142

ISEETRANSACTIONS ON MICROWAVETHEORYAND TSCHNIQURS, VOL. MIT-28, NO. 2, FEBRUARY1980

Theoretical Basis for Microwave and RF Field Effects on Excitable Cellular Membranes
~

CHARLES

A. CAIN,

MEMBER, IEEE

Abstracr-A model of a mdarrfsm for nontherroaf interaction of RF or mlcsowave fields with excitable celhfar membranes is presented. It may be poasilie for an oscillating component of membrane potential to chsnge the ctmductsnce of the membrane to all ion speeies wfdeb transverse vokageefepemkmt membrane -Is. Some specffic effects on squid gbmt axon predkte dbythernode laredkrswd.

where dn/dt dm/dt = a~(l n)~nn =a.(1 m) B.m (2) (3) (4)

dh/dt=afi(l-h)-&h and

INTRODUCTION XC1TA13LE E

CELLS

are often extremely

sensitive

to an= O.Ol(V+ 10)/

very small changes in the electrostatic field across the cellular membrane. A number of researchers have published somewhat contradictory evidence which supports [1][6] and does not support [7][10] the hypothesis that RF and microwave nonthermal provide oscillating frequencies mechanisms. electromagnetic might directly a theoretical field8 can affect excitable It is the purpose basis for an understanding cells by of this paper to of how

v+

exp~

10

1 )

(5) (6)

j3. = 0.125 exp( V/80)

Clm

=0.l(V+25)/(exp* exp( V/18)

1)

(7) (8) (9)

(10)

&=4

ah= 0.07 exp( V/20)

fields at RF and microwave affect nerve, muscle, and other In these equations V is the displacement

electrically excitable tissue, In what follows, the classical mathematical model formulated by Hodgkin and Huxley [11] for the current-voltage relations in the membrane of the giant axon L.oligo will provide a framework for further discussion, In the Hodgkin nonlinear and Huxley functions (FH-1) model, of the These certain parameters electric func%ions are field of the instantaneous nonlinear

of membrane

potential from the resting value (depolarization negative). Constants CM, ~K, ~~a, ~1, VK, VN., and VI are explained in detail in [11], The value of all constants used in the computations conductance, which follow respectively. as-egiven in Tabls are the potassium The dmensionl~ss 1, and sodium dynamical In (1) ~Kn4 and &am3h

across the membrane.

electric field provide the basis for a number microwave effects on nerve function,
THE N41MBRMTE MODEL

of possible

h 1952 Hodgkin and Huxley [11] published a set of equations based cm extensive experimental data which expressed the total squid axon membrane current as functions of instantaneous transmembrane potential and time.

quantities m, n, and h are solutions of the first-order differential equations (2)-(4), respectively, and vary between zero and unity after a change in membrane potemtial, The as and /3s in these equations depend cndy on the instantaneous value of membrane potential and are given in (5)-(10). The proper units for use of the preceding equations are as follows: pA/cm2, pF/cm2, potentials conductance are expressed in mS/cm2, in rnV, current in in capacitance

This highly achievement

predictive

empirical current

formulation

was a major model, of V is

of quantitative

biology, 1, written

In the HH as a function

and t in rns. The expressions of 6.3C.

AND RF

for the as and ~s

the total membrane

hold only at a temperature

T3i13 ~ASIS FOR
&fKXOWAW3

expressed by the following 1=

system of coupled

equations:

~IELD

~FFEGT5

CM% +-gKn4(

VK) vNa)+g,(v~1) (1)

+gNa)n%(v

The a and fi rate constants were and Huxley to be functions of the field across the membrane [1 1], We this assumption as an alternating total field across the membrane can
v(t) = V.+ Vm

assumed by Hodgkin instantaneous electric will continue to make field is applied. The be expressed as

Manuscript received January 29, 1979; revised July 12, 1979. The author is with Bioacoustics Research Laboratory Department of Electrical Engineering, University of Illinois at Urbana-Champaign, Urbana, IL 61801.

