The Center Wheel
 

I'm designing a 6-wheel drive train and I'm debating on whether or not to drop the center wheel. I know of several teams that drop the center wheel, but I also know of several teams that don't.

Dropping the center wheel can improve the maneuverability of a robot, and can also prevent "hopping" when the robot turns. However, in my opinion, it reduces the robot's stability footprint. Unless I'm convinced otherwise by this thread, I'd like to keep all 6 wheels on the ground.

I'd really like to hear from teams that don't do dropped wheel drive trains, particularly regarding how they prevent their robots from "hopping". Teams that immediately come to mind: 254, 233, 968, 25

Re: The Center Wheel
 

I would recommend dropping the center wheel. We have done so since 2005 excluding 2009 with varying amounts of success. Over the years, we have dialed in the drop to around .8" to .125". The only time we had trouble with dropping the center wheels was in 2008. I believe we had like 3/16 of drop that year and a very centrally located CG. We were to maneuverable to the point where the robot was tough to control.

On the Issue of stability, we have never had an issue. Maybe because we always try to keep our CG low.

Hope this helps.

-Ryan

Re: The Center Wheel
 

Assuming you are using 6 traction wheels, you need to drop the center wheel. The robot wont turn if you don't.

If you dont want to drop the center wheel, you can use either 2 or 4 omni wheels on the corners so that they roll instead of skidding.

Re: The Center Wheel
 

Every one of those teams bar 25 does dropped wheels.

I actually don't know of a single other team that consistently does 6WD non dropped wheels, to give you an idea of its popularity.

Re: The Center Wheel
 

The decision to drop or not depends upon the traction of your outer 4 wheels. Less traction means less need for drop, more traction means more. Thus mikelowry's comment "Assuming you are using 6 traction wheels, you need to drop the center wheel. The robot wont turn if you don't."

Make a set of side plates that have an adjustment for center drop, get a bunch of different wheels and experiment. Best way to learn.

Re: The Center Wheel
 

Which wheels do 254 and 968 drop on their 8-wheel drivetrains?

Re: The Center Wheel
 

The 4 center ones.

Drop your center wheels, unless you don't want to turn. 25 may not drop their wheels, but you'll find 25 teams who don't drop their centers and can't turn for every 1 team that can

Re: The Center Wheel
 

Don made a great suggestion (as usual), try some different things and see how it works.

And you can play engineer first, figure out what will happen using math and physics equations, then see how close your predictions are to your experimental results.

Re: The Center Wheel
 

i am 90% sure 233 drops their center wheel, just very slightly.

we have always dropped the center wheel. Depending on how far u drop it and how your CG is situated your bot will rock. I find this to be the best way to do tank drive. i personally hate the omni wheels in the corners as they make the bot far to easy to spin (from a defenders point of view). You can make a robot just a maneuverable and many times stronger by dropping the center wheel and putting real wheels on the corners.

my $0.02

Re: The Center Wheel
 

We get this question all the time, with people curious about our turning. We find there to be no real use in dropping our center wheels, because if we do, at any given time, our 6 wheel drive train could just as well be a 6 wheel drive train with two wheels in the air.

I've noticed the "hopping" robots on the field, and wondered why we've never had the problem. The skyway wheels that we tread allow us to drive with 6 as well as turn great with 6. They also have a wider width then most wheels used by FRC teams.

Re: The Center Wheel
 

yeah I was thinking if you did an 8-wheel with the middle wheels dropped you could form the equivalent of a square chassis. I may be wrong but I'm pretty sure the skidding come from the drive being rectangular, causing conflicting vectors.

I know its a bit off topic, but has anyone tried a square mecanum drive? If so did you see a difference?

Re: The Center Wheel
 

Coefficient of friction of the wheels on the carpet is one of part of the equation...normal force is another. If you have less friction, you can have more normal force, and still be able to steer as easily.

Re: The Center Wheel
 

It seems like that your problem with dropping the wheel is just because you think it will decrease the robot's stability, which in my opinion it will although I don't have an answer for you on how much, but if this is your problem I say go with your instincts and keep them at the same level. Although in order to gain some of the manueverabilty that dropping allows, make your outer wheels omni wheels and have your center wheel be traction.

On a different note and this is something I think we'll be trying in the offseason is, and we stole this from SparX (1126), do a 6 wheel with a 1/8" center drop, with the center wheel be traction wheels along with one of the outer sets. While having the other end be omnis, this will change the bot in a way that if you go in one direction, so all traction wheels are touching, you have more traction to the ground, or travel in the opposite direction, with one set of traction and one set of omni wheels touching the ground, which will give you more maneuverability

Re: The Center Wheel
 

The relatively wide wheels Team 25 uses helps avoid hopping. Their siping program is a science and an art.

