do tracks and wheels together make a better robot?

[gondorf]

is it true that having both tracks and wheels together make you more likely to be picked for the better teams. i just heard that they might be better for traction.

which brings up another rumor: do tracks increase ball accuracy?

[KenWittlief]

Tank tracks are one of those things that keep showing up on FIRST robots, in ways that I dont understand.

Friction requires a downward force. If you look at an army tank, or a bulldozer, there are idler wheels all down the side that push the track into the ground.

but I have never seen a FIRST robot with those idler wheels. If you dont have them, then what is pushing the tread into the carpet between the wheels?

ans: nothing!

I might be wrong on this, but a robot that has two wheels on each side with a tank tread, and a robot that has two wheels on each side that use the tank tread wrapped around the wheels instead (as tire material) should have exactly the same pushing force

and way less complexity

[Ken dons his fire retardant suit and hides under desk]

[sburro]

Tank tracks are one of those things that keep showing up on FIRST robots, in ways that I dont understand.

Friction requires a downward force. If you look at an army tank, or a bulldozer, there are idler wheels all down the side that push the track into the ground.

but I have never seen a FIRST robot with those idler wheels. If you dont have them, then what is pushing the tread into the carpet between the wheels?

ans: nothing!

I might be wrong on this, but a robot that has two wheels on each side with a tank tread, and a robot that has two wheels on each side that use the tank tread wrapped around the wheels instead (as tire material) should have exactly the same pushing force

and way less complexity

[Ken dons his fire retardant suit and hides under desk]
lokk at 663 robot last year
http://www.wcsrobotics.com/multi.htm

The idlers on the bottom could have been spring loaded but since it was a flat field there was no reason to.

[KenWittlief]

wow! hats off to you!

I think that is the 1st time Ive tank treads done 'right' on a FIRST robot

is your team sponsored by General Dynamics by any chance?

or Catapiller? :^)

regarding launcher accuracy - if a robot is able to turn slowly and precisely, then it is easier to target the goal

since tank treads have a lot of side friciton, they are usually geared down more than a robot with two wheels and castors (for example) - so a typical bot with treads will turn slower

and be easier to aim.

[henryBsick]

Tank tracks are one of those things that keep showing up on FIRST robots, in ways that I dont understand.

Friction requires a downward force. If you look at an army tank, or a bulldozer, there are idler wheels all down the side that push the track into the ground.

but I have never seen a FIRST robot with those idler wheels. If you dont have them, then what is pushing the tread into the carpet between the wheels?

ans: nothing!

I might be wrong on this, but a robot that has two wheels on each side with a tank tread, and a robot that has two wheels on each side that use the tank tread wrapped around the wheels instead (as tire material) should have exactly the same pushing force

and way less complexity

[Ken dons his fire retardant suit and hides under desk]

Ken,
I would like to thank you for finally bringing this into the light. I am pretty frustrated with teams that design drive-trains based on poor physics beliefs. From treads to unnecessary three motor drive trains...
Anyway, knowing vectors of force and co-efficients of friction are HUGE keys to building a competitive robot.

-Henry

[George1902]

but I have never seen a FIRST robot with those idler wheels.
You should swing by the Florida Regional this year, Ken. I know of one team there who does treads quite well.

http://www.spamrobotics.com/gallery2/build05/robot05
http://www.spamrobotics.com/gallery2/build03/2003_on_field
http://www.spamrobotics.com/gallery2/2002/pk02pits
http://www.spamrobotics.com/gallery2/2001/3_bits

[RogerR]

Ken,
I would like to thank you for finally bringing this into the light. I am pretty frustrated with teams that design drive-trains based on poor physics beliefs. From treads to unnecessary three motor drive trains...
Anyway, knowing vectors of force and co-efficients of friction are HUGE keys to building a competitive robot.

-Henry

i'm curious about what you mean by 'poor physics beliefs'. both of those examples have valid reasoning behind them. and i'm pretty confident that their designers have at least a rudimentary grasp of Co.F. and vectors.

[KenWittlief]

we were talking about tank treads that do not have idler wheels. The photos posted in this thread are the 1st bots Ive seen that do have idler wheels (the right way to use threads)

un-necessary 3 motor designs are designs that dont need 3 motors on each side -it depends on what else the drive train can do. If a bot can spin its wheels against a wall, then putting more motors on it does nothing but spin the wheels faster

and spinning the wheels faster does not give you any additional pushing force.

