Okuma 3+2 w/ offset tilt axis calculation error past 90°

[sewais]

Hello,

Currently using a modified Okuma post processer, to account for the the offset between the B axis & C axis on our Nikken rotatory table. Said offset was verified by @bob.schultz  (thank you Bob) & is manually entered in the post processer per this forum post.

This works great, until we machine past 90°, see sample below showing X value @ 75°, 90° & 105°. Our Nikken's B axis ranges from 0 to 105°

In fusion, I drew a line @ B90° & selected it for the same 2D Contour operation @ all 3 B moves above w/ Compensation Type: Off & Lead In/Out disabled. This put the spindle right on center w. the line drawn.

The X value for the 105° should be identical to the X value @ 75°, both being 15° away from 90°, however, something in the post processer is not applying a negative value once the rotation is past 90°

A full list of post modifications can be found here for reference

[sewais]

Some late night findings:

I've narrowed down the issue to the following code changes found in this post:

It seems the optimize3DPositionsByMachine has something to do w. the issue. This is also mentioned in the Post Processor Training Guide: 

I've tried to use all the above optimize types w. no success. The only way to get the correct X values is to revert to the original code shown above, but then the offset is removed.  I'm thinking it has to do w. the abc portion more than anything, since the guide mentions it being used as a vector. Here is a sample of how the X value should look, notice they are similar @ 75° & 105° as predicted:

[bob.schultz · Accepted solution]

It is expected that the rotated output coordinates would be different when both rotations are 15 degrees from 90 degrees, since the center of the rotation is not at 0,0,0.  The setRotation method gives the same value in X because it is rotating the coordinates around 0,0,0, while the optimize3DPositionsByMachine will rotate the coordinates around the center of the table, which is not at 0,0,0.

An easy way to visualize this is to say the center of the table is at 2,0,3 and the coordinate of the 75 degree point is at 2,0,3.  The output coordinates in this case would be 2,0,3 since it is at the center of the table.  A coordinate of 2,0,-3 at 105 degrees will not be at 2,0,-3 since it is not at the center of the table and will need to be rotated.  In this case both the X-axis values will be different.

If you think that the output coordinates are not correct in your case, please provide the model and post processor that you are using and I can evaluate the calculated coordinates.

[sewais]

Thank you for the reply, I think this makes sense, & I've tried to verify it w. the sketch below.

We machined our first feature below 90°, & saw an inconsistency in size of about .2mm. The first place we suspected was a post calculation error.