New twist to multi-stage design / Electromagnetic Projectile Accelerators / Forums

Forums

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

« Previous topic | Next topic »

New twist to multi-stage design

Move Thread

LAN_403

Turkey9

Sun Jan 08 2012, 05:36AM

Registered Member #1451 Joined: Wed Apr 23 2008, 03:48AM
Location: Boulder, Co
Posts: 661

I've been pondering the coilgun built by DARPA a while ago. First off, does anyone have a link to the paper? I can't seem to find it anywhere and am only going off of memory here. I was wondering why the design used a large number of thin coils instead of what the hobby community has found to be best (coil just slightly longer than the projectile). An explanation that many have commented is that multiple coils where energized at the same time. When thinking about why, I came up with a possible reason. I ran some simulations and the greatest force the projectile sees is when it is about a third of the way into a coil the same length as the projectile. With multiple coils, I'm thinking that the sections could be energized such that the magnetic field is always at this sweet spot as the projectile moves. This would also serve to somewhat linearize the forces as the projectile moves down the barrel. This also brings up the question, would feedback be needed? Feedback in the form of optical sensors is widely used in multi-stage coilguns in the hobby community due to the non consistencies from shot to shot. With the DARPA design, I would bet that the inconsistencies could be almost completely eliminated. If they could, then maybe the device can be built without optical gates and can use an open loop system.

I'm thinking about trying to build a coilgun like this myself. I don't think I'll use capacitors as my energy source. There will be many sections of low inductance coils all running off of rectified mains. I will use PWM to limit the total current drawn by all of the coils to a level that can be sustained, maybe 10A. This will also serve to make the fields inside the coils a little more linear. All the coils will be controlled by a micro through IGBTs rated high enough to give me some headroom if I want to up the power. I might also use a large capacitor bank if I need higher current than the wall can provide.

Anyway, it's been a while since I built a coilgun and this project seems like fun. Any thoughts? Suggestions? Questions? Thanks!

capacitortree

Sun Jan 08 2012, 10:39PM

Registered Member #4321 Joined: Sun Jan 08 2012, 05:10PM
Location: Illinois
Posts: 5

Is this what you're thinking of?

It has 45 stages at 15 turns each.

Sounds like an awesome project, and you might be able to raise the efficiency higher than standard coilgun designs.

Turkey9

Wed Jan 11 2012, 01:56AM

Registered Member #1451 Joined: Wed Apr 23 2008, 03:48AM
Location: Boulder, Co
Posts: 661

Thanks for that, I saw that paper but just skimmed and thought it was only about railguns. I sweat I saw another paper though, one that was more detailed.

]sheet1.pdf[/file]
This is the basic schematic of what I've designed for this coilgun. It's like a series of full bridges that are all strung together. To energize a series of coils, the top IGBT right before the series will be turned on as well as the bottom one right after the series ends. For example, 3 Top and 8 Bottom will energize coils 3 - 7. All the IGBTs will be driven by GDT to ensure proper turn-on.

I'm also building a 4 stage injector coilgun hoping that the high initial velocity will help increase efficiency. The injector stages will be high voltage, low inductance, for a very short pulse that will fire at the point where the projectile will see the most force in the coil. It's basically the same coilgun I stated to build about a year ago, the thread can be found in the project section.

Forty

Wed Jan 11 2012, 06:47PM

Registered Member #3888 Joined: Sun May 15 2011, 09:50PM
Location: Erie, PA
Posts: 649

The accelerator mentioned in that paper was an inductive design. Your idea for efficiently switching smaller coils in a reluctance design seems logical though and I imagine it's worth a try.
I also noticed that the paper's mortar projectile used an internal coil that the firing coils would induce a current into, which would apply the force to the projectile. This made me curious about the kind of current pulses that this 'coiled' projectile would work with (ie. fast pulse as in an induction launcher or slower pulse as in typical reluctance based coilguns.)
Would the coil just be shorted on itself you think?

