What is the difference between these? 2

[pmarc]

The question is simple:
What is the difference between two cases shown on attached picture in terms of process of establishing datum axis for total runout measurement? Is there any or not really?

http://files.engineering.com/getfil...-43cc-83c1-857a300493c9&file=cont_feature.pdf

I am not asking which of the approaches is better or is there any other option, I would just like to know if you see any difference and if yes, how this can be explained. Thanks a lot.

 

[ctopher]

I see it as:
Case 1, centered to the one end diameter.
Case 2, centered between the two diameters.

Chris
SolidWorks 11
ctopher's home
SolidWorks Legion

 

[fsincox]

Good question, nice picture. Wow, you know this question is going to occur. I would have seen it as the same, because it is "treated geometrically as a single feature".
Frank

 

[CheckerHater]

I say it depends on tools available.

Case 1: You have to grab BOTH end diameters with ONE chuck or collet, which may create some trouble.

If you agree to use V-blocks, then probably no difference. (Or difference between two short blocks and one really long one)

 

[pmarc]

Let me clarify:
I am interested whether both dimensioning schemes result in the same definition of geometrical relationship between datum features or not. As for now I see Frank has voted they are the same. Any other votes?

 

[CheckerHater]

No, relations between datum features are not the same.
In case 1 both features should lay within the same envelope.
In case 2 no relation between [A] and is implied on the drawing

 

[powerhound]

I'm inclined to say no, but with reservations. I believe the relationship between datum features may be affected by way of the AME being different. In case #1 both ends must conform to a single UAME whereas in case #2, each feature has its own UAME.

Again, I haven't given this a lot of deep thought but it seems like that may create different inspection scenarios and thus potentially different results.

Powerhound, GDTP T-0419
Engineering Technician
Inventor 2010
Mastercam X5
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

 

[axym]

pmarc,

This will open up another can of worms or two. But I think you already know that ;^).

In Case #1, I don't remember if the presence or absence of extension lines makes any difference to which of the OD's are included in the continuous feature. If both are included, then the datum axis would be established using the minimum circumscribed cylinder of the two OD's.

In Case #2, the datum axis would be established using simulators shrinking down on both A and B. But Y14.5-2009 is not completely clear on the details of exactly how the simulators would shrink down. Some say that the axis would be established using the minimum circumscribed cylinder of the two OD's, identical to the CF case. This would allow the possibility of an unstable axis, if A and B were different as-produced sizes. Others say that the A and B simulators would keep shrinking down independently, until a stable datum axis was established.

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[pmarc]

Evan,
For case #2, what if for as produced part, the offset of datum feature's axes was 1mm. How the datum feature simulators A and B would then look like?

 

[fsincox]

CH,
Good point, You are right, I forgot about that!
Pmarc,
Sorry, I think CH is right, meaning we would need more info.
Frank

 

[dingy2]

In my opinion, they are both the same. CF is new for 2009 and reflects both diameters while the lower pic has 2 datums, one for each end diameter.

The axis is created by both end diameters together.

Dave D.

http://www.qmsi.ca

 

[pmarc]

Frank,
Of course you would need more info because case #2 is incomplete by intent. I did it hoping someone will answer like CH and you, so my apologies for setting this small trap.

But what is the most important for me, your and CH's responses just proved that showing datum features coaxial on a print does not mean their datum feature simulators are automatically considered coaxial too. If there is no locational relationship defined between datum features, you are not able to tell how the simulators look like - what is their size, mutual orientation and location. Look again to your 100+ thread for analogy.

I just wish I was finally able to convince Evan, J-P and Dean on this. Yeah, I realize they will probably be citing p. 4.5.2 to reinforce their point of view, but I can use p. 4.9 to reinforce mine.

 

[fsincox]

I can take no credit; I have to thank CH,
I really like the fact that we get to see many points of view here. This may take forever if you had to think it through just by yourself.
You, obviously, believe this is analogous to my case, this one being coequal and the other primary-secondary, let me ponder that.
I do appreciate your not just writing me off.
Thank you,
Frank

 

[DeanD3W]

pmarc,
I agree with all the responses that have been posted... I don't agree that section 4.9 tells us that the case 2 datum feature simulators are not required to be coaxial without a coaxiality tolerance applied to datum features A & B. Thank you for getting me to read section 4.9 one more time... Now I don't like it too much . Taking 4.9 further than intended (IMO) does not provide license to set aside the very clear requirements of 4.5.2.