Cos (&w
of the membrane

(11)
potenof

where

VO is the dc displacement its resting

tial from

state and V~ is the peak amplitude

0018-9480/80/0200-0142$00,75

@1980 IEEE

CA3N : M3CROWAVE

-.

.AND RF --.----. F1B2JIEITEU1>

20, 16 . \ -~ ;B
12 ~, ~.

143

TABLE L VALUS OF CONSTANTS USED WHSN COMPUTINGA SOLUTION TO THE HH J?WJAT30NS (rAx33N FROM[1 11)
CONSTANT VALUE 1.0 UNITS pF/cm2 m

1 Vm=OOmV
------.

Vm = 250 mV V. z 250 nW, AVG VALUE ,/f .. *-------,/ ,/ ,/ ~-., \\ ..(U1I \ 1, --.,-. ..-

,-

_. -

~, . /3m(dJt)

+4 z $0

-115

+ 12
- 10.6 120 36 0.3

~ m m msl CQL2 mSf cim2 msjcll?

::[

& 04 -.+-.-+< e 02 ~ #- ---0 =

~.s--__
77 d 57r/4 3T12 27(

~
0
77/4

(rod)

the oscillating The obtained (11). becomes

component values

of applied of

membrane ~s by V(t)

potential. are then by as given

Fig. 1. Instantaneous values of rate constants /3~(@ and w(@ when VO= O and V-II= 25 mV where the membrane potential is V= Vo+ Vm cos d (-); the average values j3w,o and ~,. when Vm =25 mV and vo -O (--); and the rate Constanb when VO= v. = O ().

instantaneous

the as and

by replacing when

V in (5)-(10)

For example,

V in (8) is replacedby(11), nonnegative series as

+W

then ~~ can be

a periodic

function

which

expressed in a Fourier

where

The average value of ~~(~t)

is (14)

/3no=&~2%n(@)dwt
The quantity ing function /3~,0 is a nonnegative of V~. Therefore, monotonically an oscillating electric across the membrane

increasfield

can result in a steady (de) change in

all a and /3 rate constants.

SOME RESULTS
00

I
10

1 20

30

I 40
Vm (mV)

i 50

60

70

80

The instantaneous VO= O over a full Fig. 1 along with monotonic is illustrated have unity rate constants

values

of @w(tot) and am(ut)

when

period and V.= 25 mV are shown in the average values BM,O and a~,r The of increasing this figure, V when all rate V.= O

increase in the average value of all six a and ~ as a function in Fig. value when 2. In constant

Fig. 2. The normalized average value of alf rate constants as a function of increasing Vm when V.= O. AU average values have been normalized so that the rate constants have unity value when V. =0, e.g.,~& is the ratio of ~n,o when Vm # O to the value of /3m,owhen Vm = 0.

average values have been normalized V~ = O.

so that the constants

Fig. 3 is a plot of both am and ~m as a function of the dc membrane potential V@ Note that the average value of both am and ~~ increases with an increase in VW. The a and ~ rate constants determine both the rate of change in m, n, and h and the final these quantities. step input is
403 _ % ~q +

-,- 10~ & In z

w 1*4 a

steady-state value reached

value of after a

The

steady-state

136

Pq
is

(15)
2 -

where q is m, n, or h. The time constant for q(t) after a step input 9 1 = . ~q + Pq

-100

-20 V. (mV)

20

(16)

Fig. 3.

Average values of rate constants ~ and /3~ as a function of V. for V-=0 and Vm =25 mV.

ISEBTRANSACTIONS ON MICROWAVSTHSORYAND mCHiWQUSS,VOL. MIT-28, NO. 2, FEBRUARY 1980

I
I

I.0 ~0,9 0,8 0.7 8 ~ 0.6 6 :0,5

E 0.4 -

t \ \ , \\ \ \ \ \ \\ \

-----. Vm .0.0 -Vm.25,0mV mV

\ \ \ \ \

0,3 -

0.2 -

0,1 -

-1oo

80

60

-40 VO (mV)

-20

20

40

Fig. 4.

