Spherical wheels

[Ciro Pabón]

I was intrigued by spherical wheels on theAudi RSQ presented in the movie "I Robot":



I found several references to robots and experiments with this type of wheels, like those:

Holonomic base

Robot with three degrees of liberty for movement

The Rolling Stone

The questions:

Does anybody know a real car that uses spherical wheels?

Which could be the best approach: maglevs or air bearings like this?

[Jersey Tom]

Air bearings.. I've only seen those used on very high speed, high accuracy machine tool spindles. Don't know how well that'd work on a slow-speed wheel, especially cantilevered!

[Ciro Pabón]

Yes, I thought exactly the same. But the director of the movie seems to think it is not so farfetched.

I could not find specific comments on the Audi team about the wheels, but the site about the Audi RQS seem to imply they put some serious thought in the concept.

You are right: the first application I found, back in the 80, were only for high rpm, low load. But looking around a little, I found low rpm, high load applications in which you have to supply the air.

Here you have a transporter with 280 tons of capacity, of course without wheels. It works more like a hovercraft, but in principle, it could work. Even if this thing is hovering, imagine the inertia.



There you have another: " when compressed air is supplied to the module, the air bearings inflate and lift the load. A thin film of escaping air is formed between the air bearings and the floor, resulting in a friction-free movement in any direction". Not exactly what I am looking for, but it is a beginning:



Finally, this is a vehicle turntable. Funny. It is used in restricted spaces.



The proverbial impossibility is this: reinvent the wheel. I do not expect much from this thread, to tell the truth. But the more I try to explain, the more I learn.

[tf1]

what advantage(s) do you see from going to a spherical wheel?

[captainmorgan]

how does turning work? would there be any slip angles at all?

[RacingManiac]

Air bearings.. I've only seen those used on very high speed, high accuracy machine tool spindles. Don't know how well that'd work on a slow-speed wheel, especially cantilevered!

Calspan's Flattrac tire testing machine uses airbearing to support the belt which runs between the 2 drums(imagine a giant belt sander). It supports up to 20000lb I believe and is capable for something like 250mph....

It is the only place in the world than can do full size truck tire testing at full load....

[Ciro Pabón]

Thanks, RacingManiac, excellent reference. I do not know if this is the correct link.

what advantage(s) do you see from going to a spherical wheel?

C'mon, how could you expect me to answer that? OK, let's reinvent the wheel, then. I will play "devil's" advocate, because the wheel is like 3,000 years old and it is undefeated. You could say that modern wheels have only 100 years, but they certainly have advantage.

* "Conventional" wheels have only 2 degrees of freedom, wich means you can vary only your speed and the radius of the curve. The more obvious advantage is the ability to move with 3 degrees of freedom, which means you can rotate or move laterally the car, which in turn implies the ability to have any slip angle you wish, or none at all.

If you have seen the movie, maybe you remember when "our hero" is followed by a truck full of robots? The truck is moving sideways.

* The difference is a little deeper: to know the trajectory of a non-holonomic system, like cars and bicycles, you need to know the history of his path.

On the contrary an holonomic system trajectory is independent of its momentum or its velocity: it depends only on his coordinates and time, like, for example, a pendulum. This, in turn, means that lateral movements are not limited to parking manouvers: you can do lateral movements at any speed.

Again, if you have seen the movie, you may remember that "our hero" decides to drive manually, when, distracted, he almost hit a truck in the back: he then procedes to move the car laterally at, I do not know, 300 kph in Hollywood speed units, and the car does not slip much sideways (you could have exactly zero under/oversteer and rotation for the entire manouver).

* If, actually there is no need for an slip angle to exist for the car to displace laterally or even rotate, you do not need to create friction between the tire and the road to take a curve: essentially you could do all the time what rally pilots do when they slide, pushing the car with the engine in the desired direction. Think of an office chair with spherical wheels.

This should extend the life of the wheel "thread" enormously. For example, I did not see any flat tyres in the movie.

* Finally, they are totally cool. For example, the hub caps have a funny hexagonal design that suggest the need for the rigidity of the entire assembly over the wheel and the maximum space for air to flow. Notice the huge air intakes on the rear and front wheels and how they mix with the body. This is what made me think about air bearings. Besides, this thing was designed by Audi, apparently, in a sort of "semi-serious" mood about how cars could look in 2035. Or in 1935...

