IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES, VOL. 42, NO.

5, MAY 1994 793

FDTD Simulation of Microwave Sintering
of Ceramics in Multimode Cavities
Magdy F. Iskander, Fellow, IEEE, Ray L. Smith, Student Member, IEEE,
A. Octavio M. Andrade, Member, IEEE, Hal Kimrey, J r., and Lee M. Walsh
Abstract-Microwave sintering of ceramics in multimode cav-
ities, particularly the use of picket-fence arrangements, has re-
cently received considerable attention. Various types of ceramics
have been successfully sintered and, in some cases, a desirable
and unique microwave effect has been observed.
At present, various aspects of the sintering process such as
preparation of sample sizes and shapes, types of insulations, and
the desirability of including a process stimulus such as Si c rods
are considered forms of art and highly dependent on human
expertise. The simulation of realistic sintering experiments in a
multimode cavity may provide an improved understanding of
critical parameters involved and allow for the development of
guidelines towards the optimization of the sintering process.
In this paper, weutilize the FDTD technique to model various
geometrical arrangements and material compatibility aspects in
multimode microwave cavities and to simulate realistic sintering
experiments. The FDTD procedure starts with the simulation
of a field distribution in multimode microwave cavities that
resembles a set of measured data using liquid crystal sheets.
Also included in the simulation is the waveguide feed as well
as a ceramic loading plate placed at the base of the cavity.
The FDTD simulation thus provides realistic representation of a
typical sintering experiment. Aspects that have been successfully
simulated include the effects of various types of insulation, the
role of Si c rods on the uniformity of the resulting microwave
fields, and the possible shielding effects that may result from
excessive use of Sic. These results as well as others showing the
electromagnetic fields and power-deposition patterns in multiple
ceramic samples are presented.
I. INTRODUCTION
ECENTLY, there have been reports of a unique charac-
R teristic known as the microwave effect that may result
from microwave sintering of ceramics [ I 1-[3]. Specifically,
lower sintering temperatures, smaller grain size, and lower
activation energies were reported. With the significant interest
and continued publication of results infemng advantages of us-
ing microwaves rather than conventional sintering of ceramics,
there has been a significant need to better model a microwave
sintering process and simulate realistic sintering experiments.
There is no doubt that before the full commercial utilization of
this technology, a detailed understanding of the basic nature
of the microwave interactions as a function of frequency,
geometry, and temperature must be developed, and tradeoffs
between the use of single- and multimode cavities should be
understood. Furthermore, the effect of the sample sizes and
Manuscript received April 10, 1992; revised J uly 19, 1993.
The first three and the last authors are with the Department of Electrical
Engineering, University of Utah, Salt Lake City, UT 84112. Mr. Hal Kimrey
is with Oak Ridge National Laboratory, Oak Ridge, TN 37831.
IEEE Log Number 9216815.
shapes of the effectiveness of the microwave sintering must
be analyzed and the use of stimulus such as Si c rods in a
picket-fence arrangement must be understood and optimized.
In this paper, we used the finite-difference time-domain
(FDTD) technique to model and simulate realistic microwave
sintering experiments using multimode cavities. Although the
FDTD method has been previously used to analyze eigenvalue
and dielectric resonator problems [4]-[6], the contribution
discussed in this paper is different than those reported ear-
lier. Differences include: l ) this paper deals with sintering
experiments in multimode cavities rather than in single-mode
resonant structures; 2) a realist multimode structure that in-
cludes the feed waveguide and a ceramic base plate used
for sample placement were used in the FDTD model; and
3) the field distribution in an empty cavity was initially
adjusted so as to provide reasonably accurate representation
of experimentally measured data using liquid crystal sheets.
These, as well as the fact that the prime interest of this
paper is to model EM absorption patterns in realistic sintering
experiments, make the contribution of this paper different from
previously reported ones.
11. NUMERICAL BACKGROUND
Throughout this paper, results will be presented that were
produced using an FDTD code available at the University of
Utah [7], [8]. Although this code was developed to model
both interstitial antennas and noninvasive applicators for hy-
perthermia applications, it required simple modifications to
successfully model realistic microwave sintering experiments
[7]. The FDTD code was chosen for its ease of operation and
the ease at which it can be adapted to different geometries.