Steady-state values of mm, n~, and hm as a function of V. for and V~=25 mV. mw, n~, Rnd % comput~ from (27) using the numerically computed average values of the a and /3 rate constants as given by (14).
Vm=O

A plot of m@, n@, and h~ as a function of VOwith V~ as a parameter is given in Fig. 4. Let us define the applied potential across the membrane to be v(t)= where potential microwave this applied u(t)
is

Vmcos @t[u(t) unit

U(f

to)] Such an

(17) applied with a at to

Zo

the

step function.

~ L

0<03 -

can result

if the membrane

is irradiated

or RF pulse beginning ac field can be

at t = O and ending potential in response by numerically

i-

time to. The change in membrane

obtained

solving ( 1)( 10) with the initial conditions being the resting state values for V, m, n, and k Since no current flows through the membrane from external sources, 1 in (1) is set to zero, and the axon is assumed to be irradiated uniformly along the length such that dV/dX = O. After each iteration through the EIH equations, accurate estimates for the average values of the a(tit)s and ~(@s are obtained analogous tial
one , , I , 1 I , d t I 20 I 25

-4L L

PULSE ON (fm=250mV)

LPULSE TIME

0FF5 (m see)

numerically

using

the

appropriate field

expression potenfor is shown

to (14). The response

of Vm in Fig. 5.

of the membrane

Fig. 5. Response of model axon to a puked oscillating component of membrane electric field (10-ms pulse, Vfi = 25 rev). The membrane potential and the sodium and potassium conductartces are shown. These curves are solutions of the HH equations (see text).

to the pulse of applied

value

oscillating

The changes in membrane potential shown in Fig. 5 result from changes in both the sodium and potassium conductance. The time course of both these conductance following the pulse of applied ac field (as given in (17)) is also shown in Fig. 5. The conductance were computed after obtaining numerical solutions to the HH equations for m(t), n(t), and h(t) with Conductance gK and g~, are then
gK(~) = ~Kn4(~)

FIELD

WITHIN

THE

MEMBRANE

In estimating the level of the oscillating component of electric field developed within the membrane, the model discussed by Barnes and Hu [12] was used. Without repeating their formulation, a membrane near the surface in high water content tissue exposed to a 10 mW/cm2 (194 V/m in air) plane wave would result in an electric field of 980 V/m within the nonpolar region of the lipid bilayer (assuming a dielectric constant of 2.1 [13]). This is equivalent to an applied ac membrane potential Vn of 9.8 pV if membrane thickness is 100 ~.

(17) as the input.

(18) (19)

gl%(t) = ~N@3(t)h(t).

CAIN: MICROWAVSAND RF FD3LDEFFECTS

120

145

The effects model peak intensity the resulting This should level of functional

of such a field of 10 W/cm2 field applied

as predicted (194 kV/m would effects

by the above

,
A-

I Vm=oomv =50mv Vm=60mV Vm=250mV

are not significant.

However,

if a pulsed field with a in air) is applied, will cause of applied be 9.8 mV. significant It of
< g ~

!00 -

B-v. C-

across the membrane ac field in the model

80 -

0-

INITIAL DEPOLARIZATION 70 mV 60 -

changes

we have considered. fields

be noted

that thermal

40 -

this magnitude will be considerable for all but the lowest duty cycle pulsed waveform, e.g., a duty cycle of 0.001 will result in a field thermally equivalent, on the average, to an incident intensity of 10 mW/cm2. would Whether or not effect it is such a pulsed waveform result in significant

