Major problems with tank tread drive

[Lord Nerdlinger]

We finally hooked up our drive system today, and we can only go about 2 feet before the treads fall off the wheel. The treads are flat on aluminum rollers and they have no grove. We have a pulley that's flanged and 2 other flanged rollers (one in the front and one in the back).

The flanges are about .1 inches, and the rollers/sprockets are 2 inches wide.

Are we totatlly screwed? Or is there anythign we can do?

Would attaching circular plates onto the flanged wheels to make a really big flange do anything?

Help plzz!!!

[dlavery]

Originally Posted by Lord Nerdlinger

We finally hooked up our drive system today, and we can only go about 2 feet before the treads fall off the wheel. The treads are flat on aluminum rollers and they have no grove. We have a pulley that's flanged and 2 other flanged rollers (one in the front and one in the back).

The flanges are about .1 inches, and the rollers/sprockets are 2 inches wide.

Are we totatlly screwed? Or is there anythign we can do?

Would attaching circular plates onto the flanged wheels to make a really big flange do anything?

Help plzz!!!

Ohh, do we feel your pain! Sounds exactly like our experience last year, which was ugly enough to make us swear off treads. The problem is that as you start to turn, the side forces are pushing the treads across the face of your pulleys, and into the flanges. The side force is strong enough that one of three things has to happen (we went through them all):
- the tread climbs up the flange, and off the pulley (if the tread is loose enough)
- the tread rips and tears on the edges of the flange (if it is too tight to climb up the flange)
- the tread crushes the shape of the flanges and makes them useless (if the treads are very tight and the flanges are weak or damaged)

None of these are good conditions, and all of them will cause the robot mobility system to sieze up and stall. At this point, it is a virtual certainty that you will not be able to get new treads with self-centering ridges on the back (which is what you really need to use), unless you are incredibly lucky and can find a team with a spare set. Short of that, here are a few suggestions:

- make sure the "line" of your treads is bowed down. In other words, look along the area where the treads make contact with the floor. Ideally, the ends of the treads should be slightly higher than the middle (like a flatter version of the rocker on a rocking chair). This puts most of the weight of the robot near the center of the contact patch, where the side forces from scrubbing the treads during a turn are the smallest. This allows the ends of the treads (potentially where the side forces are greatest) to "float" a little and hopefully reduce the side forces.

- increase the size of the flanges. This will help, up to a point (if your treads are too loose they will still climb up the taller flanges and thereby increase drag on your drive motors as the tread material scrubs against the larger surface area of the flange).

- make sure you have sufficient tension on your treads. Again, this will help, up to a point (if they are too tight, the increased loading on your pulley bearings will result in extra friction in the system, and potentially could fail the bearings).

- can you convert to wheels in the next four days? If the above suggestions don't help enough, you need to take a serious look at this option. I have seen teams completely re-build their drive system in two days, so know that it can be done.

Good luck!

-dave

[Lord Nerdlinger]

Originally Posted by dlavery

Ohh, do we feel your pain! Sounds exactly like our experience last year, which was ugly enough to make us swear off treads. The problem is that as you start to turn, the side forces are pushing the treads across the face of your pulleys, and into the flanges. The side force is strong enough that one of three things has to happen (we went through them all):
- the tread climbs up the flange, and off the pulley (if the tread is loose enough)
- the tread rips and tears on the edges of the flange (if it is too tight to climb up the flange)
- the tread crushes the shape of the flanges and makes them useless (if the treads are very tight and the flanges are weak or damaged)

None of these are good conditions, and all of them will cause the robot mobility system to sieze up and stall. At this point, it is a virtual certainty that you will not be able to get new treads with self-centering ridges on the back (which is what you really need to use), unless you are incredibly lucky and can find a team with a spare set. Short of that, here are a few suggestions:

- make sure the "line" of your treads is bowed down. In other words, look along the area where the treads make contact with the floor. Ideally, the ends of the treads should be slightly higher than the middle (like a flatter version of the rocker on a rocking chair). This puts most of the weight of the robot near the center of the contact patch, where the side forces from scrubbing the treads during a turn are the smallest. This allows the ends of the treads (potentially where the side forces are greatest) to "float" a little and hopefully reduce the side forces.

- increase the size of the flanges. This will help, up to a point (if your treads are too loose they will still climb up the taller flanges and thereby increase drag on your drive motors as the tread material scrubs against the larger surface area of the flange).

- make sure you have sufficient tension on your treads. Again, this will help, up to a point (if they are too tight, the increased loading on your pulley bearings will result in extra friction in the system, and potentially could fail the bearings).

- can you convert to wheels in the next four days? If the above suggestions don't help enough, you need to take a serious look at this option. I have seen teams completely re-build their drive system in two days, so know that it can be done.

Good luck!

-dave

unfortunately that's kind of what i expected

We might try to do wheels, although i don't know how do-able that is. It was designed to be pretty much exactly the right size and deviation in height will cause us not to reach the bar, deviation in size will make us too big and be disqualified, or too small and not fit subsystems... etc.

would it help if we could get the wheels and pulleys perfectly flat or in-line? Right now they are probably within a tenth of an inch, but we could probably do it within a 100th or even less error, if it made a difference.

Thanks for the help.

We might end up trying to adapt our system to wheels but it would a last resort.

[Chris Fultz]

1. Try to machine a groove in your idler wheels to keep the track in.

2. We made small guides to keep the treads aligned. They are pretty simple.
We used lexan as a plate, and the cantilevered 2" long bolts (like #10's) from them, with metal sleeves over the bolts to act as "rollers" for alignment.
We have one set placed at the front and back of the drives.

These roller guides are spaced just a little bit wider than our track so there is not much friction. The bolts and sleeves came from Home Depot so they should be easy to find.

