Simple mechanical coilgun model
TheMerovingian, Fri Feb 10 2006, 10:04PM
Extract from the old board:
yeah, now i understand.
I need to know if this calculus that i do in my simulator is somewhat correct:
The behauviour of the LCR circuit was described before but i added the variable inductance and a backEMF generated from the moving projectile.
Now i have I on time and compute B of the lone solenoid (solenoid shell with
finite lenght) at the center of the projectile. Then i calculate the magnetization with my model (But hyperbolic tangent gives almost the same results). I calculate the dipole multiplying by the projectile volume.
knowing the dipole moving in a mag field i obtain the magnetic potential energy and taking the derivative along the moving dyrection of the projectile i get force.
I have problems in calculating inductance so for now i keep it constant
It seems to give quick'n'dirty results but the behauviour is essentually correct.
I cannot attach images to show you the GUI.
i have done some simulations for my halfbridge coilgun
(three stages, each calculated separately with the speed of the previous one)
obtaining
a speed of 46,4m/s (6g)
a muzzle energy 6,45 (with a total energy consumption of 78J)
total efficiency 8.2%
the turn off time is when the projectile releases the photogate (-6mm from coil). The suck back is present and can be eliminated controlling the pulsetime (i have a strand of 555 controlling the pulses)
The results does'n look weird for such a simple model. maybe using your reluctance model and adding eddy losses would boost the accuracy
Re: Simple mechanical coilgun model
Psyko, Fri Feb 10 2006, 11:21PM
what is this software, Mero ? Seems to be really useful. Is there a place where we can download it ?
Re: Simple mechanical coilgun model
TheMerovingian, Fri Feb 10 2006, 11:38PM
The software is still incomplete. I will release a beta when i am sure it is reliable
Re: Simple mechanical coilgun model
WaveRider, Mon Feb 13 2006, 12:56PM
Looks reasonable, Mero. Send me your simulation parameters and I'll run a camparison with my simple model...
Re:
Simple mechanical coilgun model
TheMerovingian, Mon Feb 13 2006, 05:13PM
Run a simulation with these parameters:
coil turns 320
coil lenght 30 mm
coil internal diameter 8 mm
coil external diameter 18.4 mm
no ferromagnetic shield around coil
coil inductance :388uH
coil resistance :657mOhm
Capacitance: 1000uF
initial voltage : 300V
timestep: 0.1us
topology: free running scr + diode (no quench resistor apart the coil itself)
switch voltage drop 1.4V diode voltage drop 800mV
projectile diameter: 5.5mm
projectile lenght: 30mm
projectile weight: 6 grams
saturation magnetization: 2.1 tesla
initial permeability : mu_r 5000
initial position : -30 mm (measured from center of coil to center of projectile)
initial velocity : 0 m/s
Results:
18,78 m/s
1,05 J kinetic energy
(peak velocity 21,58 m/s, suck back present)
peak force 89,3N
Greetings for your promotion
Cheers
Re: Simple mechanical coilgun model
TheMerovingian, Wed Feb 15 2006, 12:30AM
I added static friction to the model calculating it for my barrel (measuring the angle at which che projectile falls off) but it doesn't really make any difference in results. Maybe can bu used for rifled barrels since determining the coefficient is very easy.
Waverider I'm awaiting for your simulation to compare the results
Re: Simple mechanical coilgun model
WaveRider, Wed Feb 15 2006, 08:09AM
Hi Mero... Been burdened with lots of stuff to do lately. I'll try to do it today!!
Cheers..
Re:
Simple mechanical coilgun model
TheMerovingian, Wed Feb 15 2006, 01:58PM
I have compiled your program on windows but have some problems in translating the data since your simulator asks me for gap lenght and i am forced to write the difference in diameter between the projectile and the coil inner radios (my simulator does'nt threat shields
) . I recalculate Bsat with your advices in the readme. There is a little suckback but less than predicted in my simulator. Also the current predicted in your simulator is bigger (faster discharge, this explains why less suckback), i calculated the inductance using wheeler formula and the variations on ind0 with a reverse formula on the force.
Anywais your simulator says 17.5 m/s that's pretty close to my prediction but i don't really know if i calculated this correctly since there is the shield.
Re:
Simple mechanical coilgun model
WaveRider, Thu Feb 16 2006, 09:31AM
Hi Mero.
Bravo! Glad to hear that you got my code to work on your computer..
I ran a simulation with these values (based on what you gave me)
# Coil Length (m) = 3.000000e-02
# Projectile length (m) = 3.000000e-02
# Effective gap length (m) = 2.500000e-03
# Effective core radius (m) = 4.000000e-03
# Firing capacitance (F) = 1.000000e-03
# Number of turns in coil = 3.2e+02
# Low-field core relative permeability = 5.000000e+03
# Saturation flux density of core (Wb/m^2) = 1.000000e+00
# Saturation magnetisation of core (A/m) = 7.957747e+05
# Coil loss resistance (Ohms) = 6.570000e-01
# Mass of armature (kg) = 6.000000e-03
# Simple RLC (SCR-like) drive.
# Initial capacitor voltage = 3.000000e+02
# Initial coil current (A) = 0.000000e+00
# Initial armature speed (m/sec) = 0.000000e+00
# Initial armature position (m) = 0.000000e+00
# Time step (sec) = 1.000000e-06
# Print out every 10 time steps.
# Tolerance for the non-linear solution (Newton-Raphson) = 1.000000e-08
I used a slightly larger time step, but convergence was still very good. Here are the results:
# Maximum force on projectile (N) = 1.124648e+02
# Minimum force on projectile (N) = -4.859833e-01
# Average force on projectile (N) = 2.552838e+01
# Maximum projectile speed (m/s) = 1.758019e+01
# Projectile exit speed (m/s) = 1.751668e+01
# Minimum projectile speed (m/s) = 5.775584e-07
# Maximum current magnitude (A)= 3.026064e+02
# Maximum capacitor voltage (V)= 2.999992e+02
# Minimum capacitor voltage (V)= -7.553270e+00
# Maximum core flux density magnitude (T)= 3.504001e+00
# Time at which pulse is switched off (sec) = 3.349520e-317
# CPU seconds used in computation. = 0.110000
As you can see, my results came out very similar to yours. A couple of comments are in order:
1. With open coils, you can play games with the gap length to simulate the effects of the open coil. This is a bit of a fudge, though. This model specifically counts on the confinement of the field by the shield.
2. The model breaks down for thick, multilayer windings (like yours). Errors arise because the radially directed magnetic field becomes significant along the length of the core. EDIT: Furthermore: many of the flux lines do not couple into the armature /EDIT. The model assumes current is concentrated in a shell surrounding the core. The closer we come to this ideal, the better it works.
3. The specific solution will, of course, depend strongly on the magnetisation model. If you used a different saturation model, the results will certainly be different.
4. It's bloody difficult to get exactly repeatable experimental results, since even slight misplacements of the armature can vary results considerably!
Great work, Mero.. Sorry about the delay...
Cheers.
Re:
Simple mechanical coilgun model
TheMerovingian, Thu Feb 16 2006, 10:48AM
Bravo! Glad to hear that you got my code to work on your computer..
I simply compiled your sources imported in the Opensource Watcom c++/fortran compiler and gave it a go.
It would be cool to do a graphical used interface for your program. Since they are complementary (mine no shiedls, your with shields) it would be cool to see the waveforms in real time. Translating it to basic would be problematic for me
. Maybe doing some c++ dlls and calling it from basic would work.
Great work, Mero.. Sorry about the delay...
Cheers.
No problem Bill, I know that you are busy
.
Cheers..
Re:
Simple mechanical coilgun model
WaveRider, Thu Feb 16 2006, 12:50PM
Just to let you know that I plugged your geometry into my FEM based code and it's chugging away as I write. In a few hours, I'll have some results.. At this point in the simulation (with the FEM), the current-rise is much slower than the simple model, indicating significantly higher inductance as a result of the thick winding layer..
Here is a pic of the field lines early in the sim..
The armature is the thin rectangle at the bottom left. The coil is the fat block in the middle and the top rectangle is an external shield (whose permeability is set to 1.0...that of free space...it really isn't there).
With "fat" open coils and thin projectiles, many of the field lines don't seem to interact with the projectile..
Re: Simple mechanical coilgun model
TheMerovingian, Thu Feb 16 2006, 01:13PM
Heres the calculations i have done . Used the polinomial magnetization model
]1140095622_14_FT363_coilgun_force_expressio.pdf[/
file]
The inductance i calculated was derived from the wheeler formula, while the variations on the inductance are calculated in a semi-empirical way.
I had a slower current rise as you can se from the graph, that's why there is more suckback
rise rate = 5,55E05
What fem program is that?
