EV transmissions and Popular Science

[Chalo]

If I'm not mistaken, the power electronics required to render mechanical transmissions irrelevant also incur cost, weight, volume, and inefficiency-- just like mechanical transmissions. I don't doubt that either solution could be optimal depending on application and available resources, just as an e-bike direct drive hubmotor is sometimes best and a crank drive is sometimes best, depending on the specific situation.

 

[Punx0r]

Torque increases with the square of the airgap radius, so an increase of 3.16x radius (or diameter) should yield 10x the torque, if I have added up properly...

..with no change to the electrical load , amps needed , kW etc ??

Think of it in terms of an electromagnet fitted to the end of a lever arm of varying length Also, Watts is a measure of power but torque alone is only a (rotational) force, no power is consumed until things start moving apart from the resistive losses in the coil (since it's not super-conducting). Luke posted a great example once of the power his bike required to make peak torque at stall to get of the line - it was a handful of amps battery side.

 

[Hillhater]

I get the torque/ power /rpm thing, but I are you saying you could get this 10x torque with no major implications to the drive controls/inverter etc ?

 

[liveforphysics]

I get the torque/ power /rpm thing, but I are you saying you could get this 10x torque with no major implications to the drive controls/inverter etc ?

If done through magnetic geometry alone, it should indeed be possible to suffer no power draw or controller penalty VS any other system with the same final drive torque and power delivered through gearing.

If you want to get fussy about it, you should be capable of a tiny bit smaller controller and battery for the same range and performance desired because some small amount less waste is going into rubbing gear faces together.

 

[motomoto]

LFP,
I know you are the MAN when it comes to this stuff and you love the one speed direct approach. I was checking out the specs on the Formula E cars the other day. They have 5 speed transmissions
and run at 800 volts. These cars should represent the state of the art in electric drive systems. You seem to preach the opposite of gears and high voltage. The motors are 26 kgs and make 270 hp.
Please explain.

 

[liveforphysics]

LFP,
I know you are the MAN when it comes to this stuff and you love the one speed direct approach. I was checking out the specs on the Formula E cars the other day. They have 5 speed transmissions
and run at 800 volts. These cars should represent the state of the art in electric drive systems. You seem to preach the opposite of gears and high voltage. The motors are 26 kgs and make 270 hp.
Please explain.

Easy, they didn't want high performance vehicles, they wanted shifting noises for spectator acceptance. More weight is in the lossy transmission and drive system than in the motor, had they put that weight in the motor and gone direct drive they would be high performance vehicles.

High performance alternative EV drivetrain, direct drive YASA's:

http://www.yasamotors.com/the-electric-car-with-an-extra-kick/

That's what 850hp direct drive looks like, it does 204mph and the whole car is sub 1 metric ton. If they had that much power in Formula E they would only do 1/4 the laps per tiny 28kWh battery they have maternally mounted.

Formula E (while still cooler than gas racing) is some very harshly rule restricted watered down ultra mild EV builds. They crippled the cars with a tiny permitted battery volume and weight and restricted motor power and designed rules to ensure they wouldn't impact Formula1 (which you will notice is predominately sponsored by petroleum industry). On top of the harsh performance crippling restrictions, they were then selected drivetrains disregarding vehicle performance in favor of having familiarity to fans of hearing gears banging.

They already have seen what happened to racing motorcycles in such a short period of time of development achieving ~70 years of gas vehicle improvement in laptime reduction occur in 5 years. Even with the absurd performance crippling rule restrictions currently in place, it is still inevitable to eventually defeat the best ICE laptimes likely before 2020 if they were to not run it as a locked down spec-class but instead encourage free development innovation.

 

[Arlo1]

Here is an example of someone who ditched the tranny as well.
Even with gas or electric!!!! NO TRANNY. I hate trannies I like things strait
[youtube]sr1jylROsqQ[/youtube]

 

[flathill]

Yeah but if it had two speeds it could go much faster than 249mph

Way too slow. Next

Haha

Yasa motors and Croatian batteries for the win!

The answer is a one speed direct drive will always have the highest average efficiency but a two speed will always have higher performance (in theory)

But if you already have enough power to exceed the traction limit then 1=2 performance wise but two adds complexity and reduces overall efficiency on a normal drive or race cycle

Better yet just make it all wheel drive and have two different reduction ratios or two different winds with direct drive motors. Widens efficiency window and increases top and low speed accel

---
The batteries and power distribution unit were developed in conjunction with Rimac, makers of Croatia's only (why aren't there more Croatia? Step your game up) electric supercar. The motors were made by YASA, which makes some beautiful pieces of electrical engineering porn. The system makes a total of nearly 700 horsepower, which Koenigsegg says makes it the most powerful of any electric setup in production car history
-----
Must be using saft or next gen a123 to make that much power from less than 10kwh

The Regera uses Koenigsegg's 5.0 liter turbo V8 along with three electric motors. They are powered by a nine-kWh, 620-volt, water-cooled battery pack, which Koenigsegg says is the most power dense battery pack every created for a production car.

