I am trying to find the quickest way of checking the flatness of some lock plates of 0.10mm tolerance. I have managed to make a fixture to locate them in the same place each time taking out the manual alignment each part (you only have to do the first part of the batch) has anyone had success measuring flatness with the camera taking 20 or so points across 120mmx30mm square parts. I measure a plane using 9 points z level and origin and then the 20 points measuring a plane and then out put the measurement from the last plane? this is all DCC and using the Camera tesa Vision. The X,Y rotation I use a datum hole in the fixture.
after doing some tests the camera is not the way to check flatness or any z figures. It has too many variables (light , focus time, distance etc) it would definatley fail a gauge R&R in that sense. But i do think it is better for profiles if you do not have a scanning capability.
Personnaly I am not seeing the benefit of the vision side. the company would have been better spending the money on scanning equipment.
IMO
There are some measurements that you can't do with scanning head... I agree with you, vision depends on a lot of parameters, so it's hard to be accurate... But sometimes, it's better to have a measurement not accurate than no measurement !!!!!
I was able to do this and pass a GRR on a Mity QVP machine with a pattern focus option/coax light and individually tweaking every failed locations lighting and focus settings.
It was for some crazy piece of plastic for a company named Opgen. Your situation sounds almost exactly the same as what I had to do. same exact size/dimension/tolerance. Unfortunately, there was no quick way. I believe I spent a total of 40 hours tweaking and debugging each individual point and running a repeat test then a GRR for each point. For us, it was worth it considering that it was going to cost us a small fortune to outsource this.