Re:
Simple mechanical coilgun model
WaveRider, Thu Feb 16 2006, 02:23PM
What fem program is that?
It's Waverider's Magic Graphical Coilgun FEM simulator.. I like to call it "BigShot1.0"
Re: Simple mechanical coilgun model
TheMerovingian, Thu Feb 16 2006, 02:45PM
any chances to port it on windows :D ?
Re:
Simple mechanical coilgun model
WaveRider, Thu Feb 16 2006, 02:59PM
If you are volunteering.. I guess the chances would be good!
Re:
Simple mechanical coilgun model
TheMerovingian, Thu Feb 16 2006, 03:14PM
I feel an irresistible need to install Linux on my pc
Seriously this program needs a big and complex graphical interface that are way too much for my weak programming skills.
You should put an announcement to search volunteers for windows porting.
Cheers..
Re:
Simple mechanical coilgun model
WaveRider, Fri Feb 17 2006, 09:55AM
Mero, went through your force calculations.. I guess you are aware that B depends not only on the current, but also the presence of the magnetic material in the coil. Your expression for B does not include the effects of the core. This follows through in your energy discussion too. The energy of the magnetic field depends not only on the core dipole moment, but also on the field that surrounds the core... That's why the problem is not so easy...it's all coupled together!
Anyway, I had to shut my computer off last night, so the simulation did not run to completion, but it made it most of the way there! Here's a gif movie of the projectile
It's kinda interesting at the beginning..You can clearly see the effect of the armature saturating early (watch the field lines on the left being drawn in to the center early on in the simulation).
Here's the circuit dynamics:
There is a little suckback at the end and the armature exit velocity of 19.2m/s (max velocity of 20.7m/s). The timing of the current peak indicates an empty-coil inductance of about 410uH...not to far from your estimation of 388uH... Peak force was calculated to be 72N.
Re:
Simple mechanical coilgun model
TheMerovingian, Fri Feb 17 2006, 10:44AM
I know that it's not easy but i needed a simple/fast model helping me design/understand coilguns. I know that your FEM simulator is the closest to the reality and obtaining less than 20% error on final velocity makes me very happy. Also your model does consider eddy currents and residue magnetization so the results are very precise.
The projectile saturates early? This is because the current rises rapidly since the inductance is close to the coil inductance alone so the field establishes rapidly.
Your simulator is beautiful Bill, when you will release a commercial version?
Re:
Simple mechanical coilgun model
WaveRider, Fri Feb 17 2006, 06:06PM
Your simulator is beautiful Bill, when you will release a commercial version?
Thanks! Commercial version.... I don't think there would be enough buyers out there to warrant the hard work.. I may release the research code under GPL after I finish my paper and fix a few nagging bugs..if there is enough interest..
Re: Simple mechanical coilgun model
TheMerovingian, Fri Feb 17 2006, 10:47PM
I'll be happy to use your software for simulations. May need to find a linux emulator or install a linux distribution to run it.
Just for curiosity i forced your inductance value in my simulator (this feature is because it is impossible to wind a perfectly packed coil, so after building the coil i measure inductance and resistance with an RCL multimeter).
The new simulations show that:
Ipeak = 247 A (difference +3% respect to Bigshot)
Peak force = 89N (difference +23% respect to Bigshot)
Peak speed = 21,78 (difference +5% respect to Bigshot)
Exit speed = 18,78 (difference -2% respect to Bigshot)
Simulation parameters except inductance are the same.
]
1140216421_14_FT363_merossimulationforcespeed.pdf[
/file]
]
1140216421_14_FT363_merossimulationcorrentvoltage.
pdf[/file]
Re: Simple mechanical coilgun model
Evgenij, Sat Feb 18 2006, 02:07AM
Hi William, Hi Jonathan!
