IEEE TRANSACTIONS ON PLASMA SCIENCE, VOL. 39, NO.

2, FEBRUARY 2011 725

Induction Coilgun Based on “E-Shaped” Design

Laurian Gherman, Marian Pearsica, Constantin Strimbu, and Cristian-George Constantinescu

Abstract—This paper describes a device capable of throwing

aluminum rings in a range of a few meters. The device is a
modification of the jumping-ring design invented by the American
engineer and inventor E. Thomson to demonstrate his pioneering
research in alternating current. We improved the magnetic circuit
of this device and created a new induction coilgun based on
“E-shaped” design. We experimented with this new design under
laboratory conditions and simulated its behavior using an inter-
active software package. Finally, we concluded that this device is
able to achieve a higher speed than the Thomson design with lesser
power consumption.
Index Terms—Coilgun, electromagnetic launching, induction
accelerators, transformers, transient analysis.

I. I NTRODUCTION

E LECTROMAGNETIC launch technologies open a new

direction of research in increasing the range of guns
and speed of projectiles. Two main directions are taken into
consideration: railgun and coilgun technology. The induction-
coilgun technologies promise the same performance like the
railgun with no contact between the projectile and accelerator Fig. 1. Thomson design.
[1]. The most popular of these is the tubular type in which the
primary simply consists of a row of coaxial coils. All these
coils create a barrel which accelerate the projectile inside. It
is a multistage accelerator. Because of this, the complexity of
the device is great.
Another approach is an old design invented by the American
engineer and inventor E. Thomson to demonstrate his pioneer-
ing research in alternating current and high frequency, which
was well known as the “Thompson jumping ring.” Usually, this
device is used to demonstrate magnetic levitation in a steady-
state mode. However, in transient state, this device is capable
of throwing aluminum rings at a range of a few meters. We
start our work based on that old design because this approach Fig. 2. Variation of ring speed for the Thomson design obtained during
uses the properties of soft magnetic materials. With soft mag- experimentation.
netic materials, we can create strong magnetic fields using an
ordinary source of power. First, we recreated the Thompson formance to launch rings. The Thomson jumping-ring device
jumping-ring device, and we used a standard power source of used for simulation is shown in Fig. 1 and is named Thomson.
240 V 50 Hz, with the possibility to change the voltage from 0 Fig. 2 shows the variation of ring speed with voltage applied
to 50 V, to drive the coil. Above 50 V, the coil current increases to the coil measured during experimentation. The values of the
too much and can damage the coil. We experimented with this ring speed are presented to be compared with the simulation
device and conclude that this design cannot achieve a good per- results shown in Fig. 5.

II. T HEORY
Manuscript received January 6, 2010; revised May 5, 2010, August 16, 2010, The theory behind this device is based on two fundamental
and September 30, 2010; accepted November 1, 2010. Date of publication laws: the Faraday’s law of induction and Lorentz force law.
December 17, 2010; date of current version February 9, 2011.
L. Gherman is with the National Defense University “Carol I,” 050662 The Faraday’s law of induction states that the induced elec-
Bucharest, Romania (e-mail: [email protected]). tromotive force (EMF) in a closed current loop is proportional
M. Pearsica, C. Strimbu, and C.-G. Constantinescu are with the Air Force to the negative of the rate of change of magnetic flux through
Academy “Henri Coanda,” 500183 Brasov, Romania (e-mail: marianpearsica@
yahoo.com; [email protected]; [email protected]).
the loop. When applied to an inductor coil, the formula becomes
Color versions of one or more of the figures in this paper are available online
dΦ
at http://ieeexplore.ieee.org. ε = −N . (1)
Digital Object Identifier 10.1109/TPS.2010.2091650 dt

0093-3813/$26.00 © 2010 IEEE

726 IEEE TRANSACTIONS ON PLASMA SCIENCE, VOL. 39, NO. 2, FEBRUARY 2011

For our design, the rate at which the flux of the magnetic field
through the circuit changes (dΦ/dt) is produced by a source
of sinusoidal varying EMF connected to the coil and induces
an EMF in the shorted ring. This is the principle of operation
found in a transformer.
The force responsible for accelerating the ring is the Lorentz
force. The force acting on the ring is

