CAPACITIVE FLUX COMPRESSION GENERATOR

A. Cheldavi

Department of Electrical Engineering, Iran University of Science and Technology

Narmak,16844, Tehran, Iran, Fax +98-21-7454055 , Tel +98-21-7808022, cheldavi@iust.ac.ir

M. M. Danaei

Imam Hossein University

Tehran, Iran

(Received: July 2, 2002 - Accepted in Revised Form: May 15, 2003)

Abstract Conventional Flux Compression Generators (FCG's) are used to generate high power DC
pulses. A new kind of (FCG's) with series capacitance called Capacitive Flux Compression Generator
(CFCG) will be introduced and explained in this paper. This new kind is used to generate modulated
high power pulses. There are some problems to establish a capacitance in high power and high
frequency applications. In the present paper several practical methods will be addresed to make
capacitance in high power and high frequency applications.

Key Words Flux Compression Generator, High Power Sources, High Frequency Sources

‫ ﺳﺎﺧﺘﻪ ﺷﺪﻩ ﻭ ﻣﻮﺭﺩ ﺍﺳﺘﻔﺎﺩﻩ ﻗﺮﺍﺭ‬DC ‫ﭼﮑﻴﺪﻩ ﻣﻮﻟﺪ ﻫﺎﻱ ﺷﺎﺭ ﻓﺸﺮﺩﻩ ﻣﻌﻤﻮﻟﻲ ﺑﺮﺍﻱ ﺗﻮﻟﻴﺪ ﭘﺎﻟﺴﻬﺎﻱ ﺗﻮﺍﻥ ﺯﻳﺎﺩ‬
‫ ﻧﻮﻉ ﺟﺪﻳﺪﻱ ﺍﺯ ﺍﻳﻦ ﻣﻮﻟﺪﻫﺎ ﺗﺤﺖ ﻋﻨﻮﺍﻥ ﻣﻮﻟﺪ ﺷﺎﺭ ﻓﺸﺮﺩﻩ ﺧﺎﺯﻧﻲ ﺩﺭ ﺍﻳﻦ ﻣﻘﺎﻟﻪ ﻣﻌﺮﻓﻲ ﻣﻲ ﺷﻮﺩ ﻛﻪ‬.‫ﻣﻲ ﮔﻴﺮﻧﺪ‬
‫ ﺑﺮﺍﻱ ﺍﻳﻦ ﻛﺎﺭ ﺍﺯ ﻳﻚ‬.‫ﺗﻮﺍﻧﺎﻳﻲ ﺍﻳﺠﺎﺩ ﭘﺎﻟﺲ ﻫﺎﻱ ﺗﻮﺍﻥ ﺯﻳﺎﺩ ﺩﺭ ﻓﺮﻛﺎﻧﺲ ﺩﻣﺎﻱ ﺑﺎﻻ )ﺩﺭ ﺣﺪ ﭼﻨﺪ ﺩﻩ ﻣﮕﺎ ﻫﺮﺗﺰ( ﺭﺍ ﺩﺍﺭﺩ‬
‫ ﺑﺮﺍﻱ ﺍﻳﺠﺎﺩ ﺍﻳﻦ ﺧﺎﺯﻥ ﺳﺮﻱ ﺑﺎﻳﺪ ﺍﺯ‬.‫ ﻣﻮﺭﺩ ﺍﺳﺘﻔﺎﺩﻩ ﻗﺮﺍﺭ ﻣﻲ ﮔﻴﺮﺩ‬،‫ﺧﺎﺯﻥ ﺳﺮﻱ ﻛﻪ ﺑﺎ ﺳﻠﻒ ﻣﻮﻟﺪ ﺭﺯﻧﺎﻧﺲ ﻣﻲ ﻛﻨﺪ‬
.‫ﺭﻭﺷﻬﺎﻱ ﺧﺎﺻﻲ ﺍﺳﺘﻔﺎﺩﻩ ﻛﺮﺩ ﻛﻪ ﺧﺎﺻﻴﺖ ﺧﺎﺯﻧﻲ ﺩﺭ ﺟﺮﻳﺎﻥ ﻭ ﻓﺮﻛﺎﻧﺲ ﻫﺎﻱ ﺑﺎﻻ ﺣﻔﻆ ﺷﻮﺩ‬

