Winning Drive Train

[JesseK]

We had a 9.8fps linkage drive where each wheel rotation point made 1 corner of a 18x18" square.

In hindsight, we should have made it wide-drive (rotation point rectangle of 29" x 18"). Not only would we still have had dynamic, on-demand centers of rotation for various wheel orientations, with a wider base we would have also had slightly more torque in turning the trailer at low speeds. It also would have simplified the frame design, heh.

[s_forbes]

One thing I noticed: 11/12 robot on Einstein were oriented with a 'wide' drivetrain. I don't know enough about all of these robots to tell how many had traction control, swerve, 4wd/6wd, fans, etc., but that would be a nice place to start if you want to determine the 'winning' drivetrain. If you're feeling ambitious, an analysis of the drivetrains on every robot in eliminations would be pretty cool, too.

I still think that good driving and strategy are way more important than drivetrain details, but that should be obvious.

[dipmeinaluminum]

Originally Posted by s_forbes

If you're feeling ambitious, an analysis of the drivetrains on every robot in eliminations would be pretty cool, too.

Trust me, if i had the time i would. this game left me with my head spining.

Originally Posted by techtiger1

Saying this we were just fine with traction for 2009 in fact it was one of our strong suits. We made the robot weight as close as possible to the limit 119.9 and had no problems basically moved every robot we needed too at the Fl regional except 233 who had an amazing traction control program. Saying all this we could still be pushed sideways if another team had momentum no matter what happened. The fans seemed to work well for some teams at gaining speed but not really for gaining a whole lot of traction. Other then that there was no real way to gain an advantage on traction this year.

well both 354 and 2681 had the KOP drive train thinking that over designing the drive train would kill us on time and weight. but if all that does not matter then, was having a traction control program the key to pushing people around and speed? or did teams see that having a wide base with 12+ wheels the way to go? i look at teams like 2753 and they fly with no problems. being the coach behind the glass, and involoved in first for 9 years, i just looked slow. im trying to see where i went wrong. both weighed 119.8 and weight was even through out the robot.

[Tom Line]

Originally Posted by Adam Freeman

Not sure where you got your information on the quote above, but I can assure you we had some version of traction control at all events we attended this season. I can only remember one occasion when we ran a match with it turned off, and that was b/c we attempted to try a new version that was alittle too slow on response.

With regards to drivetrain, we had a wide - 6 wheel drive system with the rear wheels raise about 1/16" to allow for some "rocking" between the front and rear wheels. Not sure what effect it had on the bot, I think it just made my counterpart feel better. There was no quantitative measurements taken to confirm or deny its effectiveness.

In terms of pushing power, once we finally worked out our traction control issues, we could wiggle our way out of most issues. But, we were not a pushing robot by any stretch of the imagination.

My thoughts were why push when you could be scoring??? I guess it worked out in the end

My apologies Adam. I had spoken with one of the volunteers from Lansing, and they told me that you weren't running traction control. Mea culpa.

[TJ Cawley]

some of our mentors thought that having more contact points with the field (i.e. with the rover wheels) would provide less traction. so we went with a 4wd direct drive train holonomic drive train. we had an average speed and push power, and we had great manuverability. in our two regionals, we proved that our robot had power when we broke several strong supports with our drive train. in Atlant we changed our drive train to Direct Drive take steer with 1 joystick. the manuverability was decreased but the speed and power delivered was enhanced.

[kramarczyk]

I thought this was a good year to really nail that whole Ff=mu*Fn into people's head, but to get it past my kids we had to do some quicky tests to prove it. See attached data.

We took the kitbot stuff plus some junk around the shop and created a simple test bed to determine if more wheels would generate more force. We were less interested in pushing people around than we were in making the most of our acceleration (F=m*a). We used the kitbot frame and successively mounted wheels on one side, first 2, then 3, finally 4. We would have done more but that's all the wheels we had. We then placed this on a piece of the FRP that was spec'd and massed up the frame with stuff from around the shop. Each of these items was weighed individually (introduces error). We tried to keep the mass as close to the wheels as possible, but still needed to have a student support the outrigger side by a piece of rope with the goal being to keep the frame level (introduces more error). A fishing scale on the rope allowed us to back calculate out the load on the wheels by treating the assembly as a simply supported beam. (i.e. total mass - mass on rope = mass on wheels)

Once we know how much weight was on the wheels we moved the scale to the rear of the frame, in-line with the wheels. We applied power to the wheels and measured the force generated on the scale. Change # of wheels and repeat. Based upon the data gathered and the acknowledged crudeness of the testing we all quickly concluded that more wheels did not significantly add more force.

Based upon the test data we gathered I would say that anyway you could generate and additional lbf of thrust would be important. I was told that 45 & 469 could get 5lbf and ~4lbf respectively... That's 50-60% more thrust than me.

"Foul, foul, I saw you use 6 wheels on your wide chassis at comps!" Read Chris Hibner's white paper, "Drive Train Basics (How to Be Sure Your Robot Will Turn)" It's not just about how much force, but where you put it sometimes that matters.

[martin417]

Originally Posted by JVN

Anybody have anything quantitative to share? Anybody? Anybody? Bueller?

Here is the data we took on our bot. We were using follower wheel traction control limiting the slip ratio to ~.2. The drawbar pull was measured using a fish scale (hand held) with both joysticks firewalled, while allowing the bot to slowly travel forward over the regolith. With just the wheels, we got around 18-20 lb.f on average. If we turned on the propeller propusion, we got 26-28 lb.f.

I won't say we got 8-10 lb. thrust from the props, I am guessing there were other factors involved. The voltage drop due to the propellers running may have made the traction control work better, or the airflow may have blown away the regolith dust as it was produced, increasing traction. I don't know. I am just reporting the raw numbers.

[thefro526]

This year 816 used a Wide Based 6 Wheel Drive. We went with a wide base because it allowed us to have more of a "turret" effect with our robot while at rest and while moving. We found that a wide base was significantly more maneuverable at low speeds than a long base, and that the usual stability issues weren't a problem because of the trailer.

Our Drive train was geared for about 12fps which was more than fast enough for us throughout the season. It gave us enough speed to chase down almost any target and also helped us get out of quite a few jams. As far as pushing goes, we never had an issue pushing another robot when we really needed to. We didn't use traction control either to save valuable programming resources.

I don't have any test data on the drive, but if you're curious look at a video of us. Numerous people complemented us on our maneuverability throughout the season.

[mikelowry]

Originally Posted by JVN

Anybody have anything quantitative to share? Anybody? Anybody? Bueller?

We used fish scales to measure our pushing force. We had about 15 or so pounds with 6 wheel drop center tank drive. We added traction control that bumped us up to about 20 pounds and then we had 2 10" shrouded fans that each added another 4 pounds for a total of anywhere from 26-30 pounds of force at one time.

[ginosoprano09]

Did either 111 or 1717 have some sort of traction control?

[Jon Jack]

1717 did have traction control.

[Ice Berg]

In my opinion, the best drivetrain I saw this season was on team 743. I'm not sure on the specifics, but I know they had 14 wheels. Their defense knocked us out in semis at NY, and they made finals at both NY and CT. Watch their matches, they play great defense. Maybe someone from 743 could explain what set their drivetrain apart?

[big1boom]

I am fairly certain that 111 had traction control. From what I saw, they have a sensor wheel for each of the swerve pods.

This year, I believe that our robot was one of the best pushing robots on the fields from what I saw of our matches. We didn't have traction control, and our robot is light. (115ish) I think that the reason we had more traction, was because this year we went with a normally extremely underpowered drive system. We did the calculations, and found that 1 CIM could provide more than enough power to the wheels to break into dynamic friction (without gear efficiency factored in). So we used 1 CIM, total, this lead us to have just enough power to accelerate at nearly the maximum rate, but also allowed us to never spin our wheels, meaning we had full traction, all the time.

[Stealthguardian]

Piper Robotics Team 1802, we had 16 wheels on a floating suspension and we could push 3 robots at a time into a corner and clear the field.

even though we didnt make it to the finals at our regional we had pretty solid pushing power

[R.C.]

Originally Posted by Stealthguardian

Piper Robotics Team 1802, we had 16 wheels on a floating suspension and we could push 3 robots at a time into a corner and clear the field.

even though we didnt make it to the finals at our regional we had pretty solid pushing power

Any pics of the Drivetrain up close, this sounds very interesting. We could push 2 if we had momentum.