[deadlykitten]

hello please, is it possible to cut attached groove, width 6.3, with a mill o6, using rad comp, on a lathe ?

i wish to enable rad comp on the toolpath from image 2 ( now i code such stuff without compensation, and it requires too much time )

i know that walls won't be parallel / kindly

---

only for such operations, there is also the flexicenter from ews :

https://www.ews-tools.de/en/products...exicenter.aspx

in next video :
... 0:10 is the ews center-point : it requires good alignment between toolholder bore and spindle ( closed / clamped ); there is a special sleeve inside the toolholder, that may handle some dissalignment, but even so, is good to be sure, so to avoid the case when turret servos will be stressed, because, if things are not alligned, the clamping event will tend to move the turret and to put stress on the spindle
... 0:30 flexicenter appears, and at 0:43 it is adjusted; such toys are not balanced, but they deliver


https://www.youtube.com/watch?v=akchngDvIEE

[kurmay]

Yes, it is possible.
This is contour generation programming mode (side) function.
G132 Z C L F (for CW)
G133 Z C L F (for CCW)
G119 compensation code

[deadlykitten]

hei kurmay, pls, if you have time, will you test the code from post 2 in here ( is only half a circle diameter 20, or 2 quarters ) : https://www.cnczone.com/forums/okuma...help-g132.html

i onestly have problems with G132, but if i could make it work, then maybe also rad comp will work; otherwise, i am dead / kindly

[deadlykitten]

hello, finally, it worked : instead of G01 C... Z... i use a soubroutine CALL OSUB C... Z..., that automatically calculates feed and allows to make toolpath corections near the cnc panel, instead of going back at the pc and recreate the program.

at this moment, same operation is running on 2 lathes, and each lathe has different corections; even if all is aligned, etc, there is an error when indexing between tools : for example, if a reamer is cutting after a drill :
... reamer center line will not be coaxial with the drill center
... also reamer toolholder will not be coaxial with drill toolholder
such small differences sum up, and i don't wanna recreate the program each time something happens; i preffer to adjust it fast, near the machine; such parametric toolpaths require time to be coded, so i use them only for long-series ( > few weeks )

another thing that i encountered was that for G0 C10, VSIOC is not always 10, but 370, or -710, etc; thus even if the displayed value is 10, sometimes C axis may be at least one full revolution shifted from that value

i solved it by using MOD [ angle, 360 ], so to restrain the values between -360 and +360

in the end, all works, and i managed also to eliminate the auxiliary movement associated with entering the comp ( G42/G41 ) and exiting the comp ( G40 )

i wonder if is it possible to use imaginary vectors, but for this i will open another thread / kindly