Tank tracks vs. Wheels

My team (820) and I are considering using either Tank tracks or wheels. I have a few questions for those of you who know anything at all about tank tracks.

  1. what are the pros and cons of tank tracks?
    -we know that we’ll gain torque
    -we’ll be able to make it up the step
    -we think we won’t loose much speed

  2. Can we buy premade tank tracks? If so, from where and how much?

  3. How do we make tank tracks?

  4. How long does it take to build the tracks

  5. Any words or advice or experieces would be greatly appreciated.

Speaking from experience (we used treads in 2002):

Pros: One of the major pros of using treads is that you get a lot more traction on the bot. It will be much harder for other bots to push you around. Treads (with correct design) also make it easier for you to drive up the platform (from the 6" side panel).

Cons (Important!): Treads require a lot of torque as they produce a great deal of friction with the carpet. This is a major problem, especially if you aren’t the greatest with a drive system and gearing (just like us in 2002), as turning becomes almost impossible and causes the breakers to blow quite frequently. Depending on where you buy them (we purchased ours from BrecoFlex) they may take quite a while to get to the school. Ours took over 4 weeks and left us with almost no practice time (under a day before shipping).

I’m not denouncing treads but I would be careful before using them.

One thing you can do is mount a caster or two on pistons to deploy. This will almost emulate a 2 wheel drive and make it much easier to turn.

If you look at Truck Town Thunder’s 2002 robot, it wobbled or hopped each time they turned it.

I am speaking from experience wiht treads to

we used treads in 2001 they do have great traction and all but they also add alot of weight. We did not like are experience wiht treads we debated using them this eyar but that was quickely turned down. I think its mostly a team preference i think if we used them again we might like them but just the experience with them and wheels we like wheels better but expermient see hwo you like them

You might want to look into one of the robots at the JPL (NASA jet propulsion laboratory) that can climb stairs and whatnot. IT uses tank treds with two extra “arms” to climb over terrain…

Happy hunting

G GOLDMAN

Tank track are great.

We have had tracks for 10 years. So we know tracks and we know how far to push them.

  1. Tracks are great to climb, push, pull, and to control.

  2. Tracks are light if made right. they go a fast or as slow as you gear them.

if you look at theses pic’s you can see how much power tracks can give you

http://www.valleytech.k12.ma.us/robotics/images/2002Pictures/2002%20Robot%20Flipping%20Robot.jpg

http://www.valleytech.k12.ma.us/robotics/images/2003Pictures/UCFRegional/ClearRamp.JPG

Here is a close up of one of our track systems
http://www.valleytech.k12.ma.us/robotics/images/2002Pictures/2002%20Track.jpg

Bad thing about tracks are if you go past there point of no return they rip

http://www.valleytech.k12.ma.us/robotics/images/2002Pictures/MVC-671S.JPG

these were Breco Flex timing belts. they have made better belts for us so they don’t rip as easy. the cost is $270 per belt. the not so strong belts run around $140

If you want you can go back and look how our tracks have change in the past years

http://www.valleytech.k12.ma.us/robotics/robotics_photos2002.htm

We left out 1999 becuase that year was a good year of climbing. you have to figure that one out on your own.

If you need anything esle just e-mail me

Your assumption that treads will give you more torque is not true. You will have the exact same amount of torque as if you used wheels.

Cons: One glaring con nobody mentioned is that you have to tension them properly. If you dont, theyll fall off when you try to turn.

Perhaps the biggest con is their steep price tag. Belts are around $270 each as someone said, plus you need the pulleys, which will run you another couple hundred. All said, youre going to spend near 1000 dollars, just to get the robot running. Also, I believe many teams who used treads either tore them like crazy, or wore them down very fast. I seem to remember 226 saying they used 7+ belts over the course of the season. That adds up to quite a bit of money.

Ahh yes the age old tread vs. wheel debate. All of us (by all i mean some) from team 862 have tried to stay away from this but every year it comes back. I say wheels because its easyer, but its your choice.

