How to Measure Concentricity Correctly

[Paul F. Jackson]

Mike S,

I disagree with Jim's point that the concentricity is half the difference of the opposing max and min measured wall thicknesses.

If both features are perfectly cylindrical and are perfectly oriented (co-planar) but their centers are .007" apart then the measurement for their concentricity would be .014". Measurements of runout and total runout would also be .014" given the form and orientation conditions above.

Runout, Total Runout, and Concentricity all require that the measured feature is coaxial with the datum feature(s).

Runout captures the measurement instrument displacement each circular element relative to the axis constructed from the datum feature.

Total Runout collectively captures the sum of all circular element displacements from a mastered measurement instrument relative to the axis constructed from the datum feature.

The concentricity measurement (according to ASME Y14.5M-1994) reflects the size of the tolerance zone required to contain each measured midpoint of two opposing surface elements. Prior to 1994 the method of determining the center of the measured feature was not so defined.

Position does not require that the measured feature be coaxial with the datum features. A measurement of position would reflect the size of the tolerance zone required to contain the axis of the measured feature. That tolerance zone could be shaped like an open ended cube or a cylindrical if a diameter modifier precedes the tolerance.

The position measurement would also be .014" given the form and orientation conditions above.

What makes all of their measurements change from one to another is the way that different form and orientation characteristics or deviations affect the measurement .e.g.

A perfect ellipse that is perfectly oriented to the circular datum feature can have a .014 measurement for runout and total runout but a zero measurement for concentricity and position.

A perfect cone that is perfectly oriented to the circular datum feature can have a .014 measurement for total runout but a zero measurement for runout, concentricity and position.

A perfect cube with a height is .096" that is perfectly oriented to the circular datum feature can have a .014 measurement for runout and a zero measurement for concentricity and position.

A perfect equilateral triangle with a height is .042" that is perfectly oriented to the circular datum feature if its will have a .014" measurement for runout, total runout and concentricity and a zero measurement for position (according to ASME Y14.5M-1994) however it could have had a zero measurement for concentricity prior to the ASME Y14.5M-1994 Standard.

 

[JRKH]

Good info

Mike S,

I disagree with Jim's point that the concentricity is half the difference of the opposing max and min measured wall thicknesses.

If both features are perfectly cylindrical and are perfectly oriented (co-planar) but their centers are .007" apart then the measurement for their concentricity would be .014". Measurements of runout and total runout would also be .014" given the form and orientation conditions above.

Runout, Total Runout, and Concentricity all require that the measured feature is coaxial with the datum feature(s).

Runout captures the measurement instrument displacement each circular element relative to the axis constructed from the datum feature.

Total Runout collectively captures the sum of all circular element displacements from a mastered measurement instrument relative to the axis constructed from the datum feature.

The concentricity measurement (according to ASME Y14.5M-1994) reflects the size of the tolerance zone required to contain each measured midpoint of two opposing surface elements. Prior to 1994 the method of determining the center of the measured feature was not so defined.

Position does not require that the measured feature be coaxial with the datum features. A measurement of position would reflect the size of the tolerance zone required to contain the axis of the measured feature. That tolerance zone could be shaped like an open ended cube or a cylindrical if a diameter modifier precedes the tolerance.

The position measurement would also be .014" given the form and orientation conditions above.

What makes all of their measurements change from one to another is the way that different form and orientation characteristics or deviations affect the measurement .e.g.

A perfect ellipse that is perfectly oriented to the circular datum feature can have a .014 measurement for runout and total runout but a zero measurement for concentricity and position.

A perfect cone that is perfectly oriented to the circular datum feature can have a .014 measurement for total runout but a zero measurement for runout, concentricity and position.

A perfect cube with a height is .096" that is perfectly oriented to the circular datum feature can have a .014 measurement for runout and a zero measurement for concentricity and position.

A perfect equilateral triangle with a height is .042" that is perfectly oriented to the circular datum feature if its will have a .014" measurement for runout, total runout and concentricity and a zero measurement for position (according to ASME Y14.5M-1994) however it could have had a zero measurement for concentricity prior to the ASME Y14.5M-1994 Standard.

Paul,
You are correct and your post contains a lot of good info. But I don't think it necessarily negates my point. What Mike was looking for was a quick simple answer. My answer was based on the part he described and the desired instruments (simple).

My background is just a simple lathe operator and inspector of simple parts. The method I described has always worked for me "in the trenches". If I have made any errors in the details, Mike will be (as was I) illuminated by the detail you have provided.

James

 

[Jim Wynne]

Paul,
You are correct and your post contains a lot of good info. But I don't think it necessarily negates my point. What Mike was looking for was a quick simple answer. My answer was based on the part he described and the desired instruments (simple).

My background is just a simple lathe operator and inspector of simple parts. The method I described has always worked for me "in the trenches". If I have made any errors in the details, Mike will be (as was I) illuminated by the detail you have provided.

James

I think you're both right. Paul's explanation is technically right on, but the problem is (as I suggested in an earlier post) that most of the people who specify these things don't really understand the distinctions. In most cases, what the designer is looking for is coaxiality (two features sharing a common center axis), and the interest is in the simple linear distance between the datum axis and the controlled feature's axis, in which case JRKH's method will provide a (probably) reasonable answer.

I'll reiterate, however, that every time concentricity is specified, an attempt should be made to confirm that the specification is really what was intended.

 

[JRKH]

Right on

I think you're both right. Paul's explanation is technically right on, but the problem is (as I suggested in an earlier post) that most of the people who specify these things don't really understand the distinctions. In most cases, what the designer is looking for is coaxiality (two features sharing a common center axis), and the interest is in the simple linear distance between the datum axis and the controlled feature's axis, in which case JRKH's method will provide a (probably) reasonable answer.

I'll reiterate, however, that every time concentricity is specified, an attempt should be made to confirm that the specification is really what was intended.

My intent was certainly not to discredit anyone else's explaination. But rather to try and keep to the origional posts request for simplicity. In point of fact I have never really run into a problem using the method I described. Of course if the part tolerances were tighter, say .002" instead of .020" I would have suggested a more robust inspection method.

Those who have been around machining for a while realize that on most features that are called out, there are a multitude of variables that can effect the usability of the part. The trick is, as you point out, is to know what the use is and try to inspect for that.

Given the quality of the posts I think Mike should be most thuroughly informed, both from the "Down and dirty" to the more highly technical. It's one of the things that makes this board so useful.

James

 

[Jim Wynne]

The trick is, as you point out, is to know what the use is and try to inspect for that.

I would stress the word "know" and make sure "guess" or "assume" isn't substituted. There's nothing worse than having a perfectly functional part rejected because of bad specifications. That's why I stressed the need to make sure what the designer intended, and I'll add another caveat: get it in writing.

Given the quality of the posts I think Mike should be most thuroughly informed, both from the "Down and dirty" to the more highly technical. It's one of the things that makes this board so useful.

Yes . George Bernard Shaw said, "When two men in business together always agree, one of them is unnecessary."

 

[Paul F. Jackson]

James,
Your advice of estimating the concentricity with simple hand held instruments was good and relevant to the original question. Having been a machinest and inspector I understand the benefit of estimating the geometric form, orientation, or location deviations quickly with simple tools provided that I understand the assumptions that I am making relative to the process risks e.g. odd lobing from jaw or collet squeeze or centerless grinding etc..

The only thing that I disagreed with was halving the value. I just wanted people to understand that the answer is not the distance between centers but rather the size of the diameter zone needed to contain that center's deviation from the datum axis. So the answer is .014 not .007.

Paul

 

[JRKH]

Tolerancing methods

James,
Your advice of estimating the concentricity with simple hand held instruments was good and relevant to the original question. Having been a machinest and inspector I understand the benefit of estimating the geometric form, orientation, or location deviations quickly with simple tools provided that I understand the assumptions that I am making relative to the process risks e.g. odd lobing from jaw or collet squeeze or centerless grinding etc..

The only thing that I disagreed with was halving the value. I just wanted people to understand that the answer is not the distance between centers but rather the size of the diameter zone needed to contain that center's deviation from the datum axis. So the answer is .014 not .007.

Paul

I appreciate the feedback Paul. I wonder if we are looking at apples/oranges. Or I may be completely off my rocker!! My experience has by and large been with the old +/- tolerancing, with only occasional forays into GMT. When I was very young (couple of centuries ago) I was taught that by older machinists that TIR meant "Total Indicated Reading" and that the actual concentricity would be half of that number. Of course this assumes the concentricity tolerance given as +/-.
So in the case of the origional post I may have jumped to the conclusion that the tolerance was +/-.020 instead of .020 true position.

We Keep Learning.

James

 

[Douglas E. Purdy]

200 yrs Dealing with Parts and Drawings

No. :mg:

How did you accumulate 200 yrs of experience?

Just Wondering,
Doug

 

[Jim Wynne]

How did you accumulate 200 yrs of experience?

Just Wondering,
Doug

200 was an exaggeration based on the fact that if it's true that you're only as old as you feel, then I'm 240 years old.

 

[qualityperson]

Re: Concentricity Measurement - How to correctly measure concentricity

Hi. I'm new here and have a related concentricity question:

We have a small part with 2 stepped inside diameters in the same larger bore: .116" and .172" and .1805", which is also datum C.

There are concentricity callouts for both:
for the .116" diameter "within .008 dia. to C" and
for the .172" diameter "within .006 dia. to C"

How is concentricity measured when comparing inside diameters only?

I understand the concept when comparing an O.D. to the I.D. but I am totally confused on this.

Thanks! Julie