Question about circular runout

[SteelAndSteel]

Hi,

First of all i am using ISO. I have noticed that the meaning of circular runouts change if the arrow is not pendicular to the surface, can someone explain that to me?

My main question is that i want to put a tolerance to the cone below (the thing which are brown)

http://s11.postimg.org/a7kc6jolv/image.jpg

Could i use this option
A circular runout to add a tolerance for every circular slize, then i add straightness so the angle of the cone stays intact

or is there a more commen way to do this?

http://s11.postimg.org/a7kc6jolv/image.jpg

 

[CheckerHater]

In 2005 version of 1101 standard ISO added new clause “Circular run-out tolerance in a specified direction” (see picture)
By default, runout is measured normal to surface, which may become confusing/ difficult to achieve, especially for curved surfaces. Runout in specified direction is always measured under the same angle. The angle must be specified on the drawing even if it is 90°.
It is up to you to use runout in a specified direction, or good old runout in any direction.

Now to the second part – controlling the cone.

First, let’s get some terminology straight.
You say “circular runout to add a tolerance for every circular slize” – is it “size” or “slice”, because runout does not control size.
Also, “add straightness so the angle of the cone stays intact” – straightness will not control the angle, but it could be sufficient.
It is difficult to give advice based only on your picture. If we could see how you intend to dimension the part, that is, how angle / taper is specified and what kind of relationship between the features are you trying to achieve, it would be easier.

On the matter of “more common ways to do it”, if you were in ASME part of the world, using total runout would be the first choice.
Unfortunately ISO seems to discourage using total runout on conical surfaces.
If you really need to control your cone with one single FCF, the profile would be the second choice.

 

[pmarc]

First of all i am using ISO. I have noticed that the meaning of circular runouts change if the arrow is not pendicular to the surface, can someone explain that to me?

Well, not really. If you take a look to ISO 1101:2012 you will find following statement in clause 8.1: "The orientation alone of the leader line does not influence the definition of the tolerance zone, except in the case where the orientation of the leader line and therefore the direction of the width of the tolerance zone is indicated by a TED". [TED = Theoreticall Exact Dimension = Basic Dimension in ASME].

In the earlier version of this standard, the statement is similar (though less detailed): "The orientation alone of the leader line does not influence the definition of the tolerance."

In both versions of the standard both statements are followed by figures showing how this concept works when circular runout is applied to a conical feature.

This means that by default the width of tolerance zone always applies normal to the specified geometry, even if the arrow is not perpendicular to the surface. If you want to override this rule, TED needs to be used to define the direction of the width of the tolerance zone relative to datum.

Next, I agree with CH that technically adding straightness tolerance will not control angle of the cone, and that it would be easier to give any answer if we had more details about dimensions and tolerances on adjacent features. I also agree that profile of surface tolerance is another option to pick. One can also consider applying combination of profile of surface and runout if this satisfies his/her requirements (see fig. 8-26 in Y14.5-2009 - this technique can easily be transferred to ISO world, in my opinion).

But generally speaking, there is ISO standard, ISO 3040, that deals with dimensioning and tolerancing of cones, and some more techniques can be found there.

And finally, I have to ask this question:
CH, which paragraph(s) or figure(s) in Y14.5 make you say that in ASME part of the world using total runout would be the first choice to tolerance cones?
Please don't get me wrong, I am not saying it could not be done. In fact there is nothing in the standard, per my knowledge, that would prohibit such practice (I even used this approach a couple of times in the past). I just hardly see a trace of anything in the standard that would encourage anybody to apply total runout to conical surfaces. Thanks.

 

[CheckerHater]

I don’t know where to start…

I was not trying to create universal rules for tolerancing all kinds of cones in all the situations.
I was trying to address using runout when tolerancing cones.
I noticed that note in 8.1 says “orientation alone of the leader line does not influence the definition of the tolerance zone”. Nevertheless in illustrations to para. 18.15-18.16 leader is always shown normal to the geometry, except “tolerance in a specified direction” case where it was clearly controlled by dimension.
So, I assumed (maybe wrongfully) that it was the case that confused OP otherwise familiar with the concept of runout.

Then, I tried to address the question about alternative to combining circular runout with straightness.
Once again, I assumed that if OP was asking about runout, then some other form of runout would be the right choice.
In ASME the total runout could do the job, but the situation in ISO was not that easy.

