pic: 1010 Drive system prototype

[joeweber]

loading photo ...

[Lil' Lavery]

Simply put, my question is... why?

I can understand possibly adding a set of 3rd wheels in the middle between the mecanums for helping stability (assuming they're all coplanar), but 10-wheels is excessive.

More wheels doesn't equate to more traction (or more pushing power).

The most glaring flaw of all is that those big, heavy, expensive mecanum wheels are nothing but giant, efficiency (and "pushing power") robbing wheels in this drive system.
You have each side linked via chain, meaning each side of the drive train acts as one. In order to get strafing motion out of mecanum wheels, you need for each to be independently powered. In this configuration, each mecanum wheel is diverting some of your force in unused directions, costing you pushing power without giving you the maneuverability advantages.

[Jared Russell]

I believe the intention is that the center wheel on each side (where the chains meet) uses a ball differential. Thus, the mecanum wheels could still be independently powered and the center wheel gets the "average" speed. Likewise, the second and fourth omni wheels appear to also be linked between the left and right side with a ball differential in each wheel. These wheels would receive the "average" of the two front and two rear motors' power, respectively. When strafing none of the omni wheels would turn.

So while the kinematics work, I too have to question the benefits of this arrangement (other than being a novel use of ball differentials). More wheels != more traction. In fact, in this case, I believe that the reverse is true.

In the forward direction, the mecanum wheels will be spinning at the same rate as the omni wheels despite the fact that their rollers are offset by 45 degrees. Thus only 71% of the torque applied to these wheels is applied usefully (assuming free spinning rollers; the loss of torque, of course, is an issue with all mecanum/kiwi designs). In the lateral direction, none of the omni wheels are spinning. This means that all of the weight over those wheels (and in the center of the bot, I'd have to imagine that that is most of the weight) is not aiding in traction. Worse, this means that a bot with this drive system will be especially susceptible to pushing from the sides, as only four of the ten wheels can offer any resistance whatsoever.

joeweber - please correct me if I have misstated your design intent or made false assumptions about what you were trying to show with this drawing.

[JesseK]

Originally Posted by Lil' Lavery

Simply put, my question is... why?

I can understand possibly adding a set of 3rd wheels in the middle between the mecanums for helping stability (assuming they're all coplanar), but 10-wheels is excessive.

More wheels doesn't equate to more traction (or more pushing power).

The most glaring flaw of all is that those big, heavy, expensive mecanum wheels are nothing but giant, efficiency (and "pushing power") robbing wheels in this drive system.
You have each side linked via chain, meaning each side of the drive train acts as one. In order to get strafing motion out of mecanum wheels, you need for each to be independently powered. In this configuration, each mecanum wheel is diverting some of your force in unused directions, costing you pushing power without giving you the maneuverability advantages.

1.) More wheels can support more weight and still have omni-movement in a non-FRC application

2.) More wheels = more traction in a dynamic environment where actual amount of surface contact is unpredictable. Why? More wheels = higher probability that you will get a better mate of carpet and rubber at any given contact point. Tonight I'll try to find the link where some Japanese researchers showed this relationship when determining tread design for shoes that wouldn't slip when going up/down carpeted stairs. Essentially, the shorter the carpet and finer the threads the less this is an issue but it could explain the anecdotal evidence some teams claim they've seen.

3.) The ball differential in the middle allows for strafing as a normal mecanum would move.

4.) Why? Because he can. I think this is a spectacular representation of engineering outside the box in FRC. Pushing the boundaries of transportation systems has always been met with a 'why' or 'that's not needed', and I thoroughly enjoy seeing fresh ideas, regardless of whether or not it applies to this or previous years' games.

Though I agree, it's overkill for an actual FRC game where the conditions for every 'feature' aren't met. On the four additional omni wheel sets, the chassis may be able to get away with 1 omni instead of a pair of omnis in order to save weight. Additionally, I would recommend teflon-coating the bearings/sprockets/chains because even with 99% efficiency per wheel, 0.99^10 = 10% torque loss overall. Then take into account the fact that you can only put 1/sqrt(2) torque going in the 'forward' motion (i.e. ~70% of total torque, due to the 45 degree angle of the rollers) and it's easy to explain why this setup may not live up to pushing power and/or acceleration expectations.

[joeweber]

Thanks for the response, this system does use ball differentials where the two drive chains come together. The video and system can be viewed at http://www.team1322.org/ideas.htm . I would never use the 10 wheel design but I do like the discussion on this and how traction can be view by different people. I have heard many views on traction and multiple wheel designs. The addition of the extra wheels just popped into my mind and though I would see how it would look on paper and wanted to share. I find it interesting trying to come up with a omni direction robot that is simple to build and has an effective pushing force. Swerve drive is complex and Mecanum is simple but less traction. Maybe somebody will come up with a solution some day.

[Bryan Herbst]

I guess everyone else has beaten the "too many wheel" horse beyond death,m though I do agree.

Not sure if you were aware or not, but those mechanums are also not set up properly. The wheels should not all be oriented the same direction, but in an "x" shape.

[joeweber]

Not sure if you were aware or not, but those mechanums are also not set up properly. The wheels should not all be oriented the same direction, but in an "x" shape.
Today 09:38 AM

Yeah I'm aware I just don't feel like a redraw. If You watch our vidio the prototype is correct. http://www.team1322.org/ideas.htm

Maybee we should get away from wheels and make running and jumping robots.

