Theoretical question about motor conversions

[Janne]

Hi,

(Scroll down to question if you want to skip the intro)

I've been lately thinkin about starting "again" the project with the 8.8m wind turbine. The intention with it is at first just to heat water like before, but I'd like to reserve the option in the future to easily grid tie it, in case the feed laws ever come favourable. So what I basically already have, is a suitable tower, and a set of 3.5m blades with fixed pitch hub. What is required is a new drive line, generator, and a new control system. A more robust control system is especially important, as the system would rely on generator loading and the mechanical brake for overspeed protection. I'd like to avoid all the complexities of pitch control, if possible, even if it requires going complex with the electric controls.
The driveline and the  generator have been the biggest ?? on this project, and I've summed up my thoughts of them below;

-Direct drive axial flux
   Pros: Robust, quiet, quite easy to size up suitably from earlier designs
   Cons: Magnet cost, I estimate I would need to spend 7000€ on magnets alone for this design, as I need roughly 3x the amount of magnet in the otherpower 20' design.

-Direct drive PMG, off the shelf (http://ilmavirta.com/dokumentit/GL-PMG-15k_Horizontal_tekniset%20tiedot.pdf)
   Pros: "Easy" solution, propably quiet
   Cons: Expensive(9300€), not much experience about the unit in question about reliability etc, also the available voltage is on the high side of being suitable.

-Direct drive motor conversion
   Pros: Cheapest of the direct drive solutions, reliable and quiet
   Cons: Big machine tools required, sizing would partially be a wild quess, lots of €€€ for magnets still required.

-Geared up axial
   Pros: Sizing should be easy enough
   Cons: Eddy current losses might became a serious issue

-Geared up induction motor with capacitor excitation
   Pros: Motors available easily and no modification needed
   Cons: (Off grid)Only suitable for heating, grid tie not possible with GTI, only direct grid connection possible-->fixed speed operation. And the big pain with excitation and capacitor switching

-Geared up wound field
   Pros: Motors still available, though not easily, not much modification needed, easy load matching for heating
   Cons: Excitation takes it's share of available power, needs slip rings to feed rotor current(=unreliability factor).

-Geared up motor conversion / high speed PMA:
   Pros: All the pros of the PMA, more robust than geared up axial
   Cons: More noisy than it's direct drive colleaques.

Based on all the factors, I'm currently leaning into the direction of the geared up motor conversion. I do already have a 1:15 gearbox(transmission rate is a bit on the high side, but should work if I limit the peak efficiency operation to 10m/s wind), but not yet any suitable motor in the range of 1500rpm 15-20kW.. I assume, that it would be possible to extract about the name plate power with rated flux, speed & voltage without much of a reactance limiting problems, and hence leave some power reserve to hold the rotor speed down with extra load when necessary.

NOW FOR THE QUESTION. The motor conversion is quite well gone through in earlier projects here, but one thing is puzzling me. How to determine the magnetic flux, with which the converter motor would achieve rated voltage at rated speed? I'd like to use the original winding.. That would be an ideal goal, as it would give suitable voltage for me to work on up to about 2000 generator rpm, above which I'm not thinking of going anyway.
Is there some general flux density these motors run, or is it something individual to each motor? Intuition tells me, that if I were the motor designer, I'd design it to run near the saturation point of the stator iron. In which case I would just need to stuff the rotor with as much magnet as reasonable, and hope it is enough. But is it like that in practice?
Or would it be possible to make an estimate by measuring the current of an idle running motor, the winding inductance(with the rotor installed) and the surface area of the stator insides? Anyways, after I've found out the flux they run on, I could then build a FEMM simulation to get (at least)close the results I'm seeking without making a test rotor first.

Lots of theoretical stuff in here, but I'd rather research the subject than make a couple of non-working prototypes first, as was the case with my small axial turbine attempts

[joestue]

i think you're overestimating the eddy current losses of the geared up axial, just use thinner wire if needed, have you measured the eddy current drag yourself in a test setup?
its going to be less drag than a custom motor conversion anyway.

yes, you can calculate the flux density, its as simple as counting the turns of wire in the stator.
if you're looking to run the motor at nameplate rpm and you increase the flux 30%, you're going to be looking at really heavy losses, but it may be alright if you can figure out how to cool it and control your high end losses. you will probably need a mechanical brake for this machine.

as far as doing the motor conversion if you can cast your own steel shoes, or better yet, get pole pieces punched out from new iron, then you can then change the air gap by putting shims between the magnets and the pole piece. if you want to you can replace the magnet with a block of iron, wrap a test coil around it and measure the flux with a flux meter or use slip rings and measure the voltage while driving it at 1500 rpm. then take the amp turns needed and use that to figure out exactly how much magnet you need.

if you go for a traditional salient pole machine and use neodymium magnets, you may have enough space in there to put coils anyway, and that would give you some control over the output as well.

[SparWeb]

A good summary of the choices facing any WT generator designer.

Any of these designs could require slip-rings, so I don't think that it's a con against only the wound-field generator.

Without speed control by blade pitch you are forced to choose:  a system that regulates speed through the generator alone, or a system that can absorb wildly changing voltage and current.  It's practical to use a battery in small (<5m) systems because the energy can be absorbed by the battery at safe voltage and current levels.  Larger than that, the power is at a scale that demands a battery that will kill you or explode with one wrong move, so it's totally out of the picture.

The grid is a virtually infinite sink into which you can pour power.  So it sounds to me like your 6th choice, the geared-up motor/generator with a field winding is the way to go.
I wish I had access to something like that.  Flux has often described the practicalities of using them, and there are the old machines from the 1960's and 70's to learn from too.
My personal bias may sneak into this.  I was very interested in the work done by forum member Phil T as he restored a Jacobs a couple of years ago so maybe that's why my thoughts are leading this way.

Motor conversions.  If you're as computer savvy as me, you can play with FEMM until you have a workable generator size figured out.  When I try to take the flux measurements and turn them into voltage predictions, I'm not very accurate; maybe still calculating something wrong.  However it is a great way to make comparisons, if you are trying to choose between magnet combination A or B.   A geared-up motor conversion will be noisy, I bet.  Mine is direct-drive, I can hear it clearly.

Axial flux.  There are such generators for sale, cheaper than 9300EUR  I thought, made in China/Taiwan/Singapore.  It might be more reasonable to gear them up (or gear that Ilmavirta Hy machine down!), as much for the matching voltage range as for having the blades on a solid shaft instead of the little shaft the generator will have.  Unless you find a type that has been tested thoroughly by reputable independent agencies and seen service for decade(s), then reliability is a major risk.

I look forward to more responses to this "food for thought".