20 -

0 u-20 I o I 5 10 POSTSTIMULUS TIME (ins) 15 20

as predicted by the model awaits further analysis. Because of the large size of the giant squid axon possible, field using an internal wire electrode, to apply directly across the membrane

an RF

(e.g., see Takashima

Fig. 6. Computed membrane action potentials in response to an initial membrane depolarization of 7 mV for different values of Vm. Curves are solutions to the HH equations,

and Schwan [14].) The highly specific predictions of the model presented here may be useful in the design of experiments and the interpretation of results. For example, the effects of directly applied high-frequency fields on membrane potential conductance should as well as static or dc membrane measured. Such experiments times of the relaxation be easily

It is by no means clear that observed wave or RF radiation attributed ation with general crowave to a thermal 960 MHz heating on excitable mechanism. at levels below and general

effects of microbe

tissue can always those which while

In turtle hearts, irradiproduce mihigher

may help determine

the dielectric

can cause bradycardia, heating neurons

gating dipoles (or charged groups) which compromise the voltage sensor for membrane ionic channels (see Discussion). From this kind of information, one might determine the upper limit of frequency for application of the theory presented here.
DISCUSSION

intensities

cause tachycardia of Aplysia, a num-

[2]. In single pacemaker-type

ber of observed changes in firing patterns in response to microwave radiation were in the opposite direction to that caused by general heating [3]. Extremely low-frequency electric fields at 6 to 75 Hz, or RF radiation amplitude modulated at these frequencies have been shown to have

Hodgkin brane perrneabilities membrane

and Huxley in squid of polar

pointed of axon

out that the steep memsodium having and large potassium in the or charges the presence

potential

dependence molecules

implies

significant effects on efflux of calcium ions from brain tissue at levels which are thermally insignificant [4][6]. The model given in this paper may give a theoretical basis for some of these observed has been brane shown that oscillating of squid axon might effects. fields For example, within it the memsensi-

dipole moments [1 1]. Only through existence of such charged groups acting as voltage sensors can the cell be so exquisitely sensitive to small changes in electrostatic potential. Movement of these charged groups (associated with which the integral could proteins result of the voltage sensitive ionic chanas a channels) in protein confirmational changes

effectively

open voltage

tive potassium conductance. sodium sodium

channels resulting Furthermore, an This increase

in increased potassium oscillating field closes in in gK and steady-state of the mempotential potential. equilibrium

channels

in the steady state due to an increase

inactivation.

open (or close) the gate movement current, of recent

in the ion specific recorded

decrease in g~, results in a hyperpolarization brane (Fig. 5) because the potassium VK is on the hyperpolarized Such a membrane hyperpolarization inhibitory effect on an irradiated

nel. Measuring displacement able volume

of this charge, (see Almers

has been the subject research

of a consider[13] or Hille

side of the resting

[15] for excellent reviews). For a given value of membrane potential, there is an energy barrier separating the fully open and fully closed states of these gating dipoles. From thermodynamic principles [16], it can be shown that the probability of a dipole being in a given state is a highly applied potential (a Boltzmann tain simplifying assumptions Thus an oscillating field will average position of the dipole nonlinear distribution function of the results if cer-

would have a general neuron or muscle cell.

Thus spontaneously firing cells, e.g., pacemaker cells in neural or cardiac tissue, would be expected to decrease their impulse generating frequency in response to irradiation of sufficient magnitude. This effect is in the opposite direction to generalized heating. Such an inhibitory effect would decrease the probability of an action potential being generated for a given fixed level of excitory input. This is shown quite graphically in Fig. 6 where the solution to the HH equations is given for a fixed initial 7 mV. The absence of an O) is given in membrane depolarization (stimulus) of V= resulting membrane action potential in the oscillating field across the membrane ( V~ = trace A, Trace B shows the response to the

are made [11], [13], [17]). produce a dc change in the resulting in a steady shift in

the probability of the dipole being in a given state. It is this kind of physical mechanism which could lead to the conductance changes predicted by the modified HH model presented here.

