Re Verb
Re Verb
Re Verb
1, FEBRUARY 1998 19
(9)
with the size of the cavity. Finally, it is noted that the moment
method analysis of the monopole can also be used to compute
received fields which may be generated by an active or passive
test object and also be caused by feedback from the cavity.
Now assume that the input antenna is excited by a step
function modulated signal operating at the frequency (ra-
dian frequency ). The time domain electric field at
point may then be written as
(10)
(11)
(13)
Fig. 4. A rectangular cavity with a monopole antenna. The conductivity of
the wall is .
Since is band limited around and we can
safely assume that is a slowly varying function of
(where ) around , (10) is well approximated by
(14)
where
(15)
(16)
Fig. 5. Image sources and rays contributing at P.
Fig. 6 shows the envelope of the early time fields at a point
obtained via (16) for three values of conductivity ,
source antenna. While small monopoles easily satisfy this , and S/m (Siemens per meter). The chamber is a 1.75-
requirement, it is possible that some antennas may be too m cube and the frequency of the step-modulated signal is
large to employ ray tracing in the simplified format of (2). 10 GHz. Along with the time axis, the number of included
In this case, a generalized ray expansion may used instead images and the maximum number of ray reflections are also
[14]; however, it is expected that most antennas employed in shown in Fig. 6. Since the image sources are evenly placed in
practice as sources of EM waves would be small compared space, the number of images contributing to (16) at time is
KWON et al.: EM FIELD BUILD-UP AND QUALITY OF ELECTRICALLY LARGE SHIELDED ENCLOSURES 23
Fig. 6. The envelope of early time fields versus time: The chamber is a cube Fig. 7. Envelope function of e(P; t) in each coordinate direction: the ob-
with a = b = c = 1:75 m and the modulation frequency is 10 GHz. The servation point P is at (x; y; z ) = (0:6; 0:7; 0:8). The chamber is a cube
observation point P is at (x; y; z ) = (0:6; 0:7; 0:8). with a = b = c = 1:75 m, = 106 S/m, and the modulation frequency
is 10 GHz.
TABLE I
COMPARISON OF LOSS PER REFLECTION, NUMBER OF IMAGES AND NUMBER OF magnitude. The minimum number of reflections required for
REFLECTIONS REQUIRED FOR STEADY-STATE CONVERGENCE AND SETTLE TIMES
FOR FIVE VALUES OF : THE SETTLE TIMES ARE BASED ON A CUBIC CHAMBER OF convergence may then be calculated approximately from
SIDE 1.75 M AND THE NUMBER OF IMAGES AND REFLECTIONS ARE BASED ON (17)
(17)
and are given in Table I for each along with the corre-
sponding number of images and the time to convergence (settle
time). The table indicates that a realistic chamber (
S/m) requires an intractably large number of images and
reflections to be computed for convergence.
Fig. 7 shows the envelope functions in each coordinate
direction as a function of time for the S/m case
of Fig. 6. Plotting these field components at several different
roughly where is the volume of the chamber. locations allows one to visualize the polarization diversity
The number of reflections denotes the maximum number of throughout the chamber.
reflections any ray goes through to reach by time . The
jagged appearance of the field plots is due to the ideal step IV. SEMIDETERMINISTIC PREDICTION OF QUALITY FACTOR
input being used here; in reality, the limited bandwidth of the
source antenna and transmission line feed would cause the The most commonly used parameter for characterizing
source to have a less abrupt turn on and provide small scale closed cavities such as reverberation chambers is the quality
factor . As with other types of resonant systems, the of a
smoothing of the field plots without affecting the general shape
closed cavity will tell how effective the cavity is in building
of the curves.
up and storing energy. With a very large , the field in the
As Fig. 6 illustrates, in the lossiest case with S/m,
chamber can build up to a very large value even with small
the field reaches steady-state relatively early. This is because
input power. This is because upon turning on the excitation,
the rays that take a long time to reach have undergone a
the energy in the cavity will build up to a steady-state level
large number of attenuating reflections and become too weak
where the total input power is exactly compensated by the
to contribute significantly to the total field. As expected, it dissipated power from all the loss mechanisms combined.
takes a longer time for the field with S/m to settle The classical definition of for a single frequency of
down because there is less attenuation per reflection. For the operation at steady state is given by
more realistic case of S/m, which corresponds to a
good metallic conductor, convergence could not be reached (18)
within a reasonable amount of computer time.
