Perpendicularity

[randy64]

I am seeing a conflict on how this works.

I am in a training class right now (TCI, Al Neumann). He states that Perpendicularity can be referenced to datums that are not perpendicular to the surface being held perpendicular. He has an example in his workbook showing it that way also. He refers to Perpendicularity, Parallelism and Angularity as Orientation controls (which they are), but seems to indicate that they are really flatness controls with orientation and that Perp, Par and Ang are just our perception of the direction of orientation.

In Alex Krulikowski's book "Fundamentals of GD&T" he explicitely states that the surface being called Perpendicular must be perpendicular to the datums it is being referenced to in the feature control frame.

Which is it? I've always gone with Krulikowski's interpretation. If we didn't, we could just get rid of Perp, Ang and Par and call it all Orientation.

What says you?

Thanks.

 

[randy64]

dingy2, I wish it were just a picture that was in error. In fact, the picture represents an interpretation that states that the secondary datum does NOT have to be perpendicular to the surface in question - it is merely used to control rotation.

I believe this interpretation to be in error.

 

[vigildesign]

Look at the angularity callout in fig. 4-24 in Y14.5-1994. It is controlling perpendicular to datum "B".



Marcelino Vigil
GDTP T-0377
CSWP

 

[randy64]

vigildesign, you're right. Datum B is perpendicular to the surface in question (datum D). It is shown on the drawing to be 90 degrees from the datum B. I have no problem with a datum perpendicular to a surface controlling the perpendicularity of that surface. I do have a problem with trying to control the perpendicularity of a surface with a datum that is parallel to that surface. Just doesn't make sense to me.

 

[PaulJackson]

I don't see anything wrong or odd about Al's example and agree totally with caseynick! The caption in figure 6-35 pertains to that example since the tolerance zone happens to be oriented perpendicular to both datum features. The extension of orientation control to more than one datum feature in 6.6.1 makes no requirement for the zone to be perpendicular to all datum features in the DRF.

Paul

 

[randy64]

PaulJackson, doesn't the definition of Perpendicularity make a requirement for the zone to be perpendicular to all datum features in the DRF? It [6.6.4] states, "Perpendicularity is the condition of a surface, center plane, or axis at a right angle to a datum plane or axis." To me it couldn't be any plainer than that. Yet you are right, 6.6.1 makes no requirement of perpendicularity, but that's because it is a general statement on orientation. It is simply stating that two datums can be in relation to the surface in question, that's all. It is further defined for Perpendicularity in 6.6.4.

If Neumann's way of thinking on this subject were accepted, it would hold true for Parallelism also, i.e. you could reference a secondary datum surface that was actually perpendicular to the surface that is being held parallel. But that condition is never shown in Y14.5M-1994. Nowhere can I find a surface being held parallel to two datums. This is because it can't be done, it's the nature of Parallelism. I believe the same to be true of Perpendicularity - it CAN be perpendicular to two surfaces at once, so it can be held to two datums.

I think those that are agreeing with Neumann are creating something that is not in Y14.5M-1994. I think it goes over the line of creating something that is not really there.It is at the least a very liberal interpretation of Y14.5M-1994. When it comes to documents like this, I tend to be very conservative - I look at what is there and go within that framework, adding as little as possible.

All the responses in agreement with Neumann so far have not pointed to anything in Y14.5M-1994 that allows this, yet I've pointed out things that do not allow it and no one has directly refuted it.

Now, I try to be as open minded as possible. That's the reason I started this thread. I saw something that was different than what I had always believed and figured I had better hear both sides so I could make an informed decision. The knowledgeable people on this forum have changed my way of thinking on other subjects in the past, so I'm open to that, but so far I stand unconvinced on this one.

 

[dingy2]

Paul:

I respectfully disagree with you but, initially, I didn't think anything was wrong with the example.

The surface on example 6-35 of the 94 standard reflects a plane that is perpendicular to both datum A and also datum B.
The example in discussion is not the same. The feature is perpendicular to datum A but parallel to datum B.

Maybe there is an figure in either the 94 or 09 standard that reflects the example but I don't think so. If there is, please let me know.

Dave D.

http://www.qmsi.ca

 

[randy64]

dingy2, the 2009 standard does indeed have a picture very similar to the one I posted (one of the datums parallel to the surface in question), but all the verbage about orientation is the same as the 1994 as far as I can tell.