146

IEEE TRANSACTIONS ON MICROWAVETHEORYANO TECHNIQUES, VOL. Mm-28, NO. 2, FEBRUARY1980

same stimulus but with V~ = 5.0 mV. The peak of the action potential has been delayed by about 1 ms. An oscillating electric field equal to or greater than 6 mV completely quenches the action potential response (traces C am.3 D). The impulses shown in Fig. 6 are membrane action potentials [11]. The effects of an oscillating electric field on propagated action potentials in g~, will be discussed in a to a pulsed subsequent paper. The transitory increase

muscle cells (skeletal, smooth, and cardiac), in nerve cell axon and soma membrane, and even in the membrane of the extensive dendritic branches of some nerve cell types. Modified HH models of the type discussed herein maybe useful in predicting RF the effects systems, predicts potential of CW, For pulsed or amplion neural in the squid compoconductude modulated giant and microwave radiation example,

and neural-muscular axon, the model conductance nent of membrane potassium

that an oscillating

in response

would

increase the membrane

RF or microwave field seen in Fig. 5 is an interesting phenomenon predicted by the model which warrants further discussion. When the membrane is initially at rest (V= O), one can see from Fig. 4 that an applied ac field of 25 mV results in a substantial slight increase state value of g~. will ac field. However, before state value h(t) decrease in h and only a in m. Since gNa = m3hFN~, the final steady-

and decrease the sodium

tance in the steady state. This would have an inhibitory effect on the axon which is in the direction from effects produced by heating. ACKNOWLEDGMENT I would comments like to thank Dr. J. M. Crowley concerning for his helpful

Th> Tm so that
changes

decrease in response to the applied m(t) reaches its steadysubstantially.

Thus gNa

and suggestions

this manuscript.

rapidly increases initially but decreases as h(t) begins to change. In the squid axon, this initial increase in gN~ iS insufficient membrane. rapid initial result to have However, increase a significant excitory effect on the in other excitable membranes, a in gN. of greater magnitude might depolarization sufficient to trigger

REFERENCES [1] A. H. Frey and E. Seifert, Puke modulated UHF energy illumination of the heart associated with change in heart rate, L~e Sci., vol. 7, pp. 505512, 1968. C. E. Tirmey, J. L. Lords, and C. H. Dumey, Rate effeets in isolated turtle hearts induced by microwave irradiation, IEEE Trans. Microwace Theory Tech., vol. MTT-24, pp. 18-24, 1976. R. L, Seaman and H. Wachtel, Slow and rapid responses to CW and pulsed microwave radiation by individual Apfysia pacemakers: J. Microwace Power, vol. 13, pp. 7786, 1978. W. R, Adey and S. M. Bawin, Brain interactions with weak electric and magnetic fields, NRP Bull., vol. 15, pp. 1-29, 1977. S. M. Bawin and W. R. Adey, Sensitivity of calcium binding in cerebral tissue to weak environmental electric fields oscillating at low frequency: Proc. Nat. Acad Sci., vol. 73, pp. 19992003, 1976. S. M. Bawin, L. K. Kaczmarek, and W. R. Adey, Effeets of modulated VHF fields on the central nervous system; Ann NY A cad Sci., vol. 247, pp. 74-81, 1975. R. M. Clapman and C. A. Cain, Absence of heart-rate effects in isolated frog heart irradiated with pulse modutated microwave ertergy~ J, Microwave Power, vol. 10, pp. 411-419, 1975. L. M. Liu, F, J. Rosenbaum, and W. F. Fickard, The insensitivity of frog heart to pulse modulated microwave energy: J. Microwace power, vol. 11, pp. 225-232, 1976, C. K. Chou and A. W. Guy, Effect of 2450 MHz microwave fields on peripheral nervesj Dig. Tech. Papers (fEEE Int. Microwave Symp,, Boulder, CO), pp. 3 18320, 1973. K. R. Courtney, J. C. Lin, A. W. Guy, and C. K. Chou, Microwave effect on rabbit superior cervical ganglion, IEEE Trans. Microwaoe Theory Tech., vol. MTT-23, pp. 809-813, 1975. A. L. Hodgkin and A. F. Huxley, A quantitative description of membrane current and its application to conduction and excitation in nervefl J. Pkysiol., vol. 117, pp. 50(-544, 1952. F. S. Barnes and C. H. Hu, Model for some nonthermal effects of radio and microwave fields on biological membranes, IEEE Trans. Microwace Theoty Tech,, vol. MTT-25, pp. 742-746, 1977. W. Aimers, Gating currents and charge movements in excitable membranes, Rev. Physiol. Biochem Pharmacol., vol. 82, pp. 96-190, 1978. S. Takashima and H. P. Schwan, Passive electrical properties of squid axon membrane; J. Membrane Biol., vol. 17, pp. 51-68, 1974. B. Hille, Ionic channels in excitable membranes, Biophys. J., vol. 22, pp. 283-294, 1978. G. Schwarz, On dielectric relaxation due to chemical rate processes: J. Phys. Chem., vol. 71, pp. 4021-4030, 1967. D. Agin, Some comments on the Hodgkin-Huxfey equations; J. Theoret. Biol., vol. 5, pp. 161-170, 1963. A. F. Huxley, Ion movements during nerve activity: Arm NY Acad, Sci,, vol. 81, 221246, 1959. R. FitzHugh, Thresholds and plateaus in the Hodgkin-Huxley nerve equations; J, Gen. Physiol., vol. 43, pp. 867-896, 1960. E. P. George and E. A. .Johnson, Solutions of the Hodgkin-