[Ken hides back under his desk :^) ]

[Sean Schuff]

To accomplish a downward force on the track between the rear/front drive/idler does not require idler wheels. A rail (similar to older snowmobile designs prior to popular use of bogey wheels) will accomplish the same thing. It has to be designed right so as to not cause unnecessarily high amounts of friction. Without running numbers my hunch is that the bogey wheels are more efficient in the area of reducing friction.

With regards to the logic of using treads versus wheels - I think this year's game is a perfect example of the usefullness of tracks. Why? If your team wants to climb the ramp to score the points at the end of the match you need either ground clearance to the tune of roughly 4.5 inches on a 38" long drive with 6 inch wheels. This approach causes two problems that our team has identified - less space in the 28x38x60 box for robot "stuff" and a higher center of gravity. Tracks eliminate the need for ground clearance by providing contact all along the robot.

With that being said, our team is still using wheels this year but in a slightly different configuration than years past. Our team has not put much effort into tracks and if I keep mentioning them, Mike Johnson will kick me again! :)

My 2 cents - take them for what they're worth.

Sean

[Jonathan Norris]

With regards to the logic of using treads versus wheels - I think this year's game is a perfect example of the usefullness of tracks. Why? If your team wants to climb the ramp to score the points at the end of the match you need either ground clearance to the tune of roughly 4.5 inches on a 38" long drive with 6 inch wheels. This approach causes two problems that our team has identified - less space in the 28x38x60 box for robot "stuff" and a higher center of gravity. Tracks eliminate the need for ground clearance by providing contact all along the robot.

Sean

My team did a test today with a extra kit frame with 6" 6 wheel drive, and only needed 2.5" of clearance. This is with 6 wheels not 4.

[Sean Schuff]

My team did a test today with a extra kit frame with 6" 6 wheel drive, and only needed 2.5" of clearance. This is with 6 wheels not 4.

We ran the same caculations and came up with the same number. We are also looking at a 6 wheel configuration with the front and rear wheels driven and the middle wheels serving simply as bogeys. It definitely gives you more options with internal configurations.

[sanddrag]

If you do it correctly, you can get away with as little as 2mm clearance for 6wd. Of course that doesn't allow for wheel wear or running over dropped bolts or anything.

[Sean Schuff]

If you do it correctly, you can get away with as little as 2mm clearance for 6wd. Of course that doesn't allow for wheel wear or running over dropped bolts or anything.

Yes but can you climb the ramp AND get onto the platform that way? My guess is "no". You can get up onto the ramp with minimal clearance simply by placing the contact point between your front/rear wheels and the floor beyond the exterior of your frame. This will get you on the ramp but you will high side your robot trying to get to the top of the platform. There are two options to solve this dilemma: a) tracks or wheels all along the lower drive section of your robot or b) an area between the wheels that allows for the transition from the ramp to the platform.

[Tom Bottiglieri]

T

but I have never seen a FIRST robot with those idler wheels. If you dont have them, then what is pushing the tread into the carpet between the wheels?


You sure?
http://team195.com/gallery/albums/2005/Robot/IMG_0359%20%28Large%29.JPG
All of the benefits of tank treads. All of the benefits of 6WD. Very little hassle.

[Nuttyman54]

Team 971 took a slighty different approach in 2004. We did a halftrack (Tracks in front, wheels in back) and use the Robot Combat snowblower treads. however, we also inclined the treads quite a bit, giving the robot effectively a 6 wheel configuration. This allowed us to drive up the SIDE of the platform (We skipped the 6" steps entirely)...it also helped us get top seed and win the Silicon Valley Regional that year (thanks to 254 and 852). You could use the same concept to drive up the ramp.

[Mike Norton]

http://myweb.wit.edu/cuffiim/robotics.wmv

http://myweb.wit.edu/cuffiim/robotics2.wmv

http://www.chiefdelphi.com/forums/showthread.php?t=41300&page=2&pp=15
this came up in this thread.

To say you have not seen tracks done right you have not been looking.

you will see a 6 wheel robot trying to push us. they had a six motor drivetrain

So track done right will be fast, turn quick and be very powerful

[KenWittlief]

Yes but can you climb the ramp AND get onto the platform that way?