F = I dl × B  (2)

where I is the induced current in the ring and B is the flux

density of the magnetic field created by the primary coil.
According to the theory that is worked out by Saslow, the
force on a ring of radius r along the axis is

F = 2πrIr Brad (3)

where Ir is the induced current in the ring and Brad is the

radial component of the magnetic flux densities produced by Fig. 3. E-shaped design.
the primary coil.
Because the dimensions of the ring are fixed, the way to where Uc is the voltage applied to the coil, Rc is the resistance
increase F is by increasing Ir or Brad . The alternating current of the coil, and Lc is the inductance of the coil.
in the coil is The resulting Ic max can be two times more than Ic under
ic = Ic sin ωt (4) short-circuit steady state.
We choose to make the ring using aluminum to get a low
where resistance Rr . As a result Ir will increase, getting a stronger
force F in this way.
ω = 2πf (5) In order to get the strongest force F , we should also in-
crease Brad . However, both Ir and Brad are produced by
The system produces an axial magnetic field with a radial the same magnetic flux. It is not possible to increase Brad
component Brad without decreasing Ir , and because the force is a vector product
Nc (Brad × Ir ), for the strongest force, the magnitude of vector
Brad = μ0 KIc sin ωt (6) Brad must to be equal to the magnitude of vector Ir . Using the
lc
Thomson device, it is not possible to control the magnitude of
where K is a geometrical constant of the flaring out of magnetic vector Brad .
field, Ic is the coil current, Nc is the number of turns, and lc is We improved the Thomson device by adding an E-shaped
the length of the coil. magnetic circuit around the coil. It results in a new device,
By Faraday’s Law, the axial component of the alternat- shown in Fig. 3 and named it E-shaped.
ing magnetic field that passes through the ring induces a With a magnetic circuit, it is possible to control the mag-
current. The ring is an LR circuit with inductance Lr and nitude of vector Brad inside the air gap created. Inside, the
resistance Rr . gap in the ring and the flux lines will have only the horizontal
It can be shown that component of the magnetic flux density Bgap = Brad .
Because we power the coil with high currents, the system will
M Ic ω
Ir = (Rr cos ωt + ωLr sin ωt) (7) work deeply into the saturation region of the magnetic material.
Rr2 + ω 2 L2r
As a result, some flux lines will be in the air with a horizontal
where M is the mutual inductance component of the magnetic flux density Bair in the region
The force on the ring results, by substituting (6) and (7) into of ring.
(3) [2], to In this case, the force on the ring is
 
2πrKNc M ω 1 F = Fair + Fgap = lair Ir Bair + lgap Ir Bgap (10)
F = μ0 Ic2 R r sin 2ωt + ωLr sin2
ωt .
lc (Rr2 + ω 2 L2r ) 2
(8) where lair is the length of the ring in air and lgap is the length
of the ring in the gap
Increasing Ic will result in an increased F . In order to
increase Ic , we use the effects of inrush current under short-
lair + lgap = 2πr. (11)
circuit condition on transformers. This device is similar to a
single-phase two-winding transformer with the secondary short The system must have symmetric geometry because the re-
circuited. Under that condition, the current Ic is [9] sulting force should have only axial component. In order to find
Uc  πRc
 the force on ring in this case, we will make some assumptions.
Ic max =  1 + e− ωLc (9)
Rc2 + ω 2 L2c 1) The core is saturated only inside the coil.
GHERMAN et al.: INDUCTION COILGUN BASED ON “E-SHAPED” DESIGN 727