 

[flathill]

Its lithium metal polymer

 

[Punx0r]

but a two speed will always have higher performance (in theory)

There have been some interesting power-out graphs posted recently (sorry, I can't remember by whom) comparing different gear ratios (say, 5:1 Vs 10:1) for a given motor in a car. The lower ratio does show a performance advantage, but the speed range at which it does so is surprisingly small, like 0-10mph. After that, the higher gear outperforms it. I found this an eye-opener.

So you get a marginal power benefit, but one you almost certainly can't take advantage of due to the limits available traction from the tyres. That's also assuming a near-zero shifting time between gears. All that's before you consider the added weight and complexity of multiple transmissions. For an average car it just doesn't make sense. The discussion got to the stage of a farm tractor using an undersized motor before multiple ratios showed an actual advantage.

 

[major]

They already have seen what happened to racing motorcycles in such a short period of time of development achieving ~70 years of gas vehicle improvement in laptime reduction occur in 5 years.

Good point and we did that with a single ratio, well above unity.

 

[flathill]

When comparing a 1 speed vs a 2 speed EV design you have to pick some metric to match. Here you can see what happens when you pick the gear ratios such that the 0-100km/h time is the same. You can see with a UDDS driver cycle the single speed car goes about 12 miles farther in this case. This paper is pretty comprehensive as it takes into account shift time and adhesion limits, but I'm sure you will find some faults to pick at.

So in summary the answer is it depends. If you input all these formulas into matlab you will find the key is motor torque to vehicle mass ratio. For a performance car a single speed is probably optimal, while for a cheap car with an undersized motor more than one speed is probably optimal. A semitractor can be consider a car with an undersized motor which is why it needs gears.


Parameter Units Transmission
One-speed/ Two-speed
Optimised ratios — 6.82/ (11.47, 4.64)
Range HWFET km 157.6/ 141.8
Mean motor efficiency HWFET % 86/ 87.2
Range UDDS km 142.1/ 130.4
Mean motor efficiency UDDS % 79/ 80.2
Acceleration 0–100 km/h s 13.48/ 13.49
Acceleration 0–60 km/h s 6.48/ 5.22
Acceleration 50–80 km/h s 4.18/ 4.61
Grade climbing % 25/ 45
Top speed Km/h 151/ 222

 

[liveforphysics]

When comparing a 1 speed vs a 2 speed EV design you have to pick some metric to match. Here you can see what happens when you pick the gear ratios such that the 0-100km/h time is the same. You can see with a UDDS driver cycle the cars goes about 12 miles farther in this case with the single speed tranny. This paper is pretty comprehensive as it takes into account shift time and adhesion limits, but I'm sure you will find some faults to pick at


Parameter Units Transmission
One-speed/ Two-speed
Optimised ratios — 6.82/ (11.47, 4.64)
Range HWFET km 157.6/ 141.8
Mean motor efficiency HWFET % 86/ 87.2
Range UDDS km 142.1/ 130.4
Mean motor efficiency UDDS % 79/ 80.2
Acceleration 0–100 km/h s 13.48/ 13.49
Acceleration 0–60 km/h s 6.48/ 5.22
Acceleration 50–80 km/h s 4.18/ 4.61
Grade climbing % 25/ 45
Top speed Km/h 151/ 222

That example made up by a transmission company. What do you expect? I've never claimed you can't pick the wrong motor for the job and use a transmission as a crutch to help you get a poor motor choice to function.

Alternatively, you can pick the right motor for the application and be entirely liberated of the dragging around a ginding box of loss.

 

[flathill]

Where did you see the research was supported by a transmission company? "This project is supported by BAIC Motor Electric Vehicle Co. Ltd., the Ministry of Science and Technology, China, and University of Technology, Sydney'

There is no way you even had time to read the paper.

Let's look at the other example where they match top speed. Again single speed kicks ass range wise for a normal drive cycle, but you do give up some low 0-60 performance, but note 50-80 is better as there is no need to shift. Note the reduction in weight from switching to a single speed is taken into account. What you don't see is what happens when you use a larger motor. The gap narrows even further. The point is for some very large vehicles a motor sized for a single speed would be cost prohibitive. For your average car or truck there is no reason to even think about using more than one speed. Same goes for a motorcycle. Now with ebikes a street legal motor is quite undersized so I'm not so sure. I mean you could build a very large diameter 250W hub motor with 200nm of torque but it does not make sense cost wise. For a 250w bicycle mid drive may make more sense. For a high performance ebike a large diameter hub motor is the endgame.

Parameter Units Two-speed One-speed
Gear ratio(s) (5.7/8.17) 5.7
Powertrain efficiency — 0.8 0.9
Range HWFET km 141.5 157.3
Range UDDS km 129.9 140.2
Acceleration 0–100 km/h s 14.5 13.5
Acceleration 0–60 km/h s 6.3 6.5
Acceleration 50–80 km/h s 4.7 4.2
Grade climbing % 30 23
Top speed Km/h 180 180

 

[Arlo1]

So by using 2 speeds vs 1 speed with same motor you get better acceleration... ? Weird.