[2]

in a membrane

[3]

an action potential. In such a cell, a step of applied ac field might have a phasic excitory effect but a tonic inhibitory A wide plied model effect. range of other fields phenomena in response by the modified models paper. have in the squid they origito apHH are predicted

[4] [5]

oscillating

of the squid axon and by similar cells. A number model in a subsequent

for other

[6]

excitable

of these phenomena was published,

and models
[7]

will be discussed in detail Since the HH been modified

the equations

to fit a number

of situations

axon other than the experimental

results which

[8]

nally modeled. Modified equations can be used to describe the effects of temperature on the propagated action potential [18], the repetitive firing observed in low-calcium concentrations [18], the +rolonged action potentials produced by tetraethylammonium ions [19], [20], and the hyperpolarizing response observed in high potassium solutions [20]. In addition, modified HH-type equations have been applied with some degree of success to model other excitable tissues including myelinated nerve [21 ][23] and cardiac muscle [24], [25], Modifications of these various mathematical models to account for applied microwave or RF fields might be useful in explaining the RF and microwave field effects on nerve and muscle tissue observed in the laboratory.
SUMMARY

[9]

[10]

[11]

[12]

[13]

[14]

[15] [16] [17] [18] [19] [20]

This

ing

paper shows that it maybe possible for an oscillatcomponent of membrane potential to change the

conductance of the membrane to all ion species which traverse voltage-dependent membrane channels. Voltagesensitive channels for sodium, potassium, calcium, and chloride ions have been demonstrated in a wide range of excitable membranes. Such channels exist in membrane of

ISEE TRANSACTIONS

ON MICROWAVE

THEORYm

YRCHNIQUSS, VOL. MTr-28, NO. 2, FEBRUARY1980

147

Huxley equations for squid axon treated with tetraethylarnmonium and in potassium rich media: Austral. J. Exp. Biol. Med. Sci., vol.
39, pp. 275-293, [21] 1961.

[23]

[22]

F. A. Dodge, Iottic permeability changes underlying nerve excitation: in Biophysics of P@iological and Pharmacological Actiom, Amer. Asso6. Advart. Sci., Washington, D.C., pp. 113-143, 1961. B. Franfcenhauser and A. F, Huxley, The action potential in the myelinated nerve fibre of Xenopus kwvis as computed on the basis of voltage clamp da% .X Piysiol., vol. 171, pp. 302315, 1964.

