I think what Robert propose is great - to push current TP little bit further so planning could take care of ABC axes could be used along XYZ in inverse time (g93) mode and with g01 moves.
In my case rotational axis A is actually rotating workpiece around Y axis and radius of A is always easy obtainable as sqrt(x^2 + z^2). For passing radius of rotation to tp - I would prefer to have some dynamic method wich means that change of radius will not require restarting lcnc. I am willing to test such tp as it become available. I have machinekit simulation in place so with two gcodes results could be seen quite fast - I could record video or take output on some other intermidiate results. Test case goal would be, that also with small rotational acceleration and big rotational velocity limit, tp could plan it in a way that this large velocities are reached in such cobined small segments of gcode. I think we should focus to workpiece velocities (as I know all tool limitations like max speeds or recomended speeds are refering to tool to workpiece relationship) - but I think: *motion.current-vel* (OUT FLOAT Current cartesian velocity) refers only to vector sum of v_x,v_y and v_z. It doesnt take into account rotational angular speed of A. On Sun, Nov 6, 2016 at 10:46 PM, Chris Albertson <albertson.ch...@gmail.com> wrote: > If you can compute inverse kinematics for the general case then the > software in effect "moves" the TP at the desired rate while using inverse > kinematics to compute the required motor rates. Planning has to consider > the motor rates and if the rate is within limits. Perhaps (I've not read > the code) LinuxCNC works in a machine centric cordinate system > > I think if the goal is to limit TP rate/acceleration relative to the metal > being cut then you need work in the stock-metal's coordinate system > The hard part might be transforming the DH from machine centric to stock > centric. > > I think this is the same problem a two handed human might have if > "machining" a block of wood held in one hand with a die grinder tool in the > other hand. Best to do the computation as if the woodblock were stationary > and the human was spinning around somethings with feet in the air. > > I admit IK is a hard problem. I have a six leg, 18 DOF machine but I > cheated and hard coded the IK. > > On Sun, Nov 6, 2016 at 1:14 PM, andy pugh <bodge...@gmail.com> wrote: > > > On 6 November 2016 at 20:54, Chris Albertson <albertson.ch...@gmail.com> > > wrote: > > > There is already an industry standard for describing the general case > of > > > rotary and translational links bolted together serially and even better > > > already software for working out the kinematics and joint and movement > > > rates and accelerations > > > > LinuxCNC supports general serial kinematics defined by > > Denavit-Hartenberg parameters using the genserkins module. > > > > This is a somewhat different problem. > > > > If using genserkins (or any other TCP kinematics) then the existing > > XYZ blending is all that is needed, as _only_ XYZ moves result in a > > tool tip motion. ABC moves simply alter the tool orientation abit that > > point. > > > > The problem to be addressed here is somewhat different. > > > > -- > > atp > > "A motorcycle is a bicycle with a pandemonium attachment and is > > designed for the especial use of mechanical geniuses, daredevils and > > lunatics." > > — George Fitch, Atlanta Constitution Newspaper, 1916 > > > > ------------------------------------------------------------ > > ------------------ > > Developer Access Program for Intel Xeon Phi Processors > > Access to Intel Xeon Phi processor-based developer platforms. > > With one year of Intel Parallel Studio XE. > > Training and support from Colfax. > > Order your platform today. http://sdm.link/xeonphi > > _______________________________________________ > > Emc-users mailing list > > Emc-users@lists.sourceforge.net > > https://lists.sourceforge.net/lists/listinfo/emc-users > > > > > > -- > > Chris Albertson > Redondo Beach, California > ------------------------------------------------------------ > ------------------ > Developer Access Program for Intel Xeon Phi Processors > Access to Intel Xeon Phi processor-based developer platforms. > With one year of Intel Parallel Studio XE. > Training and support from Colfax. > Order your platform today. http://sdm.link/xeonphi > _______________________________________________ > Emc-users mailing list > Emc-users@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/emc-users > ------------------------------------------------------------------------------ Developer Access Program for Intel Xeon Phi Processors Access to Intel Xeon Phi processor-based developer platforms. With one year of Intel Parallel Studio XE. Training and support from Colfax. Order your platform today. http://sdm.link/xeonphi _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