...There are two options to solve this dilemma: a) tracks or wheels all along the lower drive section of your robot or b) an area between the wheels that allows for the transition from the ramp to the platform.

and a 3rd option: inertia !

[sanddrag]

If you do it correctly, you can get away with as little as 2mm clearance for 6wd. Of course that doesn't allow for wheel wear or running over dropped bolts or anything.
Yes but can you climb the ramp AND get onto the platform that way? My guess is "no".
Guess again. You most definitely can. The trick is in carefully analyzing how the robot gets up the ramp and onto the platform. It is all in the weight distribution. If you get it right, your clearance can be next to nothing. The end that goes up the ramp first needs to be a bit heavier. We'll call the end that goes first the "front." See, when you go up the ramp, when the front wheels pass that edge between incline and flat surfaces, the idea is for them to keep going along the imaginary incline in the air and not touch down on the platform. The, once the center wheels have reached the edge between incline and platform, then you want the weight in the front to pull the front wheels wheels down (and the back wheels up off the ramp). You'll never get hung up between the front and center, or center and rear wheels, if you get the movement right. With proper weight distribution, it is very achievable.

[Cory]

Guess again. You most definitely can. The trick is in carefully analyzing how the robot gets up the ramp and onto the platform. It is all in the weight distribution. If you get it right, your clearance can be next to nothing. The end that goes up the ramp first needs to be a bit heavier. We'll call the end that goes first the "front." See, when you go up the ramp, when the front wheels pass that edge between incline and flat surfaces, the idea is for them to keep going along the imaginary incline in the air and not touch down on the platform. The, once the center wheels have reached the edge between incline and platform, then you want the weight in the front to pull the front wheels wheels down (and the back wheels up off the ramp). You'll never get hung up between the front and center, or center and rear wheels, if you get the movement right. With proper weight distribution, it is very achievable.

The only problem with this strategy being that unless your wheels extend beyond the end rail of your frame, with 2mm of clearance you'll grind the edge of your frame into the ramp before the wheels contact the ramp surface :)

[Sean Schuff]

It is all in the weight distribution. If you get it right, your clearance can be next to nothing.


No matter how you configure it or how you distribute the weight, you will, at some point, high side your robot without the proper ground clearance. The only thing that will keep you from STAYING high sided, as Ken pointed out, is inertia. The only thing that will accomplish what you have described is a shift-on-demand center of gravity (CG).

A six wheel design (two front, two middle, two rear) still has points in between the wheels where the frame can, and if not designed properly, will contact the transition edge from ramp to platform. 2mm of clearance will not allow for the proper clearance and your CG will not shift at the right point.

With regards to your CG you've got three realistic options: in between the front and middle wheels, directly over the middle wheels, and in between the middle and rear wheels.

If your CG is in between the front and middle wheels your robot will high side itself in between the front and middle wheels as it makes the transition from ramp to platform.

If your CG is directly above the middle wheels, as your robot moves onto the ramp it will shift back to between the middle and rear wheels until the robot is on a level surface again. As the robot transitions from the floor level playing surface to the ramp, the CG will shift to the rear of the robot. Once this happens, the CG will stay between the middle and rear wheels until it is horizontal again. Since the rear wheels will have to make the transition from ramp to platform to accomplish this, the high siding will occur between the middle and rear wheels.

If your CG is in between the middle and rear wheels, your robot will high side itself in between the middle and rear wheels as it makes the transitions from ramp to platform.

Obviously there are other variables in this equation. Ken brought up inertia as being one way to overcome the need for adequate clearance but then you need enough inertia to overcome the friction between your robot frame and the ramp/platform ridge. Another variable is going up the ramp at an angle versus perpendicular to the incline. This may allow you to lower your clearance if designed properly.

For a simple topic about tracks versus wheels this has become a very detailed discussion. I love it!

Sean

[KenWittlief]

I didnt mean inertia as in scraping your way over the edge with sparks flying everywhere

I meant as in "an object in motion tends to stay in motion (in the same direction)"

in the 2002? game, with the containers and the center ramp, teams needed to be the first ontop of the ramp in auton mode. As a result many bots were airborne as they came off the upward ramp.

THAT kind of inertia ! :^)

[Nuttyman54]

in the 2002? game, with the containers and the center ramp, teams needed to be the first ontop of the ramp in auton mode. As a result many bots were airborne as they came off the upward ramp.

that was 2003 (Stack Attack). I know of at least one robot (Crabgoat, Team 190) that had their auto mode programmed to right their robot, as it had a tendency to flip over opposing robots when going over the top. Fun to watch!

