Composite callout for perpendicularity

[greenimi]

I have a part and I am looking for the right GD and T scheme. I have posted three proposed schemes and I am wondering which one will fit the best for the following scenario:
For performance reasons we must have Ø.100 size holes to be perpendicular to surface C in .010 T MMC.
At this point the drawing is incomplete by intent (surface C is not located properly, Ø.375 not located, etc), but I would like to get your pertinent opinion and to talk a little bit about my proposal datum scheme.
Datum A Ø.500±.003 bore is the one to orient the part into the assembly. Datum target B is the one to stop the remaining translation degree of freedom.
It's kind of strange, but the surface we would like to control perpendicularity from is relativly far away from the feature and because of this reason we have this issue on how to do it right.
1.) Which of these datum schemes it's the best for my goal (or maybe none)? I would say the first one, but what about the next two? Would they have the same effect?
2.) What about inspection purpose (gaging, CMM)? Which do you think is easiser to accomplish, qualify the part per the functionality/assemblability requirements and performance requirements (perpendicularity)
P.S. If I add a profile callout of surface C (relative to A and B)I was told most likely I will complicate the requirements beyond the level of expertise of some people so, I am trying to stay away from that.
And yes, ironically we preffer to use composite callout instead (but,...... if we can stay away from that is even better, so that's the reason is listed only as a third choise
Thank you for your help

 

[greenimi]

Maybe the Ø.100±.003 should read 2X for both (the size and positional callout), otherwise I think will not qualify for composite positional callout (applicable for patterns only, right?)
Thank you again

 

[pmarc]

Based on your description, I vote for the simplest one, that is scheme 2, but with tolerance value 0.010, not 0.008. As a matter of fact I am not sure where does this 0.008 come from. What is the purpose of it?

'2X' prefix should be there regardless of scheme chosen.

 

[powerhound]

I'm with pmarc. Scheme 2 has the holes located within .008|A(M)|B|C| and scheme 3 has the holes located within .010|A(M)|B|C|. Which is it? You asked which one was best for your goal but you have stated no specific goal. Scheme 1 is simply incorrect. Scheme 3 is also incorrect in that B is a point. You may be able to refine the holes' orientation to an axis but not to a point. Scheme 2 is legal but I don't know what your goal really was so I don't know if it best suits your need.

Can you please say exactly what it is you're trying to do?

John Acosta, GDTP S-0731
Engineering Technician
Inventor 2013
Mastercam X6
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

 

[greenimi]

So, let me clarify our thought process and answer your questions (pmarc and powerhound)

We want the position of these 2 holes to be in .008 relative to A at MMC and B. That is for the functionality of the product. And to address pmarc’s question where this tolerance is coming from I can say that is calculated thru some stackups and also based on some previous designs (empirical data)-so it’s a combination of those two.
Now, during the assembly of the product we have seen that if the surface C is not align properly with the two holes, problems occurs during subsequent operations and also we have performance product issues. Has been determined (research and development one hand and/or process capability on the other hand) that the proper (maximum tolerance we can live with) perpendicularity tolerance is .010.
Therefore, the question is how we want to say that in the GD and T language?
Scheme 2 (with C as a tertiary datum), was the first one proposed, and as Pmarc said with .010 (not .008 as depicted) positional tolerance, but we were afraid that we got too much of the datum shift A (A at MMB/MMC) and that translated over that distance (remember surface C is relatively far away) will put the two holes into a position we don’t want them to be in. So, we reduced a little bit (from .010 to .008) the available positional tolerance. Again, to make sure “the calculated” perpendicularity comes right. How we can calculate this perpendicularity based on the positional tolerance with A(MMB) B and C, beats me…. Any help will be greatly appreciated.
Because we don’t know to calculate this perpendicularity, the next proposal was, why not tolerance it directly—scheme 1- hole to surface C with .010. –does not matter how far away the surface are, and how much datum shift we have for A, the holes should be perpendicular in .010.
Powerhound and/or pmarc, why scheme 1 is incorrect? What could be the reason?
Somebody came up with the scheme 3 –composite positional tolerance –Reason: we want to stop only the rotational degree of freedom and we know the lower segment of the composite can do it. (and that’s why we choose composite versus two single segments). Why, again, you said that scheme 3 is also incorrect? What am I missing?
Maybe, (and I thinking out loud), do we need to add datum C at the lower segment FRTZF to get the equivalent results?
Thank you for your input

 

[pmarc]

greenimi,
I voted for #2 because your initial post mentioned nothing about 0.008 tolerance.

