Excess of poly

[happytriger2000]

Hi Johannes,
                   Have you ever encountered excess of poly stock left in between the front and the rear of a milling operation when milling a 3d head sculpture? I solved this by moving the tool length in the -z direction by 4mm where the excess of stock thickness is 8mm, but what actually causing this?
Basically the robot moves the router bits to the right side of the stock to start milling the face and once finished the robot moves to the left to finish up the rear of the head, the excess poly stock lies in between the front and rear of the head,

thanks,

[Johannes @ Robots in Architecture]

Hmm... If I understood the issue correctly, hat sounds to me like a calibration problem.
They don't show up with 3-axis milling, but with 5-axis milling (or 3 axis from both sides) these problems become more apparent.
I guess that the XYZ calibration is good, so I would look at the ABC tool angle and of course the mastering of the robot itself.

Best,
Johannes

[happytriger2000]

Thanks Johannes for the reply.
The TCP was done by the 4 pt calibration if this still cannot solve the problem what other option are there available? Would Infrared TCP calibration helps?
By the way the Metal base stand below the Robot was not leveled, the only way to have stock base bed leveled 90 degree with the spindle bit is to adjust the stock platform, will this help better result?

Thanks,

[Johannes @ Robots in Architecture]

I would guess that the issue is less the 4-point method (that gets you the XYZ coordinates), but the ABC angle of the tool.
If I want a really accurate result, I usually calibrate the XYZ tip of a very long tool and then a very short tool, get the tool axis as a vector, and then use that information to calculate the ABC angle. Geometrically that only works with processes that are symmetrical around the tool axis.

The metal stand should not really matter, as you are calibrating a base anyway.
...and of course the easiest explanation of them all is that the stock simply moved!

Best,
Johannes

[happytriger2000]

If I want a really accurate result, I usually calibrate the XYZ tip of a very long tool and then a very short tool, get the tool axis as a vector, and then use that information to calculate the ABC angle.

Can I PM you about this?

thx,

[Johannes @ Robots in Architecture]

Sure! Preferably via eMail, though!