2017 feeder Development

Ever wonder how we designed our 2017 robot's feeder? Well here's the evolution of the system in great detail! Starting at a 3:1 scaled down prototype to 18 balls per second on Einstein!

 
First pic.jpg

We found these on amazon: https://www.amazon.com/Polyurethane-Plastic-Crown-Sporting-Goods/dp/B009LNRHO4/ref=sr_1_6?ie=UTF8&qid=1511198141&sr=8-6&keywords=wiffle+golf+ball&dpID=51-jilpAp-L&preST=_SY300_QL70_&dpSrc=srch

They ended being almost exactly 3:1 scale of the official 2017 Fuel ball (2.97:1 to be exact). They even had the same number of holes and a similar seam. They were about as ideal as you could hope for scale prototyping.

j0WKlg4.jpg

 

After finding this video online: https://www.youtube.com/watch?v=R_a9QVdaMRU

We located this patent: https://patents.google.com/patent/US8047190B2/en?assignee=Dye+Precision+Inc

Using the images in the patent, we were able to reverse engineer how the thing worked, and built a solid model to print at the small fuel scale.

c1GbZSb.jpg

This was our first proto completed in the printer, still on the tray. Printed on a Fortus 400mc using PC/ABS material (black) with white SR-100 support.

StPT6fF.jpg

First printed proto pieces with support removed.

4b2mXSq.jpg

First printed proto all assembled. The hook piece loaded as a slip fit into the base cup and the webbed 'cage' snapped onto the cup, but with enough float to rotate around.

LpuJTXf - Imgur.gif

First printed proto showing the 'action' of the feeder and how it worked to push the fuel to the center. The intention in the final version was to have the hook and cage counterrotating (like its done in the Dye rotor)

jjbGTiq - Imgur.gif

Slow-mo of the first full Feeder prototype. A lot of promise, but quite a few issues with jamming. There was also a lot of hesitation with the potential complexity of having to spin the cage. We pressed through though.

7wyeIsT.jpg

We noticed an issue getting the fuel balls to pass vertically up the middle due to the tight radius. This version incorporated a passive roller to assist in getting the balls to feed from horizontal to vertical.

As time went on, we began to realize the potential for jamming and pinch points with the counterrotating cage. It's clear the Dye rotor has years of optimization in the geometry to allow it to not jam. We began looking at 'self-powered' concepts that would potentially remove the outer cage. One benefit was how tightly packed the fuel could be, which we thought was a key part to a high feedrate.

This was our first "self-powered" prototype. This proved the concept of an active roller guiding fuel up to the middle of the Feeder. Because we wanted to achieve 2x the linear speed of the tip of the Feeder hook, we need a relatively large roller. It was at this point we began to consider using a motor to power this roller.