Furthermore, the assumed highly conducting cavity walls
naturally limit the computation domain and hence minimize
computational problems associated with absorbing boundaries.
For high Q cavities, however, the convergence of the FDTD
code is rather slow and the solution may take a long time to
reach steady state.
The code uses Yees cells [8], [9] to discretize the compu-
tational domain and is, in fact, similar to the code developed
originally in [lo]. Since the code was developed for use in
unbounded space, it included the usual radiation boundary
conditions as described in [ 1 11. The extent of the modifications
made to the available code are centered around the removal
of the majority of the radiation boundary conditions (RBCs)
(except at one end of the waveguide feed region where first-
order Mur boundary conditions are used) and replacing them
0018-9480/94$04.00 0 1994 IEEE
194 IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES, VOL. 42, NO. 5 , MAY 1994
Microwave Oven
/
rofoam Supports
Front View \
Liquid Crystal Sheet
Fig. 1. Experimental arrangement used to map power distribution pattern in
a sintering microwave oven. Color temperature pattemin liquid crystal sheets
is proportional to EM power distribution pattemwhen short (typically 10 s
at 700 W) heating time is used. The dimensions of the microwave sintering
oven are 40 x 30 x 40 c d .
with metal boundaries of varying conductivities. The program
was also modified to allow for the use of a waveguide source to
feed the multimode sintering cavity. By removing the RBCs
and replacing them with conducting walls, the overall FDTD
code was greatly simplified, thus allowing for savings in both
computational time and storage requirements. These savings
were helpful in that some of the cavities have a large Q
when left unloaded, which increases the time required to
reach a steady-state condition. Besides these aforementioned
modifications, the used FDTD code was basically similar to
the one reported in earlier publications [7], [ 8] .
The majority of simulation cases were run at 2.450 GHz
with a cell size, A, of 1 cm or about 1/12th of the free-space
wavelength. Although this limited the resolution of the model,
it was helpful in identifying trends before large computational
resources were brought to bear on the model. The accuracy
of this assumption will be discussed in Section IV. The time
steps were taken to be A/2c where c is the speed of light in
free space. This led to a time step in regular computer runs
of approximately 6.67 ps. As trends became apparent, the cell
size could then be reduced to allow further investigation of
the individual phenomena. The code was run on an IBM 3090
computer and a typical cpu time averaged 1800 s.
111. OVERVIEW OF THE SOLUTION PROCEDURE
An important consideration in the modeling process de-
veloped in this paper is to provide useful information on
the relationship between different parameters that critically
influence the microwave sintering process. Many experimental
measurements were made at Oak Ridge National Laboratory
[l], [3], and the developed numerical model is expected to
confirmthese observations and hopefully lead to optimized
designs in future sintering experiments.
The first step in the simulation process was to determine
the EM power distribution pattern in a microwave cavity
typically used for sintering in our laboratory. To this end,
liquid crystal sheets were placed horizontally using Styrofoam
supports to map the power distribution pattern at successive
planes throughout the cavity. Fig. 1 shows the experimental
arrangement used in making the temperature distribution pat-
terns while Fig. 2 shows a typical result of the sintering oven
available in our laboratory. For short heating times (less than
10 s), the temperature distribution is linearly proportional to
the EM power pattern desired in our simulation. It should also
Fig. 2. Typical temperature distribution pattern obtained fromcolor patterns
measured using liquid crystal sheets. The relationship between the measured
color pattern and the shown temperature distribution was calibrated separately
using temperature-controlled water path.
100
75
50
25
0
Fig. 3. Typical steady-state electric-field distribution pattem in an empty
cavity obtained by assuming several possible modes on the excitation planes
placed within the cavity. The relative amplitudes of the various modes are
varied so as to obtain a field distribution pattern that resembles measured
values. Reported results were obtained when six modes of equal amplitude
were used on the excitation planes.
be noted that this arrangement was not intended to produce
actual field values, but only qualitative results to help in
developing a simulation with initial field-distribution values
in an empty cavity that reasonably resemble measured data.
