An unconventional car: No engine, no transmission, no differential

Nov 18, 2013 by Marlene Cimons
An experimental four wheel independently actuated, lightweight electric vehicle. Credit: Junmin Wang, Ohio State University

This car has no engine, no transmission and no differential. It weighs half as much as a conventional car. Each of its four wheels has its own built-in electric battery-powered motor, meaning the car has the ability to make sharp turns and change direction very quickly.

Without an exceptional traction and motion control system, however, this car would be quite difficult to drive, providing a driving experience vastly different from anything else on the road, and almost certainly more dangerous.

This is where Junmin Wang's expertise comes in.

Wang, assistant professor of mechanical engineering at Ohio State University, and his team are designing algorithms for the vehicle's onboard computer that will calculate and ensure motion control to keep the car stable and operating smoothly. The system, which receives and analyzes input data 100 times per second from the , the gas pedal and brake, works out how each wheel should respond.

"Without it, the car is quite difficult to drive because the wheels are not coordinated," says the National Science Foundation (NSF)-funded researcher, who also directs the university's vehicle systems and control laboratory. "You feel like you are driving something uncontrollable. You could flip over, or travel along an undesired path, or cause a crash. But when the 'controller' is active, based on feedback loops, the vehicle motion can be controlled, just as the driver expects."

With a safe and reliable control system, this new electric vehicle ultimately should make the perfect in-city car. It's efficient and maneuverable—and has no emissions. Because it is all electric, "you can use wind power or solar power, and contribute toward reducing our dependency on fossil oils," Wang says.

The computer calculates exactly how much torque the car needs for each of its four wheels. Moreover, because each wheel is independent, "one wheel can be doing the braking, while another is doing the driving," Wang says. "The computer gets signals from the driver from the steering wheel and pedal positions, then calculates the desired speed, or vehicle motion, based on a mathematical model."

Wang's work on the car began in 2009 with a grant from the Office of Naval Research Young Investigator Program. In February 2012 he received an NSF Faculty Early Career Development (CAREER) award, which supports junior faculty who exemplify the role of teacher-scholars through outstanding research, excellent education, and the integration of education and research within the context of the mission of their organization. He is receiving $400,000 over five years.

As part of the grant's educational component, Wang's lab hosted a summer program for high school students where, among other things, the teenagers disassembled and reassembled radio-controlled toy electric cars to increase their understanding of their mechanics.

Additionally, students from the Columbus Metro School, a public STEM (science, technology, engineering, math) high school open to students from around the state, participated in research internships on the experimental car in Wang's lab.

Wang's research also receives funding from the Honda-Ohio State University Partnership Program and the Ohio State University Transportation Research Endowment Program.

The experimental car weighs only about 800 kg., or a little more than 1,750 pounds, which makes it energy efficient. The researchers retrofitted a commercially available utility terrain vehicle chassis and removed the engine, transmission and differential, then added a 7.5 kW electric motor to each wheel and a 15 kW lithium-ion battery pack. A single electrical cable connects the motors to a central computer. This type of car design, where each wheel has its own individual motor, is known as "four wheel independently actuated."

The researchers tested the car and its controller on normal road conditions at the Transportation Research Center in East Liberty, Ohio, an independent automotive site for vehicle crash, emissions and durability testing. On roads with good conditions, the car followed a driver's "desired" path within four inches.

To see how it performs on slippery roads, they brought the car to an empty west campus parking lot on a snowy day. The car maneuvered with an accuracy of up to eight inches, and the vehicle traction and system prevented "fishtailing" through independent control of the left and right sides of the car.

The researchers, including doctoral student Rongrong Wang, described the car's ability to follow a specific trajectory in a paper published in January 2013 in the journal Control Engineering Practice.

Wang can't yet estimate the mileage for a single charge, since the car only has been driven during experimental testing. But he says the car provides "about 8 to 10 hours of driving on a single charge, although not continuously."

Wang thinks it will take another five- to- 10 years before the car is ready for commercial use. The researchers still must fine-tune the computer algorithms and add more safety features. Wang says it is difficult to compare their test results to a conventional car, since the latter's maneuverability is limited by the transmission and differential systems that link the wheels together mechanically.

Nevertheless, he predicts that, ultimately, the research will produce an electric that will be clean, fuel-efficient and "handle better than typical conventional cars," he says.

Explore further: Volkswagen XL1: 'World's most efficient car' makes its US debut

Related Stories

Recommended for you

Student develops filter for clean water around the world

1 hour ago

Roughly 780 million people around the world have no access to clean drinking water. According to the World Health Organization (WHO), 3.4 million people die from water-related diseases every year. ETH student Jeremy Nussbaumer ...

Minimising drag to maximise results

5 hours ago

One of the most exciting parts of the Tour de France for spectators is the tactical vying for spots in the breakaway group at the front of the pack.

User comments : 22

Adjust slider to filter visible comments by rank

Display comments: newest first

Eikka
1.6 / 5 (16) Nov 18, 2013
What is the fundamental problem they're trying to solve with the design? Why does it need computer controlled independent actuation on all wheels?

There are means to connect electric motors in series such that the rear wheels for example will automatically act as if connected to a differential. If one slows down the other speeds up because the back-EMF from the slower motor diminishes and that re-distributes the voltage, thus power, between the two motors in the same way that a differential gear would.

PPihkala
5 / 5 (8) Nov 18, 2013
Eikka: Independent control of each wheel is needed for stability control, which currently is managed by ABS brakes in ICE cars. And turning on dime needs the wheels on left and right to rotate in different direction, something quite unavailable in mainstream cars.
VendicarE
3.1 / 5 (7) Nov 18, 2013
"What is the fundamental problem they're trying to solve with the design?" - Eikka

The question is clearly implied by the first sentence.

"This car has no engine, no transmission and no differential. It weighs half as much as a conventional car" - Eikka

You seem smart enough to realize the implication of lighter designs with respect to fuel consumption efficiency.

I'm sure you can figure the rest out.

In addition, this car will be able to drive sideways (or almost so) making parallel parking much easier to achieve.

The cost is greater computational complexity, and a loss of mechanical redundancy, which I think is probably a bad idea.
krundoloss
1.9 / 5 (13) Nov 18, 2013
Yeah, PML built an Hybrid Electric Mini Cooper, with the wheels as the motors, in 2006, with over 640hp. It had a combined range of 932 miles. Maybe the creator of this electric car should have called them for information? What they created was vastly superior to this "experiment", and they did it 7 years ago. Keep trying guys!

By the way, I absolutely LOVE the idea of wheels as motors. Powerful, efficient, light, convenient, simple, just awesome!

http://www.treehu...els.html
Eikka
1.3 / 5 (14) Nov 18, 2013
Eikka: Independent control of each wheel is needed for stability control, which currently is managed by ABS brakes in ICE cars. And turning on dime needs the wheels on left and right to rotate in different direction, something quite unavailable in mainstream cars.


I repeat the question, what is the fundamental problem they're trying to solve with this design?

You seem smart enough to realize the implication of lighter designs with respect to fuel consumption efficiency.


The lighter design is more thanks to using a utility buggy as a chassis than any independent wheel design. The in-wheel motors have their own drawbacks in terms of more unsuspended weight and lower efficiency in producing torque.

In addition, this car will be able to drive sideways (or almost so) making parallel parking much easier to achieve


Not unless all wheels are also independently steering, which they aren't. They're talking about just torque vectoring.
Mike_Massen
1.4 / 5 (11) Nov 18, 2013
Surely Eikka has this back to front because the implication is electric drive is better than a non-slip differential but Eikka's suggestion is not helpful to drive out of a bog !
"If one slows down the other speeds up because the back-EMF from the slower motor diminishes and that re-distributes the voltage, thus power, between the two motors in the same way that a differential gear would."
Surely this is in no way an improvement on a classic differential ?

Wouldn't it make more sense to apply more torque to the wheel that is turning slower because surely that is the one that is slowed down *because* it hasnt enough torque to push the car.

ie. If a wheel gets bogged and it turns easily it will spin whilst the other one is slow or stationary, so if bogged you want the slow wheel to turn instead of stay still while the other spins - and this would be a worthwhile aim for electric drive to *improve* on mechanical differential and not merely copy it ?

Cheers

Eikka
1 / 5 (11) Nov 18, 2013
Surely Eikka has this back to front because the implication is electric drive is better than a non-slip differential!

Surely this is in no way an improvement on a classic differential ?


It works quite analogously to a classic differential.

Wouldn't it make more sense to apply more torque to the wheel that is turning slower


No, because if you did that in a corner the car would try to understeer and push off the road. The point of the differential is to allow the wheels to turn at different rates by redistributing power.

The drawback is that if one wheel slips, it gets transferred more power and slips more. That problem is solved by locking the differential, which in a series motor configuration could be accomplished by partially shorting out the slipping motor to divert more power to the opposite wheel.
Mike_Massen
1.4 / 5 (11) Nov 18, 2013
Eikka countered with
No, because if you did that in a corner the car would try to understeed and push off the road. The point of the differential is to allow the wheels to turn at different rates by redistributing power.
We all know what a differential does but did you notice I was offering a view of a non-slip diff, why in the heck would you want that if bogged - recall I said that.

In your paradigm the electric motor suggestion makes things worse.

In a bogged situation as I described you surely *want* the slower spinning wheel to have more torque ?

The better way to do it is by adding more power to the slowed motor and reducing the power to the easily spinning motor, no shorting or locking required - its in the algorithm applying power to each wheel...
Eikka
1 / 5 (11) Nov 18, 2013
In your paradigm the electric motor suggestion makes things worse


It behaves no differently from a regular mechanical differential.

A locking mechanism can be made as I described: short out the offending motor and all the power and torque is transferred to the other motor, or switch the motors in parallel, in which case they tend to equalize torque by transferring more current to the slower motor. A simple analog circuitry can emulate the function of a non-slip differential here by either switching the motors or by shorting out the faster motor in pulses.

But as you see, the point here is that you don't need a fancy computer controlled system to achieve stability.
Mike_Massen
1.4 / 5 (11) Nov 18, 2013
Eikka doesnt get it with
"..It behaves no differently from a regular mechanical differential.

.. short out the offending motor and all the power and torque is transferred to the other motor, or switch the motors in parallel, in which case they tend to equalize torque by transferring more current to the slower motor.

..you don't need a fancy computer controlled system to achieve stability. "
You don't have to do that. Shorting or switching motors around is so primitive, rough & a waste *because* you already have the the PWM drive for each wheel to provide speed/power control *and* motor controllers already have micros & software - its not 'fancy', its a given state of the art.

All you need is a simple equation implemented in the motor controller to treat a bogged situation differently from that of turning a corner - which you would have to correct for anyway as some suspensions are implicit under/over-steer depending on design - correction via individual wheel now expected !
Eikka
1 / 5 (11) Nov 18, 2013
Shorting or switching motors around is so primitive


It's an elegant solution to a problem.

Normal cars don't usually need a limited slip differential anyways. The existing systems that brake the slipping wheel are enough to keep them going. That way the technology would be a drop-in replacement for a mechanical differential drive, which would save enormously on cost.

*because* you already have the the PWM drive for each wheel


But you wouldn't have to have a PWM drive for each wheel. You could have a single set of control electronics for both wheels. The speed controller can be completely dumb because the motors themselves distribute the power. The motors too can be simple - you don't need fancy BLDC motors with integrated controllers because the system works with AC motors just as well.

Saves on cost and complexity again.

All you need is a simple equation implemented in the motor controller


Or a couple transistors in a box. Same difference.
24volts
1 / 5 (10) Nov 19, 2013
If you are using the 3 or more phase motors the controller gets a bit serious but if your using straight dc motors it's pretty simple to do. I've done a couple of different setups on recumbent trikes where each front wheel had it's own hubdrive motor. All I had to do was use a separate controller for each motor and tied both controllers to one throttle. I used a small pot in one of the controllers sensing line input to match the motors so they would pull straight down the road. Once done it's done. Going around a corner isn't a problem as the outside motor automatically speeds up in the corner. Both trikes worked perfectly fine that way. Using 4 motors would have been just as easy. Straight dc internally geared hub motors are not quite as efficient as the 3+ phase versions but they definitely make the electronics setup simpler.
antialias_physorg
not rated yet Nov 19, 2013
Why is ainone even discussing the bogged scenario. This is for city vehicles (as expressly stated in the article). When was the last time you got stuck with a wheel in the city (or heard of anyone). Lemme guess: never.

At 800kg the thing is so light (same as the roadster I'm driving) that you could pick up the problematic corner and heave it out of the way if you'd have to.

At city speeds the weight distribution in the wheels also has no real impact.
ubavontuba
1 / 5 (10) Nov 19, 2013
What is the fundamental problem they're trying to solve with the design? Why does it need computer controlled independent actuation on all wheels?
I suppose the wheel motors might provide regenerative braking, without the need for additional weight, but I don't know if this would be a particularly efficient design.

But unless it's designed for AWD traction purposes, I don't see the need for more than one or two motors serving just one axle (or two individual wheels) either. Why complicate things?

c0y0te
not rated yet Nov 19, 2013
Hm.... It's already there in production for couple of years now.
http://www.rimac-...toring-7
grondilu
not rated yet Nov 19, 2013
It's a bit scary, though. I mean, with a normal car, if suddenly the electronics or the engines shuts down, at least the wheel stay parallel, and the steering is still functional. So the car keeps going where you want to go and you can brake until all the kinetic energy is dissipated.

With a car like this, if the electronics has a malfunction, the wheels can suddenly stop being parallel, act in a chaotic manner and basically at high speed the car can rollover or something.

I know a lot of redundancy can be added in the electronics, but it's always reassuring when the mechanics is passively stable.

Makes me think of shopping carts: they have independent wheels and yet they are passively stable. Couldn't such a design be used?
Eikka
1 / 5 (10) Nov 19, 2013
Why is ainone even discussing the bogged scenario. This is for city vehicles (as expressly stated in the article). When was the last time you got stuck with a wheel in the city (or heard of anyone). Lemme guess: never.


Precisely my point when I asked what is the fundamental problem this design is trying to solve.

ESC is useful to get you moving if you have slippery conditions like ice on cobblestones and all you're getting is a bunch of wheelspin, but at the same time it's dangerous because it won't make you stop any better.

Makes me think of shopping carts: they have independent wheels and yet they are passively stable. Couldn't such a design be used?


You've just been very lucky with shopping carts. There's always one corner that catches and pulls you aside.
antialias_physorg
not rated yet Nov 19, 2013
but I don't know if this would be a particularly efficient design.

At the very least it would give rise to radically new interior design possibilities with gearbox and motors out of the way (and battery packs basically shapeable and placeable any way you want to.). Should also improve access to serviceable parts (not that DC motors need much servicing). Cars/car design could be a lot more modular with wheel motors - too.

if suddenly the electronics or the engines shuts down, at least the wheel stay parallel, and the steering is still functional.

True. but I have actually never had that happen (nor have ever met anyone or heard of anyone to whom that has happened). Catastrophic scenarios should be investigated - but we should also look at how likely they are. And given the simplicity of the electronics it's not hard to build in redundancy.

fortranfixer
not rated yet Nov 23, 2013
@krundoloss PML became Protean: www.proteanelectric.com
EnricM
1 / 5 (7) Nov 25, 2013
Eikka: Independent control of each wheel is needed for stability control, which currently is managed by ABS brakes in ICE cars. And turning on dime needs the wheels on left and right to rotate in different direction, something quite unavailable in mainstream cars.


And I see the applications: perfect self parking for conventional cars for instance. And not all vehicles are cars. The first thing I thought about was a fork lifter, the capabilities of such a vehicle would be perfect for that.
Eikka
1 / 5 (6) Nov 27, 2013
if suddenly the electronics or the engines shuts down, at least the wheel stay parallel, and the steering is still functional.


True. but I have actually never had that happen (nor have ever met anyone or heard of anyone to whom that has happened).


Then again, you've never seen a completely drive-by-wire car because no such things exist on the market. There have been cases like the Toyota stuck throttle issue, and I bet none of those Teslas that caught fire had any power to the motors when the main battery went poof.

turning on dime needs the wheels on left and right to rotate in different direction, something quite unavailable in mainstream cars.


And I see the applications: perfect self parking for conventional cars for instance


Cars don't do that because they'd chew the tires off the rims by dragging them sideways. You also need four corner independent steering to make it happen nicely, but that isn't a part of the design.
Eikka
1.3 / 5 (7) Nov 27, 2013
The first thing I thought about was a fork lifter, the capabilities of such a vehicle would be perfect for that.


They already have a much better solution for that, which is already in production: http://www.youtub...9TReI1aQ

It's called a mecanum wheel: http://www.youtub...o16KKm8Q