We have a KUKA Quantec KR240 R2900 with a turntable. Making a wild guess, i used the KR 90-120 R2900 Robot Type.
I'm using the custom turn table component which is perfectly oriented in A B and C to the hundredth of a degree. The X,Y, Z of the robot base is 0,0,0. The one of the turntable is -394.739,+1965.546,+591.775.
Since i want to keep working in the robot base, I have set X,Y,Z as +394.739,-1965.546,-591.775 and called it base 0. In the real world, base 0 is 0,0,0,0,0,0 as said before which is cheating, and might be the cause of the issue.
We have to mill the ends of a thousand wood battens. We place them on the turn table, which rotates to allow the robot to access difficult position. (for exemple, if the robot has to mill under the beam, it has to be positioned almost right in front of it (about -57°), while if it's on the upper part, the table has to rotate in the other direction, around +15 degrees). Reading from another post i can't find anymore, i made a formulae to calculate the angle for the turntable, based on the Z axis the mill has to come from. This means that i don't use Dynamic Value by vector (since i couldn't figure out how it worked, and provoked a number of issues) but "Same as Previous Value". The calculated angle is sent into the E01 input of the LIN Movement.
The problem is that the whole virtual trajectories move during the simulation (not just what the robot does), but the physical trajectories are those drawn in the canvas at the very beginning of the simulation, before the first rotation of the table.
What could cause it ? is it the base cheating, since base of the turntable rotates ? How could i solve that ?
I've attached 2 pictures. The only difference is the simulation slider.
Thank you very much,