FEMM based simulator's result from Russian coilgunners for you:
INPUT:
capacitance: 1000uF
initial voltage : 300V
timestep: 100us
wire diameter (copper only): 0.627mm
double wire insulation thickness: 0.05mm
coil length: 30mm
coil internal diameter: 8mm
coil external diameter: 18.4mm
projectile diameter: 5.5mm
projectile length: 30mm
OUTPUT:
Total resistance, Ohm: 1.242
Outline resistance, Ohm: 0.49
Coil resistance, Ohm 0.752
Turns of coil: 319.94
Wire length, m: 13.268
Initial coil inductance, uH: 379.74
Projectile’s mass, g: 5.559
Energy of caps, J: 45
Peak velocity, m/s: 15.697
Out velocity, m/s: 14.795
Kinetic energy, J: 0.608
Efficiency, %: 1.352
Peak force, N: 48,5 @ 900us
Peak current, A: 184,9 @ 500us
-- Evgenij
Re: Simple mechanical coilgun model
WaveRider, Sat Feb 18 2006, 09:56PM
My inductance value is a little bit high because I did not fiddle with the wire sizes to get 320 turns exactly (came out to about 348 turns). It looks like the results are quite similar.. This is good! The peak force is likely to be overestimated in simple models, particularly when the armature nearly fills the coil...it's the average force that is important... Evgenij, how are you solving the problem? My understanding is that FEMM is not too good with time domain problems...
Re:
Simple mechanical coilgun model
Evgenij, Sat Feb 18 2006, 11:18PM
WaveRider wrote ...
Evgenij, how are you solving the problem? My understanding is that FEMM is not too good with time domain problems...
FEMM has support of LUA-script language. We use a LUA-script for the description of model. By means of a script the FEMM is repeatedly started for fields calculating in every timestep.
And I compiled your program to Windows-PC, and try use it. But your soft show me more high velocity (approx. on 15%) than my real coilgun devices can reach.
We had originally such excess in our calculations too. Now we make greater work for approach of teoretical and experimental results.
Re: Simple mechanical coilgun model
TheMerovingian, Sun Feb 19 2006, 09:54PM
Two simulations. One letting alone the values calculated by the pre-simulation phase (resistance and inductance , forced additional resistance) and the second with forced initial inductance.
Same problem as waverider (the last coil layer is incomplete)
Simulation with the values not forced
INPUT:
capacitance: 1000uF
initial voltage : 300V
timestep: 1us
wire diameter (copper only): 0.627mm
double wire insulation thickness: 0.05mm
coil length: 30mm
coil internal diameter: 8mm
coil external diameter: 16.5mm
projectile diameter: 5.5mm
projectile length: 30mm
magnetization model:polynomial
OUTPUT:
Total resistance, Ohm: 1.249
Outline resistance, Ohm: 0.49 (simultated as cap ESR)
Coil resistance, Ohm 0.759 (calculated with my formula)
Turns of coil: 320
Wire length, m: 13.59
Initial coil inductance, uH: 347 (sorry, wheeler formula, the projectile added inductance is calculated and added to the initial value)
Projectile’s mass, g: 5.555
Energy of caps, J: 45
Peak velocity, m/s: 18.37
Out velocity, m/s: 16.2
Kinetic energy, J: 0.729
Efficiency, %: 1.61
Peak force, N: 57.1N @ 982us
Peak current, A: 178 @ 591us
Simulation with the values forced
INPUT:
capacitance: 1000uF
initial voltage : 300V
timestep: 1us
wire diameter (copper only): 0.627mm
double wire insulation thickness: 0.05mm
coil length: 30mm
coil internal diameter: 8mm
coil external diameter: 16.5mm
projectile diameter: 5.5mm
projectile length: 30mm
magnetization model:polynomial
Initial permeability : 5000
Saturation magnetization: 2.1Tesla
OUTPUT:
Total resistance, Ohm: 1.242
Outline resistance, Ohm: 0.49 (simultated as cap ESR)
Coil resistance, Ohm 0.752 (forced value)
Turns of coil: 320
Initial coil inductance, uH: 379.7 (forced vaue)
Projectile’s mass, g: 5.555
Energy of caps, J: 45
Peak velocity, m/s: 18.55
Out velocity, m/s: 16.38
Kinetic energy, J: 0.74
Efficiency, %: 1.65
Peak force, N: 57.7N @ 1ms
Peak current, A: 176.3 @ 630us
Initial permeability : 5000
Saturation magnetization: 2.1Tesla
The differences are quite significative but still compatible:
With the values forced:
Peak velocity, m/s: +18%
Out velocity, m/s: +11%
Kinetic energy, J: +22%
Efficiency, %: +22%
Peak force, N: +19%
Peak current, A: -4.3%
With the values not forced:
Peak velocity, m/s: +13%
Out velocity, m/s: +9,5%
Kinetic energy, J: +19%
Efficiency, %: +19%
Peak force, N: +18%
Peak current, A: -3.7%
I wondering what is the saturation magnetization in the femm program for steel
The waveforms are listed in the pdf files
]
1140386059_14_FT363_merosimulationnoforcingcurrent
voltage.pdf[/file]
]
1140386059_14_FT363_merosimulationnoforcingspeedfo
rce.pdf[/file]
]
1140386059_14_FT363_merosimulationforcingcurrentvo
ltage.pdf[/file]
]
1140386059_14_FT363_merosimulationforcingspeedforc
e.pdf[/file]
Re:
Simple mechanical coilgun model
WaveRider, Mon Feb 20 2006, 08:41AM
Hey Mero! Do you have any experimental data on the coilgun that we are all madly modeling!
