Concentricity vs. Runout
Largest Manufacturing Technology
Community On The Web

1. Plastic
Join Date
Jan 2011
Location
Orange County, CA
Posts
40
Post Thanks / Like
Likes (Given)
3
0

## Concentricity vs. Runout

Hey y'all ... I'm sure this topic has been beat to death here, I just can't find it, so please help a brother out...

People at my job throw around the term ‘concentricity’ all the time when I’m 99.9999% sure they’re really checking simple runout. I see in the machinist handbook that concentricity belongs in the “location” type along with position and symmetry. I know position, I know symmetry, I can’t seem to find a good example of concentricity.

Can anyone describe in very simple terms the difference between concentricity and runout?

EDIT: I forgot to mention that I have worked with a prime who has a spec that says any of their drawings that call out concentricity treat it as runout. Period.

2. Simplest terms I can use to describe the difference is, there ain't much. That's not to say the two words are interchangeable, but they both basically refer to the same thing.
Two circles are said to be concentric if they share the same center point, in the case of cylinders, they would share the same center line. If they don't, when the two are rotated about one of their axes, the other will run out.
If a part is held in a lathe chuck, for example, the spindle can be rotated and an indicator used to measure the concitricity between a certain feature on the part and the center of rotation of the lathe's spindle. Other features can then be checked and compared to the first and their concentricity determined. The error found can be expressed as run out. This is usually given as TIR, which would be double the difference between the two features center lines. This must be kept in mind if checking for position error this way.

3. Plastic
Join Date
Jan 2011
Location
Orange County, CA
Posts
40
Post Thanks / Like
Likes (Given)
3
0
Originally Posted by Derek Smalls
....This is usually given as TIR, which would be double the difference between the two features center lines. This must be kept in mind if checking for position error this way.
Hey Derek, thanks for the reply!

Please, why do you say double it? Is it because concentricity belongs in the position family? I know in TP if you are just dealing with a single axis deviation you would double your indicator reading as were if you had say an X and Y deviation it would be the square root of a² + b² x 2. Right? So concentricity if able to check with an indicator your result is double your indicator reading??

4. Originally Posted by PowerJunkie
Hey Derek, thanks for the reply!

Please, why do you say double it? Is it because concentricity belongs in the position family? I know in TP if you are just dealing with a single axis deviation you would double your indicator reading as were if you had say an X and Y deviation it would be the square root of a² + b² x 2. Right? So concentricity if able to check with an indicator your result is double your indicator reading??
No, I don't see the two being related. Concentricity is called out from a datum, center line of another feature for instance. True position is generally related to an x,y position, a way to tolerance a set of dimensions. My point was that IF you are checking position by sweeping a hole for instance, remember what you are seeing on an indicator is Total Indicator Reading, which is double the amount of offset (in a straight line) between centers.
The two call outs are used completely differently. I would never expect to see a call out wanting two diameters to be to true position of each other within .001". Instead, one would be established as a datum and the other would be called to be within .001" concentric to it.
On the other hand, if a hole is to be bored 3" in from each end of a block, you would not call for concentricity to it's center line, you would use either a tolerance on each dimension, or call for true position to within a certain limit.

5. Stainless
Join Date
Apr 2005
Location
Bay Area, Ca
Posts
1,341
Post Thanks / Like
Likes (Given)
11
143
There are important differences between runout, total runout, and concentricity:

Concentricity is how well a shape is centered on a theoretical axis, regardless of roundness. So a hexagon can theoretically be perfectly concentric to a datum axis while having huge runout due to the fact that it isn't round.

Runout is how much a shape deviates from a perfect circle which is perfectly centered on an axis of rotation; so runout is concentricity combined with roundness.

Total runout is how much a surface deviates from a perfect cylinder perfectly centered on the axis of rotation. So total runout is runout plus cylindricity.

In general it's easier to conceptualize concentricity than runout, but much, much harder to measure it (because you have to measure and subtract roundness, then define a theoretical axis, then locate that theoretical axis relative to the datum of choice.

I almost always prefer to use runout or total runout to tolerance a feature. In theory one might use concentricity for something like a flywheel that needs good dynamic balance, but so far I haven't run across such a situation. For equal tolerances, a runout callout is always tighter than a concentricity callout.

There are important differences between runout, total runout, and concentricity:

Concentricity is how well a shape is centered on a theoretical axis, regardless of roundness. So a hexagon can theoretically be perfectly concentric to a datum axis while having huge runout due to the fact that it isn't round.

Runout is how much a shape deviates from a perfect circle which is perfectly centered on an axis of rotation; so runout is concentricity combined with roundness.

Total runout is how much a surface deviates from a perfect cylinder perfectly centered on the axis of rotation. So total runout is runout plus cylindricity.

In general it's easier to conceptualize concentricity than runout, but much, much harder to measure it (because you have to measure and subtract roundness, then define a theoretical axis, then locate that theoretical axis relative to the datum of choice.

I almost always prefer to use runout or total runout to tolerance a feature. In theory one might use concentricity for something like a flywheel that needs good dynamic balance, but so far I haven't run across such a situation. For equal tolerances, a runout callout is always tighter than a concentricity callout.
You make some very intersting points, and I think you are technically correct in what you say. You are obviously looking at this in a more theoretical way than I was. While I agree with you technically, and many of those ideas are crucial in an enginaeering sense, I think a lot of what you say gets lost on the shop floor.
If I had a piece of hexagonal bar in a lathe chuck and my indicator read zero on the center of all the flats, I would say I had zero(to within the accuracy of the indicator) run out. I see your point, that is wrong, but that's what I would call it.

On Edit: The second half of this post made no sense! Let me regroup and I'll try again. Maybe I should just go to bed instead.

7. Plastic
Join Date
Jan 2011
Location
Orange County, CA
Posts
40
Post Thanks / Like
Likes (Given)
3
0
Humm. I'm a little more enlightened ... a little... So

Check out the attached picture. It's called a "concentricity gage" and I put tape over other mfgr markings to protect the innocent.

What we have in this gage is a simple round part. I'm holding on the OD, we'll call it -A- and indicating in the I.D. On the back side where my hand is there is a knob that I turn which turns the 2 bottom rollers and the 3rd roller on the top just holding it in place with spring tension.

I'm checking runout here, not concentricity right??

8. Yeah. You would be checking the total indicated runout of the I.D. with reference to the O.D.

9. Hot Rolled
Join Date
Apr 2013
Location
Damascus, MD
Posts
845
Post Thanks / Like
Likes (Given)
2179
390
Possibly both: If you're just recording the highest and lowest value with the test gage, yes, you're measuring runout.

But, if you were to insert a tight-fitting gage pin in your ring and indicate on the gage pin, you would be measuring the concentricity (of course, dividing the extreme spread by two).

A more complex way to measure concentricity and roundness, you may record measurements every X degrees of rotation, and plot the results.

Paolo

10. Plastic
Join Date
Jan 2011
Location
Orange County, CA
Posts
40
Post Thanks / Like
Likes (Given)
3
0
Well, that said I'm not checking "total" runout I'm just checking runout ... unless I indicate multiple places. Right?

Back to concentricity, I still would LOVE a dumb mans explanation of what exactly concentricity is, maybe with a picture and a dumb dumb explanation for someone like me

11. Originally Posted by PowerJunkie
Well, that said I'm not checking "total" runout I'm just checking runout ... unless I indicate multiple places. Right?

Back to concentricity, I still would LOVE a dumb mans explanation of what exactly concentricity is, maybe with a picture and a dumb dumb explanation for someone like me
True. TIR would be the highest range over the whole surface. I can try to make a part in solidworks that explains this afternoon if no one has already cleared it up.

In this picture you can see a 1" OD shaft with the hexagon drawn on the end. The hexagon is drawn so it's center is on the longitudinal axis of the shaft.

If you were to put this shaft in a 4-jaw on a lathe, hold on the round section(green), and indicate it as good as you can get. Minimizing the runout of that shaft with respect to the lathe. Say you got it perfect, then measured the runout of the hexagon with a dial indicator. It would be 0.058". So it has 0.058" of runout, but the hexagon is perfectly concentric with the round part of the shaft.

13. Aluminum
Join Date
Dec 2012
Location
OHIO
Posts
177
Post Thanks / Like
Likes (Given)
143
166
Originally Posted by PowerJunkie
Well, that said I'm not checking "total" runout I'm just checking runout ... unless I indicate multiple places. Right?

Back to concentricity, I still would LOVE a dumb mans explanation of what exactly concentricity is, maybe with a picture and a dumb dumb explanation for someone like me
If you want a dumb mans definition of concentricity, Im your guy. Two features sharing the same center line along an axis of rotation. We have a gauge at work that we call a concentricity gauge(not sure if thats correct but thats what we call it). We check bores in diesel engine blocks for so called concentricity, the gauge references a pilot bore on top of the block, you rotate the gauge and a DTI reads the concentricity of a receiver bore 10 inches or so below the top bore. Your basically checking to see if these two bores share the same center line.

14. Concentricity is the measure of deviation of MEDIAN POINTS from the axis of the feature to the datum axis. The tolerance zone is a cylinder around the datum axis. In other words, off center. This controls location of median points; orientation and form of the derived median lines. It is almost always preferred to use runout, total indicated runout. Nearly everytime an engineer calls out concentricity, or even cylindricity, he doesn't understand what he's trying to control.

Runout is the measure of deviation from roundness and axiality. Runout applies only to one circle of the surface AT A TIME. The tolerance zone is a hollow circle (think washer) around the datum axis with perfect form being the center of the zone. This controls location, orientation, and circularity. Note that concentricity does not control circularity but runout does.

Total indicated runout is just runout applied along an axis. The tolerance zone is a hollow cylinder with perfect form in the center (think pipe). This controls location, orientation, and cylindricity. Note that concentricity does not control cylindricity but total indicated runout does.
Last edited by Rewt; 04-20-2017 at 05:15 AM. Reason: added "total indicated" to last paragraph

15. Titanium
Join Date
Nov 2011
Location
Missouri, USA
Posts
2,782
Post Thanks / Like
Likes (Given)
1353
2707
Concentricity is the derived centerpoint/line between two objects that NOMINALLY share a center. It has no form control. Only location.

Runout is the deviation from NOMINAL form about a centerline (often not its own centerline) which controls both form and location. It also controls orientation at times, such as applying runout to a flange face about the bore centerline of a pipe flange.

They are very much different controls.

Imagine a step-pin. Two different diameter shafts in-line. One cylinder is perfectly cylindrical. The next diameter is so fucked up, it's basically a 4-lobe cloverleaf extruded along a centerline. Since the cloverleaf shape is perfectly symmetric about its centerline, it would pass the concentricity check. It's a fucked up shape that won't work as a shaft, because there is no form control on it. However, if you apply runout, you get it all.

Now... if you apply a diameter tolerance to the cloverleafed feature (that should have been a cylinder) then it can only cloverleaf within that tolerance zone. So checking concentricity can sometimes be ok, since it's size controls form to an extent.

I hate using concentricity. I avoid it unless there are no better options, which is rare. True Position often covers it better and more clearly to people. Form controls are better refined separately, usually. Other times you want to see T.I.R. control most all of it, and the diameter can be left to change. TIR will control form but not size. Depends on the application. IME, when most people say "check runout" they're talking TIR. Rarely do they mean a simple circular runout check.,

16. Originally Posted by PowerJunkie
Back to concentricity, I still would LOVE a dumb mans explanation of what exactly concentricity is, maybe with a picture and a dumb dumb explanation for someone like me
Best way I can answer is “the feature placement is on the same axis of rotation or a centerline datum if it’s bores”.

Concentric is not eccentric, meaning you want 2 bearing bores in a hub concentric so the wheel doesn’t wobble (you usually also have a squareness to flange faces noted) → so this is good. Eccentric is the feature is offset from axis of rotation like the rod journals on a crankshaft → so that is good.

Good luck,
Matt

17. Stainless
Join Date
Feb 2010
Location
ch
Posts
1,791
Post Thanks / Like
Likes (Given)
166
221
Originally Posted by Rewt
Concentricity is the measure of deviation of MEDIAN POINTS from the axis of the feature to the datum axis. The tolerance zone is a cylinder around the datum axis. In other words, off center. This controls location of median points; orientation and form of the derived median lines. It is almost always preferred to use runout, total indicated runout. Nearly everytime an engineer calls out concentricity, or even cylindricity, he doesn't understand what he's trying to control.

Runout is the measure of deviation from roundness and axiality. Runout applies only to one circle of the surface AT A TIME. The tolerance zone is a hollow circle (think washer) around the datum axis with perfect form being the center of the zone. This controls location, orientation, and circularity. Note that concentricity does not control circularity but runout does.

Total indicated runout is just runout applied along an axis. The tolerance zone is a hollow cylinder with perfect form in the center (think pipe). This controls location, orientation, and cylindricity. Note that concentricity does not control cylindricity but total indicated runout does.
why "median" points? maybe im not getting median of what.

18. Stainless
Join Date
Aug 2016
Country
ALAND ISLANDS
Posts
1,185
Post Thanks / Like
Likes (Given)
38
73
this is my 1st thougth : when a concentric movement loses precision, it shifts to an excentric position and has runout kindly !

19. Titanium
Join Date
Nov 2011
Location
Missouri, USA
Posts
2,782
Post Thanks / Like
Likes (Given)
1353
2707
Originally Posted by dian
why "median" points? maybe im not getting median of what.
Measured points on the surface.

Concentricity controls position of the centerline.

Runout controls the form from nominal condition.

20. Titanium
Join Date
Nov 2011
Location
Missouri, USA
Posts
2,782
Post Thanks / Like
Likes (Given)
1353
2707
Originally Posted by J Gilles

In this picture you can see a 1" OD shaft with the hexagon drawn on the end. The hexagon is drawn so it's center is on the longitudinal axis of the shaft.

If you were to put this shaft in a 4-jaw on a lathe, hold on the round section(green), and indicate it as good as you can get. Minimizing the runout of that shaft with respect to the lathe. Say you got it perfect, then measured the runout of the hexagon with a dial indicator. It would be 0.058". So it has 0.058" of runout, but the hexagon is perfectly concentric with the round part of the shaft.

No, no, no.

That hexagon would only have 0.058" runout if it was /supposed to be a cylinder/

Runout is the measurement of deviation from the nominal condition. Not the measure of deviation from another feature. You also wouldn't wisely use runout on a hexagonal shaft.

If you only meant for the hexagon to illustrate how concentricity can be perfect while form is horrible (like it was /supposed/ to be a cylinder but you ended up with a hexagon for hypothetical reasons) then sure, you've got it spot on.