Confusing/Odd looking datums 1

[KENAT]

Take a look at the attached sketch, it's a MUCH SIMPLIFIED & INCOMPLETE (so invoke 1.1.4 of ASME Y14.5M-1994 etc.) sketch of an actual case our contract checker came across on one of our parts today.

As far as we can tell looking at the standard B & D are valid, D is basically just used to locate the top radius centrally on the width of the item.

However, it just don't quite look right.

Anyone care to opine?


KENAT, probably the least qualified checker you'll ever meet...

 

[MintJulep]

Where is A? It's referenced in the FCFs but not identified that I can see.

I think there needs to be something to define and control the arc lengths of the R.75 bits.

 

[KENAT]

"... it's a MUCH SIMPLIFIED & INCOMPLETE (so invoke 1.1.4 of ASME Y14.5M-1994 etc.) sketch ..."

Just assume A is on a surface in the orientation of the sketch, i.e. perpendicular to B & C.

KENAT, probably the least qualified checker you'll ever meet...

 

[ctopher]

There is no location for the dim "R 2.000". There is position (FCF) tol, but no dim to locate it.

Chris
SolidWorks/PDMWorks 08 3.1
AutoCAD 06/08
ctopher's home (updated Jul 13, 2008)

 

[KENAT]

Sorry I didn't feel that was necessary for the point of my post.

Maybe I was unclear.

The issue in question was that having both B, as an edge, and D, as the width looks odd.

KENAT, probably the least qualified checker you'll ever meet...

 

[MintJulep]

MUCH SIMPLIFIED & INCOMPLETE

Yes, but we don't know what things are missing, and what things just aren't there.

Anyway, it is "interpretable", although maybe not intelligent.

Any feature located wrt D is Dependant on the actual location of D subject to whatever tolerance is allowed on the 3.000 width dimension. So it has the effect of making the tolerance zones of such features an oblong, rather than a circle.

 

[ringman]

Kenat,

How can a datum be confusing or odd? Firstly they are only theoretical, secondly they are points, lines and planes.

I think you really meant 'datum features'.

Please accept as it is intended.

 

[ringman]

As to the sketch, I do not see a problem other than incompletness. All that is in place appears to me to be legit per Y14.5.

 

[ewh]

It seems legit to me also. While they may appear odd as used, datums B and D do serve different purposes.
It is more important to locate the Ø.500 hole relative to the edge and the top radius relative to the width, for whatever reason.

When the people fear their government, there is tyranny; when the government fears the people, there is liberty. - [small]Thomas Jefferson [/small]

 

[TheTick]

Unusual but perfectly legit.

B is fine. D is fine. There is nothing that says one can't have both. Just can't use both at same time.

 

[axym]

I agree that the presence of both B and D is unusual, but I don't see anything wrong with it.

I do have a problem with the position tolerance on the 2.000 radius though. It's an example of a "partial feature of size" which is a classic pain in the CMM world. The partial arc (in this case it's around 1/4 of a circle) has no opposed points anywhere, and is technically not a feature of size. Finding its actual "axis" is highly uncertain if the feature has any form error. I am also highly suspicious of the MMC reference on the considered feature. What function would justify letting the arc center drift further away from true position if the arc was a smaller radius?

I've found that in most cases, partial arc features like this don't function as features of size. They are just surfaces that happen to be nominally cylindrical. In my opinion, they should be controlled using Surface Profile.

Sorry for the off-topic rant - in past lives I spent many hours measuring the "size" and "location" of partial arc features and arguing over the results.

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[drawoh]

KENAT,

I agree with the posts above. Datum[ ]D is legitimate. The positional tolerance referencing datum[ ]D is legitimate.

A positional tolerance referencing datums[ ]A, B and[ ]C would have been just as legitimate in my thinking, and I would have used a profile tolerance rather than a positional tolerance for the radius. I find the profile tolerance easier to visualize in a case like this, and I find axym's remarks interesting.