It was observed computationally that the field reaches its where is the total stored energy and is dissipated
steady-state value after a time when all the subsequent rays are power. A simple formula for the of a closed cavity was
attenuated by reflections to less than 6/10 of their free-space obtained in [8] by treating the cavity field as a uniformly
24 IEEE TRANSACTIONS ON ELECTROMAGNETIC COMPATIBILITY, VOL. 40, NO. 1, FEBRUARY 1998
Fig. 8. Coherent energy density build-up for five values of : a = b = c = 1:75 m. Observation pointis at (x; y; z ) = (0:6; 0:7; 0:8). Coherent
sum of ray power and exponential curve fit overlay each other.
distributed spectrum of plane waves and calculating average if one sums the power in the rays using magnitude only, the
power dissipation statistically, arriving at energy density build-up at a point may be written as
where is the relative permeability of the cavity which is expected to have a behavior similar to (21). This
wall, is the volume, and is the inner surface area of the function is plotted in Fig. 8 for various cases of values.
cavity. A time constant was suggested in [9] defined by Also plotted for each is a best curve-fit of the function
(20) (23)
where the time delay has been introduced because unlike
which determines the exponential decay rate of the stored , stays at zero until the first ray arrives at
energy when the source is suddenly turned off. Since the same . However, these two curves overlay each other to within
applies to the case of energy build-up when the source is graphical resolution [except for the small fluctuations near zero
suddenly turned on, it can be written that caused by the ideal step function in (22)], which supports the
(21) hypothesis that the point-wise energy density build-up defined
by (22) has the same functional behavior as the total energy
where is the steady-state energy. build-up in the rectangular chamber.
It is of interest to extract numerically using the ray method It remains to be seen if the time constant which gives the
presented here. However, to extract directly using (21), one best numerical curve-fit in (23) agrees with the time constant
would have to compute the total stored energy as a function in (21). Using (20) to define the -factor in terms of the curve-
of time, which would require integrating the field over the fit time constant in (23), Table II lists for the five cases of
entire volume of the chamber at each time step. Supposing values and compares it with the statistical formula of (19).
that the numerical sampling is on the order of a fraction The agreement is very good between these two very different
of a wavelength, a numerical integration scheme becomes approaches. The physical explanation of this result is that the
impractical for electrically large cavities. Therefore, the energy large set of rays involved in the evaluation of (16) and (22)
build-up at a single point or a small set of sample points is experience reflections nearly everywhere on the chamber walls
investigated instead. Sampling a small set of points randomly so that the field at any point is directly influenced by the wall
distributed throughout the chamber should indicate whether loss and chamber geometry.
the energy density build-up at a point has a form similar to For the static rectangular chamber considered here, it has
the total energy build-up in the cavity. been found that the curves of Fig. 8 are nearly independent of
As Figs. 6 and 7 show, there is no simple exponential build- the chosen point of observation in the chamber. It is known
up of the fields at a point because phase interference between that for a low-loss cavity, the steady-state field distribution
the various rays creates oscillations in the curves. However, due to a single frequency excitation is highly oscillatory as
KWON et al.: EM FIELD BUILD-UP AND QUALITY OF ELECTRICALLY LARGE SHIELDED ENCLOSURES 25
ACKNOWLEDGMENT
The authors would like to thank Dr. D. A. Hill of the
National Institute of Standards and Technology, Boulder, CO,
for his technical assistance.