For the schematic you just posted, would you still need antiparallel diodes for each coil? do the igbt's have internal diodes? or are you going to cleverly switch them somehow so that the back emf is quenched? I haven't worked with igbt's much so i don't really know.
As an added bonus, a coilgun with many short coils will potentially look like the gauss rifle from fallout 3. I hope you can make some progress on the project soon. It's all too often that we come up with cool ideas but never run them to completion.

Turkey9

Wed Jan 11 2012, 07:02PM

Registered Member #1451 Joined: Wed Apr 23 2008, 03:48AM
Location: Boulder, Co
Posts: 661

Forty wrote ...

It's all too often that we come up with cool ideas but never run them to completion.

My shop area is littered with unfinished projects.

The IGBTs I'm ordering have anti-parallel diodes but I think I'm going to add some to the coils just to be safe. The coils I'm going to use will make the system extremely over damped in hopes that the total voltage won't drop to zero before the projectile is through.

I don't know why the mortar used an internal coil, but it would have to be shorted on itself. The pictures showed copper and aluminum rings as the armature. I bet the current through the coil was used to trigger a safe mode in the internal electronics of the mortar round to keep the em from damaging it. The pulses for that gun were very fast. Around 6kV and less than 8mF if I remember correctly is what the caps were.
EDIT** Changed incorrect capacitor value. The new value is much much larger, but with such small inductance the pulses would still be very fast.

Pinkamena

Wed Jan 11 2012, 11:51PM

Registered Member #4237 Joined: Tue Nov 29 2011, 02:49PM
Location:
Posts: 117

Interesting project, but I'm sure it won't be easy to make XD
Good luck.

Saz43

Fri Jan 20 2012, 02:13AM

Registered Member #1525 Joined: Mon Jun 09 2008, 12:16AM
Location: America
Posts: 294

Forty is correct, DARPA's coil mortar was an induction design, which can use shorter coils resulting in a smoother acceleration profile (as you predicted). Coilgun Systems sheds some light on this, citing reputable IEEE sources.

I am very interested in seeing how this project turns out, since it's something that hasn't been done before.

Questions/comments:
1. Why are you limiting yourself to 10A?
2. I believe you will still need feedback. Many short coils will give you smoother acceleration and better consistency, but you will still need a way to switch accurately. According to DARPA paper, "The small velocity error is due to feedback control of the power input to the launcher and elimination of propellant variability. Infinite zoning/range adjustment comes from precise velocity control."

Good luck and keep us updated!

Turkey9

Fri Jan 20 2012, 03:04AM

Registered Member #1451 Joined: Wed Apr 23 2008, 03:48AM
Location: Boulder, Co
Posts: 661

The design has evolved since last time I posted. The main reason I wanted to limit the current to 10A was because I didn't want to buy 30 IGBTs at a couple bucks a piece. Here is the project as it stands now:

18 stages switched by 26 IGBTs with a peak current rating of 220A. Feedback for each stage using optical gates. Coils are going to be wound on a form machined from Delrin. They will be .15" thick and be repeated every 1/4" for a total length of 4.5". I'll use 22awg wire as I have a surplus of the stuff and small wire will ensure a high inductance. I'm going to control all the stages using a FPGA partially for the very fast response and partially because I have a demo board I'm using for one of my classes. The power supply will be a bank of 32400 uF charged to 160V.

I'm going to build a 4 stage coilgun to use as an injector. The reason is that with a fast initial speed, a smaller total energy will be expended. I don't want to drain the bank before the projectile is out of the barrel.

I'm going to measure the speed before and after the main section to determine delta V and work out efficiency.

I've already ordered the parts and the just got here yesterday so the build will start any day now. I will either update a current thread I have or create a new one in the projects section.

X14U2NVX

Thu Jan 26 2012, 03:11AM

Registered Member #4306 Joined: Tue Jan 03 2012, 02:42PM
Location:
Posts: 9

Am very interested, Can't wait for the results!

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