1. INTRODUCTION briefly in [15]. Some possible mechanisms of

losses have been developed in [16]-[17]. S. I.
FCG's are used extensively in several applications Shkuratov [18] focused on the development of
especially when high current pulses are needed in generators using kinetic energy of ferromagnetic
applications such as electromagnetic launchers, projectiles. [19]-[21], explained several concepts
welding metal forming, geological mapping, and for production of initial energy to power magnetic
radar. FCG's can be considered as linear electrical FCG's, and finally [22]-[24] present some special
generators. Electromagnetic energy is stored in an noval FCG's. Capacitive flux compression generator
inductor with an initial "seed current". Mechanical (CFCG) is a devise to generate high frequency
motion causes a decrease in the system inductance modulated high power pulses. Its function is the
at such a rapid rate that the flux interlinked with it subject of a limited number of researches [25-26].
has no time to diffuse out and, therefore remains Fundamentals of high explosives, CFCG principles
practically constant. FCG's have been the subject with an emphasis on the helical generator, as well
of many studies [1-20]. The principal ideas on as limiting physical mechanisms will be addressed
FCG's and its applications following the interesting in this paper.
work of Kapitza [1] have been presented in [2]-
[12]. [13] Addressed some problems regarding
microsecond pulsed high power generators. [14] 2. ONE DIMENSIONAL MODELING
Introduce a FCG code with special interesting
capabilities. Figure 1 shows the circuit model for CFCG. It is
Proximity and skin effects have been explained supposed that R and C are constants and L(t) is a

IJE Transactions A: Basics Vol. 16, No. 4, December 2003 - 337

 Differentiating 4 yields to

d 2 i di dL di i
L 2 + + (R − a ) + = 0 (5)
dt dt dt dt C

or

d 2i di i
L 2
+ (R − 2a ) + = 0 (6)
Figure 1. The circuit model for CFCG. dt dt C

The initial current and voltage stored in the

inductor and capacitor are i(0) = I0 and vc(0) = 0.
So one has

v c (0) + L(0)i ′(0) + Ri(0) + i(0)L′(0) = 0 (7)

or

(a − R ) I 0
i ′(0) = (8)
L0

Defining
Figure 2. L(t) as a function of time.

 t R − 2a  2−
R

variable as shown in Figure 2. The differential p( t ) = exp  ∫ dτ = (L 0 − at ) a (9)

equation describing the system is given as  0 L 0 − aτ 

d
[Li] + Ri + 1 ∫ idt = 0
R
(1) p( t ) 1−

dt C s( t ) = = (L 0 − at ) a (10)
L 0 − at
in which
6 can be written as
L 0 − L1
L = L0 − t = L 0 − at (2) d di  s( t )
T  p( t )  + i=0 (11)
dt  dt  C
From 1 one has
It can be shown that 11 is self adjoint. So it has a
di dL 1 unique solution as
L +i + Ri + ∫ idt = 0 (3)
dt dt C ∞
i( t ) = ∑ a n t r + n (12)
or n =0

di 1 where an and r are constants to be determined.

L + (R − a )i + ∫ idt = 0 (4)
dt C Setting 12 in 6 and choosing r = 0 gets to

338 - Vol. 16, No. 5, January 2003 IJE Transactions A: Basics

 ∞
L1
(L 0 − at )∑ a n n (n − 1) t n − 2 + 1−
n=2 a L0
So because = >> 1 , the coefficients
∞
1 ∞ L0 T
(R − 2a )∑ a n nt n −1 + ∑ antn = 0
n =1 C n =0 becomes larger and larger without limit. To
(13) overcome this problem the definition of the
coefficients in 12 must be changed as
Setting n-2=m in the first summation and n-1= m
∞
in the second summation yield to t
i( t ) = ∑ (a n T n )( ) n (19)
n+0 T
∞

