Belts are out Wheels are in?

[Kevin Ray]

"It seems to me that many teams who have successfully run in the past with belts have this year opted for wheels. Can any of you that fall into this category give me any ideas why?

I'm curious because we went with Brecoflex belts this year and are extremely happy. We are able to spin on a dime (pirouettes) and have excellent traction (a bit obvious). Probably the biggest advantage though is the fact that in Auto-mode we seem to be very accurate as to where we want to go. We seem to have very little slop and sway as we may have had with wheels. Slight bumps such as the bottom and top of the ramp have negligible effect on our dead reckoning.

Also, I would have thought that on the ramp and the HDPE teams would have wanted the extra surface area of a belt.

Answers, ideas from any team are welcome"

[The Lucas]

We went with wheels b/c our belt test showed it was ineffective at turning on the ramp due to the wire grating.

[Adam Y.]

I have a question how well does autonomous mode when the robot turns. Tank treads have a horrible track record when they turn. They have even been known to throw tracks because of there characteristic to only want to go straight.

[Bryan Faber]

Originally posted by Kevin Ray
I'm curious because we went with Brecoflex belts this year and are extremely happy. We are able to spin on a dime (pirouettes) and have excellent traction (a bit obvious). Probably the biggest advantage though is the fact that in Auto-mode we seem to be very accurate as to where we want to go. We seem to have very little slop and sway as we may have had with wheels. Slight bumps such as the bottom and top of the ramp have negligible effect on our dead reckoning.

Also, I would have thought that on the ramp and the HDPE teams would have wanted the extra surface area of a belt.

Wow, that's exactly what we thought. We're using brecoflex belts too.

[Jnadke]

Originally posted by Kevin Ray
It seems to me that many teams who have successfully run in the past with belts have this year opted for wheels. Can any of you that fall into this category give me any ideas why?

The problem with belt is that it can have a very high coefficient of friction. Most teams are going for high-speed, low torque robots this year. They wouldn't have enough torque to turn the robot without tripping breakers. Either that or you'd see a lot of drill motors frying. Not to mention most teams aren't very experienced in using treads. Without all the proper safety mechanisms in place, they are prone to fall off or rip.

Belt isn't always a good thing. The amount of traction your robot will have depends on three things, the traction material, surface area, and robot weight.

The type of traction material comes into play because if you look at the carpet, it's made up of many, many strands called fibers. These fibers need somewhere to go. If you have deep grooves in your wheels/treads, the fibers will go there. The edge of the tire treads are what actually give much of the traction.

Surface area does come into play becaus you get traction when the carpet fibers interlock with the treads on your belt/wheels. The more surface area, the more weight for every square inch of traction material touching the ground. If you put enough weight, you'll squish the carpet fibers entirely, and there won't be anything for the wheels/treads to grab onto.

In normal physics, surface area is not a factor when calculating friction. However, when surfaces have the ability to deform, surface area (more importantly, pressure) does come into play. Moving something accross a hard floor versus the carpet is different.

[Kevin Ray]

[Jdanke
In normal physics, surface area is not a factor when calculating friction. However, when surfaces have the ability to deform, surface area (more importantly, pressure) does come into play. Moving something accross a hard floor versus the carpet is different.]

Thank you Jdanke, we did understand all of the physics involved and that was exactly why we did, in fact, go with the belts.
http://www.fleef.net/robo/robo2rm.rm <-- you need RealOne player to view

I guess though, that I didn't ask the question clearly enough though; "What was it about the belts that made you shy away from them when you had success in previous years and go to the wheels? Was it the wire on the ramp? Has anyone had problems with their belts on the ramp this year?

[Kevin Ray]

[b][Wysiswyg --I have a question how well does autonomous mode when the robot turns. Tank treads have a horrible track record when they turn. They have even been known to throw tracks because of there characteristic to only want to go straight.][b]

We resolved that problem the way that most teams have seemed to do it. We lowered the center idler about 1/8" on the bottom of the track to enable us to turn on that center point. It effectively cuts the surface area of the belt as it contacts the carpet, in half. That serves to cut down on the lateral drag of the tank steering but it also cuts down on the traction (accepting the fact that the carpet "meshes" with the belt, thus creating more grip and more friction coefficient).

