Coaxiality and concentricity 5

[kakalee1]

After reading a lot about coaxiality and concentricity, I am still having trouble understanding them, and when to use one or the other.

From what I understand, concentricity seems to include coaxiality, and it is really difficult to measure. However, I still don't see what possible outcome that can happen to a concentricity callout that cannot happen to a coaxiality callout. If anyone can give me a specific example, that would be greatly appreciated.

Also, when would use concentricity rather than coaxiality?

 

[Belanger]

Concentricity is a specific GD&T symbol. Coaxiality is not a symbol, nor is it a very specific term; it is a general word used to describe the idea of two or more circles that share a common center.

The problem with that general word is that it leaves open the question of HOW to find the center of a given circle. Concentricity has very specific language that describes how to find the center. That's why you've read that concentricity includes coaxiality, and it can be difficult to measure (at least how it's defined in ASME Y14.5). On the other hand, there is no such thing as a "coaxiality callout."

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

 

[kakalee1]

Belanger,
Thanks for your input.
How exactly would you measure concentricity and coaxility? Assuming that you have a CMM machine.
Let's use the example I attached. Units are in mm. What would be the possible outcome of concentrity that can not happen with coaxiality? how would you measure them?

 

[jflongwell]

Concentrically is measured by median points of diametrically opposed elements. The cylindrical tolerance zone would be located by the datum axis. You would need two opposed indicators to measure the median points. As the part is rotated about the simulated axis, the indicators would move back and forth (dueling indicators). CMM's have difficulty with this measurement as the two measurements have to be diametrically opposed.

Coaxial is an implied relationship of a FOS datum and a FOS. The center axis of the datum and the FOS is created with a CMM by a best fit (lease squares) from all of the probe touches.

For this application, I would recommend profile of the bore to datum A. An optical comparator could do this very quickly.

John

 

[Belanger]

Kakalee1 -- In the picture you supply, there is no "coaxiality" callout. There is "concentricity" on the left-hand drawing, and then "position" on the right-hand drawing. BOTH are controlling coaxiality, which is a general verbal term for any centering effect. That was my point earlier.

Now that we know it's position being compared to concentricity, we can assess the difference. See the graphic attached below...
If we are given the concentricity symbol, then we'd have to look at the red part of my picture: the red cross-hairs would have to be within the given tolerance zone around the datum.
But if we are given the position symbol, then we would look at the black cross-hairs, which are taken from a perfect circle. We are still measuring that red part, but based on its "actual mating envelope." Thus, the part will probably fail concentricity yet pass position. (John's explanation is the same idea.)

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

 

[3DDave]

Concentricity is a terrible control to use, not because the idea is so bad, but because verification is.

Moving along the original axis, one measures from the axis to the controlled surface in opposite directions. Take the difference in the two measurements, multiply by 2, then compare that number to the concentricity limit. It is a way to control the balance of volume around an axis. It is no substitute for actual balance controls.

In this regard concentricity is like planar symmetry, and is OK when confined to revolved features.

The troublesome part is as described in the '1994 version, which includes "correspondingly-located elements of two or more radially-disposed features" which is taken to apply to non-revolved features, such as hexagons. Can it apply to gear teeth, tri-lobed features, or fan blades? Who knows as the '1994 committee included no such explanation for why it was included.


Other 'co-axial' controls don't attempt to measure the distance between axes as their primary goal. All it means in terms of position tolerancing is that the distance from the datum axis and feature axis or between mutually defined feature axes is nominally zero. The '1994 committee left/made the description confusing without adding value. I don't recall or care to look to see if it was the same in the '1982 version, but coaxial position might have been there then.

 

[powerhound]

Dave said that concentricity is a terrible control to use and then gave a good reason why it stinks so bad. My opinion is that it is a terrible control to use because it doesn't do anything that runout or position doesn't do that really means anything. You get no bonus tolerance where it might be allowed if you were to use position and you get no datum shift where it might be allowed if you were to use position. Circular runout or total runout is way better than concentricity. The only argument I've ever heard for concentricity is for dynamic balance. I've actually never heard a good argument for symmetry. If dynamic balance is that big of an issue, it needs to be handled by a balancing process. Please, put the entire idea of concentricity out of your mind forever and forget it ever existed.