A model of the cavity is then developed using FDTD to
produce an EM field pattern similar to the measured data. Two
FDTD modeling procedures were adopted for this purpose.
The first model simply consisted of choosing a set of modes
that could exist independently and simultaneously within a
small band of frequencies centered around the cavity operating
frequency of 2.45 GHz. The cavity is then forced to converge
to this set of modes at the excitation planes and the solution
is allowed to reach a steady-state condition elsewhere in
the cavity. The resulting field pattern was observed and the
amplitudes of the modes are adjusted to help match the
measured field patterns. Typical results using this method
are shown in Fig. 3, which was produced using six different
modes of equal amplitude. The second procedure involves the
modeling of the feed waveguide in the cavity. The waveguide
ISKANDER et al.: FDTD SIMULATION OF MICROWAVE SINTERING OF CERAMICS I N MULTIMODE CAVITIES 195
I I I -----.. A Insul ati on I
- 40cm h( Half Hei ht WR 340 Waveguide
Fig. 4. Model used in the FDTD simulation of the sintering experiments.
Although a typical mode stirrer was not included in the simulation, the
geometry, dimensions, and location of the feed waveguide were changed to
adjust the field patterns in the empty cavity so as to correlate with measured
results.
feed is placed horizontally along the bottom of the cavity in
an arrangement that is similar to the feed arrangement used in
the construction of the microwave oven used for sintering. A
ceramic base plate was included above the feed waveguide
first to simulate the experimental arrangement where the
ceramic plate provides a place where the sample location
is well defined and also to help improve the computational
efficiency of the FDTD code, since having additional losses
in the cavity helps shorten the time required to converge to
steady state. The feed waveguide dimensions were varied and
discontinuities (flares and posts) were introduced to provide
some degree of mode-stirring capabilities that help reproduce
the measured field-distribution pattern. Reflections at one end
of the waveguide feed were minimized by introducing a first-
order Mur absorbing boundary condition at roughly a half
guide wavelength behind the excitation plane. Fig. 4 shows the
model used in the numerical simulation, while Fig. 5(a) shows
the obtained steady-state result for the EM field distribution
in an empty cavity when a straight length of a waveguide
is used as a feed at one end and terminated at the other.
Fig. 5(b) shows the resulting steady-state EM field pattern
when the feed waveguide includes a flared transition to the
empty cavity. From Fig. 5 it may be seen that adjusting the
feed waveguide geometry and dimensions provide variables
suitable for adjusting the field patterns in the empty cavity. The
results from Fig. 5(b) also reasonably resemble the measured
liquid crystal temperature pattern shown in Fig. 2.
IV. RESULTS
After obtaining an EM field pattern in the empty cavity
that correlates well with measured data, realistic microwave
sintering experiments in a multimode cavity were simulated
using the FDTD code described above. Ceramic samples and
surrounding insulations are basically placed in the cavity as
shown in Fig. 4, and the steady-state EM power deposition
pattern is calculated in both the sample and the insulation.
The parameters expected to play critical roles in microwave
sintering experiments are then varied to determine their effect
on the steady-state EM field distribution pattern. After an
extensive number of simulation runs, observations were made
100
75
50
25
0
100

75
50
25
0
(b)
Fig. 5. Steady-state electromagnetic field patterns inside an empty cavity (a)
when a straight section of the feed waveguide was used and (b) when a flared
section of the feed waveguide was used. The difference in results of (a) and
(b) illustrate the ability to modify field patterns in an empty cavity so as to
correlate with experimentally measured results using liquid crystal sheets. The
dimensions of the sintering microwave oven are 40 x 30 X 40 cm3.
and guidelines were developed towards the optimization of
the sintering process.