[24]

[25]

B. Frankenhauser and A. B. Vallbo, Accomo&tion in myelinated nerve fibers of Xenopus laeuis as computed on the basis of voltage clamp data: Acts. Physiol. Scars., vol. 63, pp. 1-20, 1965. A. J. Brady and J. W. Woodbury, The sodium-potassimn hypothesis as the basis of electrical activity in frog ventricle, J. Physiol., vol. 154, pp. 385-407, 1960. R. E. McAllister, D. Noble, and R. W. Tsien, Reconstruction of the electrical activity of cardiac Purkiqje fibres~ J. Plysiof., vol. 251, pp. 1-59, 1975.

Short Papers
Easy Determination of the Characteristic Impedance of the Coaxial System Consisting of an Inner Regnhtr Polygon Concentric with an Outer Ciicle
KOICHI Ahtraet-Tlsfa TSURUTA m
a sfmple

RYUITI
method

TERAKADO
for the determfrdon

paper gives

of

the clwacteristic impedance of an fnner ragufar polygon cmtcentrfc with m--,mqp--moffof~itimfwfw plsmesheets ofcfsaqge *were radiafly dfapmed inthepolygon. Ihe ressdtaareissgoesf ~t with those Obt$lhWf by Laura smd hsfaonf [2], [3],

Fig. 1. Dimensions of the plane sheet of cherge.

I----J
1;1 2; ;, 1;1 1[, 1;

~ ---

sheet

charge

c:b

c+b

Riblet [1] and Laura and Luisoni [2], [3] have developed interesting techniques for the determination of the charactersitic impedance of the coaxiaf system consisting of a regular polygon conwntric with a circle, This paper gives a simpler method for an inner regular polygon of s apexes concentric with an outer circle, An infinitely long conducting plane sheet of width 2b charged with charge density Q per unit length in the z direction is situated in free space with permittivity co, as shown in Fig. L Suppose that the reference point for potential is at the conducting sheet whose potential is defined as A. The potential Vp at a point P(X, y) on the z plane can be obtained by eonformal mapping of the z plane on a w plane, whose mapping function is (1) By this mapping, equipotential lines on the w plane becomes concentric circies. Therefore, the potential Vp is easily written as follows: f
vp-A Q in ~i 2TC0 ( ; l)z+(~)

The

electric

field

produced

by the conducting of the potential.

sheet charge is

derivable

as the negative

gradient

In the proposed method an important supposition is made, that is, the distribution of charge on the conducting sheet invari. ably remain, even if other conducting sheets are set near it. Fig. 2 shows the disposition of the sheets of charge fors = 5, for example. They were radially disposed between the center and the apexes of the polygon. The potential and the field at my point can be calculated analytically by superposition for these sheets of charge. The potentials at the sides of the polygon and at the outer circle were calculated fors = 3, 4, 5, and 6, Table I shows the maximum deviations of the potentials at the sides of the polygon and at the outer circle from the average potentials ~wrso~ and v~l,, respectively, expressed m percentages of the average potentials. VPOIYgm was calculated from the potentials at 100 points which were arranged on the half-side P, Pz (Fig. 2] at equal distances and were weighted by the field strength. Vticle

+~(~+l)+(y)z 2

+~(%r+(%r-+~

{(wr+(%r+l)-d(yy 2 was calculated in the same manner as VWIYSOnfrom (2) the potenand r is the

Manuscript received December 4, 1978; revised June 13, 1979. The authors are with the Department of Electrical Engineering, Ibaraki University, Hitachi Ibaraki, 316 Japan.

tials on the arc Cl C> R is the radius of the circle radius of the circumscribed circle of the polygon.

0018-9480/80/0200-0147$00.75

01980

IEEE