02-06-2008 21:02

roboticWanderor

Re: pic: 2009 robot idea

nice. you integrated the super shifter into the chassis very well!
how are you running / tensioning your chain?

02-06-2008 21:07

sdcantrell56

Re: pic: 2009 robot idea

That looks very similar to our drivetrain this year with a few exceptions, the most notable being ours was a single speed. Some suggestions I have for you. The outer wheels look like they have bearings in the frame. You dont need the bearings as the kit wheels have bearings in them. You can use a 1/2" round standoff that is bolted in place as an axle and it will serve to strengthen the frame. Also you can have the axles on the end slide horizontally in slots to tension the chain. Another thing would be to not have the 80-20 on the ends run into the standoffs. It will require very long bolts which equal more weight. Finally it is ok to use 80-20 if you really want to but using simple 1" box tubing is much lighter and if welded will likely be stronger as well.

Overall though, it is an excellent design and you should look into building it this summer to work out some of the kinks.

02-06-2008 21:17

Andrew Schreiber

Re: pic: 2009 robot idea

Maybe use the 1 inch box tubing for the bottom structure and only have the 80-20 on the top as a mounting surface.

Oh! very nice job, it looks good.

02-06-2008 21:20

sdcantrell56

Re: pic: 2009 robot idea

I actually really like that idea. It would be the best of both worlds. Really the only reason to use 80-20 is because it is easy to build with with not much planning. It is really to heavy and not strong enough for the weight for much else.

02-06-2008 22:06

Billfred

Re: pic: 2009 robot idea

I'm all for 80/20 in the right situation. The best use of it on a drive system I've seen came from my alma mater of 1293 in 2007. Consider:

http://flickr.com/photos/billfred/1337894883/

The two bolts centered vertically with the 1293 are 1/4"-20 bolts holding each drive pod to the center rails, which appear in this photo to be 1515. Heavy, yes, but it's a pretty short length. The 1515 goes straight through to the outer plate on both pods, tying it all together (and providing pretty awesome nice mounting options for a mechanism to boot).

They haven't had any issues with that aspect of the robot to my knowledge, but I also changed teams right before that season. Someone who was active with D5 that year may be able to tell you more.

02-06-2008 22:35

pacoliketaco

Re: pic: 2009 robot idea

thanks for the suggestions. just so you know, the connection between the standoffs and the 8020 pieces was not to be a really long bolt, just a 1.25" long coupler bolt. and im not sure how much our team will be able to change in just one year. in prior years, we have used only 8020 for our entire frame, so anything else would be a major improvement. and to tension the chains, i thought to have some sort of lever arm coming off of the center of the "X" in the side panels, which would connect to a roller which would connect to a spring. but this could be overly complicated. or something like this: http://www.chiefdelphi.com/forums/sh...hain+tensioner

02-06-2008 23:01

sdcantrell56

Re: pic: 2009 robot idea

We did the slots that the axle slides in and then is locked inplace by tightening bolts on each side. This method works extremely well and does not add any weight. I would definitely recommend it over adding an idler that slides and is locked in place like the picture you posted. We used #25 chain and we never had a single problem with our drivetrain the entire season.

03-06-2008 12:36

Skifanatic

Re: pic: 2009 robot idea

Looks a lot like 25's drive train but they run it with gears instead of a chain.

03-06-2008 17:58

pacoliketaco

Re: pic: 2009 robot idea

okay, so i have been further adding features, trying to figure out how to tension the chains. im not too comfortable with moving the wheels in sliding shafts, so i added a bar of sorts to tension the chain.

the lines represent the chain, and the circles represent sprockets and tensioners. the small circles, the ones that appear to be in the slots, are composed of two tubes, the outer being of HDPE, and the inner of polyethylene. the intent was that a bolt would run through these two tubes, and would be able to slide up and down the slot, in order to tension the chains.

03-06-2008 18:00

AndyB

Re: pic: 2009 robot idea

You want to shoot for around 120 degrees of chain wrap as a minimum and I think with your current run, you will see major slippage problems with you center axle.

Try moving your tensioners inward or split it into two separate chain runs.

