Maximum possible contact and a single solution

[greenimi]

Per ASME Y14.5-2018
7.12.4 Pattern of Features of Size RMB
When RMB is applicable in a feature control frame to common datum features of size used to establish a single datum, the true geometric counterpart of each feature shall be fixed in location relative to one another. The true geometric counterparts shall expand or contract simultaneously from their worst-case material boundary to their LMB until the true geometric counterparts make maximum possible contact with the extremities of the datum feature(s). When irregularities on the feature(s) may allow the part to be unstable, a single solution shall be defined to constrain the part.
Questions:
1.) What means maximum possible contact?
2.) What means "single solution" default for unstable/rocking datum features (that replaces the former default of a candidate datum set) ?

 

[mfgenggear]

this explains very well

 

[Burunduk]

Per ASME Y14.5-2018
7.12.4 Pattern of Features of Size RMB
When RMB is applicable in a feature control frame to common datum features of size used to establish a single datum, the true geometric counterpart of each feature shall be fixed in location relative to one another. The true geometric counterparts shall expand or contract simultaneously from their worst-case material boundary to their LMB until the true geometric counterparts make maximum possible contact with the extremities of the datum feature(s). When irregularities on the feature(s) may allow the part to be unstable, a single solution shall be defined to constrain the part.
Questions:
1.) What means maximum possible contact?
2.) What means "single solution" default for unstable/rocking datum features (that replaces the former default of a candidate datum set) ?

Max. possible contact for RMB is what it sounds like - it's when the datum feature simulators reach the smallest size (for shafts/pins) or largest size (for holes) given the restrictions that act upon them (for example staying coaxial to each other) and stopped by the datum feature itself. It also means it can't stop expanding/contracting until in contact with the datum feature.

As for the single solution, search for "constrained least squares". It was discussed on this forum in the past.

 

[mfgenggear]

Greenimi
do you have access to a mechanical inspection department. go and volunteer.
to get a better grasp. some times a hands on approach really helps.

 

[greenimi]

Max. possible contact for RMB is what it sounds like - it's when the datum feature simulators reach the smallest size (for shafts/pins) or largest size (for holes) given the restrictions that act upon them (for example staying coaxial to each other) and stopped by the datum feature itself. It also means it can't stop expanding/contracting until in contact with the datum feature.

As for the single solution, search for "constrained least squares". It was discussed on this forum in the past.

So are the simulators independent from each other?

 

[greenimi]

As i stated in the previous thread:

If one simulator stops contracting to get the smallest perfectly formed cylinder circumscribed around the actual "as made" feature, what is going to happen with the other simulator? This second simulator will stop too or will keep contracting?
That was my main question about the interpretation.

Do you have any suggestions on how to interpret (per the standard) this RFS scenario?
Do we need a note to explain to the end users or the standard has already clarified this RFS case?

Or if you have a pattern of two holes modified at RMB in the callout (in the FCF) for other feature, HOW I am going to simulate this hole pattern?
With two expending pins perfectly located and oriented to each other and those pins will expand simultaneously until what?
Here is the unknown:
If one pin stop expanding what is going to happen with the other pin? Will stop expanding too? Will keep going?

The text / verbiage from the 2018, which I copied on the onset of this thread, is not clear, in my opinion, but I might very well be wrong.
Is the math standard of any help? Probably not, because the latest (2019) edition of the math standard has been released to support Y14.5-2009 and NOT Y14.5-2018, hence my dillema is not solved.

Therefore, how do YOU see and solve the issue?

 

[Burunduk]

greenimi, what implied to you that they are independent from each other?

As I said " datum feature simulators reach the smallest size (for shafts/pins) or largest size (for holes) given the restrictions that act upon them (for example staying coaxial to each other) and stopped by the datum feature itself. "

That is not independent from each other.

The goal is to make them act as equal as possible. While the standard says they "shall expand or contract simultaneously" a theoretically more exact way to achieve that would be to make them contact the datum features simultaneously.

 

[greenimi]

greenimi, what implied to you that they are independent from each other?

