datum translation questions 2

[AndrewTT]

The "Means this" portion of figure 4-32 (b) states that the datum translation modifier for datum B means "adjustable location within location tolerance".
1) Is this saying that the datum feature simulator can translate 0.1 in either direction?
2) If a gage was made for this part how much travel is allowed for the translating datum B simulator?

My understanding of datum translation is that it is used to make sure clocking datum simulators only stop rotation of the part and do not do the job of locating the part.
3) If, in figure 4-19, datum C had been modified at MMB in the hole position FCFs (∅9.2 & 4X ∅5.1) would the translation modifier not be needed anymore?
4) Same question for figure 4-32(b). If datum B had been modified at MMB for the slot position FCF would the translation modifier not be needed anymore?
5) So really, I'm asking if datum translation only gets used (or mainly gets used) when the clocking datum feature is specified at RMB?

Thank you.

 

[mkcski]

greenimi

The gages look the same, same size gages, same basic distance apart. Again, only the gages to check "B" and "C". Would you agree?

No I do not agree. The gauges are NOT the same. The pins in the first gates are fixed at the 51 BSC distance apart. The second gauge in fig 11-27 the one pin moves.

Should I understand that this case is for academic purpose only?

Yes I think this thread is academic, most of the posts on this forum are. Case studies are great for digging deeper into understanding, but in the end the function and fit-up of the "real" part is what drives the dimensioning schema, which includes datum assignment. And lets not forget to get input for Mfg and QA. You can dimension a part based on pure design criteria/parameters, but if Mfg cannot make it with confidence the drawing requirements (not making scrap or rework) and then have QA confirm that the part meets the drawing requirements too you are fooling yourself.

Certified Sr. GD&T Professional

 

[greenimi]

No I do not agree. The gauges are NOT the same. The pins in the first gates are fixed at the 51 BSC distance apart

Mkcski,
With all due respect:
I guess we have a breakdown in communication. Either you are not reading and understanding what I am saying or I am not communicating correctly and clearly.

You said (copy-paste): “In this case there are two gauges. One gauge would check the positon of the two holes - nothing else. It would have a flat plate to simulate datum plane A, and two pins at the virtual condition (8.0 dia).”……

What I am talking about is ONLY FOR THE FIRST GAGE!!!!
The FIRST GAGE looks the same either if you use perpendicularity (“B” perpendicular to A, “C” positioned to “A” and ”B”) or you use position (in the same way is shown in Mr. Meadows case) in your drawing definition.

Yes I think this thread is academic, most of the posts on this forum are.

I was talking about that the picture/ or the case shown in Mr Meadows is for academic purpose only, not necessary this thread, but I think that stands true too.
And I am NOT talking about the fact of using MMB with the translation modifier as being academic purpose only.......I AM talking about using his "shady" datum scheme (position instead of perpendicularity) as BEING for academic purpose only as the groupping and ungrouppng/ split is questionable at best. --" where the two holes are grouped together for a position callout -- implying they have equal importance -- yet they are split apart for the purpose of creating datums B and C.”)"

but in the end the function and fit-up of the "real" part is what drives the dimensioning schema, which includes datum assignment. And lets not forget to get input for Mfg and QA. You can dimension a part based on pure design criteria/parameters, but if Mfg cannot make it with confidence the drawing requirements (not making scrap or rework) and then have QA confirm that the part meets the drawing requirements too you are fooling yourself.

This will bring nothing new or relevant to the discussion…sorry

You still did not answer one of my fundamental questions: “Do you see this practice legalized somewhere in the standard”? And to repeat (to avoid any misscomunications again: What practice: The practice where the two holes are grouped together for a position callout -- implying they have equal importance -- yet they are split apart for the purpose of creating datums B and C.”)

 

[3DDave]

The practice is legal in the same way that trillions of possible chess games are legal as long as the individual rules are observed in playing them. There is no rule book that specifically lists all possible chess games in order to make all of them legal. The 'Y14.5 rule book says a hole can be a datum feature - there's no restriction on how the hole got to be there.

Whether a scheme is sensible to use vs legal is a different question. But the scheme is legal. The main excuse for splitting them is due to a failure of CMM software to cope with compound datum features. There is no reason that 10,000 holes could not be a single datum feature, for which a simple gage with 10,000 pins would be used as a datum feature simulator, but CMM operators would dislike the use.

Which is where this argument seems to always lead. Extra restrictions due to the inspection limitations instead of stickeing with the wider functional limitations.

 

[axym]

Hi All,

Sorry I'm entering into this thread rather late. I'm not quite sure what I want to say - I think that I mostly agree and partially disagree with everyone (including Jim Meadows and including Y14.5). Here are some comments. I'm going to label them, since I'm expecting that there will be responses/questioning/pushback/opposition. ;^)

a. This topic is a rough ride.

b. Y14.5 states in the MMB sections 4.11.5 "the appropriate boundary is determined by the collective effects of size and any applicable geometric tolerances relative to any higher precedence datums". So MMB calculations are the most straightforward when the tolerances on each successive datum feature in the FCF reference the higher precedence datum features. This is what is shown in the examples in Y14.5. When the tolerances on the datum features are not structured in this way, as in the Meadows Fig. 11-27 example where the secondary and tertiary are toleranced together as a pattern, then determining the tertiary MMB is not so straightforward. With no examples in the standard to refer to, we are left to debate how this should work. Or look to textbooks written by committee members and other respected experts, who may have varying opinions.

