Modeling of coilgun dynamics
WaveRider, Fri Jun 23 2006, 10:07AMJust finalised my latest paper on modeling coilgun dynamics.
It describes my finite element method where losses and material saturation are taken into account. As well as reluctance coilguns, it will also model Thompson (induction) coilguns quite easily...
If there is enough interest, I can make the finite element software that was used to generate the results for this paper available on my mass accelerators site. This will also probably give me an excuse for improving the code as well as including more geometry examples (disc launchers, etc..)
Re: Modeling of coilgun dynamics
Simon, Fri Jun 23 2006, 11:21PM
Great! I'll be giving it a read.
Are there any results you can summarise for the practical builders here?
Simon, Fri Jun 23 2006, 11:21PM
Great! I'll be giving it a read.
Are there any results you can summarise for the practical builders here?
Re: Modeling of coilgun dynamics
WaveRider, Sun Jun 25 2006, 01:35PM
Hi Simon..
I'm putting together some case studies which illustrate some of the results that might interest practical builders. This takes a bit of time, but I hope to have a draft out later this summer...
Do you have any particular questions that you'd like me to address?
WaveRider, Sun Jun 25 2006, 01:35PM
wrote ...
Are there any results you can summarise for the practical builders here?
Are there any results you can summarise for the practical builders here?
Hi Simon..
I'm putting together some case studies which illustrate some of the results that might interest practical builders. This takes a bit of time, but I hope to have a draft out later this summer...
Do you have any particular questions that you'd like me to address?
Re: Modeling of coilgun dynamics
Yohan, Wed Jun 28 2006, 02:51PM
Hey Wave, long time...
This might not be a question for the practical builder but, have you given any attention to a constant voltage source feeding an R,L circuit instead of just the R,L,C?
Yohan, Wed Jun 28 2006, 02:51PM
Hey Wave, long time...
This might not be a question for the practical builder but, have you given any attention to a constant voltage source feeding an R,L circuit instead of just the R,L,C?
Re: Modeling of coilgun dynamics
mike0t4ever, Thu Jun 29 2006, 04:47AM
just let C get very large (approximately infinity) and you get a const V source
in a small capacitance gun the capacitor is fully discharged (assuming ideal capacitor) and the inductance of the coil maintains a decaying current (R/L=time constant)
with a const voltage source (or an infinite capacitor) the current rises asymptotically to V/R (also with a R/L time constant)
with a projectile this is lower since the V= source voltage - velocity induced voltage
where velocity induced voltage depends on the speed of the projectile, its position, and the current
also if the velocity induced voltage is higher than the source voltage than the coil is outputing energy by slowwing the projectile
mike0t4ever, Thu Jun 29 2006, 04:47AM
just let C get very large (approximately infinity) and you get a const V source
in a small capacitance gun the capacitor is fully discharged (assuming ideal capacitor) and the inductance of the coil maintains a decaying current (R/L=time constant)
with a const voltage source (or an infinite capacitor) the current rises asymptotically to V/R (also with a R/L time constant)
with a projectile this is lower since the V= source voltage - velocity induced voltage
where velocity induced voltage depends on the speed of the projectile, its position, and the current
also if the velocity induced voltage is higher than the source voltage than the coil is outputing energy by slowwing the projectile
Re: Modeling of coilgun dynamics
WaveRider, Thu Jun 29 2006, 08:29AM
Indeed, I just make the capacitor very large (3e5-1e6 uF). The current rises asymptotically, as you indicated, to approx Vc / R. During the firing, current will drop to V/(R+Rd).. (Rd is the dynamic resistance that arises as a result of projectile motion.) If you think in terms of "velocity induced voltage", projectile induced EMF opposes the applied voltage when entering the coil and adds to when exiting....the reverse of the situation you describe...