ASME Y14.5-2009 Para. 9.4.2.1 says:
“Where applied to surfaces, constructed around a datum axis, total runout may be used to control cumulative variations such as circularity, straightness, coaxiality, angularity, taper, and profile of a surface.”
So ASME explicitly permits (and encourages) use of total runout on conical surfaces.

ISO 1101:2005 when making distinction between circular and total runout also seems to distinguish between features they can be applied to.
Circular runout is divided into:
Circular run-out tolerance —radial (measured perpendicular to the axis)
Circular run-out tolerance —axial (measured parallel to the axis)
Circular run-out tolerance in any direction (measured normal to geometry, straight or curved)
Circular run-out tolerance in a specified direction (measured on specified angle to the axis)
While total runout is limited to:
Total radial run-out tolerance (the tolerance zone is limited by two coaxial cylinders)
Total axial run-out tolerance (the tolerance zone is limited by two parallel planes)
And that’s it!
The opinions on the matter even between people highly versed in ISO standards differ:
G. Henzold sees no problem using total runout to control conical surfaces, but he sometimes mixes ASME and ISO applications.
Henric S. Nielsen is pretty clear that total runout can only be radial or axial, and radial TR only applies to cylindrical features.
The question of how far one may extend the principles still open.

By the way, anything new about total runout in 2012 version?

 

[pmarc]

Starting from the end of your reply.

No, there is nothing new in 2012 version of ISO 1101 regarding the subject of applicability of circular and total runout tolerances.

It is always interesting and informative to hear that there are discrepancies even amongst the greatest gurus of ISO GPS language. Luckily, I did not say anything about legality of use of total runout when conical surfaces are controlled in ISO.

As for para. 9.4.2.1 of Y14.5-2009, I agree, it can be understood as green light for total runout for conical features. But would you really call this "encouragement"? Not a single figure in the standard showing total runout applied to a cone. I would barely call it "permission". But that is just me.

 

[CheckerHater]

Para. 9.4.2.1 is not some obscure non-mandatory appendix, it's a definition of what you can do when total layout is "Applied to surfaces around an axis".
How much more encouragement do you need? I know it's a nit-picking, but that is just me.

 

[pmarc]

How much encouragment do I need?

Let me put it this way:
If fig. 9-5 in 2009 showed at least one more total runout instead of one of circular runouts, and in "Means this" area this application was clearly described, I would then agree that the standard encourages to use total runout to control features other than regular cylinders.

For the time being, I will stick to my previous opinion - the standard permits this (and I have no problems with that), but is far from making it "the first choice to tolerance cones".

 

[axym]

Hi All,

9.4.2.1 is probably one of Y14.5's more ambiguous and confusing sections. It's not surprising that there are different interpretations. Here are some thoughts:

-if cylinders were the only feature type that Total Runout can be applied to, then why would they use the term "surfaces constructed around a datum axis" ?
-if Total Runout may be applied to a cone, why is an example not shown (an example is shown for Circular Runout applied to a cone)
-There are issues with the "cumulative variations" that Total Runout can be used to control. Angularity for one - Total Runout cannot control what the Angularity characteristic does (orientation of features with some non-orthogonal basic angle to the datum). If "angularity" was intended to mean something else like included angle, then it's a very confusing choice of term. Profile of a Surface for another - the way I see it, Total Runout doesn't control the same thing as Profile of a Surface and to say it does is grossly misleading.
-There is an example in the Profile section in which Profile of a Surface is combined with directly toleranced size dimensions, on a conical feature. In this application, Surface Profile ends up providing the same control that Total Runout would. Why define this ugly mutation of Surface Profile, if Total Runout can be applied to cones?

There is evidence to support either position. On balance, I would say that there is more evidence to support the assertion that Y14.5 allows Total Runout to be applied to cones. There is nothing that says that it does not. But I don't think we could say that Y14.5 encourages the practice.

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[Belanger]

Evan -- I'm not sure why we cannot say that total runout used on a cone would control angularity (assuming that the intended angle is given as basic).
But I heartily agree with the rest of your post; profile is distinctly different from runout in that it is required to apply to a true profile (i.e., basic dims). Of course, that's been a sticking point elsewhere in these threads, so I won't derail us with the particulars.

John-Paul Belanger
Certified Sr. GD&T Professional
Geometric Learning Systems

 

[SteelAndSteel]

Thank you for your replies; you wanted more context so I have uploaded another picture:

http://s27.postimg.org/gql2o2en7/image.jpg

Kindly Paul

 

[CheckerHater]

Out of curiosity - I'd like to see opinion on illustrated process and how it would be different if we wanted to measure
a) angularity
b) straightness (and could we consider straightness special case of profile?)

