I'm measuring the material thickness on a casting with curved surfaces. I have points at the same co-ordinates on opposite sides of the material and am having issues getting the correct readings. The vectors on the opposing points are opposite, not sure if this is correct. Since these are points not along a datum so measuring parallel or perp. to them is not an option. Thanks.
The vectors should be opposite. As far as I understand you're trying to take a hit from X to the left and then from X to the right (or vice versa for both), so the vectors must be different.
Are you able to use more than one point? Perhaps define a plane or cylinder, depending on the feature. There is a width feature somewhere I've read about...
Thanks, that confirms what I'm seeing. I've been using a similar program written by the previous programmer. I'm doing everything like he did, but not getting the correct results.
Do you use vector points or surface points ?
You can try to calculate the distance along the first vector and the second, and look at the difference :
ASSIGN/V1=DOT(PT1.XYZ-PT2.XYZ,PT1.IJK)
ASSIGN/V2=DOT(PT2.XYZ-PT1.XYZ,PT2.IJK)
If V1=V2 (or very close to V2), then you've got the thickness, if not... ???????
What are "correct results" to you? It could be a menagerie of things. If you're not getting an exact number, I mean. Are you measuring using a precision machined (and preferably Hex endorsed) calibration gage, or similarly-precise, calibration-spec object? If not, it would make sense that you're not seeing what you're used to.
We need more details to help you get the correct result
I'm using a Caliper Gage with an Indicator, so my reading should be withing a mm, but I'm seeing more around 2 mm difference. The casting is somewhat round with bulges in varied places.
Based on personal experience...the most accurate way to measure thickness with a CMM:
Assuming the 2 surfaces are relatively parallel:
1. Create a Vector or Surface Point on Surface 1.
2. Create a small 3-pt Plane on Surface 2 (directly opposite the Point).
3. Dimension as a 3D Distance with the Point selected first.
Through experimentation I have found this method to be as accurate as using a ball-nosed micrometer.
