Yanom's FEMM program for simulating coilguns and calculating efficiency. / Electromagnetic Projectile Accelerators / Forums

Forums

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

« Previous topic | Next topic »

Yanom's FEMM program for simulating coilguns and calculating efficiency.

1 2 next

Move Thread

LAN_403

Yanom

Wed Jan 30 2013, 03:35AM

Registered Member #4659 Joined: Sun Apr 29 2012, 06:14PM
Location:
Posts: 158

For my high school regional Science Fair, I wrote a program that simulates coilgun shots and calculates the efficiency of the shot. It uses a three-step process:

1) Draws a coilgun using the specified parameters - innerRadius, coilThickness, coilLength, projectileLength, ampTurns (specifies the total current*turns that the coil will be powered with while in operation)
2) Iteratively moves the projectile into the coil, calculating the force at each location. This data goes into a "Force table" that correlates each location of the projectile (location 0 is where the iron is just outside the coil) with a particular force on the projectile.
3) Calculates the total kinetic energy of the projectile on coil exit. During this process, it finds the total time that the gun firing will take and multiplies this by the resistive losses of the coil, then adds this sum to the total energy in the magnetic field, to find the amount of energy expended in the shot. see the function calcTable in the code for more.

the program opens the file "a_tabulaRasa.fem" before drawing the coilgun. this file has the material properties for iron and magwire, which I learned from looking at Yandersen's Recuperational Gauss FEMM script.

the program and the TabulaRasa file are attached. you can edit the line 187 to change the parameters of the drawn coilgun

please look at my code and tell me what you think!

Yanom

Wed Jan 30 2013, 03:38AM

Registered Member #4659 Joined: Sun Apr 29 2012, 06:14PM
Location:
Posts: 158

ok, forum decided not to accept my files the first time. They're in the .zip attached to this post
]coilgunsim.zip[/file]

anyway, can anyone check my math in this program and ensure that it makes sense? especially the calcTable() function.

Yanom

Thu Feb 07 2013, 03:46PM

Registered Member #4659 Joined: Sun Apr 29 2012, 06:14PM
Location:
Posts: 158

*bump*

what does everyone think of this? Does anyone have a one-stage coilgun laying around that they can plug into this simulator and see if the simulator gives accurate efficiency values?

Yandersen

Thu Feb 07 2013, 10:26PM

Registered Member #6944 Joined: Fri Sept 28 2012, 04:54PM
Location: Canada
Posts: 340

Actually, with constant ampturns it has no use for simulations of real devices. Simple monkey-design circuit (cap, thyristor, coil and dumper diode) should be assumed, so current change must be simulated (cap discharge, inductance change, resistive losses, kinetical to magnetic energy conversions affect current change).
Will you add it? By now, it is easier to draw a coil and projectile, put it in location of interest, set current and make 1 single static simulation to have an idea of what force here will be at given current.

Yanom

Fri Feb 08 2013, 01:52AM

Registered Member #4659 Joined: Sun Apr 29 2012, 06:14PM
Location:
Posts: 158

The purpose of this was to give a rough approximation of Cg behavior at a given coil geometry and power level. I then made 4000+ setups, and got permission from my school to use all 30 workstations at once like a ghetto supercomputer. The idea was to find the most efficient setup. Results are coming soon!

Yandersen

Fri Feb 08 2013, 04:24AM

Registered Member #6944 Joined: Fri Sept 28 2012, 04:54PM
Location: Canada
Posts: 340

Some abstract results are coming soon! Get ready for random numbers! SkyNet is launched!!!

Does anybody on this planet has a coilgun driven from constant current source? It is important that current does not change in your coilgun during the shot - otherwise it's operation will give +-100500% difference from Yanom's simulation - we don't want to make him upset and his teachers to look stupid, right?..

DerAlbi

Fri Feb 08 2013, 05:27AM

Registered Member #2906 Joined: Sun Jun 06 2010, 02:20AM
Location: Dresden, Germany
Posts: 727

I must agree. You are simulating a steady state, which is NOT what a shot is. Your effieicency report will tell you exactly nothing. You cant estimate losses with a steady state. Its not even a close aproximation.

You need to use physics of the process, not the steady state.
In detail:
-use a time discrete simulation,
-calculate delta_currents and delte_voltages based on inductance / capacitance
-use the law of conservation of energy when
a) the inductance changes -> recalculate the current based on energy previously on the field.
b) the projectile acceleration. recalculate current after the accelleration. Kinetic energy is drained from the magnetic field.

In the end all energy (R-disipation, L/2*I^2 and C/2*U^2 must be close to constant. Its a good thing to keeps track of all energy-forms during the simulation to check its correctness.

DerAlbi

Sun Feb 10 2013, 06:13PM

Registered Member #2906 Joined: Sun Jun 06 2010, 02:20AM
Location: Dresden, Germany
Posts: 727

Yanom... i have done some work.
Its originally a script thats designed to simulate a multistage half bridge design. I modified it to work with the conventional circuit. The simulation seems to work, but it looses track of energy-stuff, which shouldnt matter, since its only a proof of math, thats developed for the half-bridge design. (it works in halfbridge mode)

Please feel free to modify it. I am actually not sure, if it works with just one coil. I recommend you try do understand what i am doing and build your own script.
I will share it here, so otheres can do simulation too. However i am not willing to provide extensive support. (sry).

Math looks complicated - yes it was. But its derived from known school formulas only. Its just a little messy due to time discrete stuff. Also the state machine is a little hard to follow. You can add debug-outputs for easier understanding.

Have fun!

]yanomstuff.zip[/file]

BigBad

Sun Feb 10 2013, 09:13PM

Registered Member #2529 Joined: Thu Dec 10 2009, 02:43AM
Location:
Posts: 600

On the contrary, I think this is likely to be very useful.

It would be equivalent to a coil gun where the activated coil tracks the projectile.

Yanom

Sat Mar 02 2013, 03:28AM

Registered Member #4659 Joined: Sun Apr 29 2012, 06:14PM
Location:
Posts: 158

DerAlbi wrote ...

In the end all energy (R-disipation, L/2*I^2 and C/2*U^2 must be close to constant.

wouldn't that list also include M/2*V^2, the projectile kinetic energy?

1 2 next

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