As I said " datum feature simulators reach the smallest size (for shafts/pins) or largest size (for holes) given the restrictions that act upon them (for example staying coaxial to each other) and stopped by the datum feature itself. "

That is not independent from each other.

The goal is to make them act as equal as possible. While the standard says they "shall expand or contract simultaneously" a theoretically more exact way to achieve that would be to make them contact the datum features simultaneously.

So you are of an opinion that if one datum feature simulator stops expanding or contracting then THE OTHER shall stop expanding or contracting (even the second datum feature simulator did not reach its full potential of maximum possible contact)?

Is my understanding correct?
So basically you put more emphasize on simultaneity and less on the maximum possible contact, right?

I am not saying you are wrong I am just trying to understand it.

I understand that " a single solution shall be defined to constrain the part", but I am not understanding what is the correct way (read standardized way) to get there.
Or maybe even the standard did not clearly stated how............

 

[Burunduk]

So you are of an opinion that if one datum feature simulator stops expanding or contracting then THE OTHER shall stop expanding or contracting (even the second datum feature simulator did not reach its full potential of maximum possible contact)?

Is my understanding correct?
So basically you put more emphasize on simultaneity and less on the maximum possible contact, right?

No I am not. Where did I say or imply that?
I said " a theoretically more exact way to achieve that would be to make them contact the datum features simultaneously " which cannot lead to such a conclusion.

As for the single solution, did you do the search I suggested?

 

[mfgenggear]

give an actual part or fiquire. you guys are making this way over complicated.

how flat are the parts
how straight
surface finish
how rigit
location datums primary, secondary , and so on
sometime to achieve drawing requirements
surfaces have to be machine to achieve other details. tooling purposes.
but it affects how one inspects a part.

 

[greenimi]

No I am not. Where did I say or imply that?
I said " a theoretically more exact way to achieve that would be to make them contact the datum features simultaneously " which cannot lead to such a conclusion.

As for the single solution, did you do the search I suggested?

Yes, I found this thread

Inclined datum and degrees of freedom - Drafting Standards, GD&T & Tolerance Analysis

In this mock example I have a yellow fixture with a cylindrical boss as Datum A, then an inclined Datum B at 20 deg Basic. We slip the gray part over Datum A which let's just say expands b/c it is at RFS, so Datum A is locked in solid. ' I don't think I've seen an example in the standard for an...

www.eng-tips.com

It is for the "single solution" and the " "constrained least squares""
Do you see any "other" discussions relevant to this subject?

But my main question stil remains: Is it crystal clear how to simulate a pattern at RMB?
I don't really think so.....

 

[greenimi]

Let me ask you a direct question:
Per 2018 the default stabilization procedure has been changed from candidate datum set to a single solution that minimizes the separation between the features and the TGC.
Is this new procedure more clear than the old one regarding the pattern feature at RMB?
If yes, how?

7.11.2 Irregularities on DatumFeatures Applicable RMB
If irregularities on a datum feature are such that the part is unstable (i.e., it rocks) when it is brought into contact with the corresponding true geometric counterpart, the default requirement is that the part be adjusted to a single solution that minimizes the separation between the feature and the true geometric counterpart per ASME Y14.5.1M. If a different procedure is desired (candidate datum set, Chebychev, least squares, translational least squares, etc.), it shall be specified.

 

[Burunduk]

greenimi,
See here about the single solution.

 

[greenimi]

greenimi,
See here about the single solution.

I have seen your thread, but does not solve my question about patterns at RMB.

 

[greenimi]

Burunduk,
For some reason I feel like you are beating around the bush and avoid giving me a straight answer about YOUR opinion

Are you of an opinion that if one datum feature simulator stops expanding or contracting then THE OTHER shall stop expanding or contracting (even the second datum feature simulator did not reach its full potential of maximum possible contact)?

If you don't know or don't want to take sides because the standard is not clear I am fine with that.
Pretending (that the answer is very clear and it is obvious how to simulate the RMB for patterns) will only make the matter worse.
Einstein said:
"If you can't explain it simply, you don't understand it well enough".