My customer follows 1994 right now, so that's my main concern. If ASME has adopted Neumann's take on it for 2009 then, as I stated before, it shouldn't be called Perpendicularity IMO. It should all just be called Orientation or Angularity.

 

[dingy2]

Figure 6-4 in 2009 edition is exactly a replica of your example so the example is correct to the 2009 standard.

The angularity symbol as shown in figure 6-4 as an alternative practice is much more precisely reflects the situation and should be used rather than perpendicularity.

Dave D.

http://www.qmsi.ca

 

[axym]

ASME Y14.5M-1994 states (in 6.6.1) that more than one datum feature can be specified in an orientation control. But then it only addresses "pure" cases in which the considered feature has the same relationship to both datum features. I don't think that the explanations and examples in Y14.5-1994 give a definite indication of what symbol to use in "mixed" cases like the one in the TCI example that was posted. It's open to interpretation, so I don't think we can really determine what is "correct" per the standard.

Let's revisit Fig. 4-24 on page 69, in which there is an Angularity control that references two datum features. The considered feature (a planar surface) is at an angle of about 30 degrees from datum plane A (the basic angle is not shown) and 90 degrees from datum plane B. Technically, this meets both the definition of Angularity in 6.6.2 and the definition of Perpendicularity in 6.6.4. The surface is at an angle other than 90 to a datum plane, and it's at a right angle to a datum plane. I think we all agree that the Angularity symbol is appropriate here and not Perpendicularity, but we disagree on the reason why.

What it comes down to is this question. If we wanted an orientation control for the same feature, that references B as primary and A as secondary, what symbol should be used?

One option is to choose Angularity (really, to rule out Perpendicularity) because the considered feature is not at a right angle to both datum planes. Another option is to choose Perpendicularity because the primary relationship is perpendicular (i.e. the considered feature is at a right angle to the primary datum plane).

Both options will be confusing to some. If Angularity is used, some people will be poring over the drawing trying to find the non-90 basic angle between the considered feature and datum plane B. There is also the possibility of a Perpendicularity control changing to Angularity if a secondary datum reference is added. If Perpendicularity is used, some people will find this non-intuitive because the considered feature is not perpendicular to datum plane A.

Personally, I'm most comfortable with the option of basing the symbol on the considered feature's relationship to the primary datum feature. This is the approach suggested in the TCI book, and used the most in industry AFAIK. This isn't perfect, but it doesn't make my head explode.

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[PaulJackson]

"PaulJackson, doesn't the definition of Perpendicularity make a requirement for the zone to be perpendicular to all datum features in the DRF?

No I don't think so.

"It [6.6.4] states, "Perpendicularity is the condition of a surface, center plane, or axis at a right angle to a datum plane or axis."

I have always interpreted this statement as "establishing a right angle (90 degree implied) basic relationship between the primary datum feature ("a" datum plane, or axis) and the featured controlled for orientation." No more, no less... likewise a zero or 180 degree implied for parallelism and a specified basic to the primary for all else.

Paragraph 1.1.4 covers the illustration vs. definition justification.

I would be just as happy with one symbol for orientation, having implied basics specified, and as many datum feature designations as necessary to constrain the rotational degrees of freedom that are functionally necessary for the control.

paul

 

[powerhound]

Paul,
As I mentioned in an earlier post. This seems to be another way of specifying profile of a surface, only more confusing. If the symbol were changed to profile of a surface, then it would mean exactly what you and Evan are saying except no one would be questioning it or disagreeing over it. I don't see any reason to use this method in lieu of profile of a surface. This one causes problems and the profile of a surface doesn't.

Powerhound, GDTP T-0419
Production Manager
Inventor 2009
Mastercam X3
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

 

[PaulJackson]

Powerhound Profile constrains translations as well as rotations. Orientation controls constrain only rotations.

Paul

 

[powerhound]

So are you saying that profile of a surface will mean something different than this iteration of perpendicularity?

