6 Wheel Drives Without Dropped Centers

Technical Technical Discussion
#1

I’ve been looking at different drive systems lately and I have tried to come up with something that our team could try. I have heard of teams using 6 wheel drives without a dropped center so that it is truly a 6 wheel drive and not just a 4 wheel drive with some extra stability.

I have heard that 25 has used this concept before and looking at some threads from the early 2000’s I found some information on it but I still don’t quite understand how it would work.

Can anyone enlighten me as to how this would work? Everything that I can think of says that a 6 wheel drive robot without a dropped center shouldn’t turn without a special 7th wheel mechanism or omniwheels or something of the sort.

Why can't a drivetrain turn if it is wider then it is long?
#2

Hey Cam!

Someone from 25 will probably have to pop in to clear up the details, but here is what I know. 25 does use a 6WD with no drop. In the past, they have even gone up to 8WD, still with no drop. They use what has been termed as “Brute Force” turning; no tricks, no mechanism. It’s all drive train for them. From what I recall, the secret lies in the wheels. Those massive drum wheels 25 has used in their drive system for years are from Skyway. 25 sands down the wheels so they are slick, and then cuts into them and adds tread as necessary. Here’s a little more info on it: 25’s Tires. While the thread is over 10 years old now, I seem to recall a thread in more recent years where the process is verified to still be alive.

Like I said, an actual member of 25 is going to need to provide the details, but that’s the just of it.

EDIT: Here is a more straight-forward description of the design from a member of 25 - Link

When it comes to drive systems with speed and power, you’d be hard-pressed these days to find a team who surpasses 25 in both respects.

Regards,
Leeland

#3

4 traction wheel in the back or front and 2 omnis in the front or back.
Great traction and mobility.
Our team uses this all the time. It is very effective.
I have also heard of teams that put 2 omnis in the middle.

#4

Omnis in the corners. Traction wheels in the center. Very maneuverable, fast, with a lot of pushing power.

#5

Last year I did my grade 12 physics final project on this topic. To super-simplify it (ignoring center of gravity, wheel coefficient of friction in x and y, gear ratios, etc) the rule of thumb is that you want your wheel base to be at least as wide as it is long.

There are multiple ways that a 6wd, no-drop center drive train can be effectively executed. The two most popular as far as I know are the following:

Wide Base 6wd:
This drive base is exactly what you describe, only the drive base is in wide configuration. The reason this works is because as our rule of thumb dictates, the base is wider than it is long.
This is a kit of parts base, which does have a drop center, but it looks just like one would without the drop, just pretend :rolleyes:
http://www.chiefdelphi.com/media/photos/37486

4 traction + 2 omni:
This drive train is probably more like you are envisioning, a standard long orientation, 6 wheel drive robot. The thing that allows this one to turn without issue is having omni wheels on one end. Omnis have very little friction sideways due to the rollers, essentially serving the same “cut the drive base in half” that a drop center does.

Something like this, except with omnis only at one end:
http://www.chiefdelphi.com/media/photos/31039

#6

Yes, thank you Leeland, you have come through for me again! This is exactly what I was wondering about. I hope someone from 25 could come in and provide a little bit more detail about this.

Just wait till FLR next year when we have a 6 cim, 2 speed, 6 wheel no drop drive train :wink:

#7

The process we use to cut our tires hasn’t changed significantly over the years, so a lot of the information in those old posts is still relevant. We first shave down Skyway Bead-LOK wheels using a table saw (a couple of years ago we created a machine to help us with that process - see this video), and then cut the treads using a wooden jig.

Here’s a thread from 2010 that generated some good discussion about dropping the center wheel: http://www.chiefdelphi.com/forums/showthread.php?t=85763.

In particular, this post by Jared from 341 does a good job explaining some of the specifics of our drivetrain.

If you have any more questions, feel free to ask (or come on down to Brunswick Eruption in November to check it out in person :)).

#8

Just curious if you guys have ever had experience with an 8 wheel drive train with no drop. it seems like if you are doing 8 wheels then all 8 are contributing friction while turning as opposed to 6 wheels where it seems like only 4 are.

#9

Although not quite as much pushing power as having four traction wheels on the ground.

#10

Our team had four traction wheels in the back and 2 omnis in the front. Does that count?

