I’ve also been working on adding a temperature sensor to the motor design and a forced compressed gas coolant solution. Something similar to those freeze spray cans. Spraying out a compressed gas that rapidly cools as it expands.

I don’t agree with NJay on this. I don’t think you’d want to use standard motor V/rpm values as a starting point. I’d say it would be better to go with standard V/rpm values from generators. However, I think many commercial generators are made to run both ways and as such are probably designed for compromise. If you’re going to compromise on the generator you might as well go with the same motor as the propulsion motor.

Wow, 15 phases? Why not just use a clamping circuit and let voltage vary up and down some. There must be an acceptable range of voltage to charge with. I’m not sure what the consequences of a varying power of charging are though. I know on the face of it my suggestion may sound nuts, but consider that windmills are used to generate electricity and they definitely have a varying power output. Even if the controllers are fully rectifying the windmill generator, they would still have a varying output. I would even suggest that looking at how a windmill rectifies generator output to the battery storage would be worth a look. My perspective is that as long as you aren’t stressing the batteries it would be acceptable to allow the voltage and current to varying up and down. I might even b e tempted to wire the generator as two phase and just bridge rectify it. This would be the simplest circuit providing the least lose of energy, and providing an easily calculated “average” charge rate. Now I’ve never tried anything as crazy as this. But s simple inexpensive experiment could be done to test the wisdom or folly of it.
The challenge, I think will be driving the generator in a way not to destroy the batteries. I think a generator producing maximum power output will produce too much power and would rapidly destroy any batteries it was designed to charge.

Which brings up the pertinent question Steve asked: what exactly is the best generator design? Which criteria to follow for selection?

I suggested the same as for the motor (torque), because that is an indication of how much energy effectively flows through the design; a high torque motor should be a high power generator, because both employ an intense magnetic flux. And a high efficiency motor should also be a high efficiency generator… Of course, I may be dead wrong, I don’t yet have the skills to decide. But I’ve ordered a nice book on the subject.

Now, Njay suggests the “V/rpm” constant that most motors have, as a criterion. I don’t know exactly how that maps into efficiency, but I can see how it is important in the overall system design – you’d like to know the exact voltage coming out of the generator. But as you, Celtic, have said, the question is a lot more complicated than that.

In my opinion, one of the criteria in common between my higher torque motor designs and what I think would be a good generator is the number of poles. The higher the number of magnetic poles, the better – but I’d probably separate them into independent phases on the generator. An output AC voltage on 15 phases would be a lot “easier” to rectify than on 3 phases, yielding a very smooth current and therefore lower rectifier losses. But it is a compromise between current smoothness, current intensity, and phase voltage (less phases with same pole number = higher current or voltage, depending on how we wire the coils together).

So… since this is just the tip of the iceberg, I’d say there’s a lot to digest here. In a simple DIY perspective, I’d settle for the design with the highest torque-to-electric-power ratio as a good basis for both a motor and a generator, and then build both machines and experiment and tweak them independently. Nothing substitutes effective practice.

I’m not sure I agree with best traction motor = best generator motor. That may be true, but I ‘m not certain that it must be true. For a generator you want to maximize energy output, but for traction you want torque. You may be able to get more energy out of a motor that produces less torque.
However, it may not be significantly different and it may be more convenient and cost effective to use one motor design. For starters, if you have identical power and generator motors, they are interchangeable and you have only one template.

I’ve been playing with wire winding apparatus. I built my own jig, out of mostly scraps. While the jig will need some fine-tuning it will be satisfactory for a prototype. I may eventually buy a winding machine. One thing that I have discovered is that it is absolutely critical to have some sort of tensioning device in order to produce a tight uniform coil with 14AWG solid magnet wire. This may not be as much an issue for you, as it is for me. Still if you are going to wind your own motor, then you will need to think about a winding apparatus and your windings will be in a tighter area than mine. I’m now going to work on a tensioning device. I’m thinking something simple like three rubber or plastic wheels squeezing on the wire with adjustable tension. Have to be careful though, I don’t want to deform the wire.

I also have two possible designs for an iron core for the axial motor. One involves rolling a thin laminate up into a disc on a small core, then drilling and milling out slots for the windings. This is to be followed by a careful unwinding of the laminate. Cleaning of the cut surface, applying a lamination and re-rolling it into a solid disc. Of course an outer ring must be installed against the unrolling of the disk.

Next, I need to calculate the optimum winding for the motor, well at least by the time I get a satisfactory test winding. I’m using a 16 magnet/12 coil design. I’m also considering using 36 coils.

The problem is that I ran into a really nasty geometry bug that stops me from trying the medium-to-large magnets… one of those trigonometric “error accumulation” things that make you pull your hair out. Disgusting.

But I do have a large set of data already, so I’ll give that a try. I’ll try to make the time for it asap…

In the meantime, I enhanced the results parser to include real information about magnet weight and cost. This way I can compare torque vs. (weight and cost and power).
Unfortunately there is no overall winner yet; I have to establish priorities in my criteria, and right now I settle for:
1 – Nm/Euro; 2 – Nm/kW; 3 – Nm/kg.
This means I want the biggest “bang for buck”, followed by the most electrically efficient, and lastly the lightest possible.

Keep listening…