As far as not needing the expr for the calculation I am at a loss why this worked. I probably just entered it that way without thinking and it worked.
I chose to use the 10 inch distance because it seemed to give me a path that followed the surface well. In some cases using the flute length would have given me a lot fewer points which would cause a part discrepancy. The longer the length the better the quality of the output. Ideal would be gauge length plus pivot length but that would require additional input by the programmer so I ended up using 10 inches.
And, with Post Builder, you also have the choice of linearizing moves by interpolating (distributing) angles or tool-axis vectors
I have never seen any information on this feature. That might be just the ticket.
Just so you all know, just about everything I tried made everything worse than what I had. When I got the system to offset the values by turning off mom_kin_clsf_generation then it messed up circles, helix and cycle output. It was just not worth the trade off.
So we just found a way to do it in ICAM instead.
Thanks for the ideas and help!
set mom_kin_linearization_flag "1"
set mom_kin_linearization_tol "0.001"
Does this mean NX will break up rotary moves at angle 0.001 increments (and split the XYZ change to that increment?) I will have to try that. The best algorithm (I would think) is to look at the end of the tool and head pivot lengths to evaluate (when deciding to break up moves) - but this would probably be ok.
Good to know.
That is the same conclusion I came to. It seems to be one of those hidden routines that sometimes sneek in and make life difficult.
"mom_kin_linearization_tol" is the chordal tolerance (acts like intol/outol for path generation) for breaking up the moves, not the length or delta angle of the segments. If the gauge-pivot distance is specified in the post, it will be used for linearization; otherwise the "flute length" will be phished out of the tool (as the minimum height) for interpolation.
Turbo mode must be turned off, if you ought to manipulate anything in the event handlers (Tcl). This is how turbo is achieved by bypassing most of the handlers.