Sheet Metal Drivetrain CAD Practice:
-6WD
-28lbs
-CIMple boxes
-Belt-drive
-4in Dead-axle Colson wheels
We plan to build it in December as practice for the new students.
Sheet Metal Drivetrain CAD Practice:
-6WD
-28lbs
-CIMple boxes
-Belt-drive
-4in Dead-axle Colson wheels
This looks really cool.
Hope to see pics once it is built
Looks cool! How are the pulleys attached to the colsons and how thick is the sheet metal?
My guess is that the pulleys are attached to the Colsons with this
Looks nice.
How are you handling the Colsons on dead axles? What size Colsons are you using? What pulleys are you using?
My guess is this
We are using pulley stock and turning the end down to make our own pulley/hub that will press-fit into the Colson wheel. All the sheet metal is .100" aluminum.
The Colson wheels are the standard 4" diameter, 1.5" width. The pulleys press-fitted into the wheels are 28 tooth, 5mm pitch, GT2 profile pulleys, and the ones on the output shafts are 20 tooth.
You can see here the full wheel assembly with the hub, wheel and 3/8" flanged bearings, and on the right is the pulley hub that is machined from pulley stock.
Nice, I like that custom pulley design. Slick. No flanges?
Re: pulley stock, where do you buy yours? I bought some recently, and was dismayed at how expensive it is.
Looks great!
One question for you though - how do you plan to mount bumpers? I have been practicing with sheet metal drive bases (my team plans to use one on a competition robot for the first time this year 
) and I like the idea of having the flanges of the outermost panels bent outward but I’m not entirely sure how to mount bumpers securely.
We buy all our pulleys, pulley stock, and belts from SDP-SI. I agree the pulley stock is really expensive. At $60 for a 200mm bar I questioned if it was worth the cost to make our own hubs. Although the price drops a lot per bar the more you buy, which helps justify spending that kind of money.
I haven’t put to much thought into mounting bumpers on this frame since it’s just an offseason prototype, but I think what we’ll do is something similar to 148’s 2011 bumper system; put 1" hex standoffs in the corners and then use a uni-body bumper frame and bolt it to those hex standoffs. I’ll CAD this and post a picture of it.
Here is a front view of the bumper frame and hex mount:
http://i.imgur.com/snmm5VI.jpg
You loss a bunch of strength when you don’t have the belly pan under the CIMs. try to fit the CIMs so the belly pan isn’t cut out. Anyways, great job, looks like a good start to a great bot.
I’m sorry idk if you already answered this or not i didnt see it if you did. I was wondering what the thickness of the metal was and is all the metal machined or did you get any of it assembled for you? looks awesome to see the finish product!
All the metal is .100" aluminum and is cut out using a laserjet and then bent on an automatic break, this is done by one of our machining sponsors. Then the students rivet the frame together.
As far as the overall design I owe a lot to the guys from 2168 and 228. Without their generosity and CAD files I wouldn’t be anywhere near a production ready sheet metal drivetrain. Also the design for the pulley hubs was originally from 2168’s 2013 robot and the weight reduction is heavily influenced from their 2011 base. What can I say I’m a huge fan!
$60 for an entire drives worth of pulleys isn’t crazy at all.
Well it’s actually about $230 total for all the pulley stock needed for the drive as we need 4 200mm bars of pulley stock. I only have experience using small belts on manipulators, nothing to this scale before, so i dont know if that price still a reasonable amount to spend on this drive?
If you have a CNC you could mill some GT pulley profiles. But 230$ isn’t that much for convenience in the scale of things, you can easily spend that amount on pneumatic tires or mechanums.
For that price, I’d buy these pulleys from AM and bolt them to this hub from WCP.
That’s $19 per wheel (plus wheel cost) plus tax and shipping. More importantly, it’s zero work.
This is very similar to 2168’s drivetrain this year. Check them out if you can.
I don’t think having CIM relief pockets in the belly pan is going to drastically compromise rigidity. Being able to “drop out” the transmission can be really useful. You may want to make the cutouts around that access hole a little less aggressive, and possibly consider an additional cross member if your manipulator won’t already serve that purpose. The Andymark Cross Hex stuff that comes with those gearboxes in the kit would be an easy way to add some cross support.
From searching through the forums and reading all the posts about belt drives it was my understanding the GT2 profile and 9mm wide belts were the preferred setup for a FRC application, while the AM pulleys are HTD pulleys made for 15mm wide belts I would like to use the GT2 9mm belts as their smaller but the AM solution is tempting. Do you think the different profiles would have a noticeable difference?
It is actually heavily influenced from 2168’s 2013 drivetrain. The credit for the pulley hub design belongs to them, and the weight reduction was influenced from their 2011 drivetrain. They were nice enough to send me their 2013 CAD and I learned a lot about sheet metal construction from taking it apart and figuring out how it all went together.
And that is the reason I have the pockets in the belly pan, so the gearboxes can be dropped out of the bottom of the frame for easy maintenance. As far as being a little less aggressive do you mean making the pocketing smaller?
Given that these are the pulleys that came in the 2013 KoP, I think you’ll fine 
. A team that we work with used the KoP chassis this year and had no problems that I can think of at this time with their drive. The added width is a disadvantage if you want things more compact though.
EDIT - I may have misread your post. If you were suggesting using the 9mm belts on the AM pulleys, I am not knowledgeable enough when it comes to belts to say whether the difference in profiles would have a significant effect although my gut feeling would be to avoid mixing profiles.