pic: 1323's 4x1 Wheel

[R.C.]

loading photo ...

[Andrew Schreiber]

Tread? What type, how is it held on?

I assume you are CNC-ing this. (how does one turn CNC into a verb? PM me) Do the small holes near the edge of the wheel really remove that much weight? I can't imagine they do, Try seeing how much it takes out. I know that they look cool but they probably increase your manufacturing time by a lot. Otherwise you could just drop a piece of 4" round 6061 on the lathe and have this wheel knocked out relatively quickly.

[NickE]

The wheel looks cool by itself, but pocketing doesn't always make things look good. Too much pocketing (wheels, frame, etc.) will become too busy and will take away from the beauty of your machine.

Even with a CNC, you're going to spend forever milling those tiny pockets with a tiny endmill to fit into the tiny corners of said tiny pockets. Your best bet is probably just to reduce the number of holes.

If you really want to pocket it out, you could probably work with the lightening pattern and remove more material. Try making some spokes instead of just cutting random holes.

[artdutra04]

It looks cool, but not very practical. The machine time looks like it would easily take over an hour per wheel due to the large number of small pockets and radii. If it's designed to use larger end mills to do the lightening, you can really fly through the machining. On similar (although smaller) 7075 hubs for our shooter I designed this past season, I wouldn't let myself use anything smaller than a 3/8" flat and a 1/4" ball end mill, and was able to machine each in only a few minutes. The three holes on the spokes are #10 clearance on 1-7/8 diameter in case we ever wanted to use AM sprockets (unlikely, as the timing belt works super well, and is very quiet).

Anyway, here's some pictures:

(pre broaching)

(on the robot)


As for your wheel, it would be orders of magnitude faster to lighten the wheel by using a large drill bit and changing the tiny triangular pockets around the outside of the wheel into ordinary holes. You don't need a billion triangular pockets to have a good looking robot. There is beauty in pure simplicity.

[R.C.]

Alrighty,

Adam and I were talking about these and I left out some very important details. First of all these wheels aren't designed to on a FRC bot due to the amount of run time they require. They aren't effecient. The reason for some of my crazy CAD is not to make, about 80 percent of it, is to inspire some of the members on my team.

Our team consists of 40 members, a few of them know how to do most if not all of the aspects of robotics by themselves. But this year we have a big bunch of kids that do not know what to, or what is the point of this program and Akash can tell you they don't understand FIRST History. We usually have a half and half mix. This year we have the extremes.

So showing them off the power of CAD and machining stuff for them seems to get them more motivated to DO something on their own. So far we even got funding from the district to start a summer program that is available to any team within 30 miles and is aimed at getting our members more involved.

Plus, a lot of the CAD I do is for experiments, at my work all I have to do are three CAD drawings and then I can go use the CNC mill and lathe. This wheel was made to show others what is possible. I have learned throughout my time in FIRST and CD that it doesn't matter how pretty something is, it doesn't matter. Function over Form is the big thing. If you bot or mechanism doesn't function good, what is the point of it looking pretty.

Just like Art, we used a similar 3 spoke design that used .25 thick 5.5 inch AL shooter wheels. That only used a 1/4 inch endmill and they only took 10 minutes on the CNC (less cause i got lazy clamping them). Then we wedgetop tread and used the al ends to support our whole shooter:

http://team1323.com/pictures/shooter%20wheel.jpg
http://team1323.com/gallery2/09/UC%2...218_large.html

Thanks CD for all the lessons I have been taught,

RC

P.s. Adam, thanks for that first post telling me i was doing stupid stuff a year or two ago..

[Cory]

To be honest, while complicated (and lacking some features to actually make it machinable on a mill), assuming all radii were 0.125" or larger, it wouldn't take longer than 10-15 minutes per side to run that on our CNC mill.

[R.C.]

Originally Posted by Cory

To be honest, while complicated (and lacking some features to actually make it machinable on a mill), assuming all radii were 0.125" or larger, it wouldn't take longer than 10-15 minutes per side to run that on our CNC mill.

Cory,

They are 1/16 offsets and 1/8 fillets. The reason it took me an hour to do it on the router is because I can't pocket super fast on the router. But at my work, it took 5 minutes on each side to make the pockets aka blanks via lathe. Then it was only a one side operation. The back side is plain (you only have to mill on one side, the other side doesn't have any offsets and such). It took less than half an hour at my work, but at school it would take a while due to the lack of a lathe and since our router can't go super fast in al.

[Akash Rastogi]

Just wondering, is there a lip on the wheel for tread?

Send over the CAD when you have time. You do this in 2009?

[Gary.C]

Btw, on the jpeg images refresh the link twice. It may say forbidden for some reason.

[Akash Rastogi]

I have one of the plastic wheels and its pretty sweet. Haven't gotten one of the Al ones though. The Al one did take less than half an hour so Cory is about right. (Thanks for backing us up on that btw Cory). Like RC said, only one side is pocketed like this.





I actually like the HDPE prototype better. Its cool for low stress applications on class robots plus only takes 5 minutes for RC to make at work. (I'm so mooching off of this kid all the time now lol).

[Akash Rastogi]

Here's something a little newer and more refined that we'll be working with this summer. 5"x1" made for 1771, 1323, and 11. Its made for AM sprockets. Man I hate jpgs, they take away so much of the quality from these renders.