Re:
Simple mechanical coilgun model
TheMerovingian, Mon Feb 20 2006, 10:46AM
Hhmm, ehm no wave. I'm building the halfbridge and have simulated that some time ago. The first stage is very similar to that we modelled except that there is an igbt instead a scr and it can turn off when the projectile crosses center.
maybe we should construct this device and test the models.
For the halfbridge i obtained energies up to 5.6J and speeds in the 49m/s range (total
eff 7.9%), quite disappointing since i dreamed to break the 10J barrier with my halfbridge, i hope that my simulator is wrong
.
Anyway throwing toghether this coilgun would be easy. I have tons of 2n6509 25A scrs and tons of 120uF disposable camera caps. I will search if i have the barrel.
EDIT: I found an old barrel from a previous version of the halfbridge. It has holes for the sensors but that's not important. Outer dimension 8mm, inner 6 mm, projectile 5 mm.
I should find time to wind it.
Re:
Simple mechanical coilgun model
WaveRider, Mon Feb 20 2006, 12:00PM
Evgenij wrote: FEMM has support of LUA-script language. We use a LUA-script for the description of model. By means of a script the FEMM is repeatedly started for fields calculating in every timestep.
Hmm.. I thought you could do something like this.. What assumptions did you make in terms of saturation model? How did you link the FEM to the circuit model? Your velocity seems a bit low, but the current curve looks good...
And I compiled your program to Windows-PC, and try use it. But your soft show me more high velocity (approx. on 15%) than my real coilgun devices can reach.
[
This is typical behaviour, since the coil winding model is very simple. The FEM will certainly give better results. Also, my simple model is not designed for modeling open coils (although an open coil can be "fudged"...)
We had originally such excess in our calculations too. Now we make greater work for approach of teoretical and experimental results.
My FEM-based model gives far better results for open coils. FEMM should produce good results, too.. Altho' we won't know for sure until we do the experiments!
Have fun!
Re: Simple mechanical coilgun model
TheMerovingian, Mon Feb 20 2006, 12:04PM
Waverider is right, we have no experimental reference to test the data.
Evgenij can you tell me the saturation magnetization of the armature you tested?
I think that in this thread will be more to learn with some experimental data.
I need some time to build the test device, I have exam sessions now, so need to study, no time for hand-working. Maybe the next week I will build it
Re: Simple mechanical coilgun model
TheMerovingian, Tue Feb 28 2006, 12:59PM
I have built a test device to measure speed of a given configuration and set of parameters.
The setup is simple. A single coil with SCR+Diode configuration, 920uF worth of capacitors, the coil, a multimeter, a pushbutton, a rogosky pulse amperometer (my creation)
Coil parameters:
Turns : 280
Internal Diameter : 8mm
external diameter : 22mm
lenght: 30mm
Coil electrical parameters:
297uH (measured with LCR multimeter)
417 mOmh (same as above)
Projectile Parameters:
projectile mass : 5,6g
projectile diameter : 5,5mm
starting position: -27mm
Cirtuit parameters:
Capacitance:920uF
Capacitor total ESR : 87mohm
SCR voltage drop: 1.4V
Diode voltage drop : 800mV
Remaining parameters:
Saturation magnetization(steel): 2,0Tesla (this gives the bigger problems)
Initial permeability (steel):5000
No shields
Results:
Dryfires:
7 measures, average 302,7A standard deviation 4,4A
projectile fires:
Speed:14 measures , average 14,7m/s , standard deviation 0,3m/s
current: average 299,3A,standard deviation 2,5A
Differences:
Dryfire simulation: 303,7A (+0,3%)
Projectile simulation (current) : 296,6A (-0,9%)
projectile simulation (speed) : 18,4m/s (+25%, seems very high, but consider the indetermination on weight, saturation magnetization and position repeteability, also no eddy currents considered)
Simulate and tell me your results for these parameters:
attached there are the excel sheet, and the graphs along setup photos
Re:
Simple mechanical coilgun model
Evgenij, Tue Feb 28 2006, 11:12PM
TheMerovingian wrote ...