Auto Union Type C

[pyry]

ciro, offtopic, but do you happen to be an audi fan?

[tf1]

The problem I can see is that the contact patch shrinks to an unbelievably small size. To get any kind of performance, you'd have to rely on something other than just the contact with the ground (i.e. thrust jets or something of that nature) to get the forces necessary to make rapid maneuvers. Mitsubishi already has demonstrated the capability to make a car with wheels that turn together so that it can move completely laterally. But that car uses regular tires.

[Jersey Tom]

Cool! Didn't know they had big beastful air bearings these days.

* If, actually there is no need for an slip angle to exist for the car to displace laterally or even rotate, you do not need to create friction between the tire and the road to take a curve: essentially you could do all the time what rally pilots do when they slide, pushing the car with the engine in the desired direction. Think of an office chair with spherical wheels.

There might not be slip angle, but you'll just replace it with slip ratio. One or the other has to exist for a car to go anywhere, spherical wheels or not. That's my understanding anyway.

And your contact patch won't be any smaller unless you run really high tire pressure, same as with conventional tires.

[NickT]

Some very interesting ideas, but you still need friction between the road surface and the wheels. Any mass that is moving in a uniform manner needs a force to change direction. If there is no friction between the road surface and the wheel then no change in direction can be achieved. This all assumes you want to use the wheels to change direction.

You could use jet thrusters, but it might get intersting if you are turning a corner and someone passes you in the opposite direction on the outside of the bend, you could quite literally blow them of the road if you were both using thrusters to change direction Hmmm..... I bet Michael would love this, just think he could quite literally blow his competitors of the track, in the last minutes of qualifying of course

[RacingManiac]

Thanks, RacingManiac, excellent reference. I do not know if this is the correct link.

http://www.calspan.com/tire.htm

MTS system is commercially available....Calspan's flattrac is commissioned by the US government in the 70s to do tire research for highway safety. But obviously it expands beyond that for business reason.....

[Ciro Pabón]

ciro, offtopic, but do you happen to be an audi fan?

No, I find very difficult to be a fan of a firm. They have no soul... I admire the work of the people.

BTW, I still remember the time I posted in the "Caption Competition" a photo of Caracciola shaking hands with Hitler at the first Auto Union/Mercedes show, back in the 30's (not to bash Auto Union or Mercedes at all: in South America we take more lightly this part of Europe history, actually I thought it was sort of funny)... Principessa was not amused at all.

To NICKT, JERSEYTOM and TF1:

- The tire can be a pneumatic one. There are several forms of "balloon tires", and they have a patch.

- I'd say you do not have an slip angle because you can rotate the wheel towards the direction you want to move.

- There is friction between the tire and the road, of course. This depend on the material the wheels are made.

Look, mates, climb in an office chair. Push it with your feet. You can move it in any direction you want, right? You can not do this, for example, in a bike.

Imagine that, instead of pushing with your feet you, magically, rotate the wheels until they point towards the direction you want to move, and then you make the wheels spin to move the chair. Now you see it? It is the same idea (not exactly, but it is a beginning to grasp it).

I have seen designs of this kind of wheels for robots, where its utility is evident, because they have to move in reduced spaces.

TO RACINGMANIAC:

The people at MTS explains that they have an advantage: their design allows them to make a very long moving track, that simulates the asphalt moving in relation to the car. This means you can do wind tunnel modelling of the wake of the car with greater precission.

[RacingManiac]

The TIRF at Calspan is only used to test tire, which MTS also makes a machine that does that as well. This is not for vehicle testing as a whole, but to find out the exact characteristic of the tire.

[ginsu]

Ciro I'm going to go with you on this and try to find some possible issues.

Straight line tracking. How are you going to prevent the spherical wheel from deviating in a straight line motion? It would have to be guided by some sort of magnetic ring in order to hold the wheel in the direction you intended. The problem with this, is that you would need and infinite amount of rings in order to turn the wheel. I guess you could make a sphere of magnets just below the surface of the tire.

Tire mounting. Assuming you could use a pneumatic tire, how would you mount it? Replace it? Inflate it. That would be difficult.

Handling. Setting up a camber curve would be extremely difficult. How would you control the motion of the tire in bump and rebound? Do you setup some sort of control arm with magnets on the end to interact with the magnets in the spherical wheel?

These are more questions than answers, feel free to comment.