Compulsator Discussion / Electromagnetic Projectile Accelerators / Forums

Forums

4hv.org :: Forums :: Electromagnetic Projectile Accelerators

« Previous topic | Next topic »

Compulsator Discussion

previous 1 2 3 next

Move Thread

LAN_403

Marko

Wed Feb 07 2007, 12:15PM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

I needed time to digest this thread and all links.. and I... am still boggled with some things here

Looking at jasonr's schematic, it isn't really compulsator but rather just a pulsed alternator..?

I can imagine how could a dirty, demonstrative active compulsator, or whatever it is, be mocked up from a car alternator: (please correct me if it's complete nonsense):

- One would need to remove all windings, both stator and rotor, and replace them with much fatter wire, and only a single-phase stator winding.

Then, both windings should be seriesed, rotor spun up, and then connected to a power supply at right moment.

In one part of a cycle, coils should be aligned that way that currents flow in exactly opposite directions, and big 'discharge' SCR is fired exactly at that point.

I can also see some problems in the beggining, for example, that in car alternator, some 'air gap' is left between rotor and stator armatures for mechanical reasons. This will probably cause coupling between windings to be too low and high leakage inductance will mess things up.

Another is, providing an initial current pulse to coils, wich I'm unsure how to do properly.

I guess coils would need to be pulsed while their inductance is at 'maximum' in order not to burn any power or slow the flywheel down?

And third... already figured by others... is triggering the SCR exactly at required time, wich I quess can be done using optoisolators.

And... the slip ring, wich normally carries only small exitation current, will now be torched. It will probably just set itself on fire after few firings.

\brainfart

Jasonr

Wed Feb 07 2007, 04:43PM

Registered Member #167 Joined: Mon Feb 13 2006, 06:41PM
Location: Waterloo, WI
Posts: 54

I guess my point is if you wanted to build a home built compulsator this is where â€œIâ€ would start. If you wanted to talk theory and never really build one, that is a different story. Is a car alternator the perfect platform no, is it a good starting point for an amateur who canâ€™t build a huge custom motor, sure. Is it a better platform and more common that a washing machine motor, yes.
(and reading the first post I would say he wants practical)

First "I" would
build a 3kJ cap bank~500volts
I would use a car disc brake rotor for a flywheel. And spin the alternator at 7200rpm with a belt and a washing machine motor geared up.(rotor is easily found, heavy and already balanced)
Get the platform working until I blew the windings or brushes, and do as much testing as possible. measure currents, vs rpm, play with different timings, all of the easy things

second
If it were me I would rewind the rotor with double or triple runs of 12 ga magnet wire. ( or Â¼â€ copper water pipe covered in heat shrink)
I would add a second or third set of brushes to get more surface area on the slip rings.
I would rewind the stator with as large of ga wire as possible.
I would optically fire the scr for the cap bank from a hole in the flywheel.
I would use a disc brake rotor for a flywheel. And spin the alternator at 7200rpm.

third
build a real one.

Jason
I will update this page tonight with ideas I have. http://www.rollette.com/railgun/compulsator/

Steve Conner

Wed Feb 07 2007, 05:23PM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

I agree with Jason. I think the hardest part for the amateur will be modifying the rotor to fit a compensating winding. You'd need access to a machine shop, and no small amount of precision machining skills, to modify the rotor well enough to keep it holding together at 7200rpm.

A while back, myself and some other folks did some rough calculations on this using standard power engineering theory. An average alternator has about 10% subtransient reactance, IIRC. That means it can supply 10 times its rated current into a short circuit for a few cycles, before armature reaction kicks in and the current starts to fall. 10 times the rated current of a car alternator is puny even by hobbyist pulsed power standards.

So, in modifying the alternator, you have to get the subtransient reactance down by at least an order of magnitude. That is what the compensating winding on the rotor does. If you place it right, it appears as a shorted turn in the case of passive compensation, or a mirror image of the stator winding if the compensation is active, hence lowering the reactance (=inductance) just when maximum current is needed, as explained in that earlier post with the nice diagrams.

