Making a command list that works...

Started by Woodboss, November 17, 2017, 09:53:52 PM

Previous topic - Next topic

Woodboss

I'm trying to cut out a dodecahedron from a block of foam mounted on a rotary table using a hot wire.

Cuts sequence will be: top and one upper side, then the table will rotate 72 degrees for the second upper
cut, etc.

I have defined the dodecahedron and the cut lines are good individually, but I need to get the cuts into a list in the right order and with some DIO and wait for the table. When I try to make a list the cut lines quit working.

Any help would be most appreciated...

Johannes @ Robots in Architecture

Hello,

I'm currently travelling so I cannot look at your file, but if the fabrication time is not critical you could simply work with a very generous retraction strategy, so that even if the face you cut is opposite of the robot, it will still not collide with the foam block. Ideally go up in vertical Z in addition to retracting.
With the visualization in KUKA|prc it should be easy to check. Usually there should not be an necessity to tell the turntable to "wait" - or did you build it yourself, without a KUKA motor inside?

Best,
Johannes

Woodboss

Hi Johannes,

Thanks for the quick reply!

I don't have a problem (yet) with the retracts etc. but I have so far been unable to get the cuts - linear moves- into a list format that works. When I make up the list sequence from the start and end points of each of the four cuts for example, the cut lines stop working or turning green when selected, and the command sequence is not what I would expect...

I suspect this is something simple that I'm doing wrong?

My rotary table is a full size Fibro with Esitron plc which came off a demolition site! Only signals available for robot use are DO start, and DI ready. But very versatile programming features for sequences of moves etc.

Best regards,
Charles.

Johannes @ Robots in Architecture

Hello,

I quickly added a few components how that might be approached, see the attached file.
If you want to automate it more nicely, you can use Weaverbird to create the Dodecahedron, then you get the outlines and use Boundary Surface to create a planar surface. If you "untrim" the surface, you get a square patch. Then after reparametrizing it, you can automatically generate your cutting line as a isocurve through 0.5/0.5, and get the orientation by evaluating the surface at a random point (as it's planar, it does not make a difference where, you can simply use the default 0,0)
I did not know Fibro, they look very heavy duty! As you cannot do synchronized movements with it (unless you replace your motor with a KUKA motor, that is) you then need to figure out the rotation point in relation to the robot.

The easiest way to do that, is to calibrate a tool and set the base to 0 (world coordinate system). Now move to a point on the turntable and note the cartesian position. Move the table by at least 100 degrees and repeat at the same position. Repeat again. With three points, you can then construct a circle in Rhino/Grasshopper. The center of the circle will be the center of rotation.

Hope that helps!
Best,
Johannes

Woodboss

Hi Johannes,

Your solution is far more elegant than my attempt. Though obtaining a dodecahedron in one move via Weaverbird does seem a little like cheating!

I'm restarting, and following your suggested workflow.

Thanks again for your help.

Best regards,

Charles.

Johannes @ Robots in Architecture

Hello Charles,

Great! Keep me updated, via forum or eMail!

All the best,
Johannes