Before presenting numerical results, however, it is important
that we show that the FDTD solution has converged and
does provide adequate resolution when a 1 cm FDTD cell
size is used. It is very desirable to have a variable mesh
size capability in the FDTD code so that finer mesh may be
used in the sample and insulation regions and larger mesh in
the empty cavity space. Since our code does not have the
variable mesh size capability, we decided to use a similar
but smaller 20 cmx20 cmxl 5 cm cavity model (half of the
linear dimensions of the model shown in Fig. 4) to check
the adequacy of the resolution that may be obtained from
a 1 cm FDTD cell size. Specifically, the total electric-field
distribution in the cavity was calculated when a 2 cmcube
sample surrounded by a 6 cm-thick insulation was heated
in the smaller cavity using four Si c rods as stimuli. Three
different cell sizes were used in three separate runs, and the
obtained results are shown in Fig. 6. From these results, it
may be seen that while the 0.33 cm cell size (approximately
0.15 X inside the Sic) provides an improved resolution of the
796 IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES, VOL. 42, NO. 5, MAY 1994
Fig. 7. Steady-state electric-field (magnitude) pattern in an unsintered
(green) ceramic sample surrounded by 3-cm-thick insulation. The dielectric
constant and conductivity of the sample are F, =4.13 and u =6.2 x l op6
S/m, respectively, while those of the surrounding insulation were assumed
t7. =1.537 and o =15.374 x l op6 S/m, respectively. The ceramic sample
is of cylindrical shape of 2-cmradius and 6 cmlong. The dimensions of the
microwave oven are the same as those of Fig. 5.
0
100
75
.-.
50
25
0
3 00
15
50
25
3
(c)
Fig. 6. Electric-field (magnitudc) distribution in a 20 cmx20 cmx 15 cm
multimode cavity loaded with a 2 cm3 sample surrounded by 8 cm-diameter
insulation. Four I cmx 1 cmx4 cmSi c rods were used in this simulation. (a)
FDTD cell size is 1 cm. (b) FDTD cell size is 0.5 cm. (c) FDTD cell size is
0.33 cm, which is approximately 0.15 X in Si c at the simulation frequency
of 2.450 MHz. It may be observed that the field distributions in the sample
and inwlation are generally the same.
electric-field distribution, the field-distribution results obtained
from the three cases are generally the same, particularly in
the sample and insulation regions. In other words, the 1 cm
cell-size results provide a useful general description of the
desired information. Therefore, the remaining simulation of
Fig. 8. Steady-state electric-field (magnitude) distribution pattemin an un-
sintered (green) ceramic sample surrounded by 3-cm-thick insulation. The
dimensions and dielectric properties of the sample are the same as those
of Fig. 7, while the dielectric constant of the insulation was changed from
cr =1.537 to F, =10.
the various parameters of interest will be based on a 1 cm
cell size since it provides a substantial savings on the memory
requirements of our FDTD code.
First, the effect of the type of insulation material and its
dimensions on the resulting EM field pattern in the sample was
examined. Figs. 7 and 8 show the steady-state EM field distri-
bution in a cavity in which a green (unsintered) ceramic sample
was placed surrounded by an insulation of a different value of
E, . . In many of these sintering experiments, the insulation plays
a valuable role in containing the heat and hence enhancing the
sintering process. The dimensions and complex permittivities
of the sample and the insulation used in this simulation run
are given in Figs. 7 and 8. In some cases, such as in Fig. 8,
it may be seen that the fields were actually drawn into the
sample while in others, such as Fig. 7, fields were concentrated
just outside the sample. This change in the field pattern was
clearly caused by the change in the value of the dielectric
ISKANDER et al.: FDTD SIMULATION OF MICROWAVE SINTERING OF CERAMICS IN MULTIMODE CAVITIES
~
791
100
75
50
25
0
Fig. 9. Steady-state electric-field (magnitude) distribution pattern in the
unsintered ceramic sample of Fig. 7. All dielectric parameters and dimensions
are the same as those of Fig. 7 except that the conductivity of the insulation
was changed fromo =1.537 x l op5 S/m to 0.09 S/m.
I - I
1 E46 o.oO01 0.01
1 E-05 0.001 0.1
Conductivlty of insulation (S/m)
Insulation -.-----. ceramic sample
-
Fig. 10. Steady-state total power absorbed in sample and insulation vs.
change in conductivity of the insulation. Changing the insulation conductivity
to values near 0.001 S/m does not affect the total power absorbed by the
sample. Further increase in the conductivity of the insulation effectively
shields the sample and results in a reduction in the total power absorbed by
it. The dimensions and other electrical properties of the sample and insulation
are the same as those of Fig. 7.
constant of the insulation from 1.537 to 10. The different
locations of the field maxima and minima in the green sample
result in nonuniform heating of the sample and can result in
undesirable mechanical properties or even thermal runaway
problems which may result in the destruction of the sample.
