Why don't we use existing technology for fuel efficient vehicles?

September 12, 2005 10:48 AM

MefiEngineers: Shouldn't there be a simpler solution to hybrid vehicles?

I looked at a couple of hybrid vehicles in recent weeks, and since I'm sort of a "under-the-hood do-it-yourself" type guy, I spent some time talking with folks about just what all the stuff in there was for. I was astonished at the level of complexity these vehicles exhibit. My mental model of what an energy efficient vehicle should be is based more on the diesel-electric railroad locomotive. Before you fall out of your chairs laughing, consider: A railroad locomotive can move roughly one ton of freight one mile per *tablespoon* of fuel. That's pretty darn efficient. Performance doesn't have to suffer, either. A lightly loaded locomotive can accelerate pretty handily.

My thought experiment goes along these lines: A two cylinder diesel direct-driving an fairly large alternator at a constant, efficient speed. Wire the alternator to provide 440 volts or so. Use AC traction motors to provide power to each wheel. Forget all about mechanical drive trains. The higher voltages available with this system would allow smaller motors to power such things as air conditioners, power steering pumps, etc. Acceleration should be on the sports-car level. Each wheel would have its own motor, so forget about differentials and transaxles. Dynamic braking using back EMF is easy. The base technology for this has been around since the 1930's and has been improved manyfold since then.

What am I missing, folks? Is there some fundamental constant of which I'm just not aware? I can see no technical reason why this approach won't work. Controls are a solved problem for this technology. The motors are a solved problem, (except perhaps for scaling). Why not just scale down an existing, highly efficient technology?

posted by pjern to technology (20 comments total)

This is exactly what AC Propulsion does. There's a little tow-behind trailer for the tzero that operates as a generator.

I'm guessing there are patent issues here.
posted by trevyn at 10:55 AM on September 12, 2005

Surely converting diesel fuel to energy then putting that energy through another lossy process (generation then electric motor) doesn't really make more sense thn using the diesel motor to drive the wheels directly? Trains presumably need the electric conversion to achieve speed flexibility -- a clutch doesn't do well with 1000+ tons, I guess. Probably the fuel efficiency you quote comes more from the railroad being smooth than anything else.
posted by anadem at 11:11 AM on September 12, 2005

I am not an engineer, only someone with little fear of being told I'm a moron by people who know better.

But it seems to me that running the diesel through another lossy process might well make sense if:

(1) The system solipsist (if he exists) describes is less lossy than a standard transmission and differential combined with some sort of regenerative braking system. I dunno how likely that is, but remember that normal trannies are lossy too (this is one reason why racing cars use straight-cut gears instead of synchromesh/spiral ones).

(2) The system solipsist describes is, while technically lossier than a standard setup, is sufficiently lighter than a transmission that it takes less net energy to move the car.

(3) The diesel is sufficiently more efficient because it operates at a constant speed that it's more fuel-efficient in spite of a technically lossier setup.

(2) and (3) don't seem mutually exclusive to me.

I've though the same thing, but with turbines. I'd love to see F1 or another racing league adopt a turbine-electric powerplant.
posted by ROU_Xenophobe at 11:28 AM on September 12, 2005

Disclaimer: I dropped out of the mechinical engineering department to join the computer science when I was in college.

Would that mean that all automobiles would have be on rails?
Aren't trains a lot more expensive than cars?
How does friction compare (is it higher for rubber on asphalt)?

By those questions, I'm questioning whether "one ton of freight one mile per *tablespoon* of fuel" would scale into a small vehicle carrying people at highway speeds with requisite maneuverability.

OK, now beat me about the head with my dumb questions.
posted by clearlynuts at 11:43 AM on September 12, 2005

Wow. 256 miles/gallon for a diesel train pulling 1 ton? Impressive.

I suspect that current hybrid designs are partly the way they are because they can take an existing car design, plop in the hybrid guts and be more or less good to go. This is assuming, as I think I understand it, that a hybrid car uses a gas generator to power a single electric motor that drives a traditional drive train?

Also, I wonder about the feasibility of using 4 seperate motors (or just two maybe) to drive 4 (or 2) wheels. Would it be difficult to get the rate correct? If it wasn't, the car would turn one way or the other. When turning, would it be difficult to get the difference in rotation correct to keep the car on course with no slippage?
posted by RustyBrooks at 11:53 AM on September 12, 2005

What you propose is commonly called a "series hybrid." The primary reason for the complexity of most consumer hybrid cars has to do with the wide variety operating conditions for consumer driving. According to one manufacturer of hybrid drivetrain systems, series hybrids apparently get better efficiency in town, while parallel hybrids (in which the load is split between the internal combustion engine and the electric motor) get better efficiency for cruising. The approach taken by Honda is to create a "full hybrid" that can pick the most efficient mode depending on conditions.

Railroad locomotives don't have the same operating requirements as consumer cars, and can make use of a massive "battery" of kinetic energy once they get up to cruising speed.

on preview:
anadem: Actually electrical power generation is extremely efficient because you can run the internal combustion engine at a constant speed. The real killer for ICEs are rapid changes in engine speed. Turbines at a power plant, which run at the same speed for months at a time are even better. But electric vehicles have problems with power storage and range.

Really the big challenge with developing consumer vehicles is that for about three generations consumers have been willing to sacrifice fuel economy for certain kinds of performance.
posted by KirkJobSluder at 11:59 AM on September 12, 2005

Another consideration with these locomotive-style systems is that the engine (generator) can be tuned to run efficiently at exactly one speed, whereas engines in conventional cars and current hybrids need to run efficiently -- with predictable power curves, etc -- at a wide range of speeds. This limits their peak efficiency at any one speed, so even allowing for the additional losses, the overall system may turn out to be more efficient.

Part of the reason "strong hybrids" haven't taken off may simply be driver inertia (or the perception of same). A car driven by electric motors will have a significantly different power response than one driven by internal combustion. The car companies may feel that we are not ready for that.
posted by adamrice at 12:03 PM on September 12, 2005

One downside to this system that I can see is that having traction motors at the wheel hubs would greatly increase unsprung weight on the wheels, causing greater tire and road wear and poor handling. A solution to this would be mounting the motors to the car body and connecting them to the wheels via flexible driveshafts. I don't see getting the rate correct as being too much of a stumbling block with computerized control. That's basically how all-wheel-drive systems and stability control on modern cars work.

A mechanical version of this idea would be to use a continuously variable transmission (CVT), which a growing number of cars use, including some hybrids.

I agree with adamrice that drivers' expectations of how cars are "supposed" to work will limit this. People expect to hear a motor revving when accelerating, and feeling surges of power as the gearing is changed. In the CVT Wikipedia article I linked, it mentions that some CVT systems emulate this, which seems like a bit of a waste to me.
posted by zsazsa at 12:14 PM on September 12, 2005

Stop and go driving. Period.

Rail locomotives are designed for sustained, long hauls and would not be as efficient if they had to negotiate stop & go traffic. Also, they accelerate very slowly.
posted by Doohickie at 12:16 PM on September 12, 2005

Rail locomotives are designed for sustained, long hauls and would not be as efficient if they had to negotiate stop & go traffic. Also, they accelerate very slowly.

Well, that and once a train reaches crusing speed it's going to stay at or close to cruising speed without much intervention. For trains, momentum tends to trump most everything you put in its path, including wind resistance. In fact, a primary purpose of the locomotive is not just to pull the cars, but to control the speed of the cars. As a result, a lot of their efficiency comes from regenerative braking on downgrades.
posted by KirkJobSluder at 12:35 PM on September 12, 2005

It's just mechanics. Diesel engines are heavier than petrol engines. Electric motors (with enough power to drive vehicles) are very heavy. It's all about power to weight ratio. Build something as huge and heavy as a diesel-electric locomotive power-plant and you've got huge power but also massive weight. The engine weight doesn't matter so much, however, when hauling an even more hugely heavy train at constant speed on smooth rails.

Scale that same power plant down to a car and your power to weight ratio is unfavorable, your mass and momentum is larger than a regular car (so you have to build a beefeer car chassis, adding yet more weight), you're constantly accelerating and deccelerating that mass, electric motors are inefficient when run at continually changing speed, and all the while your rolling resistance (tires with hysteresis losses on a rough road surface, especially while having to support greater weight) is orders of magnitude greater. Very inefficient.

Diesel-electric is only efficient at large scale and relatively constant power output. Great for big, heavy trains on rails. Brilliant for oil-fired power stations that don't move at all. No good for small, lightweight cars on roads in traffic.

None of this is new technology. If it worked, it would have been done long ago.
posted by normy at 1:36 PM on September 12, 2005

IANA automotive engineer, but I have to disagree with Normy.

1. Electric cars have been around as long as cars have been around. The only real impediment to their use has been batteries, not the electric motors themselves. Furthermore, a lot of clever work is being done these days on light, compact hub-mounted motors (which would presumably increase unsprung weight, but also presumably, engineers consider this a solvable problem). Electric motors have perfectly flat torque curves and can accelerate very well.

2. Any extra weight of diesels doesn't seem to be an impediment to their use in tiny, efficient cars like the Audi A2. For that matter, a gasoline engine could be used in this scenario just as well.

Ultimately, I think it boils down to design overhead, corporate inertia, and the fear of consumer rejection.
posted by adamrice at 1:53 PM on September 12, 2005

Large trucks do use this system.

I once asked why cars didn't, and was told that the diesel generators and the rest of the system doesn't scale down well - it's too bulky for a car. I was quite young at the time, so there is every chance that was a "I don't really know but I'll make a wild guess and tell him that" answer :-)
NOT an authoritive answer! :)

Maybe try to make one in the backyard? That would be an intersting project.

Also, hybrids reclaim power during braking and going downhill, AC-drive loses all that energy, because there are no batteries to store it.
AC-drive with a battery system for regenerative braking would be an interesting idea to play with :-)
posted by -harlequin- at 1:53 PM on September 12, 2005

electric motors are inefficient when run at continually changing speed

This is quite wrong, in the sense that next to combustion engines, electrics are a dream at continually changing speed.

Many (most?) all-electric cars dispense with the clutch and gearbox of a normal car, becuse electric motors are so good over such a wide range, wheras combustion engines can't even move a car without a bulky, heavy, expensive system of gears and clutches, designed to keep the motor running at as constant a speed as possible - anything out of that safe zone the motor simply stops working.

You'll often find that this is why the speed of many all-electric cars tops out at a hundred or some other fairly low figure - when no gearbox is being used, the top speed is limited by how fast the spindle can actually turn.
posted by -harlequin- at 2:03 PM on September 12, 2005

Diesel locomotives do not use regenerative braking. They use dynamic braking. The drive motors are run "backwards" as generators, but the power is just disspated in a resistive mesh.
posted by yarmond at 2:14 PM on September 12, 2005

A railroad locomotive can move roughly one ton of freight one mile per *tablespoon* of fuel.

Side note: Unless this is talking about accelerating a stationary ton of freight into motion for a mile before stopping again, then this probably says more about the famously low friction of a train when a train is in motion, than about the efficiency of the motor that has to replace the power that friction is taking out of the equation.

In which case the answer to the question might be "because AC drive isn't as efficient as you're thinking it is"

So I'm really curious - does the tablespoon of fuel move the frieght, or does it merely keep freight already in motion moving at the same speed?

I'm guessing a bit of both - the figure is probably calculated from moving hundreds of tons between cities, in which case, since the train is only stopped at each end, the longer the journey, the more the tablespoon reflects only the cost of keeping the train in motion, rather than of moving freight.
posted by -harlequin- at 2:20 PM on September 12, 2005

(non-authoritive)
I forgot to add to the above question about whether Ac-drive is about efficiency: I was always under the impression that AC drive was used almost by necessity rather than because of energy efficiency - because it is very powerful across a wide range of loads and speeds, it makes it possible to build machines that would be difficult to make with a combustion engine. (Hence the various super-machines that need enormous amounts of grunt over variable loads and speeds use AC-drive, but very little else uses it).
posted by -harlequin- at 2:27 PM on September 12, 2005

Electric cars have been around as long as cars have been around. The only real impediment to their use has been batteries, not the electric motors themselves.

Agreed, for a purely electric vehicle, but I thought we're talking about a hybrid diesel-electric system using the diesel engine to power motors via an alternator.

diesels doesn't seem to be an impediment to their use in tiny, efficient cars

Also agreed, although, however you slice it a diesel engine will always be heavier than a petrol engine of the same power output at around small car sizes. Diesel engine and materials technology has improved to the point where the extra weight can be trumped by the better efficiency of a good diesel design. Insert a strong enough motor into that transmission and you've lost that gain and then some.

next to combustion engines, electrics are a dream at continually changing speed.

Also agreed, but I was discussing the inapplicability of comparison with a diesel-electric locomotive, as discussed in the original question, where the motor is running at constant(ish) speed. An otherwise similar motor at constant speed will always be more efficient than a motor at varying speed when integrating over some total angular displacement.

I also agree that there may yet be (probably will be) all sorts of interesting developments in electric motor, fuel cell and combustion engine technology that might make small-vehicle combustion-electric solutions viable. I very much doubt it will be anything with much resemblance to a diesel locomotive, however.
posted by normy at 2:45 PM on September 12, 2005

KirkJobSluder's point about parallel and series hybrid designs seems spot on. The advantage of being able to add the power of the electric motor and the IC engine together when maximum power is needed is obvious.

As for wheel mounted motors... Along with the unsprung weight issue already mentioned, I bet the peak breaking force of a regenerative scheme will be a lot lower than for traditional disc breaks. This would be a big safety issue for high speed cars, less so for city streets only.
posted by Chuckles at 4:14 PM on September 12, 2005

Another drawback for series hybrids is that both motors, gas and electric, must be sized to handle the vehicles maximum power requirements. Whereas a parallel hybrid can have two smaller motors which work together when the going gets tough, which saves a lot of weight. (the Prius gas motor is only 70hp. Adding the 44hp electric gives you a total size not much bigger than a regular engine for a car its size).

I think the more interesting train->car technology transfer would be the direct-injection two-stroke engines. You get the unbelievable power/weight ratio of a two stroke, with the pollution control of a four-stroke. Bombardier has started marketing 2-stroke DI outboard engines, and I'm anxious to see if their design could work in a car.
posted by Popular Ethics at 8:11 PM on September 12, 2005