Re: The Center Wheel
 

If you are very adamant about not dropping your center wheel, you can use outside wheels with a lesser coefficient of friction than your center wheel.

However, this will decrease your tractive force.

Re: The Center Wheel
 

If you are also concerned about the amount of teetering you're getting from a dropped center, make sure your outer wheels are as far apart as you can go. Its an obvious solution to a very minor problem once CG is maintained as close to the center of your robot as possible (barring any intentional CG changes given certain manipulators).

.

Re: The Center Wheel
 

From the experience we've had with a dropped center 6WD the rock has never been a problem. Not once.

Re: The Center Wheel
 

Same here.

Re: The Center Wheel
 

Same here, too. You don't need a lot of "drop" to get the effect you want. And that little amount of drop results in a barely noticeable rock.

Re: The Center Wheel
 

On a flat field, at least.
With the thick plywood under the carpet near the bumps this year, our robot needed a larger drop to turn well near the bumps.

Re: The Center Wheel
 

Ditto on the drop not being a problem for 330.

We also used sticky pneumatic tires in the center and AM 6" wheels on the corners; one year we tried 6 stickies and quickly swapped at least two of them after having problems turning.

Bear in mind that 3 of the 4 years I know of us doing a 6WD drop, we had arms to place objects. We never had an issue placing or acquiring in a big hurry, or on the go.

Re: The Center Wheel
 

233 has dropped the center wheels since 03.
This year we dropped the center 4 wheels.
In my experience the hopping is caused by a flexing chassis.
We always try and build the frame as stiff as possible.
mike d

Re: The Center Wheel
 

The reason behind dropping the center wheels is due to the current drawn by the drive train with all the wheels on the carpet. If all wheels are high friction types (in all directions) then the current on the drive motors skyrockets to near stall current. With a CIM motor drive that is 129 amps per motor. With four motors pulling over 400 amps, the power supplies feeding the Crio and radio start to fail or at least disable Crio output. In some cases the hopping robots you see are due to the wheels actually breaking friction with the floor and in others it is due to the Crio shutting down the outputs. In worse case, the Crio reboots and you all know how long that takes for your individual robot design and software. The decision to keep all wheels on the same level is a risky one that each team must make based on their own robot and wheel design. You can minimize the current by altering the turning radius in software, using omni wheels, dropping the center trucks, or dropping a helper wheel to lift the front or back wheels during a turn.

Re: The Center Wheel
 

we have less of a drop, and when our robot straddles the plywood transition, it rotates on axis painfully slow. but on the flat it is fine

Re: The Center Wheel
 

excellent idea! Anyone willing to share the relevant equations?

Re: The Center Wheel
 

Whenever we've done 6-wheel designs, we've dropped the center wheel. In Overdrive, we had a forklift-like elevator that got up probably 10 feet high with dropped center wheels... and we never tipped over. We also had a giant steel plate on the bottom of the robot to lower the CG.

Dropping the center wheel a fraction of an inch does make the robot look a little unstable as it bounces between front and back wheels, this is true. But in reality, if your CG is fairly low, it introduces no risk of tipping while on flat ground (and even climbing ramps in Rack and Roll it didn't seem to matter).

You may want to create a small test bed to see the results and prove it to yourself. Create a standard robot chassis with 6 wheels, but make the center wheel's height adjustable for example, you could securely mount a screw vertically that, when tightened, pushed the axle downwards, giving you a dropped center wheel, assuming you're using dead axle's). Drive it around with both dropped center and non-dropped center, and see how it handles with each. Design some tests for stability to determine if that would be a problem. In short, get some data before you decide which way to go.

Re: The Center Wheel
 

Re: The Center Wheel
 

How about making your drop adjustable?

Also remember that you can drop the center wheel or lift the outer wheel.

When making an adjustable setup it may be easier to lift the outer wheels that probably are not interfaced with your transmission directly.......depending on the rest of your design.

Re: The Center Wheel
 

Please note I'm talking about using "standard" type (kit or similar) wheels for purposes of this discussion. The use of omni wheels, mecanum drive, holonomic drive, etc. all provide very different characteristics.

In general the "hopping" / excessive current draw / etc problem isn't limited to a six (or more) wheel drivetrain. I've seen it in four wheel drives as well - in fact our team made a design mistake in 2007 with our four wheel drive configuration which led to excessive current draw and poor turning at times.

