When jogging in linuxcnc, we logged the time for both ruckig and grotius-scurve in 2 sessions.
Grotius-scurve session:
Function duration: 10329 nanoseconds
Function duration: 2194 nanoseconds
Function duration: 2515 nanoseconds
Function duration: 6766 nanoseconds
Function duration: 1884 nanoseconds
Function duration: 1764 nanoseconds
Function duration: 5985 nanoseconds
Function duration: 2053 nanoseconds
Function duration: 1823 nanoseconds
Function duration: 8252 nanoseconds
Function duration: 1510 nanoseconds
Function duration: 1358 nanoseconds
Ruckig session:
Function duration: 21779 nanoseconds
Function duration: 9161 nanoseconds
Function duration: 10252 nanoseconds
Function duration: 20288 nanoseconds
Function duration: 8817 nanoseconds
Function duration: 10302 nanoseconds
Function duration: 52762 nanoseconds
Function duration: 42609 nanoseconds
Function duration: 32094 nanoseconds
Function duration: 24196 nanoseconds
Function duration: 41582 nanoseconds
Function duration: 32464 nanoseconds
Function duration: 45580 nanoseconds
This is good news.
Also the grotius-scurve is in principle a more powerfull scurve wich includes a linear acceleration stage between
the concave & convex curvatures, where the ruckig curve is like a spline with no linear stages possible.
Next thing to do is integrate the grotius-scurve into the linuxcnc motion planner.
The first integration is without any look ahead. Where the motion starts and stops every gcode line.
When this is working, we will work on the look ahead, and test the
angular motion
"arc" speeds.
So far so good.