2) The number of flux lines in air is very low, so we consider

all flux lines inside the core.
3) The field inside the air gap is uniform.
As a result, Fair ≈ 0, and F ≈ Fgap . Because both lateral
columns are symmetrical with the same air gap and the columns
have a square section, we can write
πr πr
F = F1gap + F2gap = Ir Bgap + Ir Bgap = πrIr Bgap
2 2
(12)
where πr/2 is the length of the arc of the ring inside the air gap Fig. 4. Variation of the ring speed for the E-shaped design obtained during
(the ring is the circumscribed circle of the square section of the experimentation.
central column).
We can write Substituting into (22), the equation of force becomes
√  
φ = Φc sin ωt (13) π 2(he − hr )ωΦ2c 1
√ F = Rr sin 2ωt + ωLr sin ωt .
2
φ 4h2e (Rr2 + ω 2 L2r ) 2
= Bgap Agap = Bgap he r 2 (14)
2 (25)
Φc
Bgap = √ sin ωt (15) The (1/2) sin 2ωt term integrated over a cycle is zero, but
2he r 2 the force is acting on the ring only for the first half-period, so
√ this component of force will oscillate with a frequency of 2ω.
where he is the height of the columns and r 2 is length of the The sin2 ωt term is always positive. As a result, for one value of
sides of the square section in the central column. voltage applied to the coil, we obtain different values of force.
According to (1) For maximum force, the sum of both terms must to be at
dΦv maximum values. We can control this with the voltage phase at
εr = = Φv ω cos(ωt − α) (16) t0 when the coil is connected to a supply source. The maximum
dt  
(he − hr )Φc hr force results to
Φv = = 1− Φc (17) √  
he he π 2(he − hr )ωΦ2c 1
Fm = ωLr − R r . (26)
where 4h2e (Rr2 + ω 2 L2r ) 2

ωLr
α = arctan . (18)
Rr
III. S IMULATIONS AND E XPERIMENTATION
hr is height of the ring, and Φv is the vertical magnetic flux The dimensions of the E-shaped device were calculated to
inside the ring surface. work on a steady-state mode at 60 V 50 Hz. This voltage gives
From (16) and (17), we can write us the possibility to drive the device into the saturated region of
(he − hr )ωΦc a magnetic circuit at 240 V.
Ir =  cos(ωt − α). (19) We used a round coil, a E 12.5 sheet of silicon steel
he Rr2 + ω 2 L2r
grain orientated with a 0.35-mm thickness. Above the coil,
Now, by substituting Ir and Bgap into (12), we can write the all three columns have the same dimensions of 25-mm width,
force on the ring as 25-mm depth, and 320-mm height. Copper wire was used for
coil winding. The dimensions of the coil are 40-mm inner
(he − hr )ωΦc Φc diameter, 52-mm outer diameter, and 58-mm height. The coil
F = πr  cos(ωt − α) √ sin ωt (20)
he Rr + ω Lr
2 2 2 2he r 2 has 300 turns.
√ Aluminum was used for the ring, having the dimensions of
π 2(he − hr )ωΦ2c 45-mm inner diameter, 50-mm outer diameter, 5-mm height,
F =  sin ωt cos(ωt − α) (21)
4h2e Rr2 + ω 2 L2r and 10-g weight.
√   An interactive software package that uses the finite element
π 2(he − hr )ωΦ2c 1 method (FEM) was used to analyze, solve 3-D electromagnetic-
F =  sin 2ωt cos α + sin ωt sin α .
2
4h2e Rr2 + ω 2 L2r 2 field problems, and simulate both Thomson and E-shaped de-
sign behavior.
(22)
In order to check the accuracy of the simulation results,
From (18), we can write we built the Thomson and E-shaped devices with the same
dimensions and materials, which were experimented in the
Rr same conditions like the simulated ones. Comparing the initial
cos α =  (23)
Rr2 + ω 2 L2r speed of the ring from experiments (Figs. 2 and 4) with the
initial speed of the ring from FEM simulations (Fig. 5), we
ωL concluded that the simulation offers a good approximation of
sin α =  . (24)
Rr2 + ω 2 L2r reality.
728 IEEE TRANSACTIONS ON PLASMA SCIENCE, VOL. 39, NO. 2, FEBRUARY 2011

Fig. 5. FEM-simulated values of the ring speed for both designs. Fig. 8. Values of the coil current at the time when the Lorentz force acting on
ring achieves the maximum value.

