General linear motor / mass launcher design rant / Electromagnetic Projectile Accelerators / Forums

Forums

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

« Previous topic | Next topic »

General linear motor / mass launcher design rant

1 2 3 next

Move Thread

LAN_403

Dr. Shark

Sat Jul 07 2007, 04:24PM

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

Amateur coil guns are stuck in the middle ages. There is one basic design, the reluctance coilgun in which a soft iron projectile moves in order to maximize the inductance of the system. This has been copied in countless variations, some of them more elaborate than others, but all suffering from low efficiency and limited projectile velocities.

I'd like ot get some brainstorming going, whether there are alternative designs that could lead to better performance. Of course there are induction coilguns with their promises and drawbacks, but I'd like to start thinking from a different point of view: Since any electric motor, be it AC, DC, of the brushed or brushless, inductance, reluctance or permanent magnet type is basically a "rolled up" coil gun, maybe better coilgun designs can be found by "unrolling" some of these again. Most of these have efficiencies approaching and exceeding 90%, mostly by virtue of all the iron that is there to direct the magnetic flux.

This is probably the fundamental difficulty with coil guns, since we only have a single "rotation" to get the motor up to speed, massive currents have to flow, making the use of iron impossible.

Still I wonder whether there are ideas out there to capitalize on some of the different motor designs to build better coilguns.

Eric

Sat Jul 07 2007, 06:14PM

Registered Member #69 Joined: Thu Feb 09 2006, 07:42AM
Location:
Posts: 116

The simplicity and performance of rotary motors has everything to do with the iron/magnets+iron and low frequency and low field operation. Once you get to high powers (fields past saturation) or very high speeds/frequencies you can't use the iron efficiently anymore. Multi-stage guns are essentially a multi-pole linear motor so that you don't have to get up to speed in 'one-turn' but you have greatly diminishing returns on the final stages due to the higher frequencies involved/iron losses.

It seems to me the only way you can go for high energy and velocity is with induction designs.

Dalus

Sat Jul 07 2007, 06:39PM

Registered Member #639 Joined: Wed Apr 11 2007, 09:09PM
Location: The Netherlands, Herkenbosch
Posts: 512

What would happen if you use a superconducting projectile?

Avalanche

Sat Jul 07 2007, 07:32PM

Registered Member #103 Joined: Thu Feb 09 2006, 08:16PM
Location: Derby, UK
Posts: 845

How about using a shorted out coil of wire for the projectile?

It'd probably have to be cast in resin or something.

Ronnie

Sat Jul 07 2007, 09:49PM

Registered Member #600 Joined: Thu Mar 22 2007, 08:41PM
Location:
Posts: 10

Another option would be to use a heavier projectile, so that greater kinetic energies can be released with the same speed. This would allow the entire system to operate at a lower frequency, although there would need to be a redesign, to be able to cope with the different projectile, which inevitably would be larger. The first problem that I can think of is that speed has a much greater influence on kinetic energy than mass.

Dave Marshall

Sun Jul 08 2007, 02:23AM

Registered Member #16 Joined: Thu Feb 02 2006, 02:22PM
Location: New Wilmington, PA
Posts: 554

I agree that the typical mode of thought on reluctance coilguns needs to be adjusted. Seven or eight years ago when the internet electronics community was just starting to take off, coilguns seemed far more popular, but were also plagued by some grave misconceptions perpetuated by one or two respected but clueless 'experts'.

One of the biggest efficiency killers that most people overlook is projectile saturation. Typical cold rolled steel, or worse, the steel used in nails and bolts has a very low saturation point. If you run a coilgun with even modest flux density, theres a good chance you're saturating the projectile and wasting alot of energy as heat.

There are really two ways to combat that. The first is to slow the discharge down (ie make the projectile slower or increase length of the projectile and or coil) to keep peak current lower, thus lowering flux density. This has some other benefits, which I'll get to in a bit.

The other way to combat saturation is to use a more suitable material. Silicon steel is probably the best option. Its damn hard to find in any quantity, and when you do find it, its bloody expensive. Finding some big scrap solenoids is a good option. Another that I'm tinkering with is laminated plates from transformers 'rolled up' into a cylinder with an insulating coating. This has the added benefit of eliminating eddy currents induced on the projectile (which is only a very small problem, but its still a problem).

I'm also looking at a similar idea with permalloy foil. The interesting benefit here is that each stage will magnetize the projectile. Alternating coil polarity could utilize this to add a small boost to efficiency as well.

Barry has done some interesting work with external iron, and it has shown promise. The problem you encounter is the same as with the projectile. You have to use a material that isn't going to saturate.

My current project is a multiple stage (20-30 stages) mains powered system. The idea is to use many smaller stages and a massive projectile to eliminate saturation, as well as suck back and the poor coupling that arises with a very high velocity projectile. It will be inductively triggered and use IGBTs to switch, allowing precise on and off switching.

Dave

Marko

Sun Jul 08 2007, 01:52PM

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

If we truly treat the thing like an unrolled motor, each next coil in a multistage design should fire at opposite polarity of the last one. This uses up the magnetization projectile got from the first coil and reverses it.

So if we don't want projectile to get saturated the next coil should give it exactly same volt seconds in order to restore the flux to zero. And since projectile is increasing in speed it isnt easiest thing to do.

This does not need to be too precise, but projectile will walk into saturation after few stages if that isn't accounted.

Only this way you could use a large number of stages like Dave implied.

Dave already mentioned the other problem, that projectile itself can take really little energy before it saturates, even if it's made of a good magnetic material.

Really efficient coilgun would, I imagine, be several meters long barrel with huge number of small coils. It's power supply from the other side wouldn't be demanding at all, you could probably get out by powering it from mains only with a tiny amount of energy storage caps.

I agree that for larger coilguns induction is a way to go. But still, why are those so inefficient?
I would imagine such a design to suffer from much less problems, and projectile could take as much energy we want from a single stage.

I imagine problem is similar as with current-status railguns.

Dalus

Sun Jul 08 2007, 01:55PM

Registered Member #639 Joined: Wed Apr 11 2007, 09:09PM
Location: The Netherlands, Herkenbosch
Posts: 512

If the projectile would be a superconductor it wouldn't be able to saturate wright? So why can't I find a single coilgun wich uses superconducting projectiles.

OZZY

Sun Jul 08 2007, 02:32PM

Registered Member #511 Joined: Sat Feb 10 2007, 11:36AM
Location: Somerset UK
Posts: 55

I think that joe is right, it`s a good idea to examine the potential of other types of linear accelerators. There are many types of electric motors and any of them could be rolled out to produce a linear version capable of accelerating a projectile.

However do not use the search for higher efficiency as a reason to do this, and do not expect to get higher efficiency as a result. All the different electric motors are capable of operating at high efficiency including the switched reluctance motor which is the rotational equivalent of the reluctance coilgun. It is not possible to transfer the high efficiency of industrial electric motors to a high velocity mass launcher, you can`t break the laws of physics.

If a motor has an efficiency of 90% this is quoted at constant speed and load, and the motor is designed to operate most efficiently at this point. If the same motor was used to accelerate a flywheel from stationary up to its no load speed the efficiency would be much lower, approximately half the rated efficiency so we are already down to 45%.

The acceleration is still too low, to achieve a high acceleration we need to increase the flux density and the current density, both these factors will have a dramatic effect on efficiency. So as you can see it is no suprise that amateur coilguns have low overall efficiency.

It is likely that different types of launcher will be strong in different areas. For example a heavy low velocity projectile may be best suited to one type of launcher but a small high velocity projectile may require a different type.

Improvements in efficiency will require a long process of development in order to optimize many different parameters. There is no magic solution to building the perfect launcher waiting to be discovered.

Dr. Shark

Mon Jul 09 2007, 02:18PM

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

Nice to see some discussion. Saturation in the projectile and the pros/cons of extrenal iron certainly deserve more attention. The first problem can only be remedied by going away from the reluctance design, e.g. using induction. Induction disk launches seem to be very efficient, and induction coil guns can probably be made equally efficient if the magnetic field is channeled properly:
External iron, I have a feeling, is the key ingredient in achiving good efficiency, and since there can be "a lot of it", saturation will not be an immediate problem.

But to get at the heart of the problem, let us look at a design example. Say we want to achive performance similar to the weakest common pistol round, the 9x19 Luger. It is typically fired trough a 0.1m barrel reaching a velocity around 300m/s with a kinetic energy of 500J. The projectile weights close to 10g.
The force required to accelerate the projectile at this rate of roughly 200k m/s^2 is around 2kN, the average power close to 500kW.

In constrast, the forces encountered in a small electric motor are more like 2N, e.g. 3 orders of magnitude less, while profiting from a 3 orders of magnitude boost in force due to the soft iron used. This means we need to make good 6 orders of magnitude. Instead of e.g. 100 amp-turns boosted by soft iron to give a field of 1T, we require a field of 1000T (wouldn't that open a black hole or something? The only time I have read of kilo-Tesla fields was in the context of atomic nuclei and neuron stars), with out the 1000-fold boost from the soft iron, i.e. a coil with 100M amp-turns. Obviously, this is very much impossible. EDIT: Spottet the error, the force goes with B^2 not B, so it is not quite as bad...

The lesson is probably that magnetics stops working well once you exceed the 1T or so that you can get for free using soft iron.

Doh, different line of thought: Ozzys remark about the efficiency of a motor during acceleration is a good point. However, I think that due to the square relationship between velocity and energy, during later stages the velocity remains "relatively constant" while the kinetic energy increases a great deal. This helps the efficiency of later stages.

From my tangential interest in RC planes and their motors I gather that the power density of a motor is roughly proportional to the speed it is operating at. The torque remains constant throughout the speed range, so the power goes up linearly. Small outrunner brushless motors seem to operate happily at speeds as high as 60.000rpms, corresponding to 100Hz mechanically or 1kHz electrically since they tend to have around 10 poles.

Both of the above facts, backed up by empirical data from Thomas Rapp here

indicate that (at least up to a point) the later stages of a coilgun are far more efficient that the early stages that basically have to start "from rest".
I guess a neat way to view this intuitively is that you supply power to the projectile in terms of providing volts to cancel the back emf in the coil, instead of burning big amps in the coils resistance to provide "stall torque" on the resting armature.

1 2 3 next

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