Who would have though your vehicle would accelerate slower if you start in second gear.
Flathill thank you for showing me what I have been doing wrong all this time. I will now go buy a motor much to small for my application and add a transmission to it so I can run out of charge earlier then I do now because of the bigger total system losses, and have less acceleration in top gear then I used to at that same speed as with my proven 1 stage of reduction.
Sometimes Luke you will be better off just banging your head on the wall.

 

[flathill]

The data in the report supports the conclusion single speed drives are all around better in a standard car.

I just hate when people talk in absolutes (I am guilty of this too). Engineering is full of compromise.

 

[Punx0r]

I'd consider a vehicle with less than 100HP/Ton to be under-powered. If you've got at least that much power I have a hunch a multi-ratio drivetrain will offer little if any advantage.

 

[liveforphysics]

Where did you see the research was supported by a transmission company? "This project is supported by BAIC Motor Electric Vehicle Co. Ltd., the Ministry of Science and Technology, China, and University of Technology, Sydney'

There is no way you even had time to read the paper.

My mistake mate, I unwisely ASSumed it was the same white paper from Zytek, where they create an example of a making a poor motor choice for the application, and then show by adding thousands of dollars of cost in a transmission (which the same money could have gone into a larger pack or proper motor) they increase vehicle UDDS range by 6% or something.

 

[major]

Ah, back on topic. I agree with liveforphysics and fathill's reference supports it as well.

.... ZF is desperately searching out somewhere to be relevant. ....

I've also noticed these electric racecars using multispeed transmissions are sponsored by those makers of transmissions in a large way judging from the size of the decals on the body.

 

[Arlo1]

Ah, back on topic. I agree with liveforphysics and fathill's reference supports it as well.

.... ZF is desperately searching out somewhere to be relevant. ....

I've also noticed these electric racecars using multispeed transmissions are sponsored by those makers of transmissions in a large way judging from the size of the decals on the body.

I think the formula E is limited in motor/controller selection. If they could get something with more power the tranny would be tossed. That's my guess as I'm working on motor controller stuff due to the fact there is very little out there for real power levels.

 

[liveforphysics]

Ah, back on topic. I agree with liveforphysics and fathill's reference supports it as well.

.... ZF is desperately searching out somewhere to be relevant. ....

I've also noticed these electric racecars using multispeed transmissions are sponsored by those makers of transmissions in a large way judging from the size of the decals on the body.

I think the formula E is limited in motor/controller selection. If they could get something with more power the tranny would be tossed. That's my guess as I'm working on motor controller stuff due to the fact there is very little out there for real power levels.

Formula E is currently a locked down spec class with no innovation permitted. If you want to participate, you can buy the single vehicle offered to use which is identical to everyone else.

You must understand, they have only motivations to keep formula E as crippled and neutered as possible.

An exciting and successful formula E would be in the interests of the fans, EV technology development, and the quality of air for all living beings, however they operate in the interests of shareholder profits of sponsors and participants ICE vehicle based interests.

Ah, back on topic. I agree with liveforphysics and fathill's reference supports it as well.

.... ZF is desperately searching out somewhere to be relevant. ....

I've also noticed these electric racecars using multispeed transmissions are sponsored by those makers of transmissions in a large way judging from the size of the decals on the body.

The current gen Honda Accord Hybrid showed the world transmissions are an unnecessary box of grinding losses and failure modes to add to even an ICE vehicle.

The writing is already clearly on the wall that if you wish to have high efficiency and/or high performance the transmission is not involved in your drivetrain. The transmission-less systems are already dominantly superior while still at infancy design maturity. Transmissions have already had >100years of maturing, there is no rabbit left to pull out of the hat to ever make them relevant again, the entire industry is a dead man walking, sell your stock if you have any, the tree will never again bare meaningful fruit and can only rot from this point forward.

The new path has been laid, and those who don't walk it are choosing obsolescence.

 

[h0tr0d]

The Regera uses Koenigsegg's 5.0 liter turbo V8 along with three electric motors. They are powered by a nine-kWh, 620-volt, water-cooled battery pack, which Koenigsegg says is the most power dense battery pack every created for a production car.

The battery is not water cooled, it's cooled by a liquid that boils 34ºC @ ambient pressure

 

[gogo]

The Regera uses Koenigsegg's 5.0 liter turbo V8 along with three electric motors. They are powered by a nine-kWh, 620-volt, water-cooled battery pack, which Koenigsegg says is the most power dense battery pack every created for a production car.

The battery is not water cooled, it's cooled by a liquid that boils 34ºC @ ambient pressure

That's 93ºF for us unwashed.

 

[flathill]

Thats interesante. In the video he mentions wet cooling and I went like huh

Vapour cooling is the third best way to cool electronics

 

[Hillhater]

Formula E is a "spec" power train ( to contain development costs) and currently supplied by McLaren.
Chassis and tires are also "spec"'
They could easily include a more powerful motor (McLaren already have one) but then the battery would have to be bigger& heavier to give them any meaningful range....but then the handling would be crap and the lap times very little improved.
In short , it's a compromise formula, devised to make it easier for teams to form , buy a pair of cars, and get in the show without needing all the tech development, engineering back up, and costs that an "open" formula would require.
I believe next season teams will have some opportunity to develop their own drive systems.