WaveRider, Thu Jun 29 2006, 08:29AM
Indeed, I just make the capacitor very large (3e5-1e6 uF). The current rises asymptotically, as you indicated, to approx Vc / R. During the firing, current will drop to V/(R+Rd).. (Rd is the dynamic resistance that arises as a result of projectile motion.) If you think in terms of "velocity induced voltage", projectile induced EMF opposes the applied voltage when entering the coil and adds to when exiting....the reverse of the situation you describe...
Re: Modeling of coilgun dynamics
mike0t4ever, Sat Jul 01 2006, 03:14AM
yeah that's right L' is +ive when entering and -ive when exiting (slipped my mind)
mike0t4ever, Sat Jul 01 2006, 03:14AM
yeah that's right L' is +ive when entering and -ive when exiting (slipped my mind)
Re: Modeling of coilgun dynamics
WaveRider, Mon Jul 03 2006, 11:42AM
In the old forum, I posted some movies of the magnetic flux behaviour during a firing cycle. I have repeated some of the runs to look at the effects of varying sheath material (or the absence of a field concentrating sheath). More specific details on the firing currents, velocities will be given on my website...
However, these are some of the early results.
Open coil, steel armature (Bsat=2.0T) no armature loss (sigma = 0)
This is a simple lossless armature accelerated by a an open coil. Flux lines penetrate the armature immediately at current switch-on (since the armature is not conductive). The upper rectangle represents the coil windings and the lower (moving) one is the armature.
Open coil, steel armature (Bsat=2.0T), conductive (sigma=4.0e4 S/cm)
Notice that the flux lines take time to penetrate into the projectile. This lowers the efficiency and the final velocity a bit. There is also a noticeable "flux-drag" effect at the end of the firing cycle.
Coil enclosed in lossless ferrite (Bsat=0.5T), lossless armature
By enclosing the coil in a ferrite sheath, acceleration performance can be enhanced. The upper "inverted U" polygon is the sheath. The coil windings are located in the rectangle nestled in the inverted "U" of the sheath.
Coil enclosed in lossless ferrite (Bsat=0.5T), lossy steel armature (sigma=4.0e4S/cm)
Like the previous example...field concentration by means of the ferrite sheath. Efficiency is reduced somewhat because of the armature losses. There is a delay in force at beginning of firing cycle because of time-delay in field penetration into the armature. There is also a drag effect at the end of the firing cycle as a result of residual field and armature motion.
Thin lossless steel sheath with lossless steel armature (Bsat = 2.0T)
Here we model the effect of a thin (2.5mm thick) steel sheath. Note the escaping field lines (because of saturation of the sheath). This is what might be expected for a coil enclosed in a sheath of packed steel rods (electrically insulated from one another).
Lossy steel sheath and armature (Bsat=2.0T, sigma = 4.0e4S/cm)
This final example coil is enclosed in a lossy 2.5mm thick steel jacket (much like a steel pipe). The armature is also lossy... Notice the finite time it takes for the flux to penetrate the shield and armature. A "flux drag" effect can be seen toward the end of the firing cycle.
EDIT 20060704
Added more detailed results to my Mass Accelerators web page. I focused on the two aspects:
WaveRider, Mon Jul 03 2006, 11:42AM
In the old forum, I posted some movies of the magnetic flux behaviour during a firing cycle. I have repeated some of the runs to look at the effects of varying sheath material (or the absence of a field concentrating sheath). More specific details on the firing currents, velocities will be given on my website...
However, these are some of the early results.
Open coil, steel armature (Bsat=2.0T) no armature loss (sigma = 0)
This is a simple lossless armature accelerated by a an open coil. Flux lines penetrate the armature immediately at current switch-on (since the armature is not conductive). The upper rectangle represents the coil windings and the lower (moving) one is the armature.
Open coil, steel armature (Bsat=2.0T), conductive (sigma=4.0e4 S/cm)
Notice that the flux lines take time to penetrate into the projectile. This lowers the efficiency and the final velocity a bit. There is also a noticeable "flux-drag" effect at the end of the firing cycle.