Simulate and tell me your results for these parameters:
Wire diameter, projectile length, initial voltage?
But if your wire is 0.8 mm, projectile 5.5x30 mm and 300 Volts then my result 14.44 m/s.
Re:
Simple mechanical coilgun model
Evgenij, Wed Mar 01 2006, 12:30AM
TheMerovingian wrote ...
a rogosky pulse amperometer (my creation)
Are you sure in linear and precision of this type of current sensor?
I use 3 resistors 0.1 Ohm 5W 1% in parallel and have 33 mOhm current sensor for DSO scope.
Re: Simple mechanical coilgun model
TheMerovingian, Wed Mar 01 2006, 01:10AM
wire diameter is 0,8mm , 300Vin , 5.5 mm diameter x 30mm
Re:
Simple mechanical coilgun model
Evgenij, Wed Mar 01 2006, 03:26AM
TheMerovingian wrote ...
wire diameter is 0,8mm , 300Vin , 5.5 mm diameter x 30mm
and
Speed:14 measures , average 14,7m/s , standard deviation 0,3m/s
Then my simulation result 14,44 has standart deviation interval. OK.
But check your current measure system. In this simulation I have 221A peak current only.
Re: Simple mechanical coilgun model
WaveRider, Wed Mar 01 2006, 09:05AM
The simple model gives a final speed of 16.5m/s.. Not too bad, given that it is not designed for open coils.
It is difficult to adequately represent open coils in simple models.. The FEM will always be the best way to go.. Evgenij's results seem reasonable to me.. I'll run it through "BigShot" later and see what happens...
Re: Simple mechanical coilgun model
TheMerovingian, Wed Mar 01 2006, 10:36AM
Evgenij take into account that the coil isn't perfetly fitted. That's i measured the inductance and resistance. Boh RCL simulation an measurements show a current in the 300A range
We need more experimental data.
Re:
Simple mechanical coilgun model
WaveRider, Wed Mar 01 2006, 08:25PM
I meant to say that Evgenij's speed calculation was pretty good.. His coil current seems a bit low.. In any case, I will believe the measurements more readily than any simulation!
Cheers!
Re:
Simple mechanical coilgun model
Evgenij, Thu Mar 02 2006, 03:48AM
TheMerovingian wrote ...
Both RCL simulation an measurements show a current in the 300A range
You have mistake in both case.
Do measuring of current as more high accuracy and you will see. When high current going to SCR, common wire, inner capacitors ESR, soldered points e.t.c then voltage drop will high and current low. Are you seriously think about 2n6509 SCR's voltage drop is 1.4V at 200-300 Amps?
Re:
Simple mechanical coilgun model
TheMerovingian, Thu Mar 02 2006, 07:25PM
Evgenij wrote ...
TheMerovingian wrote ...
Both RCL simulation an measurements show a current in the 300A range
You have mistake in both case.
Do measuring of current as more high accuracy and you will see. When high current going to SCR, common wire, inner capacitors ESR, soldered points e.t.c then voltage drop will high and current low. Are you seriously think about 2n6509 SCR's voltage drop is 1.4V at 200-300 Amps?
I understand but your simulation don't take this into account. Try to fit the coil with less density. The inductance is too high, my LCR multimeter says 297uH (5%)
Re:
Simple mechanical coilgun model
Evgenij, Thu Mar 02 2006, 10:42PM
TheMerovingian wrote ...
I understand but your simulation don't take this into account. Try to fit the coil with less density. The inductance is too high, my LCR multimeter says 297uH (5%)
INPUT:
capacitance: 920uF
initial voltage : 300V
timestep: 100us
wire diameter (copper only): 0.8mm
double wire insulation thickness: 0.05mm
coil length: 30mm
coil internal diameter: 8mm
coil external diameter: 22mm
projectile diameter: 5.5mm
projectile length: 30mm
initial position: 3mm
OUTPUT:
Total resistance, Ohm: 0.95
Outline resistance, Ohm: 0.50 <-- I thinking this parameter doing low current
Coil resistance, Ohm 0.45
Turns of coil: 273.22
Wire length, m: 12.875
Initial coil inductance, uH: 342.19 <-- projectile in coil 3mm now
Projectile’s mass, g: 5.559
Energy of caps, J: 41.4
Peak velocity, m/s: 14.559
Out velocity, m/s: 14.443
Kinetic energy, J: 0.580
Efficiency, %: 1.40
Peak force, N: 60,6 @ 700us
Peak current, A: 221,2 @ 500us
What approximate outline resistance you have?
