Shall we consider a profile tolerance with datum referenced at mmc as additional tol in stackup anal

[Babu96]

Shall we consider a profile tolerance with datum referenced at mmc as additional tol in stackup analysis

 

[greenimi]

Babu,
You might know that profile cannot be modified at MMC (or LMC), but if you are talking about modifying DATUM SHIFT (MMC in datum reference frame area of the feature control frame) that would be an acceptable option.
If you have an example, would be better.
Sometimes datum shift influence X min or X max in your stackup, sometimes does not.
Your question is way to general.....

 

[dtmbiz]

Babu

Because your post uses MMC and not MMB I would believe that you are referencing Y14.5 1994 or earlier, which actually does not affect an answer to your question. (Just an observation)

My 1st thought is "no”. Should not be considered for a tolerance stack.

The reason is that Datum Shift / Displacement is based on the Actual Mating Envelope of any produced feature.
Measurements for a feature’s produced size within its tolerance limits is unknown until the feature is produced and measurements taken.
Only then could any potential Datum Shift / Displacement be evaluated

Tolerance stacks are based on calculations of buoundaries for minimum and maximum conditions for feature sizes. Stacks should include orientation of features for consideration.
The calculations for the stack boundaries already contain any AME that would be produced that would be considered for potential Datum Shift.

Just stating this based on basic principles for boundaries vs. datum modifiers that are evaluated for produced features as a general opinion.

 

[dtmbiz]

Babu

Obviously forgot to mention tolerances of locations are considered for tolerance stacks

 

[3DDave]

It depends on what stack is being considered. For example, if remaining thickness of a wall is important then datum shift can be an important consideration.

Being more specific, as with a sketch, would help understand any particular case.

 

[dtmbiz]

Babu

I would like to retract one of my 1st post’s statements. I cannot. So, I will clarify that I was considering a considered feature’s AME and not the modified datum's AME when I posted:

The calculations for the stack boundaries already contain any AME that would be produced that would be considered for potential Datum Shift.

Based on the premise that a minimum number of DRFs should be defined in part definition and that the more features that reference a particular DRF that applies datum modifers, along with the “Simultaneous Requirement’ rule; there is less probability that any shift is actually available in reality.

E.g. if there is available datum Shift to help one feature comply with it's specifications, that Datum Shift may adversely affect other features, in a way that they now do not comply with specifications. This scenario would nullify available shift for that feature, regardless of any “calculated potential shift”. Also the more feature’s involved with a particular DRF and its modified datums; the probability for actual use of datum shift would diminish.

It doesn’t seem to me that introducing “unknowns” into a stack analysis is beneficial.

 

[chez311]

Agreed with others, specifics would help. The actual effect of datum feature shift depends on the part configuration and what is under consideration (wall thickness, etc..). Datum feature shift is not a bonus tolerance, it can be better thought of as loosening the requirement between the datum feature and the datum feature simulator - the relationship between the simulator and the associated tolerance zone(s) remains unchanged which is why its not considered a bonus.

To better understand this relationship and potential effects you might want to check out the below thread. And especially CH's post on (17 May 18 14:59) which I have attached the file he created here for simplicity - all credit goes to CH for this.

https://www.eng-tips.com/viewthread.cfm?qid=439000

 

[greenimi]

And one more datum shift tip from Tec-Ease.

 

[3DDave]

The Tec-Ease example is incorrect. "The part must be in full contact with datum feature A and C."

It doesn't need to have full contact with C. At MMB any loss of perpendicularity between B and C will mean that the C datum feature might contact the C datum simulator at only one point; likewise if C is convex, the C datum feature can only contact the C datum simulator at only one point.

In the tertiary position there must be contact, but it isn't required to be full contact. The part is allowed to shift and rotate within the restrictions of its fit within the datum B simulator.

If it was forced to be in full contact then it might be that there was enough variation of perpendicularity between the two datum features such that full contact between the C datum feature and the C datum simulator prevents the part from fitting with the datum B simulator. This is especially the case when datum feature B is at the MMB condition.

If the C datum feature contact with the C datum simulator is to orient the part while sitting on the datum A simulator it must be the second datum in the feature control frame.

 

[greenimi]

3DDave and all,

Is it true that in Tec-Ease's example datum feature C CANNOT be legally modified at MMB even 2009, 2018 standards are to be used?
That is my understanding from other discussions we had here on eng-tips (other threads) and I would like to confirm/validate it (or invalidate my understandings)

 

[pmarc]

greenimi,

Yes, I would say C shall not be referenced at MMB in this case.

I agree with 3DDave that datum feature C isn't required to be full contact with its simulator and that the part is not only allowed to translate but also rotate within the possible loose between datum feature B and its simulator.

Additionally, the way the actual holes have been shown on the left would require datum feature B's RAME size on the right to be much smaller than 50.6 to bring the holes into their position tolerance zones.

 

[chez311]

C is a nonsize feature with no location constraint to higher precedence datum features. I would agree MMB for C would be incorrect.

 

[dtmbiz]

Looking at the 2nd TecEase example from the front view for x, y, z orientation;

Datum A constrains; y translation and x & z rotation
Datum B constrains; z translation (restricted to allowable datum shift)
Datum C constrains; x translation and y rotation

A DRF defined by 3 mutually perpendicular intersecting planes does not necessarily constrain the DOF’s in the FCF’s order of those planes.

 

[pmarc]

In that example it seems like datum feature simulator B only restricts z translation because they decided to show it this way. They could have shown it rotated within the allowable loose and then it would be clear that the remaining rotational DOF might be constrained by B too.

But in general I agree - with MMB (or LMB) added to the equation it is possible that datum feature simulators will not necessarily constrain the DOFs they were supposed to.

 

[3DDave]

Since the example basically said "shall" instead of "may" for datum feature C, then that's the issue, regardless of other cases.

 

[dtmbiz]

Dont know what Y14.5 version is used for this example however, if the intended DOF contraints for Datum "B" was to constrain "y" rotation,
then the use of 2009 custom DRF could be used to identify that intent. In that were the case, a single Datum Target point (for Datum "C") might be used to relax any "y" rotation constraint that Datum "C" might affect.

The example's intent seems to be an introduction to the Datum Shift/Displacement concept.
Discussions relating to the nuances of which Datum Feature actually constrains which DOF(s) as raised in previous thread posts would be for a higher learning level than an indroductory concept example?

 

[3DDave]

Or just use C as the secondary reference. It's a bad example with bad instructions which will make future instruction harder because they were "told" how it works.

 

[pmarc]

If the part from the Tec-Ease tip is simply placed inside a mating pocket looking like the gage shown on the right (without forcing the part to contact any of the pocket walls), then using C as the secondary reference may be as bad instruction as using B as the secondary reference.

 

[3DDave]

If the instruction to make complete contact with C as given in the example is to be followed, making C secondary follows that instruction.

 

[pmarc]

Don't disagree with that. Just meant to say that there might be a case where nothing makes the considered part contact particular features of the mating part in any particular order.