Coil enclosed in lossless ferrite (Bsat=0.5T), lossless armature
By enclosing the coil in a ferrite sheath, acceleration performance can be enhanced. The upper "inverted U" polygon is the sheath. The coil windings are located in the rectangle nestled in the inverted "U" of the sheath.
Coil enclosed in lossless ferrite (Bsat=0.5T), lossy steel armature (sigma=4.0e4S/cm)
Like the previous example...field concentration by means of the ferrite sheath. Efficiency is reduced somewhat because of the armature losses. There is a delay in force at beginning of firing cycle because of time-delay in field penetration into the armature. There is also a drag effect at the end of the firing cycle as a result of residual field and armature motion.
Thin lossless steel sheath with lossless steel armature (Bsat = 2.0T)
Here we model the effect of a thin (2.5mm thick) steel sheath. Note the escaping field lines (because of saturation of the sheath). This is what might be expected for a coil enclosed in a sheath of packed steel rods (electrically insulated from one another).
Lossy steel sheath and armature (Bsat=2.0T, sigma = 4.0e4S/cm)
This final example coil is enclosed in a lossy 2.5mm thick steel jacket (much like a steel pipe). The armature is also lossy... Notice the finite time it takes for the flux to penetrate the shield and armature. A "flux drag" effect can be seen toward the end of the firing cycle.
EDIT 20060704
Added more detailed results to my Mass Accelerators web page. I focused on the two aspects:
1. enclosing the coil in a flux-enhancing shield,
2. and an initial attempt to quantify the effects of armature eddy currents on coilgun performance.
Re: Modeling of coilgun dynamics
WaveRider, Wed Jul 12 2006, 10:56AM
An example of a "short" coilgun stage...
Many launchers in the literature seem to have stages that are wider than thay are long... Thought I would model one here!
WaveRider, Wed Jul 12 2006, 10:56AM
An example of a "short" coilgun stage...
Many launchers in the literature seem to have stages that are wider than thay are long... Thought I would model one here!
Re: Modeling of coilgun dynamics
WaveRider, Mon Aug 28 2006, 12:30PM
Hi all.. Hope everyone had a great summer break!
I didn't have much time for projects this summer, but I did manage to run some simulations to study the effects of ferromagnetic shields and losses on reluctance coilgun performance... I put together a brief report on the subject found here. Or, just go to my mass accelerators page and look for the document titled “Notes on the effects of metallic coil sheaths and armature losses in reluctance coilgunsâ€. Hopefully it offers some useful insight for practical designs...
WaveRider, Mon Aug 28 2006, 12:30PM
Hi all.. Hope everyone had a great summer break!
I didn't have much time for projects this summer, but I did manage to run some simulations to study the effects of ferromagnetic shields and losses on reluctance coilgun performance... I put together a brief report on the subject found here. Or, just go to my mass accelerators page and look for the document titled “Notes on the effects of metallic coil sheaths and armature losses in reluctance coilgunsâ€. Hopefully it offers some useful insight for practical designs...
Re: Modeling of coilgun dynamics
Evgenij, Fri Dec 01 2006, 10:10AM
What shape of the coil is preferable?
This question earlier looked disputable, today like would have mathematics proved answer. Really, there is an optimum form of an electromagnet, with a known ratio of external diameter equal to three internal and length equal to two internal diameters. Such electromagnet creates the strongest magnetic field in the center.
But acceleration of a bullet in an electromagnetic field of the coil is dynamic process. In which time during each moment some parameters of system change at once, first of all a current and position of a bullet. And as the force acting on a bullet, depends both on position, and from a current, the statement about is unique a proper correlation of the sizes of the coil it not seems so obvious any more. Therefore on one of stages of modeling of multistage system I had been formulated a following problem:
There is some conditional stage of the accelerator, approximately in the middle of the barrel, possessing following data:
Diameter of a cylindrical bullet with flat end faces = 6 mm
Length of a bullet = 12 mm
Length of the coil = 12 mm
Internal diameter of the coil = 7 mm.
The coil without external iron cover.