Powerhound, GDTP T-0419
Production Manager
Inventor 2009
Mastercam X3
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

 

[randy64]

PaulJackson, I think you're reading more into it than is there with your interpretation when you say "and the feature controlled for orientation." 6.6.4 doesn't say that, it says "Perpendicularity is the condition of a surface, center plane, or axis at a right angle to a datum plane or axis." In my mind, saying "a" datum plane doesn't limit you to one datum plane. I interpret that as "any" datum plane.

Like I said above, I'm conservative in these matters. I believe my interpretation fits within what is stated within the standard, but Neumann's does not.

I don't have access to a lot of GD&T material, but I'd be curious if anybody can share other sources that interpret this the same way Nuemann does. As I stated in my original post, Alex Krulikowski states in his book that Neumann's interpretation is wrong. He has a flow chart and one of the questions is "Is the toleranced feature shown 90 degrees to the datums referenced?" If the answer is "No", then he calls it an "illegal" callout. I'll attempt to link it for you. Scroll to page 185.

 

[PaulJackson]

It would if the Al Newman example was changed to profile but not the figure 6-35 example.

Paul

 

[PaulJackson]

Randy,

Alex and Al are both on the committee that codifies the standard but neither of their own materials are the standard! It appears that one could say by Alex's flow chart of leagal/illegal and Newman's illustration that the two of them differ in their interpretation... I'll go with Al's from what I read in the standard... You can go with Alex from what you read in the standard... it appears that the we are not alone in our differing opinions.

Paul

 

[randy64]

Paul, you're no fun! I think we must stay on this forum and argue until one of us wins!! ;-)

Yeah, I don't think planes will fall out of the sky or cars will fall apart over the interpretation of this. I guess I'm just a little OCD about having consistency in the standards I'm following.

Peace.

 

[pmarc]

Randy,

As you said in one of your first posts, in Y14.5M-2009 exactly the same situation like you have is shown (fig. 6-4). I totally agree with you that it looks illogical (or at least confusing) to control perpendicularity of a feature relative to datum feature that is parallel to it, but on the other hand the standard clearly shows that this is legal, so there should be no doubt about it. In fig. 6-4 the secondary datum is invoked only to constrain an additional rotational degrees of freedom of the DRF. This leads me to conclusion that only primary datum feature from Perpendicularity callout should be perpendicular to specified feature and it is less important whether the secondary datum feauture is parallel or at right angle to controlled surface - it will always better orientate tolerance zone. I also agree that this situation can be considered as a contradiction to general definition of perpendicularity from para. 6.3.3.

Regarding the discussion who is right, Alex or Al:
Looking at fig. 6-4 from 2009 standard, I think you must admit that Alex's flow chart in step 2 can not be considered as a correct one. However Alex's book is basing on 1994 standard where such situation wasn't described so the field for interpretation has been left. If new edition of his book comes up I am pretty sure step 2 will be re-stated.

Regards

 

[powerhound]

Paul,
That's my point exactly. Fig. 6-35 has nothing to do with what I'm trying to say. I have no problem with 6-35. I do have a problem with Al Neumann's interpretation though. What I'm saying is that if you were to exchange the perp with profile of a surface -in Neumann's example- it will still mean exactly what you are saying it means as it is now, but it won't be open to any interpretations. Your interpretation that states "...and the feature controlled for orientation." is not an interpretation, but an addition to what the standard says because the standard says nothing about orienting a plane for perp. The orientation of the plane in the example is controlled by the limits of size, in this case the 11.9 dimension. If you want the plane oriented more closely than that, then you use Profile of a surface and call it out to A and B.
I guess one could look at it like this; which method would be more clearly supported by the 1994 standard in the case of a lawsuit? I see no support whatsoever for this method that Al is teaching. The support for profile of a surface in this case, is overwhelming.

Powerhound, GDTP T-0419
Production Manager
Inventor 2009
Mastercam X3
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

 

[PaulJackson]

No Powerhound if Newman's example was changed from orientation to profile the controlled surface opposite would necessarily have a basic displacement from that the profile would tolerance in addition to the orientation to [A|B]. With this change to profile size between B and the controlled surface would be dependent upon both 's Orientation tolerance to [A] via either Perpendicularity or Profile control as well as the controlled surface's Profile to [A|B].

The use of the Perpendicularity control negates any translational relationship in Newman's example so only orientation is controlled from [A|B]. Size has nothing to do with this check it could be twice size and still be oriented perfectly to [A|B].

Paul