03-06-2008 18:14

pacoliketaco

Re: pic: 2009 robot idea

yeah, i guess i didnt make that clear enough, there would be a separate chain going in each direction

03-06-2008 18:24

tseres

Re: pic: 2009 robot idea

this is the classic thread of one infected with the FIRST bug. i love it!

03-06-2008 18:25

EricH

Re: pic: 2009 robot idea

If that's one chain run, then I want to know how he got a spiral chain wrap.

It's two, with about 180 degrees for each chain and sprocket.

03-06-2008 18:30

AndyB

Re: pic: 2009 robot idea

Oh, sorry. Looks good then. I know you designed this to fit the supershifter gearbox, but it might be easier to make if you used 4" tall sides instead of 5".

With 5" sides, you would need to order plate and cut it to length and width.

With 4" sides, you can just purchase 4"x3/16" Extruded Flat Bar and cut it to length.

I guess if your gonna CNC or waterjet them anyways, it isn't that big of a deal, but I'm pretty sure flat bar is cheaper than plate anyways. (I can not find 5"x3/16" extrusion anywhere).

03-06-2008 19:12

s_forbes

Re: pic: 2009 robot idea

Looks good, nice and simple like it should be. I'd prefer just using dead axles in this particular set up for the end wheels, it would probably cheaper and easier to build. Either way, it looks like a very solid drivetrain.

Does anybody else notice the trend lately of putting the wheels between two plates held together with standoffs? I may be imagining, but there seem to be a lot of teams using this method recently...

03-06-2008 21:39

sdcantrell56

Re: pic: 2009 robot idea

I agree there has been a trend for this and we are guilty of it as well. I have tried many different drivetrains and this method (2 plates with standoffs) really seems to be the best one I have done. It is particularly good if you are sponsored by or have access to sheetmetal stuff aka Laser or waterjet because the entire frame for the drivetrain can be knocked out in a couple hours. Also it allows for really quick assembly and wheel changes as well as it is super easy to incorporate the transmission into the frame, thus reducing weight.

05-06-2008 18:38

Peter Matteson

Re: pic: 2009 robot idea

We actually originally designed our drive base to use 80/20 or 1x1x1/16 box for versitillity in case we didn't have welding capabilty anymore. I'm all for that type of flexibility.

I however disagree with your comments on weight and strength. The shape of 80/20 gives it superior stiffness to 1x1 and the weight is comparable to 1x1x1/8 tube if you actually look at weight per foot among other things.

26-06-2008 22:12

pacoliketaco

Re: pic: 2009 robot idea

so, i have been trying to design a lighter chassis than this years robot (at ~60 lbs) and i came up with this:



by my calculations, it will move at 14.82 ft/sec high gear and 5.79 ft/sec low gear. i have also figured the weight to at at around 45 lbs.

and just for future reference, 1x2 8020 weighs about .98 lbs/ft, and 1x2x.125 aluminum weighs about .88 lbs/ft. not much of a difference when the strength is taken into account.

26-06-2008 22:22

Waynep

Re: pic: 2009 robot idea

Great job here. The concept looks very similar to the 25 drive adaptation that 103 has been running with good success for a couple years. They swear by the 8020 ease of prototyping and retrofitting. You may want to get in touch with them to bounce ideas around.
-wayne

26-06-2008 22:25

Alex Cormier

Re: pic: 2009 robot idea

Looks great, good job on the improvement. I do make one suggestion, I have experienced similar type of tension systems. I say instead of a stright shot for the tension to be calibrated, I would recommend them to be holes every 1/8" or so. This will greatly improve on the idler not slipping from having too much force applied to it, and you are always sure that it did not move during a match.

26-06-2008 22:52

pacoliketaco

Re: pic: 2009 robot idea

i guess in the rails in which the tensioners ride, i could just put a few holes/dimples that allow the screws to tighten further than usual.

27-06-2008 11:40

Dan Petrovic

Re: pic: 2009 robot idea

That's actually not a bad weight for a chassis (in my opinion, of course). What I like about heavy chassis' is how your center of gravity turns out when you're all finished.

With a chassis that is less than 60 pounds, a majority of the weight is going above the chassis, raising the center of gravity to a less desirable place. Our center of gravity in 2007 was seriously at the bottom of our chassis. Teams would really have to try to get us to tip over. That chassis was probably close to 80 pounds.