The bullet going to the coil with a speed = 55 m/s
The coil is connected to the 1000 uF capacitor by a voltage 310V.
The amount of the energy spent by this step is limited, and should be 3,05J.
That is, as a result of work of a step, the voltage of the capacitor should decrease from 310V up to 300V.
The step joins at concurrence of a forward end face of a bullet to an end face of the coil and works before full retraction of a bullet. There is instant (we shall examine an ideal case) a switching-off of a current Further.
In this problem it is required to investigate coils with different external diameter, calculations have been for example made for diameters 21, 18, 15, 12, 11, 10, 9 and 8.5 mm.
For each diameter of the coil such diameter of a wire that by the moment of full retraction of a bullet, the voltage of the capacitor has fallen up to 300 Volt, that is the power consumption from the capacitor has made the set 3,05 Joules.
Further it was determined, what of coils of identical length and identical internal diameter, but with various external diameter, will give to a bullet higher speed. Below my calculations generalized results.
So, from the table it is visible, that the greatest efficiency the coil in diameter of 10 mm possesses. And efficiency of the 10-mm coil in 1,66 times above, than at the coil in diameter of 21 mm. Besides of the 10-mm coil the current approximately on 20 % is less, than at 21-millimetric. And one more important plus, the small coil of 10 mm has in six with superfluous time smaller weight, than 21 mm. The total weight of 30 stages these coils will make 160 and 1020 grams accordingly.
Excuse for not so pure translation on English. But the basic sense will be clear, I hope.
Evgenij, Fri Dec 01 2006, 10:10AM
What shape of the coil is preferable?
This question earlier looked disputable, today like would have mathematics proved answer. Really, there is an optimum form of an electromagnet, with a known ratio of external diameter equal to three internal and length equal to two internal diameters. Such electromagnet creates the strongest magnetic field in the center.
But acceleration of a bullet in an electromagnetic field of the coil is dynamic process. In which time during each moment some parameters of system change at once, first of all a current and position of a bullet. And as the force acting on a bullet, depends both on position, and from a current, the statement about is unique a proper correlation of the sizes of the coil it not seems so obvious any more. Therefore on one of stages of modeling of multistage system I had been formulated a following problem:
There is some conditional stage of the accelerator, approximately in the middle of the barrel, possessing following data:
Diameter of a cylindrical bullet with flat end faces = 6 mm
Length of a bullet = 12 mm
Length of the coil = 12 mm
Internal diameter of the coil = 7 mm.
The coil without external iron cover.
The bullet going to the coil with a speed = 55 m/s
The coil is connected to the 1000 uF capacitor by a voltage 310V.
The amount of the energy spent by this step is limited, and should be 3,05J.
That is, as a result of work of a step, the voltage of the capacitor should decrease from 310V up to 300V.
The step joins at concurrence of a forward end face of a bullet to an end face of the coil and works before full retraction of a bullet. There is instant (we shall examine an ideal case) a switching-off of a current Further.
In this problem it is required to investigate coils with different external diameter, calculations have been for example made for diameters 21, 18, 15, 12, 11, 10, 9 and 8.5 mm.
For each diameter of the coil such diameter of a wire that by the moment of full retraction of a bullet, the voltage of the capacitor has fallen up to 300 Volt, that is the power consumption from the capacitor has made the set 3,05 Joules.
Further it was determined, what of coils of identical length and identical internal diameter, but with various external diameter, will give to a bullet higher speed. Below my calculations generalized results.
So, from the table it is visible, that the greatest efficiency the coil in diameter of 10 mm possesses. And efficiency of the 10-mm coil in 1,66 times above, than at the coil in diameter of 21 mm. Besides of the 10-mm coil the current approximately on 20 % is less, than at 21-millimetric. And one more important plus, the small coil of 10 mm has in six with superfluous time smaller weight, than 21 mm. The total weight of 30 stages these coils will make 160 and 1020 grams accordingly.
Excuse for not so pure translation on English. But the basic sense will be clear, I hope.
Print this page