For quite some time I've wanted to run an electric vehicle and have been interested in renewable energy solutions and self-sufficiency. I had originally posted something similar to what follows in a thread on the EV Forums.
In the past year there have been a couple concept vehicles that utilize in-wheel hub motors. One a Mini Cooper and the other was done by Mitsubishi.
The in-wheel motors are still very expensive [read - they aren't really
that expensive to produce, but oil companies/interests still have a major share of the companies/technology that currently make them - yes this is non-researched speculation on my part, but I don't think it's too far off], that's if they are even available to mere mortals.
A more affordable solution, which I started thinking of years ago when I had a Nissan Pathfinder and revisted my idea again about a year ago for application in a Mitsubishi Eclipse Spyder, is using in-board motors hooked directly to CV half-shafts. Here is where motor design becomes a big factor. Ideally a motor should run within it's designed optimal RPM range. A small single gear transmission might be a better solution here.
The multiple motors (48-72V?) could be smaller than what is necessary for a single motor and shouldn't take up much more room than standard axle/diff drive components. This would cost more than going the single motor route, but wouldn't be as expensive as hub motors. I would think these could be whatever a single motor spec would be for the vehicle in question divided by two for driving two wheels or four for four wheels.
This would be ideal for making an AWD vehicle, which was my original thoughts for this, as each wheel could be controlled independently (think super traction control) and wouldn't have the compromises that are made with axle differentials (open, LSD, locking). Plus I think with AC motors and regen it would make for a pretty effective braking.
This type of system would need some type of central controller, which I would think could be rather simple if independent motor controllers are used. My thought is to use a linux-based central computer that handles several functions, one of which is the interface with the multiple controllers. I can certainly handle the computer side of things - programming, OS setup, etc. but I'm not so knowledgeable with the hardware side of things.
At this point I would like to use the Spyder as a test bed for this, but without some funding or sponsorship I'm thinking I will end up going the old single DC motor route as funds allow and work on developing a central computer for system monitoring, optimization, logging, etc. which could be adapted to an AWD system later.
I think the Spyder is a good platform for this - the chassis is the same as the AWD Eclipse, it's a small aerodynamic car, and would make for a great demo vehicle proving that EV's don't have to look like the ugly Hybrids (Insight, Prius, etc.).
Of course, the Tesla Roadster does this already, but they are a still at least couple years away from putting out a model that is affordable to the average person and I predict the market for EV conversions will grow considerably.
Most vehicles get retired due to mechanical failures, and I'd guess that a vast majority of such failures are related to a lack of, or poor, maintenance. The rest of the vehicle is typically in very good shape. An electric propulsion system consists of much less parts, requires considerably less maintenance, and costs less to run (gas and oil products aren't getting any cheaper). I won't get into the argument of "but you're just changing where the oil gets used and the source of pollution". I will say that I think it's a poor argument that lacks facts. Now if we consider the idea that practically every home
could provide it's own electricity supply by using solar and/or wind this argument disappears. If at the very least, every home provided some portion of it's own electric and was still tied into the grid the argument is still pretty moot.
Back to development of an EV central computer. I have sent out some inquiries to try and identify the components for such a system, but have yet to receive any responses. If I had the funding available, I would just start buying what I think would be suitable and go the trial and error route, but with AC motors and controllers suitable for EV applications being so expensive, I can't justify the potential cost of such an endeavor. My preliminary guesstimate is that a conversion system utilizing four motors with independent controllers and a central computer control would cost in the range of five to ten thousand. Ideally it should be able to be done for $5000 or less for the drive and control system, batteries being another matter of varying cost.