Singularity and position out of reach

[fengzhestanely]

Hi Johannes:
I'm trying to weld a spline with a KUKA robot arm. The position of the robot and the plate that can be welded on are both kinda fixed already. I am trying to find the tightest curve can be welded in our working cell.However, when I run the simulation, there are lots of singularities. I don't know how I could avoid these singularities.  Could you please offer me some advice?
The grasshopper files and a photo of our working cells are in the attachments.

Thanks, Stanley

[Xylotica]

Hi Stanley,

Nice welding rig you got there...
I have a similar robot, a KR6-ARC in KRC2.

I wonder what are those two Kuka|PRC components with a strange warning sign...
If you want to avoid singularities, you should try to avoid the situation where the value of axis 5 has a "Zero" value.
A good thing to do is to set a start and end position with a value other than "zero" for that axis.

I think that the trouble in your simulation comes more from the orientation of your tool.
If I flip the curve, then the path becomes valid because the tool is now on the proper side of the wrist.

You might want to define your frames otherwise, so that the orientation of the curve does not dictate so directly the orientation of the tool.



Cheers

[Xylotica]

... I'm not sure what those yellow bands stand for though.
Maybe a new feature that I don't know about yet ?

[Johannes @ Robots in Architecture]

Hello,

It's a very simple way to warn you of close singularities, in this case when A5 is close to 0 degrees!

Best,
Johannes

[fengzhestanely]

Hi Wall-E
The smaller one is the cluster I used to draw my welding head. I will try your way.

Thank you!

Stanley

[fengzhestanely]

Thank you for your reply, Johannes.
Is there any efficient way to resolve this problem or just do it by the trial-and-error way?

Cheers

[Johannes @ Robots in Architecture]

Being close to a singularity is not necessarily a problem, it might just cause the robot to slow down at the tip!
Best,
Johannes