Coming here is a description of alternators and motors, but electrostatic ones instead of electromagnetic, and though really powerful - like 2MW at 0.325Hz rotation for a wind turbine (Enercon's E82), or 800MW at 1.538Hz rotation for a dam (Itaipu of course).

Well, this topic was begun somewhere else, so you'll have to jump on the links, sorry folks... I suggest to begin the reading on page 2 of this thread
http://saposjoint.net/Forum/viewtopic.php?...fe8049a16e55235
by looking at the drawings and diagrams.

Replacing transformer oil by a liquid insulator with a higher dielectric constant, the wind turbine's alternator described there
http://saposjoint.net/Forum/viewtopic.php?...start=20#p19961
will hopefully be even smaller - it is already as big as the induction machine used up to now, but with a much better efficiency.

I wish I had data about the dielectric strength (breakdown field) of the liquids that look interesting, which shall have
- A good safety
- A high breakdown field
- A high permittivity
- A low viscosity, even in winter

but I've found nothing, and nobody gave me an input in the chemistry thread on this forum, so I will assume a liquid exists with these properties:

- Permittivity Er = 18
- Breakdown field = 15MV/m
- Viscosity 2.5mPa*s

By assuming these unambitious properties, the machine is bigger than necessary, but I hope to be going safe. For instance, permittivity around 40 is common and can exceed 90 (carbonates); 28MV/m strength (70kV over 2.5mm) is guaranteed by several insulation fluids.
Permittivity, viscosity and safety as above are met for instance by:
Pyridazine, Cyclohexanone, Benzaldehyde, Butanol
and many more, so let's hope at least one will meet or exceed 15MV/m.

And here is an improved electrostatic alternator for a gearless wind turbine - based on Enercon's E82 data, that is 2.0MW at 0.325Hz rotation.

The alternator's form is here
http://saposjoint.net/Forum/viewtopic.php?...start=20#p19961
Finger electrodes belong to combs, one for each phase, as on this drawing:
http://saposjoint.net/Forum/viewtopic.php?...start=10#p19817
and electrical power is extracted according to this latter improved scheme
http://saposjoint.net/Forum/viewtopic.php?...start=40#p20451

------------

All insulations distances are 2mm. Fingers are 1.5mm thick with hemicylindrical edges. The reasonable hypothetical dielectric liquid then allows 21.2kV between electrodes. According to the scheme 2B, the exciters have ±10.6kV, the DC output is at ±10.6kV, and fingers are charged with a maximum of 21.2kV relative to the exciter.

The alternator (its interleaved electrodes part, more precisely) has only OD=4m ID=3m. The combs are split in 3 finger widths: 22 fingers have 3mm between D=3m and D=3.3m, 21 have 3.5mm up to D=3.6m, and 28 fingers have 4mm width up to D=4m, resulting in a mean width of 3.5mm.

This allows 1884 combs on a disk side (take for instance 1884 at the stator and 1882 at the rotor to suppress cogging and get a multi-phase output), and energy is extracted 612 times per second.

Being 100mm high, each output finger has 56pF maximum capacitance with its pair of exciter neighbours, totalling 4.0nF for a comb line and 14.9µF for a two-sided disk.

Charging at 21.2kV makes 316mC, exploiting at 10.6kV gives 3.35kJ, done 612 times per second produces the desired 2.05MW.

Already a bit smaller than the existing induction alternator, and with much better efficiency. Not bad, is it?

-----------

As a variant, one may imagine to assemble 4 disks on a shaft. Assembly and service is quite more difficult; dimensions reduce then to OD=2m ID=1.5 L~1.2m, making it the smallest 2MW alternator ever built for a gearless wind turbine, and with excellent efficiency.

Or if one finds a dielectric fluid with Er=72 (and 15MV/m) or with 30MV/m (and Er=18), he gets the same diameter with a single disk.

Marc Schaefer, aka Enthalpy

The next example is an electrostatic motor. One that works within these orientable pods (or "azimuth thruster") that propel recent ships because they enable to manoeuvre without a tugboat.
http://en.wikipedia.org/wiki/Azimuth_thruster

As usual, I could only take a full-scale example. This time, it will be the Queen Mary II.
http://en.wikipedia.org/wiki/RMS_Queen_Mary_2
She uses four pods of 21.5MW each
http://en.wikipedia.org/wiki/RMS_Queen_Mar...opulsion_system
which are of the "Mermaid" series from Rolls-Royce
http://www.rolls-royce.com/marine/products...odded/index.jsp

From pictures, speed data, and ruler-on-the-screen, I tried to estimate the motors' data:
Rotates at 3.1Hz
Diameter 2.2m, length ~3m

So let's see how an electrostatic motor compares with this induction one. This application is more favourable than wind turbines to induction motors as peripheral speed isn't as low.

To be compact, the electrostatic motor fundamentally needs an optimized insulating medium. This time, I take SF6 (partly because of the higher speed), at 15b which is the maximum still gaseous at -5°C.
http://en.wikipedia.org/wiki/SF6

CF4 as well as N2+SF6 could be nice, but let's stay at SF6.

Its full 8.9MV/m/b need nice smooth electrode surfaces, with peaks under 3µm height. This would be difficult by milling alone, but subsequent polishing achieves it easily and is definitely worth it. Then, SF6 achieves a dielectric strength of 133MV/m of which we shall use 75MV/m.

