Part Alignment on a Haas - Post Error

[jorsborn]

I am trying out part alignment for the first time. I have a large weldment that I need to drill and tap holes into. Post weld it is a little wonky so I thought this would be perfect for aligning those holes on center. However, when I attempt to post for alignment I get an error. A few relevant screen shots below. Any suggestions? I downloaded the latest post processor for the NGC but that did not help.

[rengfx]

Can you share your file or more information like what the error log report says?

I use Part Alignment as a critical part of my workflow, also on NGC HAAS. I updated my post yesterday, but still have the part I had aligned end of last week in the machine so haven't run it again with updated post

[jorsborn]

See attached...

[rengfx]

Spits the code out for me no problem for me

Are you sure you're using an unmodified post?

[jorsborn]

Dang. Yes. I tried it with my usual post and when that failed I assumed it required the latest post (although I believe this feature is a few years old) so I downloaded the latest from HAAS NGC post processor from the Fusion post processor library site.

Did you use a Haas post?

Thanks for trying it out!

[rengfx]

Yes I did that with the most recent HAAS NGC Inspect Surface post ver 43920

[jorsborn]

Well, that's why. I didn't realize there was a NGC post with Inspect Surface. Downloaded that and it posts now. I now have alarms on the control about calibration. There is probably a more elaborate calibration routine which I have to perform.... but that's not a Fusion issue.

[dwilliamsFM6K4]

Sweet! Make sure that probe tip is rotating less than .0002" TIR. I run a 6"+ probe tip so that's always fun for me

Yeah do the full probe calibration again + the vector probing routine HAAS has here https://www.youtube.com/watch?v=2PKUUndD8tA + https://www.haascnc.com/content/dam/haascnc/videos/bonus-content/use-vector-probing-to-probe-a-half-...

You'll want the probe as accurate as possible prior to any alignments

[jorsborn]

Thanks! So I found that I had already completed that probing routine. I re-calibrated anyway and the alarm persisted. I dug into the Renishaw manual and found the output variables and compared it to the post where the alarm is generated (N110). Come to find out that I have always had the wrong diameter probe stylus in my library.  Problem solved.

I got all that squared away and was able to get some results (before and after shown below). I imported those and applied the alignment. It seemed to do its thing. Note the error detected is fairly small until it hits the one end and that is what I expected using a straight edge. The alignment splits that difference. Looks like a linear fit is all that you can get out of alignment. That's not ideal here but I'll take it for this exercise. Next I use the post with alignment button and look over the output. I expected to see a y axis correction at each drilled point but Y is held constant. So, it doesn't look as if it is making any alignment adjustments in the post.

[Tomek.G]

Hi @jorsborn,

The presumption behind the part alignment strategy is that the modeled part corresponds to the part mounted on the machine.
The alignment does not compensate for the deformation of the mounted part but attempts to rotate the part so that the deviations at the measured points are as minimal as possible.

In your case, I would set the coordinate system in the centerline of the holes and update it using the Probe WCS before each drill.