close

Вход

Забыли?

вход по аккаунту

?

MICROWAVE ELECTROMAGNETIC NONDESTRUCTIVE TESTING OF WOOD IN REAL-TIME

код для вставкиСкачать
Microfilmed by Univ. of Wisconsin-Madison
8 2 -0 5 5 6 4
Memorial Library. Collection Maintenance Of lice
YEN, You-Hsir* Eugene
MICROWAVE ELECTROMAGNETIC NONDESTRUCTIVE TESTING
OF WOOD IN REAL-TiME
The University of Wisconsin-Madison, Ph.D., 1931
University MicrofUms international Ann Arbor, Michigan 48io6
©
1981
You-Hsin Eugene Yen
(Thistitle card prepared by the University of Wisconsin)
PLEASE NOTE:
The negative microfilm copy of this dissertation
was prepared and inspected by the school granting
the degree. We are using this film without further
inspection or change. If there are any questions
about the film content, please write directly to the
school.
UNIVERSITY MICROFILMS
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
MICROWAVE ELECTROMAGNETIC NONDESTRUCTIVE
TESTING OF WOOD IN REAL-TIME
A thesis submitted to the Graduate School of the
University of Wisconsin-Madison in partial fulfillment of
the requirements for the degree of Doctor of Philosophy
by
YOU-HSIN YEN
Degree to be awarded: December 19 81
May 19____
August 19
Approved by Thesis Reading Committee:
/)?i,
ean, Graduate School
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1
MICROWAVE ELECTROMAGNETIC NONDESTRUCTIVE
TESTING OF WOOD IN REAL-TIME
by
YOU-HSIN YEN
A thesis submitted in partial fulfillment of the
requirements for the degree of
DOCTOR OF PHILOSOPHY
(Electrical Engineering)
at the
UNIVERSITY OF WISCONSIN-MADISON
1981
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
©
Copyright by You-Hsin Yen 1981
All Rights Reserved
.ii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ACKNOWLEDGEMENTS
The author is thankful to his major professor Ray J. King for
his continuous guidance and encouragement during the past four and
one-naif years, and appreciates his efforts to read, confirm
and correct this thesis.
Especially, the author wishes to acknowledge
with gratitude, Professor King's kindness, consideration and patience
toward the author.
Appreciation is also gratefully extended to
Professors W. P. Birkemeier, N. C. Mathur,
R. j. Vernon,
J. H. Halton, R. S. Marleau, and all other members of the graduate
faculty who contributed guidance and encouragement.
The research reported in this thesis has been financially
supported by the University of Wisconsin Research Committee for the
first year and the fall semester, 1981, by the U.S. Forest Service,
Forest Products Laboratory (FPL) for the period from July 1978 to
June 1981 and by the Engineering Experiment Station for the 1981
summer session.
In addition. Weyerhaeuser Co. assisted by purchasing
needed equipment and providing incidental expense funds.
The
Forest Service, in addition to funding, provided the objectives of
the research and technical advice on wood properties,- selection and
preparation of test material and careful supervision of the project.
Those who contributed were:
W. L. Galligan, J. H. Kaiserlik,
M. Chudnoff, M. Dean and W. L. James.
The author wishes to thank his father, Der-Ming Yen, and
family in Taiwan, his wife, Chi-nwa, and the Sung family in Madison
for their support.
iii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The author appreciates the help from C. Wang, R. Shankar,
K. Elliott, w. Meier, A. Halverson, R. Reines, W. Climovech,
K. Park and P. Leggett in the Electrical and Computer Engineering
Department, and Dr. M. B. Subrahmanyan who was in the Math Dept.
Finally, appreciation is also extended to Carol Chase, who
typed the manuscript and Mrs. Helga Fack who drew the figures.
iv
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table of Contents
Preface....................................................
v±i
Chapter 1. Introduction......................................
1
Chapter 2. Theoretical Background ofthe Problem . ............
4
Chapter 3- Instrumentation S y s t e m ...........................
16
3.1
Microwave System...................................
18
3.2
analog Signal Processing
31
3.3
Digital Signal Processing...........................
37
3.4
Servo-Control System...............................
40
Chapter 4. Experimental Results..............................
43
4.1
.......................
Relationships Between the Electrical and thePhysical
Parameters of W o o d ..................................
4.2
45
Relationships Between the Dielectric and thePhysical
Parameters of W o o d ..................................
4.3
Relationships Between the ReflectionCoefficient
73
and
the Physical Parameters of Wood......................
99
4.4
Effect of Growth-Rings onPolarization A n g l e ............ 102
4.5
Temperature Effects................................... 105
Chapter 5. Data Processing.................................... 107
Chapter 6. Discussion and Future Effort......................... Ill
6.1
Instrumentation System................................Ill
6.2
Measurement Techniques .............................
6.3
Data Acquisition and Processing........................ 117
115
Chapter 7. Applications........... .. ........................ 119
v
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Appendix A.
Extension of Jones Matrix Formal ism to Microwaves
Passing through Anisotropic M e d i a ...............
Appendix B.
2_
2q
Field, Amplitude and Phase Patterns of Polarized
W a v e s .........
126
Appendix C.
Radio Method for Measuring RF P h a s e ..........
Appendix D.
Backscattering from Modulated Electric Dipoles. . .
128
Appendix E.
Analog Circuitry for Real-Time Measurements . . . .
131
Appendix F.
Digital Signal Processing Calibrated by Digital
