8WD drivetrain?

[David Sherman]

I was wondering how to make an 8wd, having six wheels touching the floor at all times?

Do you just off set the two center wheels slightly and depending on the center of mass the robot is on the front one or the back one?

Also with an 8WD is it still easy to turn?

[petek]

Something like this?

RoboRaiders have been doing this quite well for a few years.

[Adam Shapiro]

One thing I would be concerned with in designing such a drive system would be the additional friction you'd get with the extra wheels. It would most likely make it considerably harder to turn such a system. It would certainly be a major consideration in how you'd design the new gearing system to power such a bot. Don't want to finish the drive train and find out that even light turns are blowing your breakers, and even your main breaker (like our 2002 tank)!

[Bill_Hancoc]

WOW that is really cool bot design with the 8 wheels. Do you have anymore photos of it other than what is on ur team site. I really like that.

[Alex Cormier]

We built a 8wd robot this year. We had the 4 middle wheels lowered. If i can find the post with all the details i will post it here.

Info

[Dan Petrovic]

Originally Posted by Adam Shapiro

One thing I would be concerned with in designing such a drive system would be the additional friction you'd get with the extra wheels. It would most likely make it considerably harder to turn such a system. It would certainly be a major consideration in how you'd design the new gearing system to power such a bot. Don't want to finish the drive train and find out that even light turns are blowing your breakers, and even your main breaker (like our 2002 tank)!

Yeah, that's one thing I'd think there would be a problem with 8WD. Maybe if the front and rear wheels rotated so when you wanted to turn, they would just rotate so there isn't as much rocking back and forth like we see on so many 6WD robots. Except that would be impossible, or impractical, considering the limits on CIMs.

Maybe front and rear wheels are omni-wheels?

[sanddrag]

Friction has little do do with surface area in theory. Perhaps in the real life application of FRC is different? Do more wheels equal more traction in FRC or do they just equal more weight? Will a robot with its left and right sides lined with wheels be harder to turn than the same robot with the wheels only at the ends? I'm actually not sure on this one. Simple friction theory says there will be no difference, but there is the tread/surface interaction factor to consider in real life.

[YonZ]

I believe I am right in saying that 2train, 395 has 8wd

[jgannon]

Originally Posted by YonZ

I believe I am right in saying that 2train, 395 has 8wd

You are indeed correct.
http://www.chiefdelphi.com/media/photos/24072

[Adam Shapiro]

Originally Posted by sanddrag

Friction has little do do with surface area in theory. Perhaps in the real life application of FRC is different? Do more wheels equal more traction in FRC or do they just equal more weight? Will a robot with its left and right sides lined with wheels be harder to turn than the same robot with the wheels only at the ends? I'm actually not sure on this one. Simple friction theory says there will be no difference, but there is the tread/surface interaction factor to consider in real life.

I'm not entirely sure on the matter of the robot's traction in terms of rolling motion. However, I'd be inclined to think that more wheels would produce more friction when not rolling (i.e. sideways motion, or a pushing battle). Also, the additional wheels would provide extra opposing force when the robot attempted to turn, resulting in much more work required by the motors. As I said before, we experienced this first-hand in 2002 when we used tank treads on our bot and realized that our bot couldn't turn at any quick rate without blowing breakers like mad!

[Billfred]

Look for images of 342's 2005 robot. They had 8WD (well, really 4WD with each wheel containing two), and it seemed to work well. The Magnetos did, however, have issues when they tried to adapt the design to this year's game in both durability and turning, leading them to scrap the front two sets of wheels for casters.

[lukevanoort]

Didn't 1511 use a 8WD with two high traction wheels in the center of each side and omnis at the corners?

[Rohith Surampudi]

75 doesnt use omnis or casties. All 8 are skyway 6", all 8 are powered, all 8 are mounted on trucks see our inventor/design site for CAD and inventor drawings.

Here is the website for those of you who wanted to see it
Team 75's inventor site

[DonRotolo]

Originally Posted by sanddrag

Friction has little do do with surface area in theory. Perhaps in the real life application of FRC is different? Do more wheels equal more traction in FRC or do they just equal more weight? Will a robot with its left and right sides lined with wheels be harder to turn than the same robot with the wheels only at the ends? I'm actually not sure on this one. Simple friction theory says there will be no difference, but there is the tread/surface interaction factor to consider in real life.

The independence of friction to area is valid only for hard surfaces. In cases where there is deformation (such as with carpeting or skyway wheels) then frictional force is dependent upon surface area - this is what you're getting at with your comment about "tread/surface interaction"

In FRC, more wheels means somewhat more traction, but there is a diminishing return, since the Normal force (robot weight) is constant, meaning each contact patch is being pressed into the carpet less forcefully as you add more contact patches.

As you gain too much traction, the power of the motor can become insufficient to cause sliding motion, which is often required to turn an FRC robot. (Note that cars use steerable wheels to help avoid this issue). The hyper-grippy tank tread issue.

The wheels at the furthest points from the center of rotation will need to slide the most. If a robot had all eight wheels clustered around the middle of the robot, it would be easier to turn than if the wheels were clustered in the corners. Having the wheels in-line or offset is immaterial, their distance from the center of rotation is all that matters (as far as placement goes). That is why a tank tread design needs a slight high spot to turn, and why it's generally located in the middle of the side. Having the center of gravity near there helps, too.

"Easy to turn" is so relative, but with careful design, such a bot could be indeed quite nimble.

Don

[Waynep]

So we've played around with 8 wheel configurations and here is what we found:
-As compared to a rockered 6WD system we have more consistent power delivery to the ground when pushing. Given that (hopefully) at least 6 full wheels are in contact at any point.
-Rockering the center will definitely help the robot rotate.
-Centralization of mass is a must, you really want the majority of your weight over the center of your robot for turning. The 8WD system gives you a larger range to spread out the weight than 6WD where you pretty much have to be spot on otherwise you have a forward or rearward weight bias.
-The other trick is using wheels with a different coefficient of friction. We have noprene and gum rubber treads on our robot. Ideally, you put the grippier ones in the center and the ones that slide more towards the outside. (The picture that Joey Gannon pointed to has a mismatch due to the fact that we were replacing wheels with whatever we had, we got in a good deal of pushing matches and driveoffs at Pittsburgh)
-Choose the appropriate gear ratio for such a system so you don't trip your breakers.
-There's something magical about the diameter of the wheels (this is a phenomenon I have tried to explain and still don't have a definitive answer), but it goes something like this: If you have the choice smaller diameter wheels will actually give you more bite given they will sink more into the carpet and what you get is a larger contact area than a larger diameter wheel whose contact patch can be thought of more as a tangential line. This also means that you have to correct more for the increased traction to get it to turn. If I remember correctly 229 uses smaller diameter wheels on their setups for Denominator and Irrational.

Downsides:
-Added weight and parts.
-More chains to tighten.
-Added current draw due to increased resistance and when turning
-Friction hop is an issue if turning at full speed from a stop, this was tuned out in software.
-More points of failure and things to go wrong.
-Smaller diameter wheels limits, limits sprocket size, and ground clearance .

Fellow Engineers please do correct me if I'm wrong. But these are the conclusions that my students have drawn after our testing and why we choose to go with the 8WD setup as opposed to 6WD.

Best,
Wayne