Profile (for a radius defined with ¦) used to control location, can also control the size? 1

[greenimi]

Is the profile initially designed/allocated to control location also control the size for the radius?

Has been discussed on this forum a couple of times (with even some controversy about 8.18 and 8.27 from Y14.5-2009 standard), but I would like to see how would you define the following –very simple—situation:

Radius size can be between .100/.040 (I know is defined with ± which is the controversy, but that’s the drawing I have). The form cannot be bigger than .010 (for example, if the radius on one side is .090 on any other side cannot be smaller than.080—that’s the intent)


We would like to control also the position of this radius to datum A which must be in .005.
The question is: if we add the profile control of .005 to A (intent to control the location) would that also control the size in .005 (and not in .010 which really we can live with)?

Thank you

 

[pmarc]

greenimi,

My advice. If you just have a possibility, stay away from directly toleranced radius in combination with profile callout as far as possible. You will save yourself a lot of trouble.
Even if we find agreement here on the forum on how to interpret this really weird tolerancing scheme, I am pretty sure that other readers of your drawing will face the very same dillemma later on.

 

[CheckerHater]

How about:
1. Changing your radius to BASIC .070
2. Changing upper frame in your composite to .065 (to represent virtual condition from size and location combined)
3. Changing lower frame of your composite to .060 (to represent variation of size alone)
As both requirements have to be satisfied together, your .060 zone will be floating within .065 zone in the manner shown in Fig.8-19 in 2009

 

[pmarc]

Switching to basic radius and playing with profile tolerances would be the best option in my opinion.

Question is: can we do this or must the directly toleranced radius stay on the drawing because of reasons we do not know?

These two options - 1) directly toleranced radius + profile; 2) basic radius + profile - will never result in identical geometrical requirements, so such switch should not be done just like that.

 

[greenimi]

Pmarc,
I agree with you to stay away from directly toleranced radius in combination with profile callout (ONLY IF I can).
Regarding your comment that directly toleranced radius must stay on the drawing, other than:
“If we add the basic radius and the profile callout, then nobody understands it anymore” argument, --which personally I don’t buy it--- I don’t see a reason on why the ± radius should stay on the drawing.

So, If I just make a recommendation to update this drawing, how the GD&T callouts would look like?

Let me try first and please add anything that I am missing, if you would.
Make radius R.100/.040 as a basic dimension .070. –I agree here with CH.
Add composite profile callout (or multiple single segments would work the same in this case?) as follow:
First line: profile .005 to A (intent: to control the location to A)
Second line profile .010 with no datum (to refine the form)
I know, it’s not quite right, but I don’t know, how can I make the form in .010, but the location in .005.—and I don’t know how CH came to .065 and .060.
Thank you again

 

[pmarc]

greenimi,
Having profile of .010 without datum references (as a control of form and size of the radius) and of .005 wrt A (as a control of its location to A) would be wrong in my opinion. Form and size should be a refinement of location - and in your case it will not be.

If you are fine with size & form tolerance of .010, you should be also fine with .010 of locational tolerance. Keep in mind that basic location of basic radius center relative to A will not change with the change of profile tolerance value.

So, instead of searching for overcomplicated schemes, maybe simple |prof.|.010|A| could work for you just fine?

 

[CheckerHater]

Your radius is .070 +/- .030, right?
That's .060 total tolerance zone, right?
Your .060 tolerance zone allowed to float within another .005 tolrance zone; that's in your post, right?
So, your overall tolerance zone is .065. When .060 tolerance zone is floating inside of .065 tolerance zone it's only floating .005, just what you wanted. Or did you?

 

[greenimi]

CH,
Hmmm... I understood now what you did, but I am not sure that will reflect what I described. We want the size refinament to stay in .010 and not in .060 as per your spec.

From my previous post "Radius size can be between .100/.040 (I know is defined with ± which is the controversy, but that’s the drawing I have). The form cannot be bigger than .010 (for example, if the radius on one side is .090 on any other side cannot be smaller than.080—that’s the intent)"

 

[CheckerHater]

I’ve re-read your post and it confused me a bit.
If you can hold the radius within .010, what’s the point specifying it +/-.030?
Either way, you cannot “refine” smaller tolerance with larger tolerance.
The solution could be applying third FCF, but is it worth the trouble?
.015 in upper frame and .010 in lower frame will keep your “form” within .010 and allow “location" to float .005
Is it really critical to loosen up “size”?

 

[pmarc]

greenimi,
Your requirements really seem not to be consistent.

 

[powerhound]

Shouldn't a true profile ALWAYS be defined by basic dimensions? If so, then using a +/- tolerance in conjunction with profile is incorrect and should not be considered. That being said, it is permissible to locate a true profile with +/- but the profile being located must still be defined by basic dimensions.

I believe that if profile is to be used then there is no other choice but to make the radius a basic .070. You can then decide how to define your tolerance. How about composite profile--as others have suggested--with .060 and related to datum A (and probably a secondary datum) in the upper frame, and .010 in the lower with no datum reference to control form?

You should never have larger value in the lower frame in any case though.

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

 

[drawoh]

greenimi,

I think I see what you are trying to do.

How about the following...
[ul]
[li]A basic radius dimension[/li]
[li]A sloppy profile tolerance[/li]
[li]An accurate runout tolerance[/li]
[/ul]

The radius can be sloppy, but the form must be consistent all around.

--
JHG

 

[greenimi]

pmarc,
RE"Your requirements really seem not to be consistent."
I wish I don't have to agree with you here, but I could not.... Some of inconsistencies come also from the drawing. After some discussions before the original post, I found out that we have those requirements on the drawing (± radius + profile), but in reality we really mean something different, but due to the lack of knowledge we could not place the right callouts on the drawing.

What we could not figure out how to say on the drawing is that we need the size to be between .100/.040, the location to datum A in .005 and the form in .010. --what we really mean by form is for example, we can live with the radius to be .100 on one side of the part, but no less than .080 at 180 degrees from that side, in other words we don't want to have a radius of .100 on one side and .040 at the opposite (180 degree). If it's confusing, you are not the only one.... welcome to our world....

Why " If you are fine with size & form tolerance of .010, you should be also fine with .010 of locational tolerance" ?


drawoh
On your tolerance scheme - basic dimension (I guess everybody agree with the need of basic dimension), sloppy profile tolerance and runout, profile does need datum A or not?
Runout for sure will have datum A, but what about profile? Runout will control the location to A, right? Profile will control size/form, right?

 

[CheckerHater]

Runout is very good idea, but runout will asorb both "locational" .005 and "form" .010
Is there a particular need to specify them differently?

 

[pmarc]

greenimi,
The reason I said that your requirements were inconsistent is hidden behind numbers. As it was already stated by some guys, you can not have value of tolerance responsible for surface "size" control greater than values of tolerances dealing with its form and/or location. This is against composite profile of surface concept. So either "size tolerance of .060 (.100-.040), location tolerance of .005, and form tolerance of .010" means something different to us, or you want to meet requirements that are in conflict to each other.

The reason I said: "If you are fine with size & form tolerance of .010, you should be also fine with .010 of locational tolerance" is related to above. If you are fine with actual surface of blend occupying .010 tolerance zone unrelated to any datum, how can you not be fine with the the surface falling within .010 tolerance zone related to A? I do not get it. How can you expect that actual surface disposed within .010 tolerance zone unrelated to any datum (so that it "touches" both tolerance zone boundaries) will suddenly meet .005 wrt A requirement? Again, I do not understand.

I understand what you mean by: "The form cannot be bigger than .010 (for example, if the radius on one side is .090 on any other side cannot be smaller than.080—that’s the intent)". That is the clearest requirement of all, in my opinion. And based on that (also assuming that you are fine with surface of blend falling within .040 - .100 profile tolerance zone), I think composite profile FCF:
|.060|A|
|.010|
is the reasonable option. I guess this is exactly what powerhound suggested.

You may also want to add A reference to lower segment of the FCF in order to keep the lower tolerance zone oriented to A.

 

[pmarc]

And in case if you decided to add A reference to lower segment of composite profile FCF this would work in the same way as "a sloopy profile tolerance + an accurate runout tolerance".

 

[greenimi]

Okay, Thank you everybody
I think I finally got it. I understood now, how the drawing should look like and why some (verbal/written) requirements were in conflict to each other.
Now it’s time for me, to explain it to the end user…..That’s even harder than the step we just accomplished with your help. But you know it’s a famous quote:


If you can't explain it to a six year old, you don't understand it yourself. –Albert Einstein

Thank you again

 

[drawoh]

CheckerHater,

I have to admit I do not totally understand run-out. If it controls form, what does profile do? Perhaps I should have suggested circularity.

--
JHG

 

[greenimi]

Drawoh,
Just my 2 cents: circularity does not locate, but the runout does. We need that location constrained tolerance zone.

 

[pmarc]

The thing is that even with runout tolerance applied to the surface, location of the surface, that is its distance from datum axis, would still be controlled by "sloppy" profile tolerance wrt A.