And I did not say that scheme #1 was incorrect. In the light of your additional clarification it seems reasonable choice. You control perpendicularity of holes directly to datum plane C, so the bottom callout is insensitive to any locational/orientational error of datum feature C relative to higher precedence datums A & B.

As for scheme #3, if your functional intent is to control location of holes relative to A(M)|B within dia. 0.008, the callout will not do it. It is the upper segment that controls location of the pattern to referenced datums, therefore you will allow for 0.010 wrt A(M)|B, not 0.008. Lower segment will merely tigthen orientation.

P.S.: Some food to eat. What if the callout was: |pos|dia.0.008(M)|C|A(M)|B|? Would that conflict with your functional demands?

 

[pmarc]

There should be |pos|dia.0.010(M)|C|A(M)|B| in last line.

 

[powerhound]

Clearly I didn't give Scheme enough thought. I didn't pay enough attention to that fact that the perp callout was to a different datum reference frame. Sorry about the bad answer on that one. I still think that scheme 3 is incorrect by virtue of the fact that datum point B is referenced. As pmarc said, the lower frame refines orientation and being that datum B is a point, there is no orientation to refine.

John Acosta, GDTP S-0731
Engineering Technician
Inventor 2013
Mastercam X6
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

 

[greenimi]

Pmarc,
I will answer to your question first: actually we did consider having the DRF as you described “C primary datum” (C A (M) B) before we decided to place a standalone perpendicularity control directly to C. The argument against that was made such as does datum feature C really qualifies as a primary datum? Does it have enough surface to be able to orient this part into the assembly, specially the mating part is on the other end of the part? And the answers for both questions were NO and the idea was scrapped because datum C does not do a very good job as a primary datum. If we are wrong on this assumption, yes, that’s a possibility.
Now, talking about our three schemes (proposals):
Scheme 1 is pretty clear. Those are the functional requirements.
Scheme 2: what value (calculated value) we need to have for positional callout to make this scheme 2 equivalent with scheme 1 (the requirements)? Food for thought: .008 -.006 (2x.003- total/maximum tolerance for datum shift A) = .002?
Scheme 3: Can we use Ø.008 for PLTZF (upper segment) and whatever number we get for scheme 2 (calculated one) –maybe Ø .002 for FRTZF (lower segment) to make all these three schemes having the same effect (saying the same thing in the GD and T language in three different ways)?
Maybe adding datum C on FRTZF?
Composite position: PLTZF Ø.008 A(M) B C,
FRTZF Ø.002 (?) A(M) B C(?)
What do you think? –Again, as usually, some questions are for education/academic purpose only.

Thank you very much.

 

[pmarc]

greenimi,
Yes, datum feature C does not seem to be a good feature to orient the part for measurement relative to |C|A(M)|B|. But in this case my question is, will surface C serve as good datum feature for establishing a datum for perpendicularity callout in scheme #1? Wouldn't it be primary datum feature too?

To be honest, I do not see how these 3 schemes may result in equal GD&T requirements. Just comparing scheme #1 with #2: even if you find tolerance value in #2 (let's say .002), the single FCF |pos|dia.0.002|A(M)|B|C| will automatically tighten position of holes wrt A and B within .002, which will be four times tighter than .008 defined in upper FCF in scheme #1.