However, according to (3), the Lorentz force depends on the

induced current Ir in the ring.
Fig. 7 shows the variation of the current induced in the
ring at the time when the Lorentz force achieves maximum
value, with voltage applied to the coil, for both designs from
FEM simulation. We can see that the ring current for the
E-shaped design is lower than that of the Thomson design,
as we predicted, but the force is greater. Therefore, the radial
magnetic-flux density is increased for the E-shaped design. For
a known voltage applied to the coil, we can calculate the ring
current, and we can adjust the height of the columns to obtain
the value of the radial magnetic-flux density for strongest force.
Fig. 6. Time variation of the Lorentz force for the E-shaped design (square Fig. 8 shows the values of the coil current at a time when
dots for 30 V and triangle dots for 240 V). the Lorentz force acting on ring achieves the maximum value.
The values of the coil current are almost the same for both
designs. For the same voltage and coil current used to power
both devices, the ring speed is different. We can state from the
data that the efficiency of the E-shaped design is greater than
the Thomson device.
Because the devices are working in transient mode, the in-
rush current under short-circuit conditions drives the magnetic
circuit inside the coil into the saturation region. Here, the
behavior of the device is nonlinear. However, in order to keep
things simple, we should remember the chain of dependences.
Therefore, the speed of the ring depends on the force. The
Fig. 7. Values of the current induced in the ring at the time when the Lorentz force depends on the ring current Ir and the radial magnetic
force acting on the ring achieves the maximum value.
flux densities Brad . Both of them depend on the magnetic flux
created by the coil. The magnetic flux depends on the voltage
Fig. 4 shows the variation of the ring speed with the voltage applied but not on the coil current. Keep in mind that the force
applied to the coil measured for the E-shaped design. The is acting only in the first half-period of the cycle.
change of slope around 90 V is caused by saturation of the Fig. 9 shows the time variation of the coil current at 240 V
core. In order to find the initial speed of the ring during 50 Hz, measured with an oscilloscope. Fig. 10 shows the
experimentation, we launched the ring into a vertical direction variation of the coil current observed at the beginning of the
and measured the maximum height achieved. graph in Fig. 9.
We studied the behavior of both devices at 30, 60, 90, 120, Now, because the device is working on saturated condition,
150, 180, 210, and 240 V under simulated conditions. the phenomenon observed at the beginning of the graph in
The results obtained from the FEM simulation are shown in Fig. 9 could be ferroresonance. It implies resonance with
Figs. 5–8. Fig. 5 shows the variation of the ring speed with the saturable inductance. The ferroresonance phenomenon appears
voltage applied to the coil for both designs under simulated after transient disturbances, and its effects are characterized by
conditions. It displays only the maximum values of the ring high sustained overvoltages and overcurrents with maintained
speed. levels of current and voltage waveform distortion. In our case,
The value of the ring speed depends on the voltage phase at the variation of the coil current occurs only for a very short time
t0 , when the coil is connected to a supply source (ωt0 ). For because the ring is moving.
this reason, we choose only the maximum values of the ring The interactive software package used for simulation cannot
speed to be displayed. For all values of voltage applied to the simulate the ferroresonant phenomenon. During experimen-
solenoid, the ring speed is higher for the E-shaped design than tation, we observed that the speed of the ring without that
for the Thomson design. Also, the speed of the ring is increasing phenomenon is lower than the speed of the ring with this
even for the saturated core. variation of coil current. Also, this variation of the coil current
GHERMAN et al.: INDUCTION COILGUN BASED ON “E-SHAPED” DESIGN 729