If the description doesn't make sense - PM me with an email contact and I will try and send a power point picture.

[Kevin Sevcik]

This would be our first experience with treads, as well. Pretty much the exact smae problem you describe, really. We ended up trying to live with never turning during most of the season.

If you're really, really hurting for a quick fix that miiiiiight work, I'll tell you what I recall us doing with our treads. I don't remember the specifics, as I didn't see them constructed, but here goes:

We made our own bootleg self centering belts. I'm sure this was a very painstaking process. All I remember is that we eventuallyran on our old belts, except with a ridge on the back. this might not be an option for you, though. we were running on the belts inside out, ie. the teeth were running on the ground, and the smooth side was running on our idler pulleys.

I think some of the engineers took a spare belt and cut it into a thin strip. Then they laid it on inside of our track belt, smooth side to smooth side, and somehow attached the two together. I'm thinking something like the rivets that hold jeans together, between every tooth, or every other tooth. Maybe a pop rivet and a small washer on the other side? Then, of course, we machined matching grooves in the pulleys.

*shakes head* Now that I think about it, that sounds incredibly crazy and darn near impossible. Maybe we just did the sane thing and bought some new belts in the off-season. My memory is a little fuzzy.

PS. I'm starting to loathe last minute drivetrain testing. We just recently tested our bot on carpet, and the transmission is eating gears up. I'm thinking it's not dealing well with shock loads caused by breakers tripping and then cycling on and turning on the motors at full blast. So we're looking at custom gear sources with very fast turnaround, and some programming changes and driver training to make things nicer on the trannies.

[Lord Nerdlinger]

Originally Posted by Kevin Sevcik

*shakes head* Now that I think about it, that sounds incredibly crazy and darn near impossible. Maybe we just did the sane thing and bought some new belts in the off-season. My memory is a little fuzzy.

.

Would it be legal to buy a new tread and new pulleys and wheels after we ship and then replace our current ones with them at the competition?

If they were exactly the same size and didn't change the design of the robot at all?

[Greg]

Yes, this is legal. The rules say that you can bring "identical spares" to the competition.

[Stu Bloom]

Originally Posted by Greg

Yes, this is legal. The rules say that you can bring "identical spares" to the competition.

HOLD ON A MINUTE GREG ... this is one of those situations that you shouldn't be too quick to answer. I happen to disagree ... I suppose if they were ordered before ship date then the "timing" would be OK ... but if the parts they are referring to are truly "identical" then it seems it would not solve their problem. This is really a question for FIRST.

[Eric Reed]

Originally Posted by Stu Bloom

HOLD ON A MINUTE GREG ... this is one of those situations that you shouldn't be too quick to answer. I happen to disagree ... I suppose if they were ordered before ship date then the "timing" would be OK ... but if the parts they are referring to are truly "identical" then it seems it would not solve their problem. This is really a question for FIRST.

My interpretation (someone go ahead and shoot me down) has always been that you can bring any kind of raw material to the competition that you want, if you intend to build something there. You cannot do ANY kind of alteration to this material during the time between the build period and the event.

"raw material" in this case is anything that you would normally buy rather than make. Bolts, nuts, gears, treads, etc.

I mean, you are allowed to improve your robot on site...including adding brand new mechanisms...seems to me you'd need to bring something to make that stuff from.

If I'm way off, please let me know. We are neither rule-stretchers nor rule breakers, and would regret it if we have been misinterpreting this rule.

Eric.
De Anza Robotics

[SuperJake]

We used treads in 2001 and hated them.. but Robbie Extreme (56) used treads also and had a lot of luck with them. They said that the trick is to put pop-rivets all along the tread down the center. Then on all of your wheels, cut a groove that the pop rivets will slide in. This is basically how tanks do it.

[ShadowKnight]

FIRSt actually touches on this subject in the rulebook. It is completely legal to get off the shelf parts bought after the ship dae, however you cannot modify the parts in any way between ship date and the competition. Now, if you want, you can assemble any changes you want to make to the drive train at the competition with the raw materials/ off the shelf parts you purchased between ship date and the competition

[ChrisH]

In addition to the previous suggestions, you might try putting a slight "crown" on your rollers. Particularly the ones at the ends of the robot. It doesn't take much. If the center of the rollers is slightly larger in diameter than the sides, then the belt tends to self center. Most conveyour belts do this but the crown is so slight you can barely see it. 1-5% of the diameter would probably do it.

[Lord Nerdlinger]

Originally Posted by SuperJake

We used treads in 2001 and hated them.. but Robbie Extreme (56) used treads also and had a lot of luck with them. They said that the trick is to put pop-rivets all along the tread down the center. Then on all of your wheels, cut a groove that the pop rivets will slide in. This is basically how tanks do it.

could you explain in more depth how to do this?


How do you cut the wheels in the wheel, and what are pop rivets?

[Stu Bloom]

Originally Posted by Eric Reed

My interpretation (someone go ahead and shoot me down) has always been that you can bring any kind of raw material to the competition that you want, if you intend to build something there. You cannot do ANY kind of alteration to this material during the time between the build period and the event.

"raw material" in this case is anything that you would normally buy rather than make. Bolts, nuts, gears, treads, etc.

I mean, you are allowed to improve your robot on site...including adding brand new mechanisms...seems to me you'd need to bring something to make that stuff from.

If I'm way off, please let me know. We are neither rule-stretchers nor rule breakers, and would regret it if we have been misinterpreting this rule.

Eric.
De Anza Robotics

I apologize ... I was the one who answered too quickly ...
I was thinking of a team-manufactured mechanism/alternate configuration brought to the competition which was different than what was on the bot ... I stand corrected [hangs head in shame ].

You have some good suggestions for different fixes here - good luck getting the bot up and running.