We were lucky actually because the robot is almost perfectly balanced on that point--serendipitously of course!

[Jnadke]

I've seen teams last year with pneumatic tires with deep grooves have the same, if not more, traction than teams with treads. One team that comes to mind is team 60. They had simple wheels. Yet they were considered one of the most powerful bots last year.

303, one of our partners for the Einstein finals matches, had a considerable amount of traction with simple pneumatic tires. They had one's with very deep treads. I'm not sure what kind of drivetrain they had, I'd imagine it was 4 motor because they had a lot of pushing power.

[Rook]

The game changes each year. Our robot changes with the game. Belts are more complex than the 6 wheel drive we went with this year. We wanted more maneuverability, so we went with wheels.

[sevisehda]

When it comes to traction it's 90% experimentation. Its easy to calculate friction by material and normal force(NOT pressure). Surface area can help or hurt. For deformable surfaces you want slightly more but it you have to much surface area then the pressure on every square inch is very little so you don't deform the materials. To little material and you won't maximize the deformation. So you shold experiment with the material you have to see what amount of surface area gives you the best result.

If your experiment shows that more surface area would help then build a tread system but if it shows less is better than use wheels. Width of your treads and wheels further helps you optimize this.

A general rule is wheels are faster and more manueverable and treads have more pushing power. A small minority of teams though can be fast, manueverable, and have enourmous pushing power ::cough technokats::.

[Mike Bobak]

We used belts on our robot and in the beginning we realized we had too much contact area and we would throw the belts off. So, we moved our idler around and made the once tank thread idea into a more 4-wheel drive idea. We never really had problems with throwing belts but turning on that wire mesh seemed to be a problem for us. We went with too small of a belt and it wasn't easily able to grip the wire mesh enough to turn and not slip. Next year if the game involves a ramp or wire mesh that a robot must climb up I know my team will go with pneumatic tires. But, only time and FIRST will tell what our final decision will be in the end.

Also did anyone else think about using these? http://www.goodyearindustrialproduct...le_pd_belt.pdf

We would of if we had the money. lol

[Matt Reiland]

We went with the Breakoflex belts and they work pretty good but....

1.)Pretty Expensive (200-300 ea plus over 100 for each pulley)
2.)If you get the ones with the center rib they will stay on great
3.)Tore 2 in half at Buckeye, we have new stronger ones for Great Lakes
4.)Ramp is a cheesegrator for the material little bits left under the ramp after each match.

I like them, I would use them again for sure, we have the power to spin them in high or low so power is not an issue. The lower bogey wheel helps (We have 2) to turn but it did well without it. As for traction they are great when new but they fade slowly as the material wears off, congrats to many teams who found some truely wonderful wheels this year for traction I am impressed.

[rees2001]

Kevin,
Nice video, Is that you being dragged behind your robot? This year we went with wheels again & we didn't have any problems with traction. You probably remember the huge wheels we used last year (nats). This year we just went with pneumatic 9" skyway wheels & lowered the pressure. We had more than enough traction. we can push without trouble & on top when we turn sideways we don't get pushed. We are only using 4 wheels in 4-wheel drive. Mostly, we used a tried & true chassis. We don't have $$ to experiment with different things this year.
BTW, Good luck, Its great to see that you guys are still going.

[Andrew]

We went with treads in the pre-Brecoflex era and ran into nothing but difficulties. There are cases where treads have an advantage due to continuous contact with the ground.

However, you can put on more wheels and chain drive them to achieve similar behavior.

The issues with tread slip and wear are prohibitive if you can achieve it through other means.

You can also put on as many wheels as you like, if you want to get comparable traction to treads.

In our case, we have six wheel positions and we went with duallys on four out of six. This gives us ten tires on the ground.

Due to the indeterminacy of more than three wheels supporting a vehicle, we added a suspension on our front tires to guarantee multiple wheel contact under all circumstances.

[Bduggan04]

The number one reason we went with 6 wheel drive over belts had to do with the top of the ramp. We were worried that the track drive would be no advantage at the base of the ramp because you're reduced to 2 points of contact per side and at the top it's only one per side. The top also posses the problem of teetering on the top edge, as des a 6 wheel drive robot. We remedied this with a bogey style center wheel that we can control when it is down (for a biased middle wheel) or free to move (terrain follow).