∑L a
m =0
0 m+ 2 (m + 2)(m − 1) t m +
t
Choosing = τ and anTn = bn one gets to
1 T
(R − 2a )a m +1 (m + 1) t m + amtm −
C
∞
i(τ) = ∑ b n τ n
∞
a ∑ a m +1 m(m + 1) t = 0 m (20)
n+0
m =1
(14)
So
This gets to
di di dτ 1 di
= = (21)
1 dt dτ dt T dτ
2a 2 L 0 + (R − 2a )a 1 + a 0 = 0 (15)
C
Setting 21 in 6 yields to
Satisfying the boundary conditions requires that
L0 d 2i di T
( − at ) 2 + (R − 2a ) + i = 0 (22)
− RI 0 T dτ dτ C
a 0 = i(0) = I 0 ; a 1 = i ′(0) = (16)
L0
Now the coefficients are obtained as
Using a0 and a1, a2 can be obtained as
T2
b m+2 = − bm +
I 0 − 1 (R − 2a )R L 0 C(m + 2)(m + 1)
a2 = ( + ) (17)
L 0 2C 2L 0 aTm − (R − 2a )T
b m +1
L 0 (m + 2)
Using a recursive relation one gets to the final (23)
solution for the unknown coefficients as
As it is clear in the limit when m → ∞ one has
1
a m+2 =− am +
L 0 C(m + 2)(m + 1) aT L
(18) b m+2 = b m +1 = (1 − 1 )b m +1 (24)
am − (R − 2a ) L0 L0
a m +1
L 0 (m + 2)
L1
Because 1 − ≤ 1 , bm converges. Note that to
As it is clear from 18, in the limit as m → ∞ , am+2 L0
a obtain bm we need b0 = a0 and b1 = a1T. For t ≥ T
approaches to a m +1 .
L0 the current can be obtained solving the simple following

IJE Transactions A: Basics Vol. 16, No. 4, December 2003 - 339

 Figure 3. Current for L1 = 0.5µF, L0 = 10mH, C=0.05nF and T=20µs and different R .
2a

differential equation T ∞ bn L ∞
i ′(T) = ∑ − 1
RC n =0 n + 1 R
∑b
n =0
n (27)
2
d i R di 1
2
+ + i=0 (25)
dt L1 dt L1C
3. ELECTRICAL GAIN
using the initial conditions as

∞
Electrical gain is the most important factor in FCG
i (T ) = ∑ b n (26) design procedure. This parameter is defined as the
n =0 ratio of the electrical energy (delivered to the load)

340 - Vol. 16, No. 5, January 2003 IJE Transactions A: Basics

 Figure 4. Electrical gain for different
L0 .
L1

to the initial magnetic energy stored in the So the electrical gain is obtained as
inductor. The electrical energy delivered to the
∞ ∞
load can be written as bmbn
RT ∑ ∑
We n =0 m=0 n + m + 1
1 ∞ ∞
b b = (29)
We = ∫ R 2 i 2 (τ)dτ = R ∑ ∑ m n (28) Wm 1
L 0 I 02
0 n =0 m = 0 n + m + 1 2