c. Jim Meadows is a known authority, but he is one of several known authorities in the GD&T community who don't all agree with each other. I have one of Jim's books and have learned a lot from it. But we can't take his opinion (or anyone else's) as gospel, and it's fine to respectfully question his (or anyone else's) reasoning. Jim said himself that there was a long drawn out fight between the "moving simulator" and "stationary simulator" camps, and the stationary camp happened to win in 2009. This doesn't mean that one camp was right and the other was wrong - as with most things, there are advantages and disadvantages to both approaches.

d. Y14.5 states in 4.11.6 that "the appropriate MMB for determining the size of the datum feature simulator for an internal datum feature of size is the largest MMB that the datum feature(s) of size will contain while respecting the datum feature precedence". This is illustrated in Fig. 4-16 (c), with an FCF that has an A|B(M)|D(M) datum feature sequence. The simulators would be a 10.9 pin for B and a 7.5 sleeve for C. The logic is that the B and C simulators are not sized at the MMC sizes of the datum features, they are sized at the MMB sizes (this takes into account the geometric tolerances on the datum features as well as the size). This is intended to ensure that datum precedence is not violated (datum features B and C will always fit over their MMB simulators if they conform to their tolerances).

e. This is where things get weird and confusing. There can be clearance between the MMB datum features and their simulators, so the DOF constraint becomes loose. The secondary datum feature B is supposed to constrain two translational DOF's, but because of the clearance (there could be as much as 0.2) it may only partially constrain them. Datum feature C is supposed to constrain one rotational DOF (clocking), but because of the clearance (there could be as much as 0.6) may only partially constrain it. Y14.5 calls this datum feature shift. We can use this to adjust the part on the gage, to get the considered feature to pass its position tolerance. This is often highlighted as one of the advantages of MMB references and functional gaging. But how is precedence affected?

f. We have fixed-size MMB simulators for both the secondary and tertiary datum features. There may be clearance on one of them, both of them, or neither of them. I think we can safely say that the degree of freedom constraint becomes somewhat ambiguous. A|B(M)|C(M) results in the same MMB simulators as A||C(M)|B(M) or A|B(M)-C(M). This is the situation that Jim Meadows (and many others) had issues with - if the gage is exactly the same regardless of the specified precedence, then the precedence is lost. But where did it go? I wondered about this for years - how can we have a system in which different datum feature sequences result in the same requirement? This never sat well.

g. Before thinking about the issue again as a result of this thread (thanks to greenimi for keeping it going and probing into this further), I had only thought that the MMB simulators made the DOF constraint ambiguous. But now I am thinking that the MMB simulators actually violate datum feature precedence. Some of you have already suspected this, but now I think I see why. More on this later.

Evan Janeshewski

Axymetrix Quality Engineering Inc.

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[3DDave]

f. - to continue with the chess analogy - with no other pieces having moved, there is a pawn at BK4 and a pawn at WK4.
In one version of the game each player advanced their pawn exactly one position over a total of 4 turns.
In another version, each player moved their pawn 2 positions over a total of 2 turns.

How can a differing number of player turns result in exactly the same result?

When one considers material boundaries, consider the case that B is referenced RFS.
What is the MMB for C(M) under that condition relative to [A|B]?

There are two components - distance from B and tilt relative to A.

The result is not a cylinder. For inspection, making a rule about diameters is easier, even though such a calculation doesn't represent the variations that are allowed by the initial position callout.

 

[greenimi]

Excellent write-up Evan, Thank you
Just a couple of small issues:

Part d: You said:

The simulators would be a 10.9 pin for B and a 7.5 sleeve for C. The logic is that the B and C simulators are not sized at the MMC sizes of the datum features, they are sized at the MMB sizes (this takes into account the geometric tolerances on the datum features as well as the size). This is intended to ensure that datum precedence is not violated (datum features B and C will always fit over their MMB simulators if they conform to their tolerances).

For correctness, shouldn’t be a replacement between "C" and "D".
In other words, “D” instead of “C” in your sentence?
“D” datum feature simulator is a sleeve.
“C” datum feature simulator , a gage block / MMB is 1.9 not 7.5.
“D”/MMB is 7.5 (case/scenario c)

Part e: Same mixup between “C” instead of “D”.

Datum feature C is supposed to constrain one rotational DOF (clocking), but because of the clearance (there could be as much as 0.6) may only partially constrain it. Y14.5 calls this datum feature shift.

Should have been: “Datum feature D is supposed to constrain one rotational DOF (clocking), but because of the clearance (there could be as much as 0.6) may only partially constrain it. Y14.5 calls this datum feature shift.”

And by the way, datum feature “C” will have of maximum 0.2 (not “as much as 0.6”). Am I right?

On both parts of your post (part d and part e) you are referring to fig 4-16/case c.) and not to Meadows example. Correct?

Part f: Now, Mr. Meadows picture came into play for datum feature precedence violation.

Sometimes I am (too) stickler to details. Thank you very much for your review and pertinent opinion.

 

[axym]

greenimi,

Thanks for pointing out these details - my errors really added confusion. Wow, I was all over the place in that post ;^).

Yes, in those two comments I wrote C when it should have been D. Sorry, I guess I'm accustomed to A-B-C examples.

Datum feature C (the parallel-plane slot width) would have a maximum clearance of 0.2 and not 0.6.

In Part d. and Part e. I was referring to Fig. 4-16 (c) of Y14.5-2009, and not to the Meadows example.

In Part f. I kind of drifted over to the Meadows example (with the translation modifier removed).

EJ

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca