Electric Car Shortcut

[paultrafalgar]

Rationale for Electric Car Development:
http://www.youtube.com/watch?v=Dag56PuC8DE
and a patented in-wheel motor:
http://www.e-traction.com/TheWheel.htm

 

[njs]

A gear box losing 50% of mechanical energy? LOL

 

[paultrafalgar]

Let's expand a little the part of the second reference I gave which (I think) gave rise to your comment:
"
Conventional electric vehicles customarily make use of the same energy transfer systems found in vehicles with fossil fuel engines, which utilizes multiple cogwheels and bearings embedded in gearboxes, differentials, and shafts to power the wheels. Together these conventional components represent a formidable resistance usually absorbing roughly 50% of the energy input under nominal driving conditions.
"
The figure IIRC for the overall efficiency of conventional vehicle transmissions is 15%, that is, put in 100Kwatt and get out 15Kwatts at the road. So I think 50% is a bit optimistic. That's why the In-wheel motor is such a breakthrough. Don't scoff!

 

[swbluto]

The figure IIRC for the overall efficiency of conventional vehicle transmissions is 15%, that is, put in 100Kwatt and get out 15Kwatts at the road.

(Do you mean 85%? Or are you referring to the average system efficiency (Including gasoline engine efficiency losses)?

 

[paultrafalgar]

The figure IIRC for the overall efficiency of conventional vehicle transmissions is 15%, that is, put in 100Kwatt and get out 15Kwatts at the road.

(Do you mean 85%? Or are you referring to the average system efficiency (Including gasoline engine efficiency losses)?

Not 85%, 15% and yes IIRC overall losses. As I said, put in 100Kwatts get out 15Kwatts. That's what makes sense to compare with an in-wheel motor.

 

[liveforphysics]

In the racing world, we measure the difference between the engine dyno readings, and the power delivered at the wheels when the engine is in the chassis.

A FWD manual gear box generally has a difference of 8-12% between the power it lays at the wheels on the chassis dyno rollers vs a direct measurement made at the crankshaft on an engine dyno.

A RWD manual gear box generally has 12-18% loss.

An AWD manual has always been >20% in my experience, and AWD is extremely poorly suited for racing, so I don't have much experience with them.

Something interesting is the tire loss alone. It can be a heck of a lot higher than any rolling resistance calculators would lead you to think. SpeedFactory (my favorite dyno shop) has a Dyna-pack type dyno and DynaJet type dyno. One mounts the car by removing the wheels and bolting the hub of the car directly to the dyno brake, and one works by strapping the car in place with the wheels on rollers. Tire differences can make a 3-6% difference alone in the readings! The dragslicks we have to run to get the needed traction to apply the power we make have 6% losses! This explains why they go so damn hot when you're running around on them on the street, not even spinning them, just driving on them at freeway speeds and they get smoking hot. Street tires have quite a bit less loss than slicks, but it's still substantial, and a hubmotor still must deal with this loss.

I'm curious where the hubmotor guys like to figure in the roughly 1-2% electrical/mechanical efficiency that is inherent in a non-geared hubmotor at crawling speeds like 1-20rpm.

 

[swbluto]

The figure IIRC for the overall efficiency of conventional vehicle transmissions is 15%, that is, put in 100Kwatt and get out 15Kwatts at the road.

(Do you mean 85%? Or are you referring to the average system efficiency (Including gasoline engine efficiency losses)?

Not 85%, 15% and yes IIRC overall losses. As I said, put in 100Kwatts get out 15Kwatts. That's what makes sense to compare with an in-wheel motor.

When you say "put in", you mean strictly the transmission, right? Which doesn't include the thermal energy losses of burning gasoline nor the efficiency of the motor.

If you included the efficiency of the motor, I would find that believable.

 

[liveforphysics]

Of course. I'm talking the difference between horsepower at the crankshaft vs horsepower at the wheels. AKA, the difference between BHP (brake horse power) and WHP (wheel horse power).

 

[liveforphysics]

Gasoline is roughly 12.7kwhr/Kg.

A proper NA gasoline racing engine can do 0.26lbs/hp-hr BSFC in a normal performance tune. Which would be 0.159Kg/Kw-hr.

That would mean 6.27kwhr-kg.

Let's throw it in a FWD car with a manual tranny. at 10% engine crankshaft to wheel drivetrain efficiency.

This would yield a total system efficiency from gasoline to mechanical energy at the surface of the tire of 44%.

