Radial Measurement vs. Radius

[[Li...]]

I just recently started using radial measurements to inspect certain areas of castings we get.  We do this to verify that after CNC machining we will have enough wall thickness.  I did one yesterday that gave decent data of 5 points. Only 2 of the points taken around this feature that nominally has a diameter of 1.620" / R .81"±.030" were oversized.  The other check I was tasked in reporting was the location of the diameter/radius.  This is where things get weird.  It reported that the radius was 2.5601" and the location was in spec only because of LMC modifier on a .084" true position.

Can someone help me with what is happening here? How are the radial measurements [Size-Standards] in spec to slightly out, and using Radius in [Size/Standards] drop down showing a huge deviation from nominal and the data reported in the radial measurements? Not to mention the true position.  I will say that, yes these are castings, and this cast feature does look as though the supplier did some grinding or even added material(aluminum) somehow. Which is not uncommon in the casting world.  

[[Ma...]]

If you have constantly problem in this area, then i would blame small angle of circle/cylinder. It looks around 30°. This will often report wrong radius.

Do recheck. If you will get similar results, then use curve callout to see deviations.

[[To...]]

Please sign in to view this username.

 When you use Radius Measurement, the radius's origin is constrained to its' nominal location and the distance is to the point.  Similar to constraining a circle's origin in the evaluation window.

A common workaround is to set the evaluation to constrain the origin to get a better radius value but this circle can't be used for position.   What I do is go into Evaluation of the circle, check Constrain X and Y, click OK and use this circle for the Radius.  Next, create a copy of that circle, select Re-call Feature Points, un-check "Re-Calculate Nominal Geometry", select the original circle, click OK.  Click Evaluation button.  Un-check Constrain X & Y, and check Constrain Radius.  Use this circle for the position.

Edited May 30

[[Li...]]

Thanks for the help on this. I was trying to explain this to my SQE evaluating data for FAI.  I just couldn't for certain tell him how radial measurement reports the data. 

So, with the circle path I used, touch points not "scan" will individually align to nominal location? 

Please sign in to view this quote.

Thanks. Yes, I was able to touch 5 points of a circle path at about 35°

[[To...]]

Please sign in to view this quote.

It reports the distance from the nominal center of the radius to each individual point.

[[Li...]]

Please sign in to view this quote.

But when reported as a true position ("Item 89 TP) it is way out of location. I am going to work on this some more, but I checked this on a totally different CMM (not Zeiss, not Calypso) and it is out of location on that machine as well. 

[[Ky...]]

I'm honestly not surprised that you would be having issues with that particular feature. The repeatability (and accuracy) of circles goes down as their arch length decreases. It's an unfortunately fact of life. If you have taken the CALYPSO Advanced course a few years ago, the book actually has a nice chart on it.

 

The "easy" answer is to constrain the element that you are not concerned about (e.g., for radius constrain the position, for the position, constrain the radius). It at least gives you more repeatable answers. The better, but more complicated, answer is that would probably be better as a profile callout rather than a position/radius callout.

[[Ma...]]

Also something to consider is Space Point Distance. If you have a CAD Model take several Space points along the radius and report out the Space Point Distance. The nominal will be 0.0000 the actual will be the deviation or “T” value, the distance along the normal vector of the point. CAD Evaluation can be used to graphically display the variation along the model.