I much prefer that establishing a datum be a process that is separate and independent of any tolerance evaluation of the datum features. I don't like the fact that portions of chapter 4 now tend to link the two processes.

Within reason, no matter how imperfect a set of datum features are, one should still be able to establish a datum reference frame. There should be no linking of tolerance evaluation and datum reference frame establishment processes.

My vote is that section 4.5.2 clearly tells us that the case 2 datum feature simulators for A and B are related in a "perfectly" coaxial way.

As Frank pointed out, a coaxiality control needs to be added to case 2 anyway... When this addition is made, absolutely no change should be made in the process or simulators for establishing A-B for case 2.

The confusion that Evan points out regarding potential simulator behavior differences between cases 1 and 2 is unfortunate, but real, with Y14.5 as it is now. It is currently a point of discussion within both Y14.5.1 and Y14.5 committees. I don't think those issues affect the argument you are making here, or my response to it.

Dean

http://www.d3w-engineering.com

 

[pmarc]

Dean,
Sorry but this still does not convince me. In addition to my discussion with Jim Meadows, I also spoke with Alex Krulikowski on Frank's thread (or actually on my second sketch that I created in that thread) and he also supported my standpoint. I am not saying that based on this I am 100% right, but at least I have a right to think that your interpretation is also doubtful.

 

[drstrole]

Section 4.5.2 tells us that the datum feature simulators are in-line, it is not discussing the datum features. The datum features need to have the appropriate control to locate the to each other. The CF in the one example does that, the other example is incomplete.

Drstrole
GDTP - Senior Level

 

[DeanD3W]

pmarc,
Isn't 4.5.2 in Y14.5-2009 pretty clearly stated? Where do you see the weakness in its wording?

Dean

http://www.d3w-engineering.com

 

[pmarc]

Dean,
In the other thread you said:

1) Please read section 4.5.2 of Y14.5-2009, then
2) draw an axial direction view of the datum feature simulators, then
3) draw a part with a nicely oriented, but poorly located slot, then
4) rotate the part to bring the center of that slot up to fit around the datum feature B simulator...

This descripition perfectly matches to the part shown in my second sketch in Frank's thread:

http://files.engineering.com/getfil...cf-012349c1f676&file=keyway_parallelism_2.JPG

and is absolutely correct under one fundamental condition which is the key here - the distance between datum feature simulators A & B is defined by basic zero dimension

This is actually how I understand 4.5.2. You know I have no authority to go into semantics here, but for me the word "shall" means "must". So there must be something that defines this basic relationship (by use of a method listed in para. 4.9), otherwise it does not exist. If the relationship was implied, the very first sentence of 4.5.2 would be something like: "Datum feature simulators have the following requirements" and not "shall have".

That is why I keep saying the first two cases in the link above are incomplete. That is why I keep saying case #2 from original sketch in this thread is incomplete too. They all do not meet requirements of 4.5.2 and 4.9, so the simulators B can be located anywhere relative to datum axes A and it is not possible to calculate their boundaries wrt A in any manner. It is not even possible to build a physical datum feature simulator B properly because you do not have product positional tolerance between features A & B based on which you would calculate 10% positional tolerance between simulators A & B in the fixture.

So I absolutely agree with your statement from the other thread that confusion comes from the fact that we have different interpretation of how datum feature simulators work. I am just afraid that we will not find an agreement on this because the standard is not precise enough to indicate which of us is right.

-----------------------------------------------------
There is a Tec-Ease tip showing something similar to our situation here. Notice this one extremely important detail which does the difference - it is a position callout applied to a slot width defining basic location between datum features A & B and in consequence between their datum features simulators and making the Tec Ease's example similar to my case #3. Without this callout, you would not even need basic 14 dimension, so you would not know the basic distance between simulators A and B, so you would not know how a datum reference frame for position tolerance of small hole looks like.

http://tec-ease.com/gdt-tips-view.php?q=104

 

[fsincox]

pmarc,
Your last posted examples do not contain the original condition? Have we moved on and I just did not catch it?
I am sorry for my confusion
Frank

 

[DeanD3W]

pmarc,
There is no reason that a basic dimension could be considered to exist in the absence of a tolerance specification. I can't see how you can possibly read 4.5.2 in the way that you're describing. Deleting "shall" has no effect upon the requirements in this case, as far as I can manage to comprehend (then again, I am in China at the moment, so maybe I'm reading things upside down ).

Do others see pmarc's point? I may be missing something.

Dean

http://www.d3w-engineering.com