How about "correct, but poor practise."?

JHG

 

[ctopher]

Datum B is not needed, D takes care of that surface. It's similar to double dimensioning, you have two datums referencing the same surface...or plane.

Chris
SolidWorks/PDMWorks 08 3.1
AutoCAD 06/08
ctopher's home (updated Jul 13, 2008)

 

[powerhound]

I agree with Evan on this. The position tolerance on the R2.000 is not good. A radius is not a feature of size and thus not subject to material modifiers nor should it be used with position. Profile of a surface or line is appropriate in this case. The MMC on datum D is legitimate even with profile.

Powerhound, GDTP T-0419
Production Supervisor
Inventor 2008
Mastercam X2
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

 

[KENAT]

Actually Axym, when I first looked at it my first thought was surface profile for that partial radius. You may make a good point on the MMC too, it's function is basically as a cover and I think it may be recessed (not sure) if so that may be an issue.

I'll bring those points up with my colleague.

KENAT, probably the least qualified checker you'll ever meet...

 

[KENAT]

Posts while I was typing again!

Ctopher, the hole related to B need to be related to the edge of the part. The large rad at the top needs to be centered on the width. Basically ewh got it.

That's why we have both datums, it looked almost like "double dimensioning" to me till I really thought about it and even then still looked odd, hence the post. However, others here seem to agree that it's fundamentally OK if a bit unusual.

The MMC is functionally OK looking at the function with my checker. Powerhound/axym, could you clarify why you say it's not a feature of size, you had me convinced but then I looked at 1.3.17 and I don't see that it precludes it, it is a cylindrical surface, albeit not a complete cylinder. I can see how it may not be a good feature of size to support inspection but, as we barely do inspection that wont be an issue;-).

We talked over the surface profile too, I still lean toward surface profile but it's going to a sheet metal shop and he's concerned it would confuse them, not sure I buy it but still, he has a lot more experience than I. The current vendor is already making good parts to the current drawing, incomplete/poorly toleranced though it may be. However, we're tidying the pack to potentially allow outsourcing abroad (as mentioned in other thread of mine) so I’m not sure it’s a valid reason for not doing it correctly, if indeed position truly is wrong.


KENAT, probably the least qualified checker you'll ever meet...

 

[btrueblood]

Kenat,

Excuse please, this '82-trained engineer who is not completely at ease yet with '94 datum identifiers.

But, Datum D to me at first glance looked like the left edge of the part, i.e. the opposite face to datum B. If D is intended to be the width (both faces considered together), then it would be more clear to me if the datum "suction cup" was attached to the dimension line for the 3.000 width, and not from a single edge. Is such a representation legal in '94? My CAD software allows it, so I've never questioned it.

 

[KENAT]

The examples in -94 show it pretty much as in my sketch.

(b) placed on an extension line of a feaure of sie when the datum is the axis or center plane.

My contract checker is more familiar with 82 and asked a similar question. We talked about moving the D identifier but seemed like as shown in my sketch was the preffered way of doing it. I suppose I could make an extension of the dimension line and then come off that so the 'plunger' is vertical but I'm not sure if it warrents the effort.

KENAT, probably the least qualified checker you'll ever meet...

 

[ringman]

Is the post questioning how to dimension the part or if the selected datum features are correctly defined?

 

[KENAT]

The original question was basically if anyone knew of a reason why having both the width as a datum (feature) for some features and the edge as a datum (feature) for other features was OK, based purely on the fact it looked odd not because I could find anything in the standard that said it was wrong.

I believe I've received adequate responses on this to confirm my belief that it's acceptable, if unusual.

The issue of if positional tolerance is the correct control for the radius came up in axym/powerhounds response. We're sticking with positional on this drawing but I'm interested in hearing any more arguments in preference of profile for future reference.


KENAT, probably the least qualified checker you'll ever meet...