[3] V. Matos, L. Silva, and J. S. Esteves, “Induction coil gun,” in Proc. 3rd Int.
Conf. Hands-On Sci., 2006, pp. 165–168.
[4] D. Vachtl, D. Kováč, and D. Mayer, “Forces calculation method of
Thompson levitating ring during its non-harmonic feeding,” Acta Elec-
trotechnica et Informatica, vol. 6, no. 2, pp. 5–10, 2006.
[5] H. D. Fair, “Progress in electromagnetic launch science and technology,”
IEEE Trans. Magn., vol. 43, no. 1, pp. 93–98, Jan. 2007.
[6] A. Balikci, Z. Zabar, L. Birenbaum, and D. Czarkowski, “On the design of
coilguns for super-velocity launchers,” IEEE Trans. Magn., vol. 43, no. 1,
pp. 107–110, Jan. 2007.
[7] A. C. De Azevedo, A. C. Delaiba, J. C. De Oliveira, B. C. Carvalho, and
H. De S. Bronzeado, “Transformer mechanical stress caused by external
short-circuit: A time domain approach,” in Proc. IPST, Lyon, France,
Jun. 4–7, 2007.
[8] B. Marder, “A coilgun design primer,” IEEE Trans. Magn., vol. 29, no. 1,
pp. 701–705, Jan. 1993.
[9] M. B. B. Sharifian, R. Esmaeilzadeh, M. Farrokhifar, J. Faiz, M. Ghadimi,
and G. Ahrabian, “Computation of a single-phase shell-type transformer
windings forces caused by inrush and short-circuit currents,” J. Comput.
Fig. 9. Variation of the coil current at 240 V 50 Hz for the E-shaped design. Sci., vol. 4, no. 1, pp. 51–58, 2008.

Laurian Gherman was born in Romania, in 1972.

He received the B.Eng. degree from the Politehnica
University of Timisoara, Romania and Ph.D. de-
gree from the Transilvania University of Brasov,
Romania.
Currently, he is a Major Assistant Professor with
the Regional Department of Defense Resources
Management Studies, National Defense University
“Carol I,” Brasov, Romania. His teaching and re-
search activities include electromagnetic-field the-
ory, magnetic levitation, and electromagnetic launch
systems.

Fig. 10. Variation of the coil current at 240 V 50 Hz observed at the beginning
of the graph in Fig. 9. Marian Pearsica was born in Romania, in 1964. He
received the B.Eng. degree from The Military Tech-
nical Academy of Bucharest, Romania and Ph.D.
depends on the voltage phase at t0 when the coil is connected to degree from the Politehnica University of Bucharest,
the supply source (ωt0 ). Further investigation should be done Romania.
in order to determine the real causes of this phenomenon. He is currently LTC Professor with the IT and
Finally, by adding a magnetic circuit around the coil, we can Electronic Department, Air Force Academy “Henri
Coanda,” Brasov, Romania. His teaching and re-
increase the performance of the Thomson jumping-ring device. search activities include electromagnetic-field theory
and electric motors.

IV. C ONCLUSION
This paper has presented a new E-shaped design that is
Constantin Strimbu was born in Romania, in 1964.
superior in performance to the original Thomson jumping- He received the B.Eng. degree from the Politehnic
ring design based on experimental measurements and FEM Institute of Cluj-Napoca, Romania and Ph.D. de-
simulations. gree from the Transilvania University of Brasov,
Based on these results, the next configuration for study is Romania.
He is currently an Assistant Professor with the
a multistage coilgun that utilizes similar E-shaped magnetic IT and Electronic Department, Air Force Academy
material for enhancing launcher performance. Because the ring “Henri Coanda,” Brasov, Romania. His teaching and
is not a projectile, two or more rings can be attached to a real research activities include microwaves, simulation,
projectile. If each ring is launched by an E-shaped coilgun, and modeling of electric circuits.
we can obtain a parallel multistage coilgun. Synchronizing all
launchers will be the big problem.
Another direction of research is to find the right magnetic Cristian-George Constantinescu was born in
material, dimensions of magnetic circuit, and the frequency of Romania, in 1959. He received the B.Eng. and Ph.D.
power source to achieve maximum speed for the ring. degrees from the Transilvania University of Brasov,
Romania.
He is currently an Assistant Professor with the
R EFERENCES IT and Electronic Department, Air Force Academy
“Henri Coanda,” Brasov, Romania. His teaching and
[1] E. R. Laithwaite, “Induction coil guns for hypervelocities,” Proc. Inst.
research activities include probabilities, electric, and
Elect. Eng.—Elect. Power Appl., vol. 142, no. 3, pp. 215–221, May 1995.
electronic circuits.
[2] J. Hall, “Forces on the jumping ring,” Phys. Teacher, vol. 35, pp. 80–83,
Feb. 1997.