You are right Cory, we spent a literal fortune on the tracks last year and it would make me think twice about doing it again (Can’t remember if it was seven or eight ripped belts). As Cory stated above the self aligning pulley’s were $100 each ($400 total) and the belts were close to $150 for serrated and almost $300 for the solid center. As for shipping them, Brecoflex ran pretty far behind in getting us replacements, and our use for them is way out of their design parameters as far as Brecoflex was concerned. Our gearboxes may have been too much for them last year. Also, don’t think that because you have treads you can climb anything, it all depends on the type of treads and their orientation compared to the chassis (Look at most tanks and the front has a steep angle instead of a big pulley)

I haven’t checked with the company that Andy Baker posted but they may be better price or delivery than Brecoflex and they sell self aligning tracks there also.

On a side note, we noticed no difference in the tearing between the regular Brecoflex belts with the serrated center rib and the solid core which cost twice as much. Our failures started on the edge of the tread and continued across the belt and appeared to have nothing to do with the center. Many teams also had de-lamination problems with the tracks.

Treads are definately cool, they suck power out of the drivetrain like no other on turning, and you can get great traction, but, don’t sell pneumatic tires short. I was VERY impressed with the performance of some of the six wheeled bots out there and their pushing force





Only when our arms were up. If you lowered them, it wouldnt. LOL, it was funny to watch though.

Thanks to everyone for answering my questions. I have a few more for you guys.

  1. How much roughly did it cost your team to build the whole tread system?

  2. Did you have to have the inner wheel machined? Did you have anything machined?

  3. How many of the belts did you guys rip?

  4. How would my team go about making a tank tread?

  5. I’m worried about turning with the treads, I think that it’ll be a lot harder with treads and that we’ll probably trigger the breakers.

  6. Is it at all possible to buy a premade system and then just install it on our robot?

Our team (1177) went through this analysis and decision making process today. We eventually created a decision matrix listing our options which ranged from 2 wheeled, 2 motor drive up to multi-wheel/tread and multi-motor drive. We ended up deciding to use wheels and multiple motors in a 4 wheel drive configuration for a combination of reasons.

The main reason is as stated in the replies thus far. Treads are great for traction, however that traction comes at a price which is increased power draw. Along with that power hungry traction system, your speed/current ratio is lower. We resisted the urge this year to use treads, and will utilize the 4 wheel drive.

When you guys ripped your tread, how tight were the belts?

We used an alligator clip in 2001 which in my opinion is significantly weaker then a solid belt and those did not break, Im really at a loss for why your belts would snap like that?

the easiest way to make turning easier on a tracked robot (or any robot for that matter) is to lower the center idler wheel down about 50 thousandths of an inch. This decreases your effective track length while turning, and still allows you to keep most of your traction.

a slightly more complicated, yet more effective solution is to have a mechanism that lifts one end of your robot off of the groung using skids or casters. when you get yourself into a situation when you need your traction again, you just plop the full length of your treads back onto the ground.

Here is something I found the other day when investigating drivetrain solutions. We still might use these, but maybe somebody else can as well.

It’s pre-made, relatively cheap, can easily made lighter, and all you have to do is bolt some motors in to get it to work. Maybe it will work for you. I called them and they have a lot on stock (plus some other neat things on the site):

http://www.robotcombat.com/store_tanktreads.html

Only $140 + shipping for the set. They are only 20" long, but if you put them on the front of a bot with casters or the like on the back for stability (or even some kind of suspension), it could be a good combination.

Good luck.

Be very careful with casters! Even with treads you will lose a measure of control with your casters (although the surface area from the treads will most likely override this). We had that problem in 2001 (our rookie year) in which we used a combination of wheelchair wheels and casters; with that combination I believe our robot drove in a straight line only once or twice during the season…

Very good point…but hey…I’m an electircal enginer and my solution is, “We’ll fix it with software!” :wink:

Honestly, it’s a good point. Casters were just a quick suggestion.

I love it! I’m actually an EE/programmer by “trade” but I’m also co-captain of our team so one of my jobs is to know everything in a general sense as well as to know all of our past failures!

castors work great if you close the steering control loop with a yaw rate sensor - then the SW can tell how much the driver wants to turn, and how fast the bot actually IS turning, and correct itself.

We did this last year - it works extreemly well!

See…fixing it with software!!! :stuck_out_tongue:

Good suggestion.