Can symmetry be implied without GD&T?

[cwdaniel]

A part has a defined center line, features are symmetrical about that center line. Distance between the symmetrical features (tapped holes) is important. Is a dimension mandatory from the center line to a feature on one side?

Or is it allowable to leave that dimension out to imply symmetry and not a tolerance stack to one side of the center line?

 

[MechNorth]

It's been common for many years to just dimension between the two points and put "= =" under the dimension line (I believe there was a thread on this a while back). It's not part of any current standard that I've heard of, so it may not be recognized by the shop. A second consideration is that a centerline doesn't exist on a physical part, so how are you going to check those positions without defining that centerline, which essentially is the domain of GD&T.

In other words ... do the GD&T!

Jim Sykes, P.Eng, GDTP-S
Profile Services
CAD-Documentation-GD&T-Product Development

http://www.profileservices.ca

 

[cwdaniel]

The centerline is extended from a hole in the center of the part so there is an actual feature to use as a datum. Trouble is they frown on the use of GD&T unless absolutely necessary here. Otherwise I'd use it.

 

[drawoh]

cwdaniel,

If your part looks symmetrical and the drawing is specified as being done to ASME Y14.5M-1994, then symmetry is implied. This is true, GD&T or no GD&T.

Without ASME Y14.5M-1994, there is no standard to interpret your drawing by, and you should make things explicit.

JHG

 

[aardvarkdw]

If you are concerned about reducing dimensions on a symmetrical part, you can place a centerline on the part with "=" at each end as shown in ASME Y14.2M-1992 (referenced by ASME Y14.5M-1994). This is just a visual way of saying that the dimensions and tolerances apply idenically and individually to both sides of the part. You will want to verify that your tolerance stack ups are still valid if you do this. The better method is to, in Jim's words, "...do the GD&T!".

David

 

[GDTcoach]

Keep in mind that there is a difference between dimensioning (the goal) and tolerancing (how much you can blow it by). With dimensioning, a symmetrical relationship my be implied. But, features shown symmetrical must be controlled with geometric tolerances according to Y14.5. In other words, listen to Jim. By the way, GD&T doesn't make parts expensive--tight tolerances do.

 

[ctopher]

ditto!
What the others wrote.

Chris
SolidWorks 06 4.1/PDMWorks 06
AutoCAD 06
ctopher's home (updated 10-27-06)

 

[KENAT]

cwdaniel

The centerline is extended from a hole in the center of the part so there is an actual feature to use as a datum.

Some of the above are sharper on GD&T than I but to me if you don't use it then you haven't really told them to use the hole as the datum. They may be just as valid using the outside dimensions or some other symmetrical feature.

I've seen implied symmetry cause problems before so I'd always make it explicit.

By the way, GD&T doesn't make parts expensive--tight tolerances do.

If only we could convince everyone of this!

 

[drawoh]

KENAT,

You are right. The datum must be called up explicitly on the drawing. It can be the hole. It can be the two outside edges. In either case, your datum is a feature of size. Unless it is very accurate, you should call it up at MMC. This may be an issue with the functionality of the part.

I have an issue with the idea that GD&T is somehow more complicated and expensive than no GD&T. A recommended procedure on a part like this is to apply a composite positional tolerance. One true position attaches the pattern to the face sloppily. The other true position locates the holes accurately with respect to each other. This is an ideal solution when you have to match the holes in the mating part, but you are not concerned about the overall location.

In effect, the symmetry is nominal, not actual.

I am not sure what happens when you do not apply basic dimensions. The ASME standard goes very quickly into positional tolerances and true position. I would regard non-basic dimensions as ambiguous.

JHG

 

[aardvarkdw]

All that making a dimension basic does is tell the person reading the print that that dimension does not follow the standard tolerences set up in the tolerance block and that the tolerance to apply is shown elsewhere. It tells them to assume the dimension to be theoretially perfect.

David

 

[drawoh]

aardvarkdw,

I am assume that when you do not make the dimension basic, you also do not apply the positional tolerance.

Let's assume our part is a rectangular plate, let's continue to interpret the drawing through ASME Y14.5M-1994, and let's assume the tolerrance block in millimeters says X=+/-0.5, X.X=+-0.2 X.XX=+/-0.1, and angles are +/-1deg.

The sides of the plate are allowed be one degree out of perpendicular to the base. Are orthogonally dimensioned holes measured from the angled side, or from the corner at the base?

If both sides are one degree out of perpendicular in the same direction, how do you determine where symmetry applies?

If you ask yourselves these questions, the GD&T positional tolerance becomes simple.

JHG

 

[aardvarkdw]

Drawoh,

I agree that using the GD&T is the best way to go. I was just trying to clarify your question. You said,"I am not sure what happens when you do not apply basic dimensions. The ASME standard goes very quickly into positional tolerances and true position. I would regard non-basic dimensions as ambiguous.".

For instance you would not make a hole diameter a basic dimension if your only feature control frame is a positional tolerance to locate the hole. That would be ambiguous because now you don't have a tolerance on the hole diameter. This is true for any feature of size. Only the position tolerances effected by the positional FCF would be basic, all others would not be basic.

David

 

[MechNorth]

Actually, aardvark, you can't use a positional on a non-toleranced feature, so there can be no ambiguity on that point. There must be a tolerance of size applied, and the general tolerance block could provide that ... hate it, but it is an (unpleasant) option. A conflict would arise where a note appears on the drawing to the effect of "ALL NON-TOLERANCED DIMENSIONS ARE BASIC", along with traditional general linear-tolerance catch-all (based on # of decimal places) that reads "UNLESS OTHERWISE STATED". Seen that a number of times, and it's hard for people to recognize that there is a conflict.

Jim Sykes, P.Eng, GDTP-S
Profile Services
CAD-Documentation-GD&T-Product Development

http://www.profileservices.ca

 

[ewh]

... but you can use a positional tolerance of 0.000 at max matl on the hole.

 

[aardvarkdw]

Jim,

My point is that you can't indiscriminatly make all dimensions basic. If they are all basic they must all have a FCF that they refer to. If they don't they can't be basic and therefore must either have a linear tolerance attached to the dimension or in a generic tolerance block. By making a dimension basic you are removing all linear tolerances from the dimension and saying that you have toleranced it in a FCF.

David

 

[MechNorth]

Exactly right. Tks for clarifying the statement, Dave.

EWH, you can use 0@MMC positional tolerance on a FOS that is linearly toleranced, but not on a feature that has a basic size dimension. That assumes that you can live with the trade-off of size vs position.

Jim

Jim Sykes, P.Eng, GDTP-S
Profile Services
CAD-Documentation-GD&T-Product Development

http://www.profileservices.ca

 

[drawoh]

When I said you should apply basic dimensions everywhere, I was refering to hole co-ordinates. Holes mostly are features of size. I should have been more clear. The discussion was about hole positions.

You can use basic everywhere if you use surface profile tolerances to control the holes. Usually, this is not convenient. I have done it once in a while.

I have been using the zero true position tolerances lately, primarly because they are an exact description of what I require from clearance holes.

Eg.

M6X1 hole located at 0.2DIA of true position.

The clearance hole is 6.2DIA +0.8/-0 at 0 DIA MMC.

There is a 6.2mm diameter potentially occupied by the screw, which must be kept clear. The allowable positional error increases as the hole gets bigger. The 6.8mm diameter hole may be off by 0.4mm (0.8mmDIA TP). This is all extremely simple to calculate. You do have to add some clearance for perpendicularity through thick parts.

Note also that 0 true position is valid, fabricatable and inspectable, even if you do not specify MMC.

JHG

 

[MechNorth]

OK, drawoh, I'll bite...how is 0-Position (without MMC) valid, fabricatable and inspectable?

Jim Sykes, P.Eng, GDTP-S
Profile Services
CAD-Documentation-GD&T-Product Development

http://www.profileservices.ca

 

[ewh]

I am confused by this also. If the condition is RMC, then there are no tolerances to take into account. I can see how tolerance would be available at LMC, but have never come across a situation where that would be desireable.

 

[drawoh]

The profile of the hole must be contained within the MMC outline and the LMC outline. Thus, the maximum position error is allowed when the hole is half way between MMC and LMC.

You control the MMC that allows clearance around a fastener or other features. You control LMC, for whatever reason you would want to control LMC. Perhaps you have a single big hole through a piece with a slightly bigger OD.

I wasn't saying you have to do it. It is an option.

JHG