Friday, January 30, 2009

18> PRT Motor-in-the-Wheel Design

This illustration shows how a motor can be mounted inside of a PRT “Pod-Car” wheel.
I was first inspired by a design from an electric “concept car” where the wheels were actually motors. The idea is simple. If you plug in a motor but hold the shaft, it will spin, (winding up the cord until it unplugs itself). So why not feed the cord through the shaft somehow? That way the motor itself would be a wheel, instead of mounting a wheel on the shaft? (Or more likely, a sprocket, gear or pulley) It turns out this idea is actually widely used. A Google search revealed that the concept is commonly employed to move conveyor belts, (drum motors) and to electrify bicycles in China (hub motors) and in fans, especially computer and ceiling fans.

Despite the obvious simplicity (and efficiency) of the idea, apparently the motors don’t perform well in terms of torque, as compared with a much higher speed motor fitted with a reduction gear. Also they are not exactly cheap or easy to find, so I bit the bullet and researched some more conventional motors.

My switching design calls for hard wheels, so I put in a bunch of them to have the traction to climb steep smooth slopes. I don’t really know how many is really optimum, at this point, but I checked out some “NEMA” standard sized brushless motors, and found that the NEMA 42 standard comes with sufficient torque that if each or the wheels had a motor within they could pull (together) with between 300 and 800 lbs. of forward power, while having a top speed of 60 mph. (Sorry, younger and euro readers, I am not natively metric and feeling lazy) Obviously, reducing the top speed increases torque proportionately, but it also worthy to note just how much of a trade-off is required between the weight, the steepness of the climbs and the top speed. There is no free lunch. For example a combined vehicle/payload weight of 1600 lbs. will probably not make a 45-degree climb.

It is also worth noting that the separate motors solve the fact that my system has no differential gears. For those who do not know what a differential gear is, here is a brief explanation – On car, when turning a corner, the outside wheels have further to travel than the inside wheels. Therefore they have to revolve faster, making more revolutions than the inside wheels. The differential gear accommodates this, preventing one or both wheels from slipping on the pavement, which would create wear and partial loss of control. Even though the right and left wheels in my PRT design are fairly close together, the same phenomena would apply to a lesser extent. With separate right and left motors, however, the RPMs of the wheels can be precisely controlled to actually create the appropriate steering forces.


Mr_Grant said...

You wouldn't need the complication of precise control of differential gears if propulsion is not done through the wheels. Have you ruled out linear motors?

Dan said...

Dan the Blogger Responds:

Welcome back, Mr. Grant. As I have mentioned, I believe that the track needs to be very cheap. PRT without a large network is, in my opinion, not economically feasible or, at least, not the best choice. I firmly believe that the track needs to be able to pay for itself with as little traffic as possible. How little traffic? That’s a question deserving of it’s own post. Anyway, as I understand it, linear motors require magnets in the track, and that is obviously an expensive proposition. I envision a system that could be extended well out of town, meaning that traffic could be quite light. The prospect of trying to pay for thousands of magnets that are only used for, perhaps, a couple of seconds per minute would seem to be impossible to justify compared to the magnets arranged in circle (regular motor) and in almost constant use generating revenue.

Part of the appeal of linear motors is in aspects of control, particularly centralized traffic control. As PCs have become dirt cheap and networkable, however, the model of the big mainframe controlling everything is seeming less attractive. Linear motors also require very close dimensional tolerances between the magnets in track and in the part that propels the car. This would seem problematic with the designs I have proposed, particularly in switching tracks. I hate to rule anything out, particularly designs that reduce moving parts, but I must say I can’t see using them from what I know. I did pick up one interesting advantage from Wikipedia however. It seems that in Moscow icing on the wheels of their monorail would cause slipping on starts and stops, so they use linear motors to help out where needed. Perhaps we should consider them in a similar capacity, such as the slope-climbing example I sited.

Mr_Grant said...

LIMs do not require magnets in the track.

With a vehicle-mounted LIM, the track contains a metal reaction plate (a strip of steel+alu). When power goes to the LIM coils, a reactive or attractive force is created.

Vectus has done the reverse, mounting the LIM in their test track, but the company says it can do LIMs-in-vehicles if that's what customers want.

Anyway, they think they have solved the tolerance problem. There was once a paper by Taxi 2000 that talked about how the optimum gap is maintained, but I don't recall what it involved at the moment. It may consider it to be proprietary these days.

The problem I have with gears is again based on what I know from bicycle tech -- small gears that spin fast wear out fast. Going gearless kills 2 birds -- cost and the materials in the spare parts.

Dan said...

Dan the Blogger Responds

I tried to access the Vectus site but I got a virus alert. Anyway, about the magnets in the track. O.K. I lied. It is true that a linear motor may be built without opposing magnets if one side (movable or stationary) induces magnetism on the other. I would think that there would have to be a tradeoff, however, power-wise. In regular servo and stepper motors they use powerful, rare earth magnets and hair thickness tolerances between rotor and stator to get good torque. (Very important, since magnetism gets exponentially more powerful with closer tolerances). I do not know the specifications of the systems promoted by linear technology proponents, but I have a hard time imagining a successful system that enables sharp turns (up and down as well as cornering) while still being suitable for linear motors. Frankly, I like the next logical step in linear motors, Meglev. But, personally, I see ideal PRT performance as less like a train and more like a sports car.

The real problem is this; Modern electric motors are wonders of engineering and are dirt cheap, very efficient, easy to replace. They are lightweight and compact. They are standardized and the characteristics (wattage, torque, heat dissipation, etc.) are all known. They are time tested and trusted. It would cost tens of millions to engineer one from scratch without the knowledge of previous designs. That’s a formidable head start.

From the beginning I have tried to de-mystify PRT technology and use, where possible, off the self –parts. My aim is to be practical and get people from here to there quickly: To enable the largest, least-likely-to-fail, network possible. To be the voice for the Volkswagen approach when the Lamborghini approach can’t muster venture capital.
Linear motor technology is, for the time being, for our purposes, without measureable dimensions that can be accommodated for, and doesn't it have torque or horsepower ratings that can be relied on. There are no competing manufacturers, no consumer base. It is not a technology that has reached a maturity wherein it has become a commoditized product, with the cost savings that go with mass production.

I hate to say “No”: to any reasonable point of view, especially since this concept has acceptance within this field. I have an axiom, “There is nothing more efficient than absence” and eliminating all moving parts is certainly attractive. I also hate to rule on it categorically since that runs counter to any kind of democracy within the debate. It is my hunch, however, that rotary motors will give a more robust performance.

Finally, with respect to the durability of the gears themselves, I would note the normal means of extending their lifespans, those being having the smaller gear made of a harder material than the larger, and to have the gears bathed in gear oil. I did not design in such a system, and it should be there. It should not leak or admit contaminants. This being so, I believe a life span of a half million miles could be expected between replacements.

This has been a most thought provoking exchange! Thanks Mr. Grant!

Dan said...

One more thing.. Baldor Motor makes a heavy duty LIM, and I checked out it's ratings. (I was wrong to assume that nobody has them with specs and all..) Bad news is that that at 100% duty cycle it can only lift one third its weight. 304lb. coils, 100 lbs. pull. At 15% it's 500lbs. I have ridden these before. At the Houston (IAH) airport they have a train under the terminals. There's a coil every half dozen ft. No wonder they put them in the track. there is only a minimal duty cycle that way. Gap is 1/8th", (half power at 3/8") Reacts with 1/4" steel plate covered with 1/8" aluminium. Coils can be faced to each other for more power,runs on 460 vac three phase. Requires a fan for 100% duty.

You know, in a different country or time there might be the long term vision to put down the cash for extensive LIM track, magnets and all. And it might just be the way to go. But right now I see dirt cheap, risk free track, and cars that pay for themselves after half their estimated lives and then start paying down that track. Plus I think I get 4 times more torque per weight with rotary motors.

Bengt Gustafsson said...

Yes, one of the best advantages of linear motors is the no-slip part. Vectus in Uppsala, Sweden use linear motors in the track for their test track and it runs nicely in the winter. There are films on YouTube.

If you have the active part of the linear motor (with the coils) in the vehicle you don't have to have magnets in the track, short circuited coils are enough.

Dan said...

Thanks Bengt, I guess I kinda let this thread go dead. I just thought I would mention that the illustration for this post shows a design with a pretty big flaw. There is no way to bath the gears in oil without some insanely complex vessel with huge O-rings and bearings and all kinds of other garbage. The solution probably takes a picture to explain though, so I'll put it in a post.

Amelia Wilson said...
This comment has been removed by the author.
Amelia Wilson said...

This information is very useful. We can't stop innovating and improving in every field in our life. But it's always important to find an insurance company that you can trust sooner rather than later. Believe me, a good insurance policy will last and you'll certainly be thanking yourself later. If you are looking for the best insurance company I would recommend checking out Hadley Insurit Group. They offer home, business, life, and car insurance . I would highly recommend them for your insurance needs.