Storage Oscilloscope.........................
127
132
Appendix G.
Circuitry for Plotting Two-Dimensional Scans. . . .
137
Appendix H.
Instrumentation Operating Procedure .............
138
Appendix I.
Programs for Real-Time Data Processing........
Appendix J.
Modified Analog Circuitry for Real-Time
Measurements.................................
Bibliography............
141
156
157
vi
I
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Preface
Historically, various physical properties of wood and wood pro­
ducts
have been difficult to measure.
Some measurements of useful
properties have been destructive and time consuming.
Other measure­
ments have been nondestructive but inaccurate or complicated [Tiuri,
1979 and 1980].
The forest products industry would greatly benefit
from the development of a simple, fast, nondestructive, accurate
method for measuring the physical properties of wood.
In particular,
development of a method for estimating moisture content (M.C.), spec­
ific gravity (S.G.), and grain angle (8) is desirable, and a method
for measuring all three simultaneously would be particularly useful.
These physical parameters are important in processing and determining
end use.
There are currently no devices that can automatically and
rapidly monitor these parameters simultaneously with the desired
accuracy.
Electromagnetic dielectric properties are useful diagnostic in­
dicators of the physical and mechanical properties of wood.
In par­
ticular, the complex dielectric constant is a strong function of the
specific gravity (or density ) and moisture content.
wood is anisotropic,
Also, since
the dielectric properties are described by a
tensor which can be used to characterize the change in polarization
as the electromagnetic field propagates through or is reflected from
the wocd.
Thus electromagnetic methods are potential tools for non­
destructive testing of wood.
Moreover, such methods lend themselves
vii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
to automation wherein the physical and mechanical properties of wood
and wood products can be monitored in real-time.
Ultrasonic and x-ray methods are useful for detecting the pres­
ence of air voids and foreign materials in dielectric materials.
But
for measuring water content and hardness, microwaves are much more
useful.
The principle of microwave method for measuring moisture is
based on the fact that the permittivity of water is much higher and
often more lossy than that of most dry substances.
For industrial purposes, the absorption of infrared radiation,
the scattering of y-rays, and the measurement of the DC or AC conduc­
tivity can be used for on-line measurements. In many applications,
the accuracy of these methods is not sufficient and in all cases, an
additional weighing is necessary to determine M.C.
Furthermore, in­
frared systems only measure the surface layer of the material and the
conductivity strongly depends on the salts dissolved in the water.
For microwaves, the ionic conductivity which strongly depends upon
the chemical composition of the host material is decreased.
Micro­
wave absorption is predominantly due to free water relaxation plus a
minor contribution from the bound water molecules, and is rather inde­
pendent of the dry material composition.
This advantage, together
with other well-known capabilities of microwave diagnostic techniques
(e.g., being non-destructive in nature, contactless and continuous in
principle) have made microwave transmission measurements an extremely
competitive tool in industrial moisture determination.
viii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The objective of this research is to investigate the use of a
Microwave Homodyne System to estimate moisture content, specific
gravity and grain angle with an acceptable degree of accuracy by in­
dependently measuring the loss (aw) , phase shift (£&), depolarization
ratio (y) and polarization angle (0) when microwaves are transmitted
through dimension limber, using an empirical method [King 1978, King
and Yen 1980 & 1981].
The physical properties of wood, i.e., moisture content, specific
gravity and grain angle, are related to the complex dielectric tensor
in a very complicated way.
A further objective of this research is to
understand these relationships and to develop suitable microwave
measuring systems capable of discerning the individual effects of
moisture content, specific gravity and grain angle (as well as growthrings), either directly through the use of empirical data or through
the use of the complex dielectric tensor.
ix
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
v
List of Symbols
v
S:
electric field
H:
magnetic field
2|l (Ej): the electric field is parallel (perpendicular) to the wood
grain
Emax (3nun
. ):
the amplitude of electric field along the major (minor)
axis of polarization ellipse
oi: angular microwave frequency
Xgj
microwave free space wavelength
fSI: electrical modulation frequency
a:
amplitude modulation index
fs: scatterer spinning frequency
5'(t): instantaneous rotation angle
u(t): instantaneous direction of spinning dipole
w:
thickness of wood
M.C.: moisture content
S.G.: specific gravity
9:
grain angle of wood
Ci*w: measured substitution loss of C2
)
relative to the
max wooct
absence of wood, i.e. (E
)„
.max 0
(a*w)jj C(a-wJ^):
a:
a»w for Ej| (3^)
attenuation constant (unit thickness loss) of (3
)
,
2I32C WOOCl
Ojj (aj_): a for 3j, (3^)
Ai:
measured substitution phase shift of (3
,, relative to
max wood
the absence of the wood, i.e. C3
r
L
max 0
r
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
A$|j (A<^): A(f> for E|j CEj^)
S: phase constant, (unit thickness phase shift} cf (E
)
max wcