Help with runout measurement

[joaomsv]

My name is João Vitório and I work as Quality Engineer in Portugal. I have now some doubts exposed in the next :

In the company were I work we are machining Cast Iron parts, with a considerable weight and size. The drawing of the parts has a specification of tolerance for axial and radial run-out.

I was checking about run-out measurement and it seems that we need to messure the part in rotation with the gauge in a fixed position.

We have a CMM (Brown&Sharpe) and our supplier is assuring us that it is possible to measure the run-out with the part fixed on the machine.

How can be explained this diference between the reference standards and the CMM measuring method? This method is acceptable and accurate? There is some scientific documentation or standards to make the validation of this method?

Thank you in advance for your help.


Best regards

João Vitório

 

[dingy2]

Circular and total runout is usually performed in a chuck on small cylindrical products with an indicator but not on large product like the one you are talking about.

There is a theoretical way of performing this on a CMM but not done in reality.

Since circular runout comprises of positional (off location) and roundness combined, I would suggest that the drawing be changed to positional in RFS and then add a roundness tolerance. This makes more sense and I believe the you could do that with a lot of confidence.

Dave D.

http://www.qmsi.ca

 

[PaulJackson]

I disagree with Dave!!! CMM software estimates both circular and total runout very well. The more points that are taken in a circular cross-section perpendicular to the datum feature(s) axis and the more cross-sections the better the estimation for circular runout will be. With total runout the points need not be in individual circular cross-sections rather the greater the the numuber of points probed on the feature the better the estimation. Each of the individual points will be figured for its radius to the datum axis and the difference between the max and min in any one cross-section will estimate the circular runout likewise the difference between the max and min radius of all points probed on the feature will estimate the total runout.

The same holds true for the physical inspection with test instruments the greater number of the infinite number of circurlar traces that are performed the better the estimates of circular an total runout will be with the tools.

There is one advantage to performing the inspection with test instruments and that is that the human eye and brain can typically hunt and estimate the extreme minimum and maximum point more efficiently than programs that are either limited to a fixed pattern and number of points or those more sophisticated that can iteratively hunt for extreme trends. That human influence advantage however can also be corrupted by the goals of the interrogation where a source inspection seeks to prove that the measurement passes and stops when the objecticve is achieved while the buyers inspection seeks to discover what's wrong and stops when his objectives are achieved.

CMM's estimate runouts very well every day of the week!!!

Paul

 

[dingy2]

Paul:

I have posted the question "Using a CMM to confirm circular or total runout" on the other site that we both view and I might have to "eat crowe" on this one. You may end up with it? Let's see how the question makes out.

Dave D.

http://www.qmsi.ca

 

[TheTick]

It's all just math. Once measurements are taken at points, the postprocessor can calulate the resulting axis between datums and determine runout of other measured points. The quality of the end result depends on the quality of the programming, number of points, and accuracy of point measurements.

My concern would be over how the multitude of minute errors in CMM point measurement affect the final result.

Put the machine to the test. Get some parts measured on a mandrel/dial indicator setup, then on CMM, and compare.

Honesty may be the best policy, but insanity is a better defense.

http://www.EsoxRepublic.com

-SolidWorks API VB programming help

 

[dingy2]

TheTick:

There is absolutely no doubt that if one used tool room equipment on a small cylindrical features, one would have 100% contact around the feature as the part is rotated. One would have infinite contact using this method.

As Paul stated "CMM software estimates both circular and total runout very well" depending upon the number of points but it is still an estimate. It will never give the same confidence level as the dial indicator. The indicator method would supersede the CMM method since the CMM would need infinite contact to be in the same level.

Now we come to large products. Is circular or total runout appropriate? I think we can achieve whatever we are after using positional in which the CMM is most appropriate.



Dave D.

http://www.qmsi.ca

 

[PaulJackson]

Dave,

If one posed a GD&T question to a forum…the answer to that question would typically be biased by the expertise of the audience that it is presented to. There are a number of different forums that specialize in discussing topics according to the occupation of their contributors.

From a Designer and Engineer’s perspective… this forum is the place.
From a CMM measurement perspective…cmmtalk.com or pcdmisforum.com.
From a Quality oversight and control perspective…elsmar.com.
From a Standard’s discussion perspective… tech.groups.yahoo.com/group/Y14-5_User_Group/

You’ll have people in each of those forums that have expertise in multiple perspectives but the majority of the respondents in each will answer the question according to the paradigm of their profession. I saw the parallel question you posed and I will refrain from responding to it there, however, I think that you could have chosen a better forum to find out how CMM’s evaluate runout.

It is important to note Dave that all inspections “estimate.” None…not even attribute gauges can determine absolutely “without error” whether the imperfect surfaces “pass or fail” to the specified “perfect” tolerance zone. You responded to “TheTick” that “One would have infinite contact using this (indicator) method” but one would also have an infinite number of cross-sections to scrutinize to “estimate” an answer for circular runout.

The question of whether runout or position and form control is appropriate for the feature should primarily be driven by the feature’s function and only when necessary by subsequent drivers to address needs like measurement stability, existing or established tool and gauge strategies, etc.

Paul