The video can be viewed in much higher quality and less distortion from Youtube. Click on this link:
Latest video and click on the "HQ" button (high quality). Analysis of the video is below.


How it was done

The CSUCI thresher was filmed with a Photron FASTCAM-X 1280PCI model 4k high speed video camera generously loaned by a local company.



The first attempt was at 4000 frames per second (fps) and 1/8000 second shutter. At this speed, the image size is only 640x128 pixels. This you-tube video is stretched 2x vertically, and plays 267 times slower.

Threshing on the wheel rim proved difficult to film, so instead we threshed using the spokes in the wheel, filming through the spinning wheel. NOTE: every 2nd spoke doesn't contact the millet -- it's on the other side of the wheel. The impact of a spoke causes a pressure wave to propagate through the millet, but this wave damps rapidly, before the next spoke hits.



We also attempted to thresh fresh millet. It took greater effort, and more florets came off with the seed, but the thresher proved somewhat capable even with fresh, wet millet! (Note, we didn't rotate the stalk, but only attempted to thresh the one side because we were filming).



The second filming attempt used an improved millet feed mechanism (i.e., a semi-rigid piece of cardboard, vs. a hand-held one). The frame rate was reduced to increase image resolution, with the shutter kept at 1/8000 sec to manage blur. We filmed threshing of a few stalks and will post additional snippets over the next few days as we analyze them. A 2000 fps video, slowed down 133 times, is here on you-tube (Note: click on the "watch in high quality" link below the video to see it in higher resolution).

Preliminary analysis of the high speed video:
We've only begun to look at the movies and a few points caught our attention:

1) The millet's motion is highly damped. Energy from consecutive impacts doesn't accumulate, as in a resonant system. This may be why the thresher is so robust and works over such a wide range of speeds, pressures, millet variability, etc.

2) Millet seed often flies off on the rebound from the spoke impact, and the seeds often are spinning -- so angular momentum and the anisotropic strength of the floret stalk play a role.

3) The florets that get knocked off the panicle seem to come from an end or damaged area. The dense packing of interior florets protects them from bending too far and breaking off.

4) This slide is a possible explanation:



See also: High Speed Video at Hampshire College.