[JesseK]

Originally Posted by joeweber

...
I find it interesting trying to come up with a omni direction robot that is simple to build and has an effective pushing force. Swerve drive is complex and Mecanum is simple but less traction. Maybe somebody will come up with a solution some day.

Just thought about something after reading a couple of posts. In the few instances where the robot is trying to move forward, the middle omni wheels would essentially give the drive train 100% torque moving forward, and 71% torque moving sideways.

384 & 1086 seemed to have massively powerful mecanums in 2007 & 2008, and what I've gathered from them is it's all about gearing and how you try to push 'bots around. If your bot is square to the bot you're trying to push, so long as your bot is geared properly there shouldn't be a problem in pushing the other bot. Traction (as defined by the ability to convert torque into forward movement, which takes the roller angle into account)) is still a matter of what the rubber rollers are made out of and not the angle of the rollers.

Furthermore (and I just had this thought....), if your bot tries to push on the diagonal, (which should have the code only driving two motors at full speed), you only have 0.5 * 0.71 = 35% of the available total torque with the other two wheels acting as castors. This situation is probably where most people get their eye-witness accounts of 'mecanums can't push'.

[Lil' Lavery]

Originally Posted by JesseK

1.) More wheels can support more weight and still have omni-movement in a non-FRC application

2.) More wheels = more traction in a dynamic environment where actual amount of surface contact is unpredictable. Why? More wheels = higher probability that you will get a better mate of carpet and rubber at any given contact point. Tonight I'll try to find the link where some Japanese researchers showed this relationship when determining tread design for shoes that wouldn't slip when going up/down carpeted stairs. Essentially, the shorter the carpet and finer the threads the less this is an issue but it could explain the anecdotal evidence some teams claim they've seen.

I was viewing this through the limited lens of FRC, and there are situations where having more wheels/greater contact area could improve traction. In an FRC game (at least like every one we've seen so far), the benefit would be minimal in terms of traction, and quite possibly harmful in terms of "pushing power."

A middle set of driven wheels could be beneficial for tasks such as a ramp and stair climbing, but on a planar surface I question the advantage it gives.

Originally Posted by JesseK

Just thought about something after reading a couple of posts. In the few instances where the robot is trying to move forward, the middle omni wheels would essentially give the drive train 100% torque moving forward, and 71% torque moving sideways.

Assuming equal coefficients of friction and even weight distribution (neither of which are likely to be true in the real world), wouldn't forward/backward motion actually utilize 88.4% of the total torque? The three sets of omni-wheels would have a total of 60% of the weight and (theoretically) 100% efficiency in that direction, while the mecanums would hold 40% of the weight and have 71% efficiency.
In perpendicular (strafing) motion, it would (theoretically) have only 28.4% of the total torque, as the omni-wheels (which hold 60% of the bots weight) are only acting in the forward/reverse direction.

Originally Posted by JesseK

Furthermore (and I just had this thought....), if your bot tries to push on the diagonal, (which should have the code only driving two motors at full speed), you only have 0.5 * 0.71 = 35% of the available total torque with the other two wheels acting as castors. This situation is probably where most people get their eye-witness accounts of 'mecanums can't push'.

On a standard mecanum you'll have 50% of total torque. 2 wheels operating at 100%, and 2 wheels operating at 0% (casters).
In this drive (once again assuming equal coefficients of friction and even weight distribution), 20% of the robots weight will be acting at 100% (driven mecanums), 40% will be operating at 71% (four driven omni-wheels operating at 45º), and 40% will be operating at 0% (undriven omnis and mecanums). That theoretically means you're effectively using 48.4% of your torque. Or just under what a standard mecanum would be using.

Originally Posted by JesseK

384 & 1086 seemed to have massively powerful mecanums in 2007 & 2008, and what I've gathered from them is it's all about gearing and how you try to push 'bots around. If your bot is square to the bot you're trying to push, so long as your bot is geared properly there shouldn't be a problem in pushing the other bot. Traction (as defined by the ability to convert torque into forward movement, which takes the roller angle into account)) is still a matter of what the rubber rollers are made out of and not the angle of the rollers.

I think you've hit a good note, though from a slightly different perspective. In general, I've found, that a robots anecdotal "pushing power" has every bit as much, if not more, to do with how it's driven than the mechanical features of the bot. Many of the 8-wheel defensive monsters out there seem to have more pushing power than a 4-wheeled bot because they drive more aggressively, and actively use their momentum and defensive driving techniques to their advantage.

[EricH]

Remembering the previously posted sketches/pictures of the prototype, this drivetrain is rigged for keeping mecanum capability on ramps, as one end hinges. (Note: the hinge is right around the center axle). Doubles as suspension, too.

Now, here's the question: Are those axles running across the width of the robot? If so, is there any particular reason for them? I can see that one is rigged for rotational motion up and down, but not the other--was that just "I'm too lazy" or simply the way it's designed?

[Jared Russell]

Originally Posted by EricH

Remembering the previously posted sketches/pictures of the prototype, this drivetrain is rigged for keeping mecanum capability on ramps, as one end hinges. (Note: the hinge is right around the center axle). Doubles as suspension, too.

Now, here's the question: Are those axles running across the width of the robot? If so, is there any particular reason for them? I can see that one is rigged for rotational motion up and down, but not the other--was that just "I'm too lazy" or simply the way it's designed?

I believe that the axles running the width of the robot serve to link together the ball differentials of the 2nd and 4th wheels on each side. The result is that these wheels receive the "average" power of the two front or two rear motors. Thus when strafing, the 2nd and 4th wheels do not turn.