 

[greenimi]

Pmarc,
Yes you are right, and that is the main reason we came up with scheme 2 (to not have surface C as a primary datum for perpendicularity, as this surface C is not good as a primary datum nor for position and nor for perpendicularity).
In other words the question is: how we can say in GD and T language that we want these two holes be perpendicular to surface C without breaking any rules, regulations or common (good) practice (such as primary datum to be "big" enouugh to be capable of orienting the part into the assembly as in the real world)?
I guess that saying, for example composite position PLTZF Ø.008 A (M)B (not C) , and FRTZF Ø.002 A(M) B and C (for orientation/perpendicularity ONLY) is illegal.
Also in my opinion illegal is using two single segments tolerance such as:
first segment position Ø.008 A(M) B
second segment position Ø.002 A(M)B and C
So, again, the dilema is how to say we want the requirements are position of.008 wrt A(M)B and perpendicularity of .010 to C, without using C as a primary datum? (assumming .002 --when C is used as a tertiary dautm--is the caluculated value driven from perpendicularity minus datum shift from A)

 

[CheckerHater]

If I had to machine your part, I would first put your DIA .375 into vee and clamp it down.

Then I would try to machine both flat “C” and 2 DIA .100 holes FROM SINGLE FIXTURING without re-clamping. That would result in the best possible relationship between flat and holes that you can produce.

To make DIA .500 bore I would hold DIA .375 surface in a chuck.

In this process (yes, process!) your DIA .375 will become primary machining datum for flat, bore, and pair of holes. This is called SIMULTANEOUS REQUIREMENT.

On the drawing I would mark DIA .375 as primary datum and position the flat, the bore, and the holes within .005 wrt DIA .375 RFS. It’s do-able. This will create SIMULTANEOUS REQUIREMENT on your drawing. It will hold your .500 bore and .100 holes together as well as provide proper clocking between .100 holes and flat.

And find something better for datum B. Sharp point derived from a cone in not the best choice, even if it’s “functional”.

 

[greenimi]

CheckerHater,
Thank you for your input from the manufacturing perspective. Sometimes I am stuck into the design paradigm. I think a middle balance (input from all sides involved, design, quality, manufacturing) is needed. I know the subject matter experts and the applicable standard/ available books endorse the functional criteria. I don't want to start again this discussion.
As a matter of fact pmarc had a great thread awhile ago about this, called Functional vs. Process Driven GD&T. Thank you pmarc.
At this point, I am interested more in a mathematical relationship (if it's one available) to make those three schemes have the same effect (produce equivalent results and have the same meaning)

 

[CheckerHater]

I have even less desire to start a discussion, but I would like to make a short note to defend my choices, which, by the way were made based on “applicable standard / available books “and “matter experts” opinions, not to mention “functional requirements”.

Let’s start with the books.
Y14.5 paragraph 4.7 DATUM REFERENCE FRAME states:
“Sufficient datum features or designated portions of these features are chosen to position the part in relation to a set of three mutually perpendicular planes, jointly called a datum reference frame.”

Datum features have to be “sufficient” and your “C” flat definitely isn’t.
It is not my fantasy; in your own words:
“Does datum feature C really qualifies as a primary datum? Does it have enough surface to be able to orient this part into the assembly, specially the mating part is on the other end of the part? And the answers for both questions were NO and the idea was scrapped because datum C does not do a very good job as a primary datum”
You are absolutely right; you cannot drill holes in Detroit perpendicular to plane in Chicago.

Now, what are actual functional requirements?
Again, in your own words:
“Now, during the assembly of the product we have seen that if the surface C is not align properly with the two holes, problems occurs during subsequent operations and also we have performance product issues.”
So, we have a shaft with different features located at different ends of it. Those features have to be “clocked” to each other, otherwise part doesn’t work. Let’s take a look:

http://www.tec-ease.com/gdt-tips-view.php?q=123

Apparently, when you want two features on the ends of a shaft to be perpendicular to each other, simultaneous requirement is THE TEXTBOOK EXAMPLE according to well respected expert on the matters of GD&T.

And about “mathematical relationship”. (Your own words, you guessed it):
“…that is the main reason we came up with scheme 2 (to not have surface C as a primary datum for perpendicularity, as this surface C is not good as a primary datum nor for position and nor for perpendicularity).
In other words the question is: how we can say in GD and T language that we want these two holes be perpendicular to surface C without breaking any rules, regulations or common (good) practice…”
You are actually asking “How to say ‘feature A is perpendicular to feature B, but feature B is not a datum for feature A’”.
I would say it’s mathematically impossible. Unless you can say that both A and B are SIMULTANEOUSLY perpendicular / parallel to something else.