From that point of view, a compulsator is really just a FCG that spins instead of exploding, and the field coils provide the seed flux to be compressed. With a car altenator, the existing field coil on the rotor would probably do fine to "seed" it.

Madgyver

Wed Feb 07 2007, 07:09PM

Registered Member #177 Joined: Wed Feb 15 2006, 02:16PM
Location: Munich, Germany
Posts: 214

I admit, I haven't though about iron losses but I understand and agree to you points. I just thought that because he said:

wrote ...
small scale demonstration device (i.e. NOT suitable for powering something)

that he just wants a small tabletop display, where the effect of compensation can be demonstrated.

Sure, a 3kJ cap bank sounds very interesting and I don't doubt that it would be a very impressive setup which could power a Railgun. But I think it is overkill, compared to what Joe seems to want.

Of course if we are talking about serious power levels suitable for amateur railguns, there are no other ways possible.

I wonder if it would be possible for an amateur to fabricate a new rotor out of composite material, making it air core.

Marko

Wed Feb 07 2007, 07:49PM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

I think the hardest part for the amateur will be modifying the rotor to fit a compensating winding. You'd need access to a machine shop, and no small amount of precision machining skills, to modify the rotor well enough to keep it holding together at 7200rpm.

Afaiu, the compensation winding is on stator in case of active compulsator:

connected in series with the (also stationary) compensating winding.

so modifying is a bit easier, but still a lot of trouble.
With passive, copper sleeves on rotor rather than stator are a better idea.

I had an idea to use the stator winding both as compensator and field winding, but I see that it didn't make much sense.

I figure that those two coils need to be 90 degrees out of phase to each other, so compensation occurs exactly at the voltage peak.

Could maybe, a stator of a single phase motor be used for that application? It has two windings out of phase as such and would be perfect to be rewound for a small compulsator. But, now it would need a custom rotor.

With a passive compulsator, rotor could be used as a field winding, wich would greatly reduce stress on the slip ring, and make the stator carry the heavy output current.
(I guess that was what steve talked about).

I imagine this would be hard because short-circuit rings again need to be 90deg out of 'phase' with rotor windings, and I haven't seen such a rotor anywhere.

I'm sorry if I just repeated everything that has been said.. repeating just seems to make me understand things better..

I wonder if it would be possible for an amateur to fabricate a new rotor out of composite material, making it air core.

I can imagine that air-cored rotor would simply have too much leakage inductance; coupling with compensation coil would be poor and so the output would.

What else except iron could be used anyways? What do those multi-megajoule compulsators use as their core material?

Madgyver

Wed Feb 07 2007, 08:16PM

Registered Member #177 Joined: Wed Feb 15 2006, 02:16PM
Location: Munich, Germany
Posts: 214

Firkragg wrote ...

I can imagine that air-cored rotor would simply have too much leakage inductance; coupling with compensation coil would be poor and so the output would.

What else except iron could be used anyways? What do those multi-megajoule compulsators use as their core material?

Well actually they use fiber composite rotors, I just don't know if an amateur can build something like that. The low coupling is the price to be paid for reduction of iron losses.

Steve Conner

Wed Feb 07 2007, 11:41PM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

The basic principle is that the flux gets caught and pinched between some rotating windings and some stationary windings. So, passive compensation needs the compensating winding to be on the rotor, if the output winding is on the stator.

In order for the flux to get smashed between them, the two sets of windings have to be carrying equal and opposite (and insanely large) amp-turns. The active design achieves this by connecting them in series, the passive design does it by induction.

The newer machines made for military railguns are all composite and air cored. The drawback of an air cored machine, AFAIK, is that it needs a much higher field current to excite it. I read about one machine that had about 2500A field current flowing for 1 second before the discharge, which was something like 500kA. It used the same winding for field and discharge.