REFERENCES
[1] P. Corona, G. Latrmiral, E. Paolini, and L. Piccioli, “Use of reverber-
ating chamber for measurements of radiated power in the microwave
range,” IEEE Trans. Electromagn. Compat., vol. EMC-18, pp. 54–59,
Fig. 9. Coherent energy density build-up for each polarization component of May 1976.
the = 106 (S/m) case: Observation point is at (x; y; z ) = (0:6; 0:7; 0:8). [2] M. T. Ma, M. Kanda, M. Crawford, and E. Larsen, “A review of electro-
Coherent sum of ray power and exponential curve fit overlay each other. magnetic compatibility/interference measurement methodologies,” Proc.
IEEE, vol. 73, pp. 388–411, Mar. 1985.
[3] D. I. Wu and D. C. Chang, “The effect of an electrically large stirrer in
a function of location. This oscillatory behavior is caused by a mode-stirred chamber,” IEEE Trans. Electromagn. Compat., vol. 31,
constructive and destructive phase interference among rays. pp. 164–169, May 1989.
[4] D. A. Hill, “Electronic mode stirring for reverberation chambers,” IEEE
However, once the phase information is taken out as in (22) Trans. Electromagn. Compat., vol. 36, pp. 294–299, Nov. 1994.
so that the power of individual rays are summed to form a [5] B. H. Liu, D. C. Chang, and M. T. Ma, “Eigenmodes and the composite
quality factor of a reverberating chamber,” NBS Tech. Notes, no. 1066,
monotonically increasing function with time, (22) is expected Aug. 1983.
to be a slowly varying function of position. More complex [6] C. T. Tai, Dyadic Green’s Functions in Electromagnetic Theory. Scran-
cavity geometries are expected to show more spatial variation ton, PA: Intext Educat. Publ., 1971.
[7] R. E. Richardson, “Mode-stirred chamber calibration factor, relaxation
in energy density than the simple rectangular cavity. time, and scaling laws,” IEEE Trans. Instrum. Meas., vol. IM-34, pp.
Fig. 9 shows the energy density curves for each polarization 573–580, Dec. 1985.
component of the S/m case. They also have the [8] J. M. Dunn, “Local, high-frequency analysis of the fields in a mode-
stirred chamber,” IEEE Trans. Electromagn. Compat., vol. 32, pp.
form of (23) and are expected to have similar time constants. 53–58, Feb. 1990.
(However, it has been found that the time constant for the [9] D. A. Hill, M. T. Ma, A. R. Ondrejka, B. F. Riddle, M. L. Crawford, and
R. T. Johnk, “Aperture excitation of electrically large, lossy cavities,”
cavity is most accurately computed using total field values.) IEEE Trans. Electromagn. Compat., vol. 36, pp. 169–178, Aug. 1994.
These curves will change for different observation points and, [10] G. A. Deschamps, “Ray techniques in electromagnetics,” Proc. IEEE
therefore, may be used to characterize the polarization diver- vol. 60, pp. 1022–1035, Sept. 1972,.
[11] R. F. Harrington, Time-Harmonic Electromagnetic Fields. New York:
sity throughout the cavity more clearly than the fluctuating McGraw-Hill, 1961.
field plots of Fig. 7. For example, it is clear from Fig. 9 that [12] C. T. Tai and P. Rozenfeld, “Different representations of dyadic Green’s
the component dominates at the given observation point. functions for a rectangular cavity,” IEEE Trans. Microwave Theory
Tech., vol. MTT-24, pp. 597–601, Sept. 1976.
[13] H. Ling, R. Chou, and S. W. Lee, “Shooting and bouncing rays:
Calculating the RCS of an arbitrary shaped cavity,” IEEE Trans.
V. CONCLUSION Antennas Propagat., vol. 37, pp. 194–205, Feb. 1989.
[14] P. H. Pathak and R. J. Burkholder, “High-frequency electromagnetic
It has been shown that the deterministic early-time field scattering by open-ended waveduide cavities,” Radio Sci., vol. 26, no.
build-up at points within a closed cavity may be computed 1, pp. 211–218, 1991.
26 IEEE TRANSACTIONS ON ELECTROMAGNETIC COMPATIBILITY, VOL. 40, NO. 1, FEBRUARY 1998
[15] C. A. Balanis, Advanced Engineering Electromagnetics. New York: Prabhakar H. Pathak (M’76–SM’81–F’86)
Wiley, 1989. received the B.Sc. degree in physics from the
[16] S. Chen and S. Jeng, “An SBR/image approach for radio wave prop- University of Bombay, India, in 1962, the B.S.
agation in indoor environments with metallic furniture,” IEEE Trans. degree in electrical engineering from the Louisiana
Antennas Propagat., vol. 45, pp. 98–106, Jan. 1997. State University, Baton Rouge, in 1965, and the
[17] R. F. Harrington, Field Computation by Moment Methods. New York: M.Sc. and Ph.D. degrees in electrical engineering
Macmillian, 1968. from The Ohio State University, Columbus, in 1970
and 1973, respectively.
He has been with The Ohio State University
since 1973 and is currently a Professor there. He
Do-Hoon Kwon (S’94) was born in Seoul, Ko- has participated in invited lectures and several short
rea. He received the B.S. degree from the Ko- courses on the uniform geometrical theory of diffraction and other high-
rea Advanced Institute of Science and Technology frequency methods, both in the United States and abroad. He has authored
(KAIST), Korea, in 1994, and the M.S. degree from and co-authored chapters on the subject of high-frequency diffraction for five
The Ohio State University, Columbus, in 1995. He books. Currently, he is serving as a member of the editorial board of the
is currently working toward the Ph.D. degree at The International Series of Monographs on Advanced Electromagnetics (Tokyo,
Ohio State University. Japan: Sci. House). He has dealt primarily with the development of uniform
Since April 1994, he has been a Graduate Re- asymptotic solutions that improve and extend the geometrical theory of
search Associate with the ElectroScience Labora- diffraction solutions for solving antenna and scattering problems associated
tory, Department of Electrical Engineering, The with complex structures, such as aircraft and spacecraft. In addition, he has
Ohio State University. His main research interests been involved with the development of efficient hybrid methods of analysis
are high-frequency electromagnetic scattering and computational electromag- for reflector and microstrip-type antennas and, more recently, for dealing
netics. with electromagnetic wave propagation in the presence of complex radiating
structures such as those involved in shipboard and urban environments.
His work also includes the areas of geometrical theory of diffraction and
asymptotic methods and the analytical inversion of the solutions obtained
therefrom into the time domain to arrive at a progressing wave picture for
Robert J. Burkholder (S’85–M’89–SM’97) transient radiation and scattering and he is involved with the analysis of
received the B.S., M.S., and Ph.D. degrees electromagnetic penetration into and scattering by deep as well as shallow
in electrical engineering from The Ohio State open-ended cavities and the development of Gaussian beam techniques for
University, Columbus, in 1984, 1985, and 1989, antennas and other applications. His research interests include electromagnetic
respectively. theory, mathematical methods, antennas, and scattering.
From 1989 to 1994, he was a Postdoctoral Dr. Pathak is a member of Sigma Xi and a member of the U.S. Commission
Research Associate at The Ohio State University B of URSI. He was named an IEEE AP-S Distinguished Lecturer for a three-
ElectroScience Laboratory, where he is currently a year term beginning in 1991. He is a former Associate Editor of the IEEE
Senior Research Associate. His research specialties TRANSACTIONS ON ANTENNAS AND PROPAGATION.
are high-frequency asymptotic techniques and their
hybrid combination with numerical techniques for
solving electromagnetic radiation and scattering problems. He has contributed
extensively to the electromagnetic analysis of large cavities (such as jet
inlets/exhausts) and is currently working on the more general problem
of antenna radiation, propagation, and coupling in complex multibounce
environments.
Dr. Burkholder is currently serving as Associate Editor for the IEEE
TRANSACTIONS ON ANTENNAS AND PROPAGATION and is Chairman of the
Columbus Joint Chapter of the IEEE Antennas and Propagation and
Microwave Theory and Techniques Societies. He is also a full member
of URSI Commission B, and a member of the Applied Computational
Electromagnetics Society (ACES).