The conductivity of the insulation can have a significant ef-
fect on the EM field distribution pattern in a typical microwave
sintering experiment. Fig. 9 shows the effect of increasing
the conductivity of the insulation from 1.537 x S/m
to 0.09 S/m, while leaving the other properties the same as
in Fig. 7. From these results it is clear that the conductivity
of the insulation plays an important role and, in particular,
the increase of the conductivity results in an increase in the
scattered fields from the sample and insulation arrangement.
The effect of changing the conductivity of the insulation is
further investigated in Fig. 10, which shows the variation
100
75
50
25
0
Fig. 11. Steady-state electric-field (magnitude) distribution pattern in a green
ceramic sample surrounded by 3-cm-thick insulation when four Si c rods
were placed in the insulation to stimulate the sintering process at lower
temperatures. The dimensions of the sample and the insulation are the same
as those of Fig. 7. The Si c rods are of 0.6 cm radius and 10 cm long. The
dielectric constant and conductivity u of the Si c rods are 29.363 and 0.9
S/m, respectively.
of the total power absorbed in the ceramic sample and the
insulation versus the increase in the conductivity of insulation.
From Fig. 10 it may be seen that increasing the conductivity of
the insulation up to values of 0.001 S/m may help draw more
power from the feed waveguide while maintaining the total
amount of EM power absorbed by the sample constant. If the
conductivity of the insulation exceeds 0.001 S/m, on the other
hand, the absorbed EM power by the sample is reduced while
the power absorbed by the insulation continues to increase.
One of the more successful microwave sintering experi-
ments involves the use of Si c rods in a picket-fence ar-
rangement to help the heating process, particularly at lower
temperatures. In some sintering experiments, the number and
arrangement of these Si c rods is critical to the success of the
experiment. The developed FDTD code was used to examine
various aspects of the use of Si c rods to stimulate sintering
of ceramic samples. Fig. 11 shows the obtained results for
the EM field distribution pattern when 4 Si c rods were used
as a stimulus in the same arrangement as in Fig. 7. It can
beseen that the EM fields tend to concentrate around the Si c
rods. However, as more Si c rods are added to the model, as in
Fig. 12 where 12 rods were used, the electric-field distribution
in the sample decreases as the energy is mostly absorbed in
the Si c instead of the ceramic sample. Although such a trend
may be difficult to see by comparing Figs. 11 and 12, careful
examination of numerical values shows that the electric-field
value at the center of the sample is 4.9 x V/min Fig. 12
as compared with 11.05 x
A drawback of the use of the stimulus arrangement is,
therefore, related to the fact that an excessive use of Si c
rods and the fact that they have high electrical conductivity
may hinder the penetration of the EM energy to the ceramic
samples. If such reduction occurs, the sintering will be mainly
due to conventional heating (heat transfer from Si c rods
to sample), and this may in turn result in the reduction in
the often-desirable microwave effect. Another drawback of
using Si c rods is related to the lowering of the Qs of the
V/m in Fig. 11.
798 IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES, VOL. 42, NO. 5, MAY 1994
100
75
50
0
Fig. 12. Steady-state electric-field (magnitude) distribution pattern in the
same sintering arrangement as that of Fig. 11 except that 12 (instead of 4)
Sic rods were used as stimuli.
25
0
Fig. 13. Steady-state electric-field (magnitude) distribution in four ceramic
samples placed in the microwave sintering cavity of Fig. 7. Each sample is
cylindrical in shape with a radius of 2 cm and 6 cm long. The insulation is 2
cm thick. No Si c rods were used in this simulation. The dielectric properties
of the sample and insulation arethe same as those in Fig. 7.
various cavity modes in the multimode microwave sintering
cavity. This will clearly result in reduced electric-field intensity
and to a reduction of the efficiency of the sintering process. It
is, therefore, suggested that a balance should be observed and
the number of Si c rods should be kept to a minimum, say 4
to 6 rods, depending on the size of the sample.