Sadly, the engines in passenger cars, even my Insight, have terrible efficiency compared to proper racing engines that use extremely high compression, extreme measures to reduce friction, and extremely minimal pumping losses, which all make for very high efficiency. Typical production engines are lucky to get 0.32-0.36lbs/hp-hr, which in a manual tranny FWD manual tranny car would make the total system efficiency ~35%.


To get 15%, you would need to be looking at some 1960's cars, or some rubbish made by the big 3 (GM, FoMoCo, Mopar).

 

[jag]

and a patented in-wheel motor:
http://www.e-traction.com/TheWheel.htm

I think the future promise of in-hub motors for cars and heavier vehicles is great. However technology might not yet be there. Looking at the .pdf spec sheet of this company's car-size hub, it has a peak torque/weight unit of only half that I get in my 9C bike hub...

The Mitsubishi iMiev originally used hub motors at a prototype stage, but the hub motors were dropped for the released production version.

The e-traction company seems more focused on heavy vehicles though. There is an example of a forklift. I would imagine large unsprung weight doesn't matter so much on a flat factory concrete floor or storage yard.

B.t.w. Looks like their web site is made by PR people skilled at obfuscating facts. Of course ES readers aren't their target audience. It is more likely aimed at the politicians and/or whoever provides their funding.

 

[swbluto]

Typical production engines are lucky to get 0.32-0.36lbs/hp-hr, which in a manual tranny FWD manual tranny car would make the total system efficiency ~35%.

Sounds true. I've heard that efficiency is usually sustainably achievable under high power like highway driving. At lower power, it seems they're lucky to eke out 15-25% as most of the energy goes towards heating the relatively bulky engine designed for higher max power. I think I calculated the typical energy efficiency of an "average car" driven by an American to be about 19% in one example.

 

[TylerDurden]

Expect to see individual pancake motors on the inboard end of halfshafts... like Tesla's inboard disc brakes. No unsprung weight and more out of harm's way.

No new tech required.

 

[liveforphysics]

Expect to see individual pancake motors on the inboard end of halfshafts... like Tesla's inboard disc brakes. No unsprung weight and more out of harm's way.

No new tech required.

As a guy who has exploded perhaps 40-50 axles and driveshafts over a 10 year period, the idea of in-board brakes seems a little spooky... thank God its only on the rear brakes of the Tesla. Its really tough to even tell if a sportscar has rear brakes hooked up or not when driving, so I don't see anyone wrecking over a blown axle removing the rear brake function on a Tesla (unless they have 2-way LSD? Anyone know?)


I like the simplicity of an in-board pancake, but anything direct drive at wheel speed is inherently going to have the low power density and awful low speed efficiency of an ungeared motor at low RPM. Having something like a little 10:1 gearbox somewhere in the picture just about makes for a motor that can have 10x the power density of a direct drive setup.

 

[TylerDurden]

I like the simplicity of an in-board pancake, but anything direct drive at wheel speed is inherently going to have the low power density and awful low speed efficiency of an ungeared motor at low RPM. Having something like a little 10:1 gearbox somewhere in the picture just about makes for a motor that can have 10x the power density of a direct drive setup.

Agreed.

Just thinking 'shortcuts'.

 

[liveforphysics]

Typical production engines are lucky to get 0.32-0.36lbs/hp-hr, which in a manual tranny FWD manual tranny car would make the total system efficiency ~35%.

Sounds true. I've heard that efficiency is usually sustainably achievable under high power like highway driving. At lower power, it seems they're lucky to eke out 15-25% as most of the energy goes towards heating the relatively bulky engine designed for higher max power. I think I calculated the typical energy efficiency of an "average car" driven by an American to be about 19% in one example.

Yep, that's efficiency at WOT of course. It's lower for partial throttle.

In the case of a hubmotor though, its also got its own efficiency limitations. It may have some speed where its 70-80% efficient, but at near zero RPM, its of course near zero percent efficient, and te curve doesn't climb up that quickly. If its going to be optimized for freeways speeds, then its around town or in-traffic efficiency is also going to suffer badly. You should work the numbers out on it, but I bet due to no gearing, it takes a much harder efficiency hit for low speed in-traffic driving conditions than an ICE vehicle takes.

 

[jag]

In the case of a hubmotor though, its also got its own efficiency limitations. It may have some speed where its 70-80% efficient, but at near zero RPM, its of course near zero percent efficient, and te curve doesn't climb up that quickly. If its going to be optimized for freeways speeds, then its around town or in-traffic efficiency is also going to suffer badly. You should work the numbers out on it, but I bet due to no gearing, it takes a much harder efficiency hit for low speed in-traffic driving conditions than an ICE vehicle takes.

