Meas distance differences

[PowerJunkie]

I've got this situation where I'm meausing the distance from C/L of a hole to the end face of a part. Simple really, only the results are not correct. I did a manual measurement with height stand and indicator got it spot on at .5997.

Depending on how I measure it, either circle OR cylinder, and picking up the face at 2 different probe angles, I get 2 readings I call good that are within .0001 and 2 readings that I call bad that are -.0005 and +.0004 (see pic's attached).

It seems to me that it should be most accurate while measuring as a cylinder and shouldn't matter which probe angle I pick up the face on. ?

Fresh calibration of all probe angles as of this morning (D2L95), max StdDev .0001 on 28 different probe angles.

I have looped this program 5 times and all the #'s repeat within .0001

Any ideas of what I'm doing wrong to get all this deviation?

The whole reason I did this test this morning is I was measuring another part last week with similar features that did not give correct results.

[Wingman]

I've got this situation where I'm meausing the distance from C/L of a hole to the end face of a part. Simple really, only the results are not correct. I did a manual measurement with height stand and indicator got it spot on at .5997.

Depending on how I measure it, either circle OR cylinder, and picking up the face at 2 different probe angles, I get 2 readings I call good that are within .0001 and 2 readings that I call bad that are -.0005 and +.0004 (see pic's attached).

It seems to me that it should be most accurate while measuring as a cylinder and shouldn't matter which probe angle I pick up the face on. ?

Fresh calibration of all probe angles as of this morning (D2L95), max StdDev .0001 on 28 different probe angles.

I have looped this program 5 times and all the #'s repeat within .0001

Any ideas of what I'm doing wrong to get all this deviation?

The whole reason I did this test this morning is I was measuring another part last week with similar features that did not give correct results.

I'll take a stab here.
1st, if you want the indicator and the CMM to report the same values you must hit the same points. If you want the exact same values you must hit the exact same points and align the exact same way.

2nd, are the hit points the same with both circles and both planes? Same with both Cylinders? Deviations in the hit points between the two features will account for some error.

3rd, the form of the cylinder could move the centerpoint as much as you have shown 0.0004" (4 tenths) and account for the differences between the CL of the circle and the CL of the Cylinder.

4th, there is 'some' repeatability error with all articulating heads. Getting close to a tenth here is not uncommon.

In summary, if you want the numbers to match you have to use the exact same setup and touch the exact same points in BOTH methods.

HTH,
Bill

[Cumofo]

nvm



G

[John Kingston]

I have to agree with Wingman. Based on where your planes are shown, the hits are not in the same place. You could be picking up higher/lower spots on the plane surface with one method. Using the surface plate and height stand is getting the highest spots on that surface.

[PowerJunkie]

Yes, I see your points!

I made a new program setting the alignment to the face, and set the 2 planes at the same points theo, and they were very close @ actual. The cylinder is a construction from the 2 circles.
After dimensioning it again, there is now .0004 total deviation which is about 1/2 of what I was getting before, that's cool!!

It's still showing on the high side, worst case .0007 off what I get on the rock, and I leveled that part on the rock dead.

I'll keep plugging away at this, try and get it perfect.

Thanks for you comments