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

 

[Tunalover]

95% of the applications I run into can be handled by position, profile, or flatness. Since I don't work with rotating equipment, run out is not something I have to deal with. I always control coaxiality between cylindrical features using positional tolerance control. For reasons described in this thread, I have NEVER used cylindricity or concentricity.

Tunalover

 

[CheckerHater]

I think I posted this link before. Either way - enjoy.

This one may be useful as well.

 

[pmarc]

In order to have full picture of the situation, I think it would be good to ask one more question:
Which GD&T standard are you, kakalee1, using - ISO or ASME?

All the answers given so far assumed that this is ASME. If, however, by any chance ISO in charge, the answer will be different, that is, there is no difference between concentricity and position in ISO GD&T (GPS) world.

 

[CheckerHater]

Good point, especially considering that the term "coaxiality" actually does exist in GPS

 

[fsincox]

Exactly, pmarc beat me to it! ISO or ASME?
Frank

 

[fsincox]

The dimensioning style in his example suggests ISO.
Frank

 

[CheckerHater]

And opinion "From what I understand, concentricity ... is really difficult to measure." suggests ASME

 

[Belanger]

That's true -- things change if we are talking about ISO.
So the terminology of the OP might indeed be correct, but the difference illustrated in my graphic still applies.

I don't agree 100% with those who say that runout is better because it encompasses everything done by concentricity. Yes, but did you know that it actually does more than concentricity, thus it might be more expensive and difficult to hold?
Certainly runout is cheaper to inspect, but actually it's more expensive/confining to mfg.

Of course, the bottom line is that all this should be driven by the functional requirements.

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

 

[dgallup]

Frequently when I put concentricity or position on a drawing QA measures runout anyway! Doh!

----------------------------------------

The Help for this program was created in Windows Help format, which depends on a feature that isn't included in this version of Windows.

 

[fsincox]

JP, thank you.
dgallup,
If it passes they are fine.
Frank

 

[powerhound]

The reasons I think runout is a better control is because:
1) Runout controls form, where concentricity doesn't.
2) Runout is what they want anyway.
3) Runout is what will be checked anyway.

The arguments to these will be:
1) What if form control isn't desired? Answer: The location of median points of diametrically opposed elements are not really desired either so may as well go with the easier check.
2) How do you know they really want runout and not concentricity? Answer: Neither in my experience, nor the experience of anyone else I've ever asked this question of (who really knew the difference), has concentricity ever been the real goal. It was always really runout or position.

Actually, this question is open to all of you too. Have you ever, in your experience, seen an instance where concentricity--as defined by ASME--was really what was needed and runout/position or even a balancing process was not a better option? Please leave out the "No, but that doesn't mean someone else hasn't." I'm aware of that and am still looking for that individual.

3)How do you knw they will check runout and not true concentricity? I refer you to dgallup's post. It has been my experience as well. We actually had a device in the QC department where I used to work called a "concentricity checker". It actually checked runout but no one knew the difference.

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

 

[kakalee1]

To all,

Thanks for all your input especially Belanger and Powerhound. They are really helpful.

I don't really have any background as a QA inspector or drafting, so all of these info are a bit too much to digest at once. I'll do more research to understand concentricity completely.

As for standard, my company follow the ASME y14.5m-1994.

One main reason why I created this post is because I don't know if our QA department inspect concentricity correctly. I can't really say that they are wrong or right when I don't fully understand it myself, and I have a feeling that no one from my company do.

Reading all the feedbacks here, it seems that my customer's concentricity callout was really unnecessary, and I believe all what they really need is a position call out.

I'll comeback to this post later when I have more question.

 

[pmarc]

To powerhound:
How will runout work if a part's feature to which concentricity callout is applied is not nominally cylindrical (see attachment)?

http://files.engineering.com/getfile.aspx?folder=4495c2a2-de48-42be-99b2-8c6796b4872b&file=fig.png

To kakalee1:
First thing you need to understand the concept of concentricity (and especially to grasp the difference between concentricity and position [coaxiality] at RFS) is to analyze figures 5-55 through 5-58 and associated text in Y14.5M-1994.