different variations that spin off of the composite idea.
What you want is essentially shown in Fig. 7-38 of the 2009 standard(the 4-hole pattern in that part is pretty much what you're asking about), and described in paragraph 7.5.1.

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

 

[powerhound]

Thanks JP. I'm just now getting back to this thread and see you and the others have pretty much addressed it well.

MrGearhead,
There are definitely several ways to do this and be legal. My input was based on the datum reference frame as shown in scheme 2. Changing the datum reference frame can definitely open up the possibilities. I was just trying to stick with what you had shown.

Hopefully we've given you some good info to be able to specify exactly what you need.

John Acosta, GDTP S-0731
Engineering Technician
Inventor 2013
Mastercam X6
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

 

[Vimalmechs]

Hi I got a doubt here,

Primarily, as per my understanding from MrGearhead, Hes with a base plate having a reference "key hole". and other 5 holes are referring the key hole.
And i am guessing up, the functionally is, there is a top plate for this bottom plate.The key hole in the bottom plate serves the purpose of maintaining the same reference with the top plate.

Secondarily, The dimensioning scheme in the fig "SCHEME 1" serves the functionality well.But the GD&T applied is confusing me. It doesnt refer the pilot hole. but it refers the edge. edges are not an important datum here.
In "SCHEME 2" the dimensioning is refers from edges. Not from the key hole.

Thirdly, when i am applying the composition tolerance (based on powerhound and belanger replies) on these patterns (6 holes).Still it refers the bottom plate edges and i am considering my "pilot hole" under the same "composite tol zone", like a normal hole.

In my personal opinion pilot hole is the master reference. but if your considering one hole, the datum deriving from the pilot hole can rotate. to prevent the rotation either you should use any edge or any other hole axis.but here edges are with least priority.
I mean to say why cant you drill one more pilot hole to achieve the secondary datum for all these hole patterns ?.

 

[CheckerHater]

I used to be big fan of Scheme 2, because even with lack of basic dimensions tying hole positions back to edges relations between the holes still controlled and orientation of hole pattern to edges is controlled too.

Today I would rather go with composite for the holes, or loose profile requirement for the edges with just a small note. Is everybody comfortable with self-referencing datum B in Peter's example?

 

[fsincox]

My position is, "in the end", datums on the edge is still an old school manufacturing processing method and not a "pure" functional requirement in many cases. I prefer a use functional features approach!
Frank

 

[drawoh]

powerhound,

I have no objections to scheme[ ]1. It shows the design intent that the holes are related more to each other than to the datums. I would use a composite tolerance if I felt that the fabricator would have an easier time grouping the holes together than locating from the datum. Perhaps this is a crude weldment.

I got into datum dimensioning in the old days when machinists zeroed their milling machines on a corner, then drilled everything at coordinates. Now, just about everybody programs the thing into CNC.

Yet another dimensioning scheme would be to assign holes as datums[ ]A and[ ]B.

--
JHG

 

[powerhound]

The problem with scheme 1 is that the holes are not tied back to the datums via basic dimensions. This is contrary to what the standard says.

John Acosta, GDTP S-0731
Engineering Technician
Inventor 2013
Mastercam X6
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II