I calculated as 0.35 Ohm SCR and solder and wire resistance + 0.15 Ohm capacitor's ESR = 0.5 Ohm
Re:
Simple mechanical coilgun model
WaveRider, Fri Mar 03 2006, 10:15AM
BigShot reports an initial inductance of 304uH assuming 265 turns (only way to reduce winding density is by reducing the number of turns in the coil area). Multilayer coil inductance is indeed strongly dependent on the distribution of the windings!
I'll try to do a full motion simulation over the weekend.
Ciao!
Here's my simulation. I used the following parameters..
1. Capacitor Value (uF) 920.0
2. System Resistance (ohms) 0.0
3. Winding Conductivity (S/cm) 4.0e5
4. Quench Resistance (ohms) 1.0
5. Initial Capacitor Voltage (V) 300.0
6. Initial Coil Current (A) 0.0
7. Initial Armature Velocity (m/s) 0.0
8. Initial Armature Position (cm) 0.3
9. Initial Time Step Size (msec) 0.000025
10. Print out every ? times steps 1
11. Newton Method Tolerance 5.0e-3
12. Output Data File MERO2
13. Z-direction plot points 150
14. R-direction plot points 100
15. Number of contours in plot 15
16. FEM interpolation order 2
17. Exterior boundary type PEC
18. Armature triangle area factor 0.001
19. Sheath triangle area factor 0.09
20. Coil region triangle area factor 0.005
21. Air region triangle area factor 0.09
22. Switching position (for HB, D&Q) 3.0
1. Winding Length (cm) 3.1
2. Winding Thickness (cm) 0.7
3. Armature center hole radius (cm) 0.05
4. Sheath thickness (cm) 1.0
5. End Cap Thickness (cm) 1.0
6. Air Gap (cm) 0.125
7. Coil-sheath Air Gap (cm) 1.0
8. mu_i for Sheath 1.0
9. mu_i for Armature 5000.0
10. B_sat for Sheath (T) 0.5
11. B_sat for Armature (T) 2.0
12. Sheath conductivity (S/cm) 0.0
13. Armature Conductivity (S/cm) 4.0e4
14. Coil Wire Diameter (mm) 0.92
15. Armature Radius (cm) 0.275
16. Armature Length (cm) 3.0
17. Armature Point Length 1.0
18. Armature mass (g) 5.6
I fudged the coil wire diameter and coil length a tiny bit to get the required number of turns. I also made the armature conductive (although I ran a non-conductive case, and the difference was minimal).
The peak current is in line with what Mero has reported from his experiments (around 300A.. the plot shows about 320A peak current). The force is in the ballpark of Evgenij's results. I compute the max. force to be 77N (he computed 60.6...expected since his computed current is lower.) Final velocity is computed to be 15.6m/s...a little bit high, but I have probably underestimated your firing circuit resistance... With a bit more resistance in the firing circuit, it should give a pretty good answer...
Plot of current, force and armature speed.
Animation of the projectile motion and field lines.
[Edit: Fixed double post and oversized image. Remember 400 pixels is the maximum.]
Re: Simple mechanical coilgun model
TheMerovingian, Sun Mar 05 2006, 10:35PM
Are eddy currents considered in this simulation? Also you should add 87mohms of capacitor ESR. I think that the speed will fall into the 14-15m/s region. Good work Bill
Re:
Simple mechanical coilgun model
Evgenij, Mon Mar 06 2006, 03:14PM
WaveRider wrote ...
I fudged the coil wire diameter and coil length a tiny bit to get the required number of turns.
Add the thickness of insulation to simulate too:
d,mm - D,mm - DD,mm
0.300 - 0.334 - 0.352
0.500 - 0.544 - 0.566
0.800 - 0.855 - 0.884
1.000 - 1.062 - 1.094
1.250 - 1.316 - 1.349
1.500 - 1.570 - 1.606
d – copper only wire diameter (without insulation)
D – one-layer emaled magnetic wire diameter
DD- double-layers emaled magnetic wire diameter