11-21-2005 09:23 AM - edited 11-21-2005 09:23 AM
Message Edited by TommyBoy on 11-21-2005 04:24 PM
11-22-2005 11:28 PM
11-23-2005 07:42 AM
Hi!
thanks for your quick answer.
I can't give you very much numbers, but I can say that the application is a big hydrolic machine that
punches holes in the I beams. These move with a irregular speed on the machine. There a 2 motors to
move the beams: at the start the first motor is used to push the beam in the machine, and in the second half
of the application the 2nd motor starts pulling the beam. Max speed is about 5 cm/s. The machine is built to avoid
rocking of the beam. The beam is plain steel. If necessary a second camera shouldn't be a problem.
Thanks again!
Thomas
11-23-2005 09:12 AM
Laser doppler velocimetry may suit your needs, you may wan to Google it. Here is an example of such a unit with quadrature outputs:
http://www.betalasermike.com/Literature/USLiterature/ls4000_series_brochure.pdf
I presume mounting an encoder to the drive motor shafts is out of the question (i.e. slippage)?
11-26-2005 04:35 PM
Given your description of the application - do you need to be so fussy. A spring loaded rubber roller, friction driven by the beam and coupled to a standard encoder would give reasonable accuracy of feed rate. Enough to do any surface inspection with a Linescan camera - it all depends on the relative sizes of features you are trying to detect on the beam. I would go for the simple solution first - grab some images and then see if you can reliably work with them - but perhaps you have already done this?
Charles