 

[3DDave]

I'm thinking that runout doesn't necessarily control angularity, which seems to be a zone that is bounded by two planes inclined at a basic angle, which a revolved shape is not.

Profile of a line? Sure, but angularity, maybe not so much.

Checkerhater - I think the indicator in the diagram should be perpendicular to the axis, unless it is specified on the drawing otherwise.

 

[Belanger]

3DDave -- I think it was assumed that the angle would be given as a basic dim, thus making total runout legal on that one (and keep the indicator stays normal to the surface, per paragraph 9.4.2). I would say that total runout applied a cone controls angularity. Circular runout would not.

CH -- certainly, straightness is a special case of profile. It is profile of a line applied to straight surface elements with no datum references.

John-Paul Belanger
Certified Sr. GD&T Professional
Geometric Learning Systems

 

[3DDave]

I am concentrating on the difference in tolerance zone form - total runout is generally a curved tolerance zone while angularity is always a planar tolerance zone.

 

[CheckerHater]

runout doesn't necessarily control angularity, which seems to be a zone that is bounded by two planes

Angularity is not necessarily a zone bounded by 2 planes. See ASME Y14.5.1M-1994 FIG. A-19a

indicator in the diagram should be perpendicular to the axis, unless it is specified on the drawing otherwise

Indicator is normal to the part surface. In ASME there is no way to specify position of the indicator on the drawing. In ISO they have "circular run-out tolerance in a specified direction" where you can specify the angle.

Either way I only provided the diagram of the fixture. How do you know what was specified on the drawing?

 

[3DDave]

I was going by the '94 standard's "means this" where runout was applied to a curved profile. Based on what I read, the indicator is positioned perpendicular or parallel to the axis of rotation in order to measure runout.

If orientation was specified on the drawing, you should have said so in describing your diagram. For all anyone knows you are holding back on various notes that make your special case setup correct.

 

[Belanger]

No, even in 1994 the indicator was normal to the surface. See Fig. 6-47 for an example of circular runout, and 6-50 (mid-right hand side) for an example of total runout.

John-Paul Belanger
Certified Sr. GD&T Professional
Geometric Learning Systems

 

[pmarc]

Out of curiosity - I'd like to see opinion on illustrated process and how it would be different if we wanted to measure
a) angularity
b) straightness (and could we consider straightness special case of profile?)

For angularity (I assume you are thinking about angularity of line elements of conical surface) the dial indicator has to be oriented relative to a datum simulated by two centering devices, which seems to be the case on this illustration. For straightness all you have to do is to "scan" each linear element lengthwise, but the dial indicator does not have to be oriented to the datum axis.

On the picture attached below, the red linear element will easily satisfy straightness requirement, but most likely will not meet angularity callout (unless the angularity tolerance is extremely loose).

http://files.engineering.com/getfil...20-92b9-53f03589e1e3&file=tot_runout_cone.png

Regarding orientation of dial indicator, per Y14.5 in any kind of runout check it must stay normal to true geometric shape of a feature, as described in "Means this" sections of figs. 9-2 and 9-3 in 2009 or figs. 6-47 and 6-48 in 1994. This will perhaps sound like nitpicking, but in my opinion "normal to true geometric shape of a feature" is not the same as "normal to the part surface".

 

[SteelAndSteel]

About the picture i posted; i guess you can't use angularity to define a cone? (wishful thinking)
Could you use the position tolerance to define the cone?; i have only
seen the position tolerance for hole center and flat surfaces

About pmarc's comment:
ee fig. 8-26 in Y14.5-2009 - this technique can easily be transferred to ISO world, in my opinion

I can't see how

 

[pmarc]

SteelAndSteel,
The technique shown in fig. 8-26 could unambiguosly work in ISO provided that entire toleranced contour, together with adjacent cylinders, was defined by basic dimensions. As long as you don't have it basic, I would recommend to stay away from this approach.

Could you use the position tolerance to define the cone? Well, in ISO theoretically yes, because ISO treats cones as features of size. I am saying "theoretically", because I am not aware of any figure in any of ISO standards showing this technique, but I can imagine it working. Position would control location of what is called extracted median line of the cone (see ISO 14660-2:1999). But even then you would still need additional tolerance(s) to control form and size of the cone.