Factors affecting efficiency / Electromagnetic Projectile Accelerators / Forums

Forums

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

« Previous topic | Next topic »

Factors affecting efficiency

1 2 next

Move Thread

LAN_403

Camel

Sun Dec 14 2008, 09:47AM

Registered Member #1694 Joined: Sat Sept 13 2008, 09:13AM
Location: Australia
Posts: 108

Hi,

I'm writing some software that will help coilgunners calculate mathematically optimal guns given a number of parameters. I understand the relationship between voltage, capacitance, impedance, number of turns ect, but what I'm struggling with is this.

How can I determine the force applied to the projectile? I'm guessing it has something to do with amp turns and the distance between the projectile and the centre of the coil.

El_Roberto

Sun Dec 14 2008, 10:36AM

Registered Member #1774 Joined: Wed Oct 22 2008, 02:51AM
Location:
Posts: 135

While I cant answer you question, are you going to make this work for multistage coilguns? If so it would be very easy (Once you've got a way of determining the force applied) to use information such as distance between coils and sensors, initial speed, and coil length to work out the time the projectle will take to reach the centre of the coil and therfore work out the inductance. This could save people allot of calculating, because as soon as I get a microphone and a good supplier of copper wire Im going to be doing all of this manually for my coilgun (And with 10-20 stages it could save allot of time).

Camel

Sun Dec 14 2008, 11:49AM

Registered Member #1694 Joined: Sat Sept 13 2008, 09:13AM
Location: Australia
Posts: 108

No, single stage but users would be able to enter initial kinetic energy.

The more I think about this, the more difficult I think this is going to be. :S

Barry

Sun Dec 14 2008, 03:43PM

Registered Member #90 Joined: Thu Feb 09 2006, 02:44PM
Location: Seattle, Washington
Posts: 301

Camel wrote ...

How can I determine the force applied to the projectile? I'm guessing it has something to do with amp turns and the distance between the projectile and the centre of the coil.

Bad news ... the force is highly dependent on 3d geometry and nonlinear material properties, along with amp-turns and position. That's why we need a finite-element magnetics solver (such as FEMM) to estimate a force profile.

I wish there was a useful shortcut, but haven't come up with any myself. All I can think of is to provide a few pre-canned force profiles that would represent a range from low-energy to high-energy coilguns where saturation is progressively more important. But this is hard for a user to know what to choose, and still wildly inaccurate. I hope someone else has a better idea!

Cheers, Barry
If a 911 operator has a heart attack, who do they call?

rp181

Sun Dec 14 2008, 03:49PM

Registered Member #1062 Joined: Tue Oct 16 2007, 02:01AM
Location:
Posts: 1529

There is simply to much factors to determine the optimum setup.

The force depends on eddy currents, mass, size, density, permeability, saturation, starting position in coil, makeup (sold or epxoy?), and the coil, which leads into a loop. You can try it, but personally i dont think you will be able to do it.

This is what i made, just for quick refrence. You put in projectile mass, in energy, chrono'd speed. It gives out effeciency, energy, and ft/lbs.

TodX

Sun Dec 14 2008, 08:00PM

Registered Member #1118 Joined: Wed Nov 14 2007, 09:42AM
Location:
Posts: 5

Finite element analysis (FEA) is the best way to go about it, but you don't have to rewrite FEMM, it is open source. You could setup a frontend that generates FEMM geometry based on some typical coilgun guidelines. To take it a step further you can then make your program add the 4th dimension to the FEA by doing small time unit extrapolations from force, current, etc that was outputted by FEMM.

What I would personally find very useful is if a tool was made that could do all this, and also take into account the EE side of things aswell: capacitor discharge, current ramp up, energy lost in resistance, etc. If all of these were integrated into one big CG simulator that would give a much more accurate way of testing CG designs without the time/effort/cost of building them first. However, that would basicly mean integrating FEMM and a spice package together, which is no small task =/.

Camel

Mon Dec 15 2008, 01:53AM

Registered Member #1694 Joined: Sat Sept 13 2008, 09:13AM
Location: Australia
Posts: 108

Yeah this is looking like an insanely difficult task. Of course it must be entirely possibly, but perhaps I'll leave it for now . . . :P

Barry

Mon Dec 15 2008, 05:44AM

Registered Member #90 Joined: Thu Feb 09 2006, 02:44PM
Location: Seattle, Washington
Posts: 301

Shucks, I was so hoping you would find a slick way to integrate all these elements together.

Have you heard the phrase "paralysis by analysis"?

Your assessment is spot on: it's an insane amount of work, lol. Especially considering the eddy current aspects, without which the resulting velocity is only estimated within an order of magnitude.

Before FEMM was available, I had some limited success with obtaining force profiles through building the coil and physically measuring the current for 1G acceleration. That is, measuring the current required to lift the projectile against gravity, repeated at 1mm intervals throughout the range of motion. Although it's useful for spreadsheet computation and estimates, it is a low power static test and neglects saturation and eddy currents.

Still, a force profile like this can tell you a great deal about expected behavior according to applied waveform. It could conceivably be used in simulators, if the user will input the data for you.

Cheers, Barry
So, I see this frisbee and ask myself "hmmm, why is it getting bigger?" and then... it hits me...

Camel

Mon Dec 15 2008, 07:59AM

Registered Member #1694 Joined: Sat Sept 13 2008, 09:13AM
Location: Australia
Posts: 108

It looks like this has already been done.

WaveRider

Mon Dec 15 2008, 02:14PM

Registered Member #29 Joined: Fri Feb 03 2006, 09:00AM
Location: Hasselt, Belgium
Posts: 500

And here as well. Perhaps if you ask nicely, he will share his code with you!

It has eddy current models incorporated as well as core saturation...

1 2 next

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