6WD Pros and Cons

[Madison]

Originally Posted by Daniel_LaFleur

Kevins comments about being able to turn better are a function of wheelbase and not traction or CoF. Because of this I commented on shortening the wheelbase of a 4WD system would (by driving 36" wide rather than 36" long), in effect, create an equivelent wheelbase on a 4WD robot, and thus the same turning advantage of the 6WD system.

Wheelbase, while taking about turning, acts only as a short hand description of considerations of friction, coefficient of friction and torque. The behaviors described by Kevin's comments are absolutely a function of coefficient of friction and resulting traction. There are a handful of ways to make your robot turn more easily -- one is shortening the distance between the center of mass and a wheel's contact point, reducing the torque required to rotate the robot as a result of reducing the distance variable in your torque equation. Conversely, you might reduce the force variable in the equation by using wheels with a lower coefficient of friction; it is by this principle that omniwheels work. You could, with more difficulty, also increase the available torque.

I'm sure you understand this, but some of the folks coming down the line later may read this without some of the inherent assumptions I think you're making -- namely, considerations for all other things being equal. I just want to be sure that we avoid having someone think that they can use wheels with a higher coefficient of friction and simultaneously shorten the wheel base and have everything come up roses.

[Daniel_LaFleur]

Originally Posted by M. Krass

Wheelbase, while taking about turning, acts only as a short hand description of considerations of friction, coefficient of friction and torque. The behaviors described by Kevin's comments are absolutely a function of coefficient of friction and resulting traction. There are a handful of ways to make your robot turn more easily -- one is shortening the distance between the center of mass and a wheel's contact point, reducing the torque required to rotate the robot as a result of reducing the distance variable in your torque equation. Conversely, you might reduce the force variable in the equation by using wheels with a lower coefficient of friction; it is by this principle that omniwheels work. You could, with more difficulty, also increase the available torque.

I'm sure you understand this, but some of the folks coming down the line later may read this without some of the inherent assumptions I think you're making -- namely, considerations for all other things being equal. I just want to be sure that we avoid having someone think that they can use wheels with a higher coefficient of friction and simultaneously shorten the wheel base and have everything come up roses.

You are right, I was making a few assumptions (gotta stop doing that ).

corpralchees' first post only talked about a 6WD vs 4WD high traction setup

Originally Posted by corpralchee

Sorry if this is the wrong thread for this question but seems like it would fit. My team is wondering if going with a 6wd traction wheels all on the same level would be better than a 4wd traction wheels all on the same level.

To me, this sounded like if he just added 2 more powered wheels would it be better. The answer to this is just adding 2 more of the same type of wheels, at the same level, without changing wheelbase will not assist in turning nor will it (in and of itself) give you more traction.

So now that we are knee deep into the discussion, let me ask you what you mean by 'better'. Are you looking for more pushing force? better manuverability? Easier turning?

Edit: Also corpralchee, Wider tires do not give more traction, in and of themselves. So Widening the center tires, alone, wont directly help. There's a whole science around tire width, contact patch, durometer and sidewall height ... etc.

[Kevin Sevcik]

Originally Posted by Daniel_LaFleur

To me, this sounded like if he just added 2 more powered wheels would it be better. The answer to this is just adding 2 more of the same type of wheels, at the same level, without changing wheelbase will not assist in turning nor will it (in and of itself) give you more traction.

Daniel,

You're correct that adding 2 more wheels at the same level won't help traction, but it will help with turning. It won't help as much as a pair of lowered wheels to shorten the wheelbase, but assuming equal weight distribution on all six wheels, then each wheel has 33% less weight on it than a 4wd configuration and will be 33% easier to scrub sideways. Assuming the robot turns about the midpoint of your two center wheels, you still have only 4 wheels scrubbing but now with 33% less weight. So the robot is theoretically at least 33% easier to turn. That's not counting gains in the torque the middle wheels put out since they're operating under static friction instead of dynamic friction. So adding 2 more wheels certainly does help turning a little.

I'll admit the above all happens in the land of point masses and spherical chickens, but I don't think reality is THAT far off in this case.

[kramarczyk]

Originally Posted by Kevin Sevcik

but assuming equal weight distribution on all six wheels, then each wheel has 33% less weight on it than a 4wd configuration and will be 33% easier to scrub sideways.
...
I'll admit the above all happens in the land of point masses and spherical chickens, but I don't think reality is THAT far off in this case.

Your point is more valid than you think. Doing some reseach previously I found that adding wheels in the center leads to them carrying almost 2/3 of the total weight. See the supporting Machinery Handbook snippet posted here http://www.chiefdelphi.com/forums/sh...d.php?p=605839

I'd also like to point out for the silent masses that a white paper is available that explains, in detail, the physics of skid steering.
http://www.chiefdelphi.com/media/papers/1443
There is also a spreadsheet based upon this paper for 4x4 chassis.
http://www.chiefdelphi.com/media/papers/1917

[Daniel_LaFleur]

Originally Posted by Kevin Sevcik

Daniel,

You're correct that adding 2 more wheels at the same level won't help traction, but it will help with turning. It won't help as much as a pair of lowered wheels to shorten the wheelbase, but assuming equal weight distribution on all six wheels, then each wheel has 33% less weight on it than a 4wd configuration and will be 33% easier to scrub sideways. Assuming the robot turns about the midpoint of your two center wheels, you still have only 4 wheels scrubbing but now with 33% less weight. So the robot is theoretically at least 33% easier to turn. That's not counting gains in the torque the middle wheels put out since they're operating under static friction instead of dynamic friction. So adding 2 more wheels certainly does help turning a little.

I'll admit the above all happens in the land of point masses and spherical chickens, but I don't think reality is THAT far off in this case.

Hmmm ... maybe I'm being a dolt here and not understanding.

Each axle on a 6WD robot has 33% of it weight over it (assuming level loading).
Each axle of a 4WD robot has 50% of it's weight over it (again assuming level loading).

A 6WD robot has to overcome the traction of 2 of it's axles in order to turn, or in otherwords overcome (66% of it's weight over wheels x CoF).
A 4WD robot only needs to overcome the traction of 1 of it's axles (or 50% of it's weight over wheels x CoF).

Again, maybe I'm being a dolt here and not seeing the obvious, but why would it be easier to turn against more weight on wheels?

** as a note: this is all theoretical. It assumes many things which in nature just don't happen such as weigh being perfectly distributed over each wheel, the plane that the wheels sit on is perfectly flat, all wheels touch the plane in 1 (and only 1) point, the effective suspension for all wheels are identical, etc **

I

[lukevanoort]

Originally Posted by Daniel_LaFleur

Hmmm ... maybe I'm being a dolt here and not understanding.

Each axle on a 6WD robot has 33% of it weight over it (assuming level loading).
Each axle of a 4WD robot has 50% of it's weight over it (again assuming level loading).

A 6WD robot has to overcome the traction of 2 of it's axles in order to turn, or in otherwords overcome (66% of it's weight over wheels x CoF).
A 4WD robot only needs to overcome the traction of 1 of it's axles (or 50% of it's weight over wheels x CoF).

Again, maybe I'm being a dolt here and not seeing the obvious, but why would it be easier to turn against more weight on wheels?

** as a note: this is all theoretical. It assumes many things which in nature just don't happen such as weigh being perfectly distributed over each wheel, the plane that the wheels sit on is perfectly flat, all wheels touch the plane in 1 (and only 1) point, the effective suspension for all wheels are identical, etc **

I

You would be right, but 4WDs turn around their center, and not one pair of wheels. (unless there is something funny with the wheels)

[Madison]

Originally Posted by Daniel_LaFleur

Hmmm ... maybe I'm being a dolt here and not understanding.

Each axle on a 6WD robot has 33% of it weight over it (assuming level loading).
Each axle of a 4WD robot has 50% of it's weight over it (again assuming level loading).

A 6WD robot has to overcome the traction of 2 of it's axles in order to turn, or in otherwords overcome (66% of it's weight over wheels x CoF).
A 4WD robot only needs to overcome the traction of 1 of it's axles (or 50% of it's weight over wheels x CoF).

Again, maybe I'm being a dolt here and not seeing the obvious, but why would it be easier to turn against more weight on wheels?

** as a note: this is all theoretical. It assumes many things which in nature just don't happen such as weigh being perfectly distributed over each wheel, the plane that the wheels sit on is perfectly flat, all wheels touch the plane in 1 (and only 1) point, the effective suspension for all wheels are identical, etc **

I

In the instances above, you're assuming a different center of mass in each case. If the weight is distributed evenly across the robots, the center of mass will be in the center of the robot and it will, thus, turn about that point.

[Kevin Sevcik]

Originally Posted by Daniel_LaFleur

Hmmm ... maybe I'm being a dolt here and not understanding.

Each axle on a 6WD robot has 33% of it weight over it (assuming level loading).
Each axle of a 4WD robot has 50% of it's weight over it (again assuming level loading).

A 6WD robot has to overcome the traction of 2 of it's axles in order to turn, or in otherwords overcome (66% of it's weight over wheels x CoF).
A 4WD robot only needs to overcome the traction of 1 of it's axles (or 50% of it's weight over wheels x CoF).

Again, maybe I'm being a dolt here and not seeing the obvious, but why would it be easier to turn against more weight on wheels?

** as a note: this is all theoretical. It assumes many things which in nature just don't happen such as weigh being perfectly distributed over each wheel, the plane that the wheels sit on is perfectly flat, all wheels touch the plane in 1 (and only 1) point, the effective suspension for all wheels are identical, etc **

I

The scenario you posit may be the case, but it still works out in favor of the 6WD. The difference is that you're interested in the torque necessary to turn the robot. The torque would be coming from the wheels applying force around the center of rotation. So if the 6WD is turning about the center of the robot, you have 33% weight x 2 axles x 1/2 wheelbase = 33% weight. (I'm ignoring the CoF). For your 4WD scenario, you'd have to turn around the midpoint of an axle. So, 50% weight x 1 axle x 1 wheelbase = 50% weight.

This is, in fact, just a very rough approximation, as the width of your wheelbase affects how much skidding happens and what the actual torque arm for the friction is and what the torque arms of the wheel turning forces are, etc. etc. etc. I'm assuming all that is covered in much more detail in those whitepapers kramarczyk linked.

Also, for the record, I have not convinced myself that a traction wheeled robot necessarily turns about its center of mass. My intuition tells me it's a bit more complicated and has more to do with what wheels are slipping and how much.

[kramarczyk]

Originally Posted by Kevin Sevcik

This is, in fact, just a very rough approximation, as the width of your wheelbase affects how much skidding happens and what the actual torque arm for the friction is and what the torque arms of the wheel turning forces are, etc. etc. etc. I'm assuming all that is covered in much more detail in those whitepapers kramarczyk linked.

Yup, adds up the contribution that each wheel adds to the turning moment (M=f*d) less the resistance of each wheel to the turning moment.

Summary

Ever wonder why your four-wheel-drive system won't turn? You'll never need to wonder again. Drive Train Basics gives a full explanation, complete with mathematical equations, to allow you to design a drive train that will turn without a problem.

[nuggetsyl]

If you gear it right, 6 wheels are better for most first games. But not always.


P.S. do not lower the middle wheel on a 6 wheel drive robot.

[Cory]

Originally Posted by nuggetsyl

If you gear it right, 6 wheels are better for most first games. But not always.


P.S. do not lower the middle wheel on a 6 wheel drive robot.

C'mon Shaun. You have to know this statement is totally misleading.

The vast majority of teams who do 6wd need to drop the center wheels (even if only slightly) to allow their robots to turn. In the past, we have had turning issues, even on a robot with dropped center wheels.

As stated before, by dropping the wheels you effectively have shortened the wheelbase to 50% it's previous length, as your weight is being supported by 4 wheels at any given time. By shortening the wheelbase you make it much easier to turn.

Every year you guys claim your robot turns perfectly without a dropped center wheel. You are known for your high traction wheels, and you have a wheelbase length that is twice that of most other teams. While many teams have low gears for turning, you have only a single speed which is geared to 9+ fps.

Until you can show everyone the physics or math that validate that system, stating that no team should run a 6wd system with dropped center wheels just means that anyone who takes your advice will find themselves wondering what good it did to add 2 more wheels, when their robot still won't turn. Clearly your exact setup works well for you, but to claim categorically that you should not drop the center wheel is just plain wrong and totally misleading.

[ChuckDickerson]

Originally Posted by nuggetsyl

If you gear it right, 6 wheels are better for most first games. But not always.


P.S. do not lower the middle wheel on a 6 wheel drive robot.

So why? This is exactly the sort of blanket statement I was trying to avoid when I started this thread. As Cory indicates in the previous reply, Team 25 is known for their awsome 6WD setup. I was hoping that teams with experiance in 6WD setups (like Team 25) would post the pros and cons of the 6WD systems they have developed over the years rather than making a blanket statement that this one or that one is "better" with no supporting evidence as to why. Come on nuggetsyl! You can do better than that.

[Donut]

I know there was a discussion on this the previous year, but the topic of 25 not lowering center wheels reminded me of one of the more interesting ideas, like in this post by Joe Johnson.

It's a possible explanation of 25's drive system, and even if it's not exactly right, is an alternative for dropping the wheel on a 6 wheel drive system. The use of kinetic friction to turn is a novel idea, as you get the turning advantages when you want to turn, but your opponent still has to fight the full force of static friction when your robot is stationary.

[nuggetsyl]

Originally Posted by DeepWater

So why? This is exactly the sort of blanket statement I was trying to avoid when I started this thread. As Cory indicates in the previous reply, Team 25 is known for their awsome 6WD setup. I was hoping that teams with experiance in 6WD setups (like Team 25) would post the pros and cons of the 6WD systems they have developed over the years rather than making a blanket statement that this one or that one is "better" with no supporting evidence as to why. Come on nuggetsyl! You can do better than that.

I am trying to keep the answers short and to the point. K.I.S.S. The point of having 6 wheels is for traction. By lowering the middle wheel you have just defeated the purpose of using 6 wheels. And make it much easier to be pushed from the side corner of your robot.

here is you pros and cons

pro- much harder to push, drive stright lines better, easier to control in tank drive.

cons- turns choppy, heavy if weight is a concern, need to put alot of time into gearing it right.

[s_forbes]

Originally Posted by nuggetsyl

I am trying to keep the answers short and to the point. K.I.S.S. The point of having 6 wheels is for traction. By lowering the middle wheel you have just defeated the purpose of using 6 wheels. And make it much easier to be pushed from the side corner of your robot.

As I see it the point of lowering the center wheel is to make it easy to turn your robot but still maintain the ability to push people when you want to. Turning your robot works both ways. If it's easy for you to turn, then it will be easy for others to turn you. If you build your robot so it can't be turned by other bots, it will be very hard for your robot to turn by itself.