Have a nice day

 

[greenimi]

CheckerHater, Thank you again for your input.
In our meetings we have considered adding a profile of .008 wrt to A(M) and B--if you see my original posting is even mentioned there--
Now, I guess, if we add the above mentioned profile callout, that will take care of the location of surface C and will bring simultaneous requirements in effect because the two holes position will also be wrt A(M) and B. Therefore, a mutual relationship between the two holes and surface C is going to be established and consequently a perpendicularity relationship.
So, again, position wrt at A(M) and B and profile wrt to A(M) and B, means perpendicularity between holes and surface C. I agree here. But, again, stated in my initial post--"I was told most likely I will complicate the requirements beyond the level of expertise of some people so, I am trying to stay away from that". That's of course, not the GD and T language fault, it's our fault. We plead no contest.
That's being said, I am trying to find an alternative way to say what we want with different "words" (since GD and T it's a continuously developing language) our people/vendor/supplier will understand. if it's not possible (and the mathematically relationship is impossible) that's fine and we have to take it as is.
Have a nice day too.

 

[CheckerHater]

Here is the thing:

I cannot think of many different ways to call two features perpendicular.

Number 1: Direct dimensioning. 90 DEG +/- 0.5 DEG.
Out of question because you want GD&T control.

Number 2: Direct geometrical tolerancing. One feature is a datum, other is called out perpendicular to that datum.
You mentioned flat “C” being no good use as a datum.

Number 3: Indirect geometrical tolerancing. Both features have geometrical controls that allow us to imply basic 90 DEG angle between them.
This is where sim. reqt. magic kicks in.

I am really interested to hear about Number 4.

I understand your concern about sim. reqt. being not well understood. I gues you have to fight this battle uphill. Maybe show your shop that tec-ease tip
Good luck! After all, what’s the point of using GD&T assuming that no-one understands it anyway?

 

[greenimi]

That was exactly my orgininal intent: " to find Number 4" on your list, so I have started this thread here knowing a lot of people on this site are experts/certified professionals on GD and T.
I thought composite callouts (or even two single segments) will help somehow (that's why I called the thread Composite callout for perpendicularity), knowing that the lower segment will refine the orientation on both of these callouts (composite and/or two single segments).
We were just trying to find the right tool for the job, but looks like we haven't found it yet.

 

[bxbzq]

Why is 2 single segments position tolerance not in your list?
pos|dia.010|A(M)|B|C|
pos|dia.008|A(M)|B|

 

[greenimi]

Bxbzq,
Actually, if you are following the thread, you will see that I proposed two single segment control either (not among the schemes I draw, but in the verbiage). But looks like nor the composite callout and nor the two single segments will provide equivalent results to the original intent (functionality).
I have shown composite (and not two single segments) because composite on the lower segment can stop only rotational degree of freedom (not translation) and what is needed for our part is to stop rotation (the other degrees of freedom are already being stopped by higher precedence datums A and B).
On your proposed scheme,
pos|dia.010|A(M)|B|C|
pos|dia.008|A(M)|B|

we talked about that too, and again, as stated before, we were afraid that we got too much of the datum shift A (A at MMB/MMC) and that translated over that distance (remember surface C is relatively far away) will put the two holes into an orientation relative to surface C we don’t want them to be in (way over acceptable .010). So, we have asked for so called “calculated value” (.002 for example)—equivalent mathematical value. Therefore, your scheme will become:
pos|dia.002|A(M)|B|C|
pos|dia.008|A(M)|B|

which position wrt A and B is (like pmarc said) four times tighter than the original one.
Looks like we don’t have a “legal” option (legal= to follow all the GD and T rules, abide by Y14.5 standard)

 

[bxbzq]

I don't know the function, but possible to assign the two holes as datum, say D, and the surface C being perpendicular to datum D at .002? Maybe use profile tolerance .002 relative to A|D? Of course the two holes are positioned ø.008 relative to A(M)|B.