Dr. Shark

Thu Feb 08 2007, 09:57AM

Registered Member #75 Joined: Thu Feb 09 2006, 09:30AM
Location: Montana, USA
Posts: 711

Wow, it's good to see the amount of discussion that has developed here! The first time in the history of man, amateurs are beginning to challenge the hegemony of the US military. Nah, 99% of what I know still comes from the great source that Madgyver pointed out: "COMPULSATOR RESEARCH AT THE UNIVERSITY OF TEXAS AT AUSTIN--AN OVERVIEW", I'd definitely suggest everyone to go over it. If you don't have an IEEE subscription please PM me, maybe there is something I can do

Contrary to what some have said, I belive passive compulsation is the way to go for a first design: The Reduction in inductance is exactly the same as for an active circuit, the only difference lies in the additional "flux compression" that happens in active compulsation and still boosts the output somewhat, giving a shorter, higher power output pulse. Too short for EM-guns if you believe the Austin Texas guys.

I have now come to think of compensation in terms of a transformer: Short the secondary and the primary current goes through the roof, as the impedance the mains socket sees drops from the mutual inductance of the transformer to the leakage inductance. The same would happen if an identical winding was connected in anti-parallel to the primary (active compensation), and the effect would be exactly the same.

I have now sourced a small shaded pole motor that might be suitable to use for a small compulsator:

It only has a single pole which dramatically simplifies the design, and the magnetic window area of this small transformer, being 5cm^2 is larger than that of my 48pole washing machine motor! I am planning to implant some permanent magnets into the rotor, but leaving the "squirrel cage" intact to act as the compensation winding. That way no slip rings will be required at all, again simlifying the design for a proof-of-concept device.
The design parameters I am eyeing are something like this: I want to store about 2kJ of mechanical energy in a 10cm diameter aluminium flywheel, rotating at a whooping 30000rpm. I have spun up CD-ROMs to that speed with a Dremel tool and they hold together (unless you touch them that is - then they explode) so it should be ok for an aluminium disk.
This would give me a pulse-width of about 1ms, during which I hope to extract 5kA at 50V or so. While this is only a couple of J, it could still drive a small coilgun, and it would be an ideal basis for further experimentation.

Marko

Thu Feb 08 2007, 11:24AM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

I think it can be sumarrized as:

Passive compulsators - slip rings carry only small excitation current, output winding is stator,
but losses in copper sleeves on rotor will be high due to termedous currents flowing in them, so they can never actually fully compensate stator's inductance.

Active - need for a massive slip rings, but since exactly the same current flows in rotor and compensation winding, compensation should be better than of passive compy.

Now it all depends between tradeoff of between large slip rings or lossier short-circuit rings.

I also think that passive approach may give much more chance for a ghetto version of compulsator to work, simply due to problems with slip rings.

I am planning to implant some permanent magnets into the rotor, but leaving the "squirrel cage" intact to act as the compensation winding. That way no slip rings will be required at all, again simlifying the design for a proof-of-concept device.

Magnets could be a good start, although probably not as powerful as exscitation winding would be.

I can imagine that biggest trouble is going to be to plant them at a correct phase in the rotor toj coincide minimal inductance and max voltage output...

Dr. Shark

Sun Feb 11 2007, 01:09PM

Registered Member #75 Joined: Thu Feb 09 2006, 09:30AM
Location: Montana, USA
Posts: 711

Theory aside, I made some "not quite progress" on my model:

Firstly I implanted some small NIB magnets in the rotor, but then I decided that bigger is better, and went ahead with 20mm diameter magnets

which I epoxied into the rotor and heat-cured on my soldering iron

I proceeded to sand off the rough edges, so the rotor would still rotate, and put the thing back together:

Alas, when spinning the generator at 2500rpm with my power drill, not a single mV of EMF was induced!
I don't have the slightest clue as to what I did wrong: The permanent magnets are plenty powerfull, in fact cogging is so bad that I cannot turn the shaft by hand. The armature winding is only 5 turns of magnet wire now, put that should produce enough voltage to give a reding on my DMM! In fact, estimating the B-field to be 1T (yeah, maybe an order of magnitude less, I don't have a clue as to calculated this for a permanent magnet) and the window area at 5cm^2, I get a flux of 5e-4. Since the EMF is the time derivative of this, and I have a rotational frequency of 40Hz, the EMF should be around V=5e-4 * 40Hz = 20mV per turn, or something like that.
Mysterious....

previous 1 2 3 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob