Looking for ideas for a manually written angled-probing routine

[tilikoom]

Hello Community,

Caught in a bad situation where we were asked to write a manual probing routine to check the alignment of two angled holes on a massively warped fixture that somehow slipped through QC. Currently our CAM software does not support probing. The fixture shown is a mock-up. Confusing matters more is that the fixture has compound angles. It is not the same overall thickness all the way across, although difficult to see in the screenshot.

The manager is inexperienced at probing and refuses to make another fixture. He assumes that we can align the part to a machine axis even though it is twisted to all hell. Therefore he insists on writing a manual probing routine that finds the misalignment of the two angled holes as mentioned, and manually adding that discrepancy to a G68.2 plane or G54.4 correction offset in the machining program.

Here's a rough draft of our current plan:

1. Create an initial starting point on the bottom corner, and set that as G54
2. Hard code the exact (or as close as possible) location of the first hole and set that as a G68.2 TWP, then probe the first hole.
3. Move to the next hole in Y+ and probe that hole
4. Document the misalignment of the two holes
5. Add the discrepancy to the G68.2 TWP in the machining program

Any feedback would be highly appreciated.

Thank you

 

 

[Tim Johnson]

Can you use the machine as a sine-plate? (if you don't have one) That seems like that would be easier and faster.

[tilikoom]

We discovered that just probing the first hole and using it as the machining origin may suffice - sheer luck. So hopefully we will not have to do any complex error compensations. Also got lucky with some other "emergency" jobs that pushed this aside!

To answer your question since I was not at liberty to show the real part, this is just a quick mock-up. Unfortunately the real part is mounted to a fixture that would not permit the use of a sine plate.

[cncappsjames]

With probing, you can solve linear/positional discrepancies as well as pitch, roll and yaw with WSEC (G54.4 on a FANUC control). Fixture Twist/Bow error is a FAR more complicated endeavor usually requiring some sort of error mapping.

[civiceg]

Trust me I am far from a Autodesk fan, but Fusion's alignment probing routines and ability to adjust the nc code to that alignment is awesome. Probing with skip signals for 3d probing is typically all I ever use Fusion for, unless a contract forces me to make a part in Fusion. Hour and a half mark talks about alignment. Granted this is for a Haas, but really any control with DPRNT capability could make it work with minor tweaks to the inspection post.