SingArea\LockAxis4: Very weird behaviors

DenisFR

Hello,

Can someone test this code on an IRB 4600-60/2.05 with RW 6.13.01.00:

MODULE Module1
  TASK PERS tooldata tGripper_LS:=[TRUE,[[-162.201,7.50024,492.289],[0.653041,-0.653041,-0.271179,-0.271178]],[1,[0,0,1],[1,0,0,0],0,0,0]];
  TASK PERS wobjdata wLTrfRWay_LS:=[FALSE,TRUE,"",[[1329.56,-494.855,690],[1,0,0,1.78363E-8]],[[420.69,-24.75,-40],[1,0,0,0]]];
  CONST robtarget pEntryLTrfRWay_LS:=[[-600,-250,200],[0.707106781,0,0,0.707106781],[-1,0,-1,0],[9E+9,9E+9,9E+9,9E+9,9E+9,9E+9]];

  PROC main()
    !Add your code here
    MoveAbsJ [[-9.56,-50,20,0,30,90],[9E9,9E9,9E9,9E9,9E9,9E9]]\NoEOffs,v1000,z50,tGripper_LS\WObj:=wobj0;
    SingArea\Off;
    MoveL pEntryLTrfRWay_LS,v1000,z50,tGripper_LS\WObj:=wLTrfRWay_LS;
    MoveAbsJ [[-9.56,-50,20,0,30,90],[9E9,9E9,9E9,9E9,9E9,9E9]]\NoEOffs,v1000,z50,tGripper_LS\WObj:=wobj0;
    SingArea\LockAxis4;
    MoveL pEntryLTrfRWay_LS,v1000,z50,tGripper_LS\WObj:=wLTrfRWay_LS;
    MoveAbsJ [[-9.56,-50,20,0,30,90],[9E9,9E9,9E9,9E9,9E9,9E9]]\NoEOffs,v1000,z50,tGripper_LS\WObj:=wobj0;
    SingArea\Wrist;
    MoveL pEntryLTrfRWay_LS,v1000,z50,tGripper_LS\WObj:=wLTrfRWay_LS;
  ENDPROC
ENDMODULE

SingArea\Off; Works fine
SingArea\Wrist; Works fine, but tool orientation is not kept (I need it)
SingArea\LockAxis4;
The robot move strangely to, in final, trig an error...

The thing is that All my position are calculated and in some case, I need this SingArea\LockAxis4.

Thanks

lemster68

Denis, the first thing that I notice is the lack of fine points in the moves.  I suggest to try again with fine points.

DenisFR

Hi @lemster68

It's not about final point but how robot move to it.

lemster68

Yes, but I am thinking that it may be reading ahead, not having the actual setting at the time you think it is.

lemster68

BTW, good to see you back here Denis.

DenisFR

Even If I Run it Step By Step, robot move really weird...

lemster68

Hi Denis.  I built a station with that robot model and RW, tried your code.  Is see the peculiarity that you mention, and it errors with axis 5 at joint limits.  The tcp trace did not show for that move, but I suspect that, with axis 4 locked, that the robot is trying to force axis 5 through the wrist to remain on the path with the correct orientation.  Another strange behavior that I saw was that when I changed that motion to joint, the robot locked axis 4 and also moved strangely until it faulted out.  SingArea is for linear and circular interpolation, not joint.
Finally, from the manual is this:  If the argument \LockAxis4 is specified, then axis 4 is locked to 0 or ±180 degrees to avoid singular points. The TCP will in general follow the correct path, but the orientation of the tool will deviate if the position was not programmed with axis 4 at 0 or ±180 degrees. For paths with large reorientations, the TCP may deviate from the programmed linear path.