27-06-2008 11:56

Tom Line

Re: pic: 2009 robot idea

One thing - I always like keep things simple and easy with few parts.

Have you considered just using aluminum C-channel (or even square tubing) for the left and right areas where the wheels are? A hole saw can make quick weight savings holes, and you don't need the fasteners, the machining, etc. It's stronger, lighter, and easier to make. Just a suggestion.

27-06-2008 12:05

sgreco

Re: pic: 2009 robot idea

This conversation about weight is interesting. My team's did a swereve steer this year, it weighed a little over 90 pounds. Once we added the electronics and compresor it weighed 100 pounds. Then with the battery it was closer to 110 pounds. All of this was about a half inch from the ground(maybe an inch).

27-06-2008 14:23

=Martin=Taylor=

Re: pic: 2009 robot idea

Great Job! The level of detail in the CAD is amazing. How'd you make the chains?

This is VERY similar to my team's last two drive trains. This kind of system is very adjustable and easy to build. It's worked great for us these last two years.

2007:


2008:


Some things to consider:
5.79 fps is probably a little two slow. We ran low gear at about his speed this year and it was useless. The slowest I would go is 7 fps.

You could save more weight with #25 chain. This system makes chain tensioning a breeze, and you could easily implement #25. We used #25 this year and never lost a chain.

I'm curious as to why you put the axles on the center line of the extrusion. It would be a lot easier to put the pillow blocks on the top or bottom. You could save a lot of machining, and replace the 1020 with 1010.

Altogether great work though. Are you gonna build a prototype?

27-06-2008 21:19

whytheheckme

Re: pic: 2009 robot idea

A note on the weight.... Remember that this year's control system will weigh significantly (3-8 lbs., if I recall correctly (someone correct me)) more than in previous years. It also will require quite a bit more (continuous) space than in previous years.

Love the design. We need an overhaul on our concept of the idea of a 'working drivetrain' (currently my team is having trouble with that concept), and this is a great place for us to start!

Jacob

07-07-2008 16:29

team2061

Re: pic: 2009 robot idea

i saw that u guys used pneumatic wheels on the 2007 robot...we were going to use that this year but we couldnt find a way to mount sprockets on. Were those sprockets custom or were they bought?

07-07-2008 18:08

pacoliketaco

Re: pic: 2009 robot idea

thanks for all comments/suggestions. i guess i could change the sprocket on the two wheels powered directly from the transmission to something smaller. the main reason that i wanted to use 1020 and mount the axles in the center was for strength. the past three years, we had a very similar drivetrain, which i felt was excessive, but then again, it was very robust. the reason i want a light drivetrain, is that in the past three years my team had been in FRC, we havent had a complicated robot. our other systems end up to be too heavy, and we have to make accommodations by making simpler mechanisms. this year i dont want to have to worry about weight, and starting off 10+ lbs lighter will help.
2008:

2007:


i wanted to make something that was only one level of 8020, to reduce weight. in the past i noticed that a single 1010 rectangle bends too much, so i decided to use 1020. using only 6in wheels, by putting them on either the top or bottom, there was 7/16" of clearance (too little) or 3.5" (too much)

about the weight of chain, #35 chain weighs only .21875 lbs/ft, and with 9.5 feet of chain that is only 2.078 lbs. switching to #25 chain, which weighs .085 lbs/ft, would save 1.27 pounds, which isnt much for all of the added trouble. the purpose of all of my designing is prep for next year, as our team doesnt have the time/materials to build a chassis during the summer. just to see that your team seems to have built the same thing that i wanted to make, shows that i am headed in a good direction.

as to how i make the chains in autodesk, i just use the KOP download, and modify each link a little bit. i added a VERY small rectangle in the center of them, in order for an angle constraint to hold them together. or you can ground one end of a set of links, and simply drag the other end away from it, until it is straight. once you do this, you have length of straight pieces of chain, but you still need the curves. to do this, i find the distance from the center of each sprocket to each other, and make a new assembly to work out of. i then add in the links one by one, making them fit around the sprocket. then i add in the straight pieces, and put the whole assembly on the finished robot. its a lot of work, but it looks awesome =]