Ebikes.ca's simulator shows the efficiency curve for hub motors. For a car size hub the shape of the curve will be similar, but the units (torque/speeds) higher.
http://www.ebike.ca/simulator/

I think a small, lightweight city vehicle is a much more reasonable target for a hub-motor based car design. Likely some of the hub motors already in production for scooters and motorbikes could be used. Then one can optimize efficiency for typical city driving at speeds 30-50km/h instead of having to deal with also highway speeds.

 

[paultrafalgar]

Guys, we're getting bogged down in the details of efficiency here. The title of the thread and the subject of the video I posted is that small operators can accelerate the jump to electric vehicles via the in-wheel motor route. Much easier for them than designing a Prius hybrid vehicle. The point of posting the in-wheel manufacturer's link was to say (as it were to the small guys) that such technology is available for them to adopt and to use to adapt current vehicles. Do you agree with that?

 

[Malcolm]

such technology is available for them to adopt and to use to adapt current vehicles. Do you agree with that?

E-traction, as already mentioned, seems to be aiming at heavy-duty vehicles and I can't find any pictures of actual products on their site.

The only in-wheel motors I know of that are readily available now are the motorcycle hub motor from Enertrac and various scooter-type hub motors from suppliers such as Kelly. The most promising contender for larger motors seemed to be PML Flightlink, which then turned into HiPa Drive, but their website is currently down.

Unless there are any other suppliers I've missed this means you're limited to building buses or very lightweight (sub-500 kg) vehicles.

 

[paultrafalgar]

How about a project here, called the In-Wheel Motor Project. Take the commonest size of wheel and make a prototype in-wheel motor from it. Recumpence, Thud, you mechanical geniuses, what do you think? It wouldn't hurt to thrash out some CAD, would it - TD, Miles?

 

[Malcolm]

How about a project here, called the In-Wheel Motor Project

There's a long thread following this idea here: http://www.diyelectriccar.com/forums/showthread.php/open-source-hub-motor-wheel-motor-14185.html?highlight=wheel+motor
but they don't seem to have got very far.

I got the impression that the geniuses (genii?) are more interested in using lighter, faster spinning motors and gearing down for maximum power and efficiency. To use a motor like this in a motorised wheel you would need to incorporate a gear reduction, which means more complexity.

I like the elegance of the wheel motor and the possibilities it offers for changing layout, I just don't see how it can be done without considerable compromises on handling, motor dimensions, rotational inertia etc. It would be relatively easy to put a couple of pancake motors in the place of the diff and drive the half shafts independently, without requiring much other modification on most cars. To improve (electrical) efficiency you could add a planetary gear reduction between each motor and (shortened) half shaft. The resulting drive system wouldn't be as compact as a wheel motor system, but it would probably be more reliable and easier to service by the average mechanic. A sophisticated wheel motor assembly is likely to be very complex and require specialist tools and knowledge to fix.

 

[paultrafalgar]

Not qualified to critique your analysis Malcolm, but people are making in-wheel motors. In searching for a link to the electric Mini that was supposed to have them I came across this for example:
http://www.elec.eng.osaka-cu.ac.jp/~aevc/823-827.pdf
Edit:
see also:
http://www.driving.ca/ottawa/Electric+cars+make+inroads+Detroit/2433391/story.html

 

[paultrafalgar]

How about this, from a UK company:
http://www.lightningcarcompany.co.uk/hipadrive.php

 

[paultrafalgar]

The Electric Mini does not have in-wheel motors after all, but you might find this interesting:
http://www.stadtdatenbank.de/bundesland_bayern/by-muenchen/mini-e-switched-on-l32868.html

 

[Malcolm]

Sorry Paul, I didn't mean to sound so negative about wheel motors. I'm sure they will become available in a few years' time, but they won't be simple or cheap. It will require exotic materials to keep weight down and specially developed motors (ironless?). I did hear of a few people contacting PML Flightlink (which supplied the motors for the electric Mini I think you're referring to). The price they were quoting at the time was something like £20,000 per wheel... Having said that, maybe we could get a team together. I'm handy with a grinder and I can make a good pot of tea

 

[swbluto]

It seems a smaller wheel would reduce the need for gearing reduction, but it seems a larger wheel would be favored by in-wheel technology since the area is squared to the radius (And thus continuous current and torque limits, which can be thought of as corresponding to power capability) while the thrust is inversely linear to the radius.

So, I'd imagine it might be a practical with a large enough wheel for an "in-wheel" motor. Wider wheels wouldn't hurt, I'd imagine.