TCP 5-axis kinematics

Id like to ask if the question if 3+2 / tilted workplane (hiedenhain G7 or fanuc G68.2) has been solved?
I see this is vital for 5 axis machining. Canned cycles and 2.5 D machining off-axis (undercuts in holes). Also for recovery as manual movement is in line with the spindle vector. This type of machining makes up 95% of muli axis machining, with rtcp rarely required other than some finish (swarf) machining. I know this because I am employed by the 5th largest aerospace tooling manufacturer globally and have machined many freeform tools with vacuum grooves and holes normal to the surface.
Some machines are delivered 3+2 only and require an upgrade (pay) to be allowed to do 5 axes simultaneously.

To add to the above post. My machine configuration is head/table the example max5 is basically my machine. 3+2 is necessary on a machine like this so that you can retract the tool along its axis regardless of the angle of rotation. with 3+2 active Z becomes a interpolation of X and Z and X becomes an interpolation of X and Z.

i think that i have read that UVW is the coordinate system relative to the tool but am unsure how to use them or weather they are relavent.
Heidenhain G7 uses Eulers rule to re-position work-plane/ref frame with or without rtcp active (movement of rotation axes around tool centre or not)
i hope this makes sense to somebody as i would really like to have this feature or something similar. The trig is simple enough would it be possible to use the userk function ?

on another note has anybody added switchkins to max5?

Sorry if i am all over the place but i have a lot to learn about linuxcnc.

Cheers

Matt

With switchkins you can rotate your table to the required angles and then switch to tcp which aligns your Z axis with your tool axis. I use this with a mechanical (indexing) swivel head to machine pockets in handrails for a stair maker.
I'm not sure if max5 kinematic has been adjusted for switchkins yet.

this gantry mill has part of the function i am trying to describe in that z axis becomes w and i assume that u and v can behave in the same manner ie perpendicular and at right angles to each other then be substituted as x y z when switched to 3 + 2

Ah, yes that's true, I had forgotten about the U,V,W axis implementation. However I'm uncertain as to how it is implemented. I wonder if W is handled in the kinematics of that gantry simulation.

Hello all,
I am a newbie on linuxcnc, Ihave made and configure a 5axis (XYZAC) machine, made an ATC,, auto tool length sensing, etc, etc, program everything and that works on the main branch in "normal" kinematic, with TCP configure, a limitation appears ( couldn´t move the rotary like i wanted to acheive what i needed), so i found this nice branch, very pleased with everything just I have troubles with more axes ( like controlling my ATC carousel ), well don´t work at all, before homing everything ok, I can move every axis linear and rotary, after homing, I lost my 6th axe , the W ( the carousel). I was trying to find something to send HAL command to move it, and no, i can´t find any info on that.
UVW are completely disabled with the switch kins mod I tried the dgarr/s_kinsv2 thinking V2 thinking it would be even better ?
any idea people ?
thank you !
Guillaume

UVW are completely disabled with the switch kins mod I tried the dgarr/s_kinsv2 thinking V2 thinking it would be even better ?

In dgarr/s_kinsv2 xyzac-trt and xyzbc-trt kinematics do handle all optional axis letters. (As long as they are specified in the ini file.

But maybe we could continue about dgarr/s_kinsv2 over here:

forum.linuxcnc.org/38-general-linuxcnc-q...witchable-kinematics

I'm using switchkins. I am impressed with your efforts.
I made a machine with Y-axis gantry.
I want to make the Y-axis home switch separate on the left and right to make it square automatically, but it seems that I can not increase the number of joints.
What should I do?