#11

Cheap option: zip ties around the treads of the 4 corner wheels?

#12

We ran an 8 wheel drive with no drop in 2007, which was detailed in the Behind the Design book from that year. From a driving perspective, it was noticeably harder to turn than the six wheel configuration, but not by much. Since Rack N’ Roll was such a defensive game, the added stability from the longer wheelbase and extra resistance to turning worked to our benefit. It was fine for that game/year, but we haven’t returned to an 8WD of the same type since (even when we built Evil Machine 5’s twin in 2011).

As a fun side note, we blew one of our drive speed controllers at Battle O’ Baltimore that year (and didn’t have a replacement with us to swap it out) and were still able to drive around pretty well.

#13

The turning ability of a robot with no drop-center and all traction wheels is dependent on the normal force experienced at each wheel. That is to say, it depends on how much force each of the wheels on the robot supports.

As Jared mentioned in his post that was linked to, a 6 wheel drive with no drop center is statically indeterminant. This is because if you isolate the physics on one side of the drive train (assuming a perfectly symmetrical robot), you have the force for 3 wheels as variables but only the sum of moments about the wheels’ axes and the sum of forces in the vertical axis as equations. That’s 3 variables and 2 independent equations, so this mathematical model cannot solve for how much weight each of the wheels supports.

This makes it complicated to approach from a design perspective, which is why the over simplification of “a long wheel base robot without drop center wheels will not turn” is often made. In reality, as you approach the limit where all of the weight is supported by the 4 corner wheels, the robot will behave like a 4WD long robot and have difficulty turning. As more and more of the weight is supported by the center wheels, this reduces the normal force at the corners, thereby reducing the tractive force applied at the corners and reducing the robot’s resistance to turning. More weight at the corners = Harder turning. Less weight at the corners = Easier turning.

When the center wheel is dropped on a robot, much of the robot’s weight is supported by the center wheel at all times in addition to shortening the wheel base, which makes turning easier (another generalization, but a widely accurate one so this is part of why drop centers are so popular).

Team 25 REALLY knows what they are doing, which is why they can pull this off. My guess would be that criteria that lead them to this type of design decision are:
-Makes it harder for defense to turn them
-Eliminates rocking, so stability isn’t an issue

I personally cannot perform the math necessary to prove that a drive train without a drop-center would turn. I also have not done the necessary testing and experimentation to have the experience needed to design such a drive train that I will guarantee can turn well. Therefore, I would hesitate to build this type of drive train. This is where it’s important for teams to build within their means and recognize them.

#14

Does anyone know if versa wheels would work for this sort of drive train? They have really good traction going forward and backwards but not the best laterally because of the W tread on them. I think that this sort of wheel would almost be perfect for this scenario.

#15

I think the diagonal pattern that 25 uses is important to their success. This is obviously something that versa wheels don’t have.
It might be worth trying with colsons though…

#16

Someone on CD said versawheels tend to dig in to the carpet like cleats when pushed laterally

#17

I heard other wise from the highest traction wheels thread:

#18

What T^2 said is true given your robot remains completely flat on the floor, parallel with the carpet, and doesn’t weigh too much. In the event you’re not parallel (ie. being pushed from the side and slightly tipped upwards on the pushing end), the W tread will grip to the carpet like cleats. It won’t need to be a significant lift, but anything to change the robot from parallel with the floor will engage cleat mode.

Source: Experience with cleat mode this year.

#19

Quite interesting, Miss Daisy used the VersaWheels this year, and I witnessed them getting pushed sideways. However, I’m not certain if they dropped the center wheel or not (6wd, 6cim), It would be interesting to hear about their drivetrain.

#20

48 has been running no drop center, 6WD West Coast setups since 2011.

2011: two 4" omnis up front, four 4" performance wheels in rear, roughtop tread. Long wheelbase.

2012: six 8" AM pneumatics, square footprint

2013: six 6" dia. 1" wide aluminum performance wheels, blue nitrile roughtop tread. Very slightly wide-biased footprint.

All three were powered by some variation of the A-M SuperShifter, 2 CIM’s per side.

2012 was not surprisingly the least-maneuverable of the three setups, but it still was good enough to get the job done in one of our most productive seasons ever.

The 2013 iteration is very nimble yet able to provide enough punch to apply solid defensive pressure when needed.