And I agree I am the uneducated and ignorant individual in this discussion

 

[Burunduk]

greenimi,
I'm not a politician, and I don't dodge direct answers - at least not when it comes to this topic. The reason why you don't accept my answer and repeat your question, is that you've made a presumtion that you refuse to let go although I challenged it.
As I said and repeated: " a theoretically more exact way to achieve that would be to make them contact the datum features simultaneously ". In other words, there should be no condition in which one component of the common datum feature simulator has stopped contracting and the other one is able to keep contracting. They need to come into contact with the two separate portions of the physical common datum feature simultaneously.

I will give you an example, but first I want to remind you the goal. The goal is to "... simultaneously constrain ..." as outlined here:


The exact technique is not the goal itself, it is secondary to the goal and should attempt to realize it.

Now the example:
Suppose each of the coaxial (per design) shaft portions designated and referenced as a common datum feature A-B are dimensioned by dia. 10+/-0.1 and toleranced as a 2X pattern by a datumless position of 0 at MMC. And suppose that in reality they came out almost perfectly aligned (the measured position error is 0). But one came out at dia. 10.05 while the other at dia. 9.9.
If both coaxial components of the simulator start contracting from the MMB of 10.1, then to come at contact with the actual datum feature at the same time, they should be contracting during the same time-window ("simultaneously" as the standard specifies in 7.12.4 Pattern of Features of Size RMB), but at different rates.

If you accept that this is what should happen or at least that there should be an attempt to achieve that, then, your question about what happens to one contracting cylinder when the other stopped becomes irrelevant.

Now the above may sound technically complicated and mfgenggear may say it is a "PITA", , but we all know that RMB (and RFS) is always more complex to inspect than MMB (and MMC).

As for the single solution,
In the thread I linked to you'll find a snippet made by SeasonLee from a pretty good explanation by GeoTol about the constrained least squares, which is the new stabilization procedure for a rocking datum, and that is exactly what you asked about. I still don't own a copy of the latest math standard, but I have seen this method mentioned only in the context of planar datums. However as far as I understand it, I don't see a reason why the same principle can't be used to accomodate the establishment of a common datum from a pattern of features of size.

I also like the technique provided in SeasonLee's GeoTol reference in the more recent thread (where this discussion originated) utilizing the two Unrelated AMEs to obtain separate axes, and then deriving a circumscribed cylinder containing these two unrelated axes, which in turn provides a new common datum axis. It could probably be incorporated in CMM calculation algorithms (may be already).

My own way to tackle this would be to simply say that the common axis should be established by two Related AMEs, which are "Related" not because of any preceding datums in the datum reference frame, but because, although forming a primary true geometric counterpart, they must be perfectly coaxial to each other.

 

[3DDave]

Varying rates leaves open the chance that one feature is locked in before the other and therefore will not represent the installed condition. Interference fits and the like don't have adaptive rates.

Tossing in different rates suggests that there is an additional step to find some ratio of rates that produces some best condition, a condition that isn't specified beyond the vague "maximum possible contact."

Maximum contact would be, in any other engineering area, the result of material plastic deformation with such as a hydraulic crimper to remove all gaps and high points of contact to reach the lowest ones.

Least squares is only available to CMMs, a clear marketing lock-in move and a gift to CMM makers.

At least if there aren't answers defined in the standard, falling back on platitudes is a position to take.

Give them the benefit of the doubt. Assume whatever is necessary to cover the gaps. It's not like they have had decades to sort this out. Once more into the breach. Ad hoc solutions to problems not dealt with in the "standard."

 

[Burunduk]

Varying rates leaves open the chance that one feature is locked in before the other and therefore will not represent the installed condition. Interference fits and the like don't have adaptive rates.

The varying rates could be part of the solution only in the exact case I provided as an example. Had the two cylinders been produced very close to the condition of being coaxial and same size, equal rates would be the way to go. It is all matter of technique, which is always imperfect and the Y14.5 standard is not expected to cover. I would say they should have just stuck with "simultaneously constrain" in all relevant paragraphs. The wording they used in 7.12.4 was unecessary.