IJE Transactions A: Basics Vol. 16, No. 4, December 2003 - 341

 4. EXAMPLES AND RESULTS Conversion”, New York, McGraw-Hill, (1966).
9. Levi, E. and Zabar, Z., “Basic Design of Electrical
Consider a circuit as shown in Figure 1. It is Machines”, IEEE Trans. Power App. Syst., Vol. PAS-
104, (1985), 1513-1518.
supposed that the initial current in the circuit is 1 10. Levi, E., Zabar, Z. and Birenbaum, L., “Basic
Ampere. The current and capacitor voltage Performance of Flux-Compression/Expansion
R Electromechanical Converters”, IEEE Trans. Plasma
waveforms for different values of , L0 = 10mH, Sci., Vol. 20, No. 5, (Oct. 1992), 554-561.
a 11. Engle, T. G., Becker, J. E., Nunnally, W. C., “Energy
L1 = 0.5µH, C = 0.05nF, and T = 20µs are shown Conversion and High Power Pulse Production Using
in Figure 3. The electrical gain ratio is shown in Miniature Magnetic Flux Compressors”, IEEE Trans.
Plasma Sci., Vol. 28, No. 5, (Oct. 2000), 1342-1346.
R L 12. Novac, B. M. et al., “A Novel Flux Compression/Dynamic
Figure 4 for different values of and 0 . Transformer Technique for High-Voltage Pulse
a L1
Generation”, IEEE Trans. Plasma Sci., Vol. 28, No. 5,
(Oct. 2000), 1356-1361.
13. Turchi, P. J., “Problems and Prospects for Microsecond
Pulsed Power Above Ten Mega Ampers”, IEEE Trans.
Plasma Sci., Vol. 28, No. 5, (Oct. 2000), 1414-1421.
4. CONCLUSIONS 14. Cash, M. A. et al., “Benchmarking of a Flux Compression
Generator Code”, IEEE Trans. Plasma Sci., Vol. 28, No.
A simple one-dimensional approach is given for 5, (Oct. 2000), 1434-1439.
analysis and design of CFCG. The current gain and 15. Novac, B. M., Smith, I. R. and Enache, M. C., “Accurate
Modeling of the Proximity Effect in Helical Flux-
the electrical gain are obtained. This approach is
Compression Generators”, IEEE Trans. Plasma Sci., Vol.
based on the circuit principles, so it can be used for 28, No. 5, (Oct. 2000), 1353-1355.
the design of CFCG up to few hundred megahertz. 16. Neuber, A. A. et al., “Optical Diagnostics on Helical Flux
For higher frequency design and analysis two or Compression Generators”, IEEE Trans. Plasma Sci., Vol.
three-dimensional methods must be used which is 28, No. 5, (Oct. 2000), 1445-1450.
17. Shkuratov, S. I. et al., “Pulsed Power Generation Using
the subject of the future work of the authors. As it Open and Closed Ferromagnetic Circuits”, IEEE Trans.
is clear from the figures, this device can be Plasma Sci., Vol. 28, No. 5, (Oct. 2000), 1347-1352.
designed for electrical gain around 10000. 18. Altgilbers, L. et al., “Magnetocumulative Generators”,
New York, Springer-Verlag, (2000).
19. Fowler, C. M., Garn, W. B. and Caird, R. S., “Production
of Very High Magnetic Fields by Implosion”, J. Appl.
Phys., Vol. 31, No. 3, (1960), 588-594.
5. REFERENCES 20. Fowler, C. M., Caird, R. S. and Erickson, D. J., Eds.,
“Megagauss Technology and Pulsed Power Application”,
1. Kapitza, P. L., “Further Developments of the Method of New York, Plenum, (1987).
Obtaining Strong Magnetic Fields”, Proc. Roy. Soc., 21. Golovina, V. V. et al., “Magnet-Powered Initial Stage of
A115, (1927), 658. Magnetic Flux Compressor”, in Megagauss and
2. Lewin, J. D. and Smith, P. F., “Production of Very High Megaampere Pulse Technology and Applications, Russia,
Magnetic Field by Flux Compression”, Rev. Sci. Inst., Sarov VNIIEF, (1997), 333.
Vol. 35, (1964), 541. 22. Sheppard, M. G. et al., “MC-1 Generator Performance
3. Kuoepfel, H., “Pulsed High Magnetic Fields”, North with Higher-Energy Explosives", Tenth IEEE Pulsed
Holland, (1970). Power Conference, Albuquerque, NM, (July 1995).
4. Pande, H. C., “Linear Analysis of Electromechanical 23. Fowler, C. M. et al., “The Russian-American High
Pulser”, AIEE Trans. pt. I, Vol. 80, (1961), 309-315. Magnetic Field Collaboration”, Tenth IEEE Pulsed
5. Pande, H. C., “Current-Source Excitation Type Power Conference, Albuquerque, NM, (July 1995).
Electromechanical Pulser”, AIEE Trans. Communication 24. Fowler, C. M., “Four Decades in the Megagauss World”,
and Electronics, Vol. 74, (1964), 528-535. Physica B, 246, (1998), 158.
6. Pretap, S. B. et al., “A Compulsator Driven Rapid-Fire 25. Onochin, V. V., “Analysis of the Equivalent Circuit of
EM Gun”, IEEE Trans. Magn., Vol. 20, No. 2, (1984), Explosive Magnetic Generator of Frequency”, http://www.
211-214. laboratory.ru/articl/phys/eap050.htm.
7. Driga, M. D. et al., “Advanced Compulsator Designs”, 26. Soshenko, V. A. and Ivanov, V. C., “Investigation of the
IEEE Trans. Magn., Vol. 25, (1989), 142-146. Parameters of Explosive Driven Magnetic Generator of
8. Levi, E. and Panzer, M., “Electromechanical Power Frequency”, Proc. of EUROM Conf., (Aug. 1998).

342 - Vol. 16, No. 5, January 2003 IJE Transactions A: Basics