[Sean Schuff]

in the 2002? game, with the containers and the center ramp, teams needed to be the first ontop of the ramp in auton mode. As a result many bots were airborne as they came off the upward ramp.

THAT kind of inertia ! :^)

Imagine the surprise on the faces of the drivers, operators and coaches as your robot went airborne off the ramp! :) Sweet butta!

[KenWittlief]

Imagine the surprise on the faces of the drivers, operators and coaches as your robot went airborne off the ramp! :) Sweet butta!

no surprize - if your bot is going uphill then it MUST get airborne when the angle of the ramp changes suddenly.

The question we all had to calculate was 'how airborne'. If you went fast enough it was possible to fly over the top of the ramp completely, and land on the other side.

Most bots only got a couple inches of air.

[gondorf]

so just drawing a conclusion from these answers yes treads do help if you have a shooter, it does increase ball accuracy, and it helps to get the robot up the stupid ramp that is impossible from my teams perspective.

*** writes note to tell team to use treads***

[Sean Schuff]

so just drawing a conclusion from these answers yes treads do help if you have a shooter, it does increase ball accuracy, and it helps to get the robot up the stupid ramp that is impossible from my teams perspective.

Don't limit yourself to just treads. They are an elegant solution to the problem if you can implement them well. I have seen quite a few teams who have done a bang-up job with treads. Just look through some of the pictures pages for past competitions and you will see what I mean. But that doesn't mean that it is the only solution to the problem. Remember, one problem, 1,000+ teams and 1,000+ different solutions to the same problem.

You can accomplish the same accuracy with wheels it just requires the appropriate implementation of them.

Good luck!

[greencactus3]

we were talking about tank treads that do not have idler wheels. The photos posted in this thread are the 1st bots Ive seen that do have idler wheels (the right way to use threads)

un-necessary 3 motor designs are designs that dont need 3 motors on each side -it depends on what else the drive train can do. If a bot can spin its wheels against a wall, then putting more motors on it does nothing but spin the wheels faster

and spinning the wheels faster does not give you any additional pushing force.

[Ken hides back under his desk :^) ]
but it lets you gear higher and = faster top speed. depending on how well you can control it.. but same with cars. who needs a car that goes 200mph... but people still buy because it has a high top speed. the potential. and just because they can... which we cant.. boo

[Jay Trzaskos]

http://firstrobotics.uwaterloo.ca/drivetrainsgallery.php

Take a look around on this site to see how tank treads have been implemented over the years. Actually, this (http://firstrobotics.uwaterloo.ca/resources.php) is a great resources for any part of your robot.

JT
229

[gondorf]

tank treads may be a good thing for a robot but they are a pain to make.

from how i tried to do it tank treads are sort of like 4 wheel drive with a chain connecting it.

my team is having trouble with just 4 wheel drive, let alone a tread

[Mike Norton]

http://www.valleytech.k12.ma.us/robotics/images/2004Pictures/2004VIEW5.jpg

http://www.valleytech.k12.ma.us/robotics/images/2004Pictures/2004track3.JPG

This is hard ? It is pretty easy to do. with BrecoFlex pulleys and timing Belt

[gondorf]

http://www.valleytech.k12.ma.us/robotics/images/2004Pictures/2004VIEW5.jpg

http://www.valleytech.k12.ma.us/robotics/images/2004Pictures/2004track3.JPG

This is hard ? It is pretty easy to do. with BrecoFlex pulleys and timing Belt

you know if my team knew what brecoflex pulleys and a timing belt were we would probably pat ourselves on the back and act all proud.

i could probably figure out a tread if it wasnt for the fact that it has to be a team decision and othere dont agree on treads

[Jay Trzaskos]

Hey guys, let's stick to the discussion at hand.

If you would like to learn more about brecoflex timing belts take a look here (http://www.brecoflex.com/) . If you would like to learn more about how treads have been used in the past try searching the forums (http://www.chiefdelphi.com/forums/search.php?) for old threads discussing the issue. I would also advise everyone to take a look at this (http://firstrobotics.uwaterloo.ca/resources.php) website. It's a really great resource to use when dicussing designs, as I stated earlier.

JT
229