Different methods to measure

[[Ro...]]

Hello Programmers from around the globe,

I hope you all are having a wonderful day/night

Just stumbled on an issue which has been a pain in the butt, just looking to pick other brains.

Below you will see an image attached(just a snipit), it is a stainless steel stamped and bent part.

1P, 2P ,3P- Plane 1,Plane 2 and Plane 3 respectively and 1C and 2C- Circles 1 and Circle 2 respectively



Just a few pointers
1. 1 and 2 are never perpendicular to 3, they can be open either ways
2. R1's are never an one


Right now I have 3 ways to measure it.

1.Go from P1 to P2
2.Go from Max of C1 to Max of C2 ( 90 degree scan path for my circle)
3.From intersection point of 1P and 1C to intersection point of 2P to 2C

As per the print I will get same results on all 3 methods but that never happens.

I might be wrong in the way I do it and I am getting a lot of kickback because I can not get it close to what they got on the comparator.

Thanks in Advance

[[Da...]]

How are they checking it on the comparator? Use the same method. The numbers will be different because the CMM will pick up flaws that the comparator won't.

[[Cl...]]

Honestly, it's a little difficult to understand everything you trying to explain.
Does the print state anything about measuring the part in a free-state, or constrained?

I don't know what your alignment looks like. That could be part of the problem. For the
radii, probe single points (thank Roberto Flores for that strategy).

Scan/probe lines on those outer surfaces and intersect them. Simple distance
point to point for those X &Y values.

[[Ro...]]

Please sign in to view this quote.

Hi Clarke,

Thanks you for your reply!

It's in free state, so we built a fixture to hold it

My alignment is using for Z and Y axis P3 and P1 with Z being the spatial rotation

I will try to make the circle using single points. Thank again for your help Clarke

[[Ro...]]

Please sign in to view this quote.

I've tried it that way. We have like 20 different variations of this part and they run on different machines due to this I have issues matching it with the way they check on the comparator.

"The numbers will be different because the CMM will pick up flaws that the comparator won't" I totally agree with you on this David but my production team isn't that accommodating.

[[Ow...]]

Are they putting it in the same fixture when they measure it on the comparator? Probably not and that there in itself could cause a significant difference because, the fixture constrains movement.

Plus, as you've mentioned, the comparator measures edges and the CMM measures the surfaces, two different things.

If the fixture simulates how the part adapts to the mating component (if it has one) then that's definitely how it should be measured.