Oh?

Let's assume that there is a motor at each wheel, but they mounted to the unibody and connected to the wheel via a short CV shaft, such that the motor does not contribute to unsprung weight.

So what are the disadvantages?

The only other thing I can come up with is that you haven't left a spot for a gear reduction. I figure it'd be easy enough to affix a two- or three-ratio planetary gearset between each motor and its CV shaft. The nature of the powerband on an electric motor is such that one speed is often sufficient, so we don't need to worry about six or seven speeds.

I see several advantages.

First of all, three of the four motors could fail and the car would still be mobile in most situations.

You can use smaller motors (because there are more of them), which allows you to mount them lower. That leads to a lower center of gravity, which improves several aspects of ride and handling.

Such a setup offers lots of opportunities for torque vectoring. When cruising down the highway, the car could power just one wheel, letting the others spin freely (w/ planetaries disengaged). Let off the gas and all four can engage and go into regen mode.

Packaging opportunities abound when the only thing you need to route power to a wheel is a one-inch-thick flexible cable. No need to worry about routing exhaust, fuel lines, etc over and around axles or gearboxes.

So what am I missing?

 

Well for one, I would think the control logic would be very complex. With a motor at each wheel, you'd have to vary the speed of each motor when cornering to act as a differential. I woudlnt even want to think about desining that system.

 

I think it's easier than you think

With four wheels that are entirely independent of each other from a mechanical standpoint, there should be no problem going around a corner with equal torque applied. The outer wheels will spin faster of their own accord.

You could certainly increase the power to the outside wheels to help push a car through a turn. There is potential to set something up where you could spend a lot of time tuning parameters based on input from the steering column, yaw and roll sensors, etc. Alternatively, you could simply design it so that anything more than 20 degrees of steering input results in a 10% power boost to the outside rear wheel.

I don't think there's significantly more complexity in such a system than there already is in a modern ECU.

 

I'm pretty sure that glossing over the engineering of such a control system as simple and no problem is the funniest thing i will hear all day.

 

Was that a counterpoint?

 

Okay ... but "unibody mounted" is not what jg95z28 said; he said "wheel mounted" which is very different, and what I was responding to. I totally agree with mounting the electric motors on the chassis connected via halfshafts. Whether that means one motor per wheel or a larger single motor through a differential, I think will take some time to sort out which is better.

 

A V8 rumble and the delicious aroma of exhaust, you hippie!

I guess my question is, what is the benefit of having an electric motor at all 4 wheels? Most cars aren't even 4wd anyway...If it's a 2wd car, what's the benefit of having a motor at each corner instead of one in the middle of two half-shafts?

 

Oh? When do you expect to hear that?

I know you didn't hear it (or read it) in my post... because I didn't say it.



Well, of course. I meant more from an engineering standpoint.

I think you'll be hard pressed to find anyone who can give a good argument that 4WD isn't the best, all other things being equal.

Most cars aren't 4wd because it adds too much cost/weight/complexity/packaging issues to make it worthwhile (in other words, all other things aren't equal).

What I'm suggesting is that if the cost/weight/complexity/packaging issues are reduced or mitigated by a system like the one I described, maybe the benefits will start to outweigh the costs.

Putting a motor in the middle of two half-shafts would work fine, but then you need a differential, and you've now guaranteed that the space between the wheels is spoken for from a packaging standpoint. With individual motors, you don't need a differential. Does a differential weigh more than the second motor? Does it cost more? Maybe; it depends on a lot of factors (power and durability in particular).

As electric motors become more common in cars, I think we'll see costs go down as a function of supply and demand. I think that electric motors have been used long enough in other areas that we are probably not going to see much improvement in efficiency (measured as power/weight, power/size, power/input energy ratios -- I'm thinking other industries have maximized those attributes already and the automotive world can simply take advantage). Anyway, I'm guessing the falling costs will eventually swing the cost-benefit analysis in favor of more motors.

 

Fair enough. So when you said there were "too many" disadvantages, you were referring to the two disadvantages I listed in the first place?

 

I don't think 4 separate motors would be as efficient as 1 larger one. You would have the overhead of extra wiring, separate cases, shafts, gears, etc.

 

I'll give you wiring and cases, but I think those are pretty small.

Shafts? I'm pretty sure there are going to be at least as many shafts on a one-motor setup as on a 4-motor setup, especially if we're still doing 4WD on the one-motor setup.

Gears? Only if my planetary gear reduction at the wheel proves necessary/useful. The Tesla Roadster and the Volt both have single-speed transmissions. I don't know if they use direct drive or if there's a gear reduction, but the gears would/could be pretty simple.

Don't forget all of the stuff you don't need.

 

I'm sorry, you said it was easy, no problems, and can be simply designed.

Yes, i butchered you with snippery, but i felt like jumping on the program/project management talk of how trivial the job is.

The part i left at the end is key. Maybe there's not more complexity than in a modern PCM/BCM. In fact, i bet there's not. But i'm not going to ignore that the modern PCM has been evolving for generations and is STILL evolving to this day.

Now certainly today's hardware means we're not going to go back to the days of the earliest ECMs and limited capabilities. But at the same time, managing parts tolerances, different stresses and operating points, fine details of torque management and regenerative braking are whole new problems that will take time and resources to figure out. You've got to optimize driving experience and performance as well as parts lifetime and reliability.

Sorry, i wouldn't call that easy.

As for no problem going around a corner, it will be more involved than simply providing equal power to each wheel. Especially if you want regenerative braking to work in turns too.

As for 'simply designing' a first order rough approximation of add power during turn > X°, i guess you don't actually intend this to be a pleasant to drive commuter vehicle since you'd constantly be fighting weird, disproportionate thrust in turns.

 

In my defense, I said "the most efficient electric cars will have motors in each wheel", I never said "wheel-mounted". What I meant was "at each wheel" and was referring to what Jake explained, unibody mounted at each wheel. I was thinking of GM's skateboard chassis when I wrote that.

 

You will need to package the batteries and most likely some sort of cooling system still.

 

I see that you acknowledge that you butchered me with snippery, but just in case others aren't reading so closely:

* I said it was easier than SteveO thinks. He implied that it would be a nightmare to get the control system right. I think it would be no more complex than a modern ECU.

* I said that there would be no problem caused by the lack of a differential due to a characteristic inherent in electric motors.

* I offered a simpler approach to the controls programming than the one I offered first.

I'm not saying it's trivial by any means. I'm just drawing a parallel to another control system already included on all modern cars (the ECU) and saying that if we can do that, we can certainly do this.

I used the word "easy" one time, saying that it would be "easy enough" to put a planetary gearset between the motor and the output shaft that goes to the wheel. Yes, it has to be designed, validated, calibrated, etc., but so does everything else on the vehicle. This is not significantly more complicated that any other modern automotive powertrain system.

Fine, so it wouldn't be an on-off situation. We're still not beyond what any competent engineering team with the right resources can do.

I'm not claiming that my ideas are production ready without any further thought. You seem to be taking them that way. Stop it.