In some sintering experiments, multiple samples might be
used. Fig. 13 shows the steady-state electric-field distribution
when four samples were used. From Fig. 13 it may be seen that
for such an arrangement, excessive electric field is localized
in the insulation layer and that the inclusion of appropriately
located Si c rods may result in more favorable distribution of
the electric-field pattern. From Figs. 13 and 14, we conclude
that Si c rods may play a role in the sintering experiment
both by absorbing EM power and subsequently conventionally
heating the sample until sufficiently higher conductivity of
the sample is achieved at elevated temperatures, and also by
modifying the field distribution in the sintering cavity so as to
provide better uniformity in the field values at the locations
of the ceramic samples.
100
75
50
25
0
Fig. 14. Steady-state electric-field (magnitude) distribution in the four ce-
ramic samples of Fig. 13. Four Si c rods were used in this case and a significant
change in the field distribution pattern was observed.
V. CONCLUSIONS
An FDTD code was developed and successfully used to
model some of the many factors that influence a realistic
microwave sintering process in multimode cavities. The sim-
ulation procedure is based on an initial field distribution in an
empty cavity that is similar to an experimentally measured
pattern using liquid crystal sheets. The dimensions of the
waveguide feed were adjusted and discontinuities were added
to provide some controls on the resulting EM field distribution
pattern in the empty cavity so as to resemble the measured
data. The accuracy and adequacy of the 1 cm cell size used
in the calculations were also checked by making calculations
for 1 cm, 0.5 cm, and 0.33 cm cell sizes.
The developed FDTD code was then used to simulate and
examine several parameters of interest in a typical microwave
sintering experiment. This includes the conductivity of the
insulation surrounding the sample and the role of the Si c
rods in modifying the EM field distribution pattern in the
sample. It is shown that while an increase in the conductivity
of the insulation may stimulate the sintering process by
raising the temperature of the insulation and subsequently the
sample, excessive increase in the conductivity on the insulation
decreases the penetration of the EM fields to the sample and
hence reduces the desirable microwave effect. It is also
shown that a similar effect may result from an excessive use
of Si c rods as a stimulus.
The above obtained results as well as others including
the simulation of multiple samples in a microwave sintering
experiment certainly helped in identifying some trends that
are important in understanding and optimizing the microwave
sintering process. Future plans include the continued use of the
FDTD code to develop guidelines regarding the optimization
of the sintering process in multimode cavities. Additional
efforts to integrate the FDTD code with a heat-transfer com-
puter program to calculate the temperature distribution pattern
in a sintering experiment are also underway. This latter ef-
fort should also help make quantitative comparison between
simulated and experimental results.
ISKANDER rr a/: FDTD SIMULATION OF MICROWAVE SINTERING OF CERAMICS I N MULTIMODE CAVITIES 799
REFERENCES
M. A. J anney and H. D. Kimrey, Microwave sintering of alumina at 28
GHz, in G. L. Messing, E. R. Fuller, and H. Hausner, Eds., Ceramic
Power Science I I , Westerville, OH, American Ceramic Society, 1988,
pp. 919-924.
Y. L. Tian, D. L. J ohnson, and M. E. Brodwin, Ultrafine microstructure
of A1203 produced by microwave sintering, in G. L. Messing, E.
R. Fuller, and H. Hausner, Eds., Ceramic Power Science 11, American
Ceramic Society, Westerville, Ohio, 1988, pp. 925-932.
H. D. Kimrey and M. A. J anney, Design principles for high frequency
microwave cavities, Proc. Mater. Res. Soc., vol. 124, p. 367, 1988.
D. H. Choi and W. J . R. Hoefer, The finite difference time domain
method and its application to eigenvalue problems, IEEE Trans. Mi-
crowave Theory Tech., vol. MTT-34, pp. 1464-1469, 1986.
A. Navarro, M. J . Nunez, and E. Martin, Finite difference time domain
FFT method applied to axially symmetrical electromagnetic resonant
devices, IEE Proc., vol. 137, Part H, pp. 193-195, 1990.