The motor has this form
http://saposjoint.cjb.net/Forum/viewtopic....start=20#p19961
except that it has several disks at the rotor and the stator.

Now the less good news: assembly and servicing need to thread several disks on the shaft and in the stator's casing. But these diameters allow precise machined parts at least.

Insulation distances are 1mm as well as finger thickness, resulting in a field concentration of 1.315 and maximum voltage differences of 57.0kV. In the scheme 2B, this allows exciter voltages of ±28.5kV, and the same as input finger voltage and DC bus voltage.
http://saposjoint.cjb.net/Forum/viewtopic....start=40#p20451

Fingers are 2mm wide from D=1.6m to D=1.9m and 2.5mm wide from D=1.9m to D=2.2m, totalling per comb line 75 fingers of 2.25mm mean width. A disk side holds 1676 comb lines, giving 5196 charging half-cycles per second.

Such thin fingers can still be 50mm high, so each has a maximum capacitance of 2.0pF with its surrounding pair. A comb line has 149pF and a two-sided disk 501nF. Then, the scheme 2B (in its motor variant) gives per disk 29mC and 814J per half-cycle and 4.2MW per disk, needing only 5 disks at the rotor.

So with the same OD=2.2m and with ID=1.6m, the electrostatic engine is only ~1.0m long, clearly smaller than the induction equivalent, and with an excellent efficiency that doesn't need cooling.

Of course, the engineer's choice will be to augment the motor's length and reduce its diameter in order to improve water flow in the propeller's stream. For instance, keeping L=3m would allow OD=1.3m.

Marc Schaefer, aka Enthalpy

I've also checked how the scheme 2B improves the electrostatic alternator designed for Itaipu: very nice.

Basic design is kept from there, but dimensions are smaller:
http://saposjoint.net/Forum/viewtopic.php?...start=20#p20022

The active region, with interleaved rotor and stator fingers, has now just ID=8.0m and OD=9.6m. With 2mm spacings everywhere and fingers 2mm thick and 150mm high, it works with ±53.2kV (exciters, output fingers, DC output) at a maximum field of 70MV/m in vacuum.

Each 4mm wide finger has at maximum 5.3pF to the exciter. The 100 fingers of a comb have 529pF, and the 2*4188 combs of a disk have 4.4µF. Five stacked disks (totalling some 1.5m height) have 22µF, store at maximum 2.4C and 125kJ which are discharged 6441 times per second, giving 808MW.

As a comparison, Itaipu's electromagnetic alternator have >16m OD, and I hope to gain 1% efficiency.

Marc Schaefer, aka Enthalpy

One more advantage of the electrostatic alternator or motor is that it can be built very lightweight. As opposed to electromagnetic machines, which need big sections of ferromagnetic material, electrostatic machines need just a skin of conducting material; only the dielectric medium must have some volume, but it can consist of a gas or vacuum.

Designers may think of fingers made of hollow injected plastic with a metal film evaporated on them (or within them if the plastic must bring its dielectric properties - deposited possibly by a catalytic process). A more extreme possibility is to make them just of a very thin nickel skin, like in coupling bellows for instance, 3D shaped for mechanical stability (adding bends to get two-dimensional curvature if necessary).

Whether engineers want to use high voltages on a plane is questionable. On a satellite, it would be nice. Are we eventually replacing the stupid solar cells by a solar gas turbine on satellites?

Marc Schaefer, aka Enthalpy

As a variant, alternators and motors can also be built where capacitance between electrodes vary because movement changes permittivity.

For instance, fingers of ceramic or plastic with a high permittivity can move in a medium (vacuum, gas, liquid) of low permittivity. Or fingers with low permittivity move in a high permittivity liquid. Or moving parts can alternate high and low permittivity.

Or electrodes can move near such dielectrics of different permittivity.

I've seen no clear advantage over moving one electrodes set against another set. A well-tailored combination, where permittivity ratio matches breakdown field, could be slightly more compact than moving electrodes, but it brings certainly more worries with charges stored at the interfaces and leakage.

I had a look at an electrostatic motor to replace the turbine, and possibly the gear, of an aircraft's turboprop. The goal being, from an existing turboprop: 2.5MW at 20Hz (1200rpm) in OD=500mm and L=3m.

Of course, an electric motor would combine with fuel cells to fly the plane, as already discussed there:
http://www.physforum.com/index.php?showtopic=15490

And the answer is: the electrostatic motor would need one gear stage and though it would have to run in secondary vacuum - or maybe in high-pressure insulating gas. And then I hate the idea of putting seal rings on a shaft, having neither access nor time to repair in flight, and depend on secondary vacuum to get thrust from the power plant.

Much more so since a nearly-classical electromagnetic motor would do the job properly. Build it with Nd magnets, about 21 pole pairs, ID=350mm GapDia=400mm OD=500mm and L=1m, solid copper bars as 1-turn windings, and you get the 2.5MW at 20Hz with <3% losses and weighing about 800kg.

So this case is very different from a satellite where vacuum is very reliable. Or where the complete alternator can sit within the hermetic casing filled with the working gas that makes the thermodynamic cycle.

Marc Schaefer, aka Enthalpy