Turning torque (also called turning scrub) is the culprit and is largely affected by the wheel base width vs. length. Yes wheel types, etc can have an effect too, but as a general rule in standard tank style drives (4,6,8, no matter how many wheels) you'll want your drive base width (distance between wheels left to right) to be equal to or wider than the drive base length (distance from center of front wheel to center of back wheel).

The VEX Classroom Competition teacher's handbook, the Inventor's Guide, and the Autodesk curriculum all carry more detailed information on the topic - diagrams, etc.

In an FRC robot, if you build your chassis in the classic long rectangular configuration (ala 254 most years - ~38" long x ~28" wide), a six wheel drop center configuration provides huge maneuverability advantages and cuts out excessive current draw because only 4 wheels are ever on the ground at one time - cutting the effective front to back wheel base in half, making your effective wheel base on a flat surface wider than it is long.

If you build your FRC robot in "wide body" configuration (~28" long x ~38" wide), there's most likely (depending on other variables) no need to drop a center set of wheels as the wide configuration and shorter front to back wheel base/footprint provides for easier turning all by itself. Team 25 has built some great 6-wheel wide body robots and some long ones as well. In those machines (long), as stated before, clearly they are using means other than a drop center wheel to overcome the turning scrub/current draw issue and they'd be best suited to answer the question as to how they get it done.

If your drop distance (my team has always used .125") is in a decent range, rocking should never be a problem. We did encounter an issue this year like one other post here stated. We started out long configuration, drop center and because our CG was so low and our mass was so evenly distributed around the center, we had a difficult time driving the robot because we were effectively turning on two wheels most of the time. So we took our drop shims out in Washington, DC which made us easier to control and almost impossible to turn (dropping to like 7 volts on a fresh battery). After that we put a pair of omni wheel on one end making turning much easier, but maintaining our control while driving/turning.

Whatever you decide, just be sure you understand the way turning torque/scrub changes with the configuration of your robot chassis and wheel position. Those simple factors should enable you to make sound decisions without things getting too very complicated.

Re: The Center Wheel
 

Team 1323 did this and you can see the rails that they used here
http://www.chiefdelphi.com/media/photos/32378

Re: The Center Wheel
 

First step: Read the paper that Joe Ross linked to. It explains perfectly how the statics of a 4WD robot works. For a dropped-center 6WD, plug in Lwb = the distance between your center wheel and either outside wheel (so a 38 inch long 6WD wheelbase is effectively 18 inches long as opposed to 38 inches for 4WD). This shows why dropped center 6WD bots can spin on a dime, as Lwd > Ltw.

On how team 25 can turn without a dropped center wheel - there is no magic here. Just look at the math!

• A robot with 3 wheels per side in contact with the ground is an example of a statically indeterminate system - we need to look beyond statics to figure out how much load is being supported by each wheel. Clearly, the more weight that is supported by the center wheel, the better you will turn (assuming your CoM is on or near the axis of the center wheel). By using information about chassis stiffness and the moments of mass distribution about the chassis, you could solve this problem to figure out just how much of the weight the center wheels are supporting vs. the outside wheels.
• Assume that team 25's chassis is perfectly rigid and their mass is uniformly distributed. In this case, the 6WD closely resembles a tank tread system. You'll notice that in Chris Hibner's white paper, this very case is discussed! It turns out that for a tank tread, the effective Lwd is actually only 2/3 of the distance between the outside wheels. Of course, there IS no such thing as a perfectly stiff chassis, so the effective Lwd is going to have a bit of a fudge factor.
• Assume team 25's actual wheelbase length is 32 inches (38 inches minus half a wheel diameter - approx. 4 inches - on each side). Assume their wheelbase width is 24 inches (28 inches minus half their tread - approx. 2 inches - on each side). 2/3 of 32 inches is 21.3 inches, which is less than 24 inches! So the Drive Train Basics white paper says that turning is possible.
• The other piece of the puzzle is that team 25's wheels do not have the same traction in the X and Y axes. The siping pattern is designed to grip more in Y than in X (similar to wedgetop conveyor belting). Moreover, at least on their 2010 machine, there is a denser siping pattern (and therefore more grip) on their center wheels than on the outer ones.

So, the short version: Team 25 can turn because a non-drop center 6WD is still going to have a shorter effective wheelbase than a 4WD bot, and their wheels are grippier in the forward/back axis than left to right. Balancing these factors to effectively mix speed, maneuverability, stability, and pushing power is somewhat of a black art, which is why there is only one team 25.

Re: The Center Wheel
 

thanks!