-, Study of TEo and TI10 modes in dielectric resonators by a
finite difference time domain method coupled with the discrete Fourier
transform, IEEE Trans. Microwave Theory Tech., vol. 39, pp. 1417,
1991.
M. F. Iskander, Computer Modeling and Numerical Techniques for
Quantifying Microwave Interactions with Materials, in W. B. Snyder,
W. H. Sutton, M. F. Iskander, and D. L. J ohnson, Eds., Microwave Pro-
cessing of Materials II. Published by the Materials Research Society,
vol. 189, 1990, pp. 149-171.
P. C. Cherry and M. F. Iskander, FDTD analysis of power deposition
patterns of an array of interstitial antennas for use in microwave
hyperthermia, IEEE Trans. Microwave Theory Tech., Aug. 1992.
K. S. Yee, Numerical solution of initial boundary value problems in-
volving Maxwells equations in isotropic media, IEEE Trans. Antennas
Propagat., vol. AP-14, p. 302, 1966.
A. Taflove and K. R. Umashankar, The finite-difference time-domain
method for numerical modeling of electromagnetic wave interactions
with arbitrary structures, in M. A. Morgan, Ed., Finite Element and
Finite-Difference Methods in Electromagnetic Scatter. Amsterdam: El-
sevier, 1990.
G. Mur, Absorbing boundary conditions for the finite-difference ap-
proximation of the time-domain electromagnetic field equations, IEEE
Trans. Electromag. Compat., vol. EC-23, p. 377, 1981.
Magdy F. Iskander is Professor of Electrical En-
gineering at the University of Utah He is also the
Director of the Engineering Clinic Program, which
he established in 1986 He is also the Director of
the NSFOEEE Center for Computer Applications
in Electromagnetic Education and Director of the
State Center of Excellence for Advanced Computer-
Aided Science and Engineering Education He has
received the Curtis W McGraw ASEE National
Research Award for outstanding early achievements
by a univmity faculty member, the ASEE George
Westinghouse National Award for innovation in Engineering Education, and
the 1992 Richard R Stoddard Awad fromthe IEEE EMC Society
He edited two special issue\ of the Journal of Microwave Power, one on
Electromdgnetics and Energy Applications, Mach 1983, and the other on
Electromagnetic Techniques in Medical Didgnosis and Imaging, September
1983 He also edited a special issue of the ACES Journal on computer-aided
electromagnetics education He authored one book on Electrornagnetic Fields
and Waves, (Prentice-Hall, 1992), edited the CAEME Software Book, Vol
I , 1991, and coedited two book$ on Microwave Processing of Materials,
both published by the Materials Research Society in 1991 and 1992 He
ha5 published over 100 paper\ in technical journdls and made numerous
presentdtions in technical conferences
Dr lskdnder is the editor of the J oumal Computer Applications in Engi-
neering Education, publi5hed by Wiley He is a member of the National
Resedrch Council Committee on Microwave Processing of Materials His
research intere3t include the use of numerical techniques in electromagnetics
Ray L. Smith (S92) was born in Salt Lake City,
Utah. He received B.S. and M.E. degrees i n electri-
cal engineering fromthe University of Utah in 1991
and 1993, respectively.
From 1991 to 1993, Mr. Smith worked as a Re-
search Assistant. His research interests are numer-
ical techniques for solving engineering problems,
particularly as applied to electromagnetics.
A. Octavio M. Andrade (S70-M72) was born in
Sao Paulo, Brazil, on May 23, 1945. He received the
B.Sc. degree in Electrical Engineering fromEscola
de Engenharia Maua, Sao Paulo, Brazil, in 1969. He
received the M.Sc. in Electrical Engineering degree
in 1972 and the D.Sc. degree in 1981, both fromthe
Polytechnic School of the University of Sao Paulo,
Brazil.
In 1970 he was awarded a student fellowshio
fromthe Foundation for Research Support of the
State of Sao Paulo (FAPESP) for research work on
his M.Sc. Thesis. From 1970 until 1972, his work was concentrated on the
design and construction of TEM-mode stripline directional couplers, as well
as teaching undergraduate courses on Transmission Lines and Microwaves.
Beginning in 1973, he has directed his research interests to the area of
Power Applications of Microwaves, especially in developing novel techniques
for the determination of the dielectric properties of materials, which led
to his D.Sc. Thesis. In 1975, he was on a short stay at the University
of Manitoba, Winnipeg, Canada, where he worked on the development
of cavity resonator methods for the determination of complex dielectric
constant of materials, and designed a partially-filled resonator technique for
such measurements as a function of temperature. His research and teaching
activities at Escola de Engenharia Maua until 1989 continued to be directed
towards the implementation of dielectric constant measurement techniques,
along with the design and construction of numerous high-power microwave
applicators and components. From 1982 to 1984, he directed part of his
research interests to the area of biological effects of microwave radiation. In
1984 he received a post-doctoral fellowship fromthe British Council and was
on a short stay at the Cambridge University Addenbrokes Hospital School
of Clinical Medicine, England, where he followed the investigations being
conducted on microwave hyperthermia for cancer treatment.
His present fields of interest include the use of numerical techniques in
electromagnetics and the use of microwave methods for materials character-
ization and processing, and he has published and co-authored many papers
and made numerous presentations on the subject i n technical conferences.
In 1989 Dr. Andrade was Chairman of the Steering Committee of the
1989 SBMO International Microwave Symposiumheld i n Sao Paulo, Brazil,
and sponsored by the Brazilian Microwave Society (SBMO), of which he is
a founding Member. From 1990 to 1993 he was Visiting Professor at the
Department of Electrical Engineering of the University of Utah, Salt Lake
City, where he was engaged in numerous teaching and research activities in
the areas of Antennas and Microwaves. He was advisor of the Oak Ridge
National Laboratories and Hewlett-Packard Clinics of the Engineering Clinic
Program for three consecutive years. He also coauthored a chapter on the
first CAEME (Computer Applications in Electromagnetics Education) Book
in 1991.
Dr. Andrade is also a member of the Review Board of the Joumal on
Computer Applications in Engineering Education, published by Wiley. He is
presently Professor of Electrical Engineering at Escola de Engenharia Maua,
Sao Paulo, Brazil.
800 IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES, VOL 42, NO. 5, MAY 1994
Harold D. Kimrey, Jr. received the B.S. and M.S.
degrees in electrical engineering fromNorth Car-
olina State University in 1980 and 1982, respec-
tively.
From 1981 to 1991, he was with the research
staff in the Fusion Energy Division at Oak Ridge
National Laboratory, where he lead the efforts for
developing microwave technology for ceramic pro-
cessing. He was also involved in research on mod-
eling of gyrotron ceramic output windows, devel-
opment of microwave systems for several fusion
research machines including the Advanced Toroidal Facility, the Radio
Frequency Test Facility, and the Elmo Bumpy Torus. He was involved in
the National Gyrotron Development Programas technical monitor. In 1991,
he joined the Metals and Ceramics Division at Oak Ridge National Laboratory
as a Development Staff member. His responsibilities are focused on program
development in the microwave processing area. He is the author of over
40 publications in journals, conferences, and symposiumproceedings. He
is the holder of four patents with seven pending in the area of microwave
devices and applications. He is the principal investigator on a microwave
diamond synthesis project using high-frequency microwaves at 28 GHz. He
is the programmanager for the ORNL Engineering Clinic at the University of
Utah, where a modeling effort has begun to better understand microwave
interactions with materials. He is the principal engineer for microwave
equipment development in the division including a 5000 I, 2.45 GHz and
28 GHz microwave furnace.
In 1989, Mr. Kimrcy cofounded a new company, Microwave Materials
Technology, which specializes in the application of microwave energy to
materials processing. Products include conventional and custommicrowave
equipment, consulting, toll processing of materials, and contract research.
ii I
Lee M. Walsh was bomin Madison, Wisconsin, on
October 14, 1956. He received his B.S.E.E. from
the University of Utah in J une 1993.
In J une 1993, Mr. Walsh started his career in elec-
trical engineering with Motorolas Semiconductor
Products Sector in Austin, Texas. He is a Product
Engineer for the Customer Specified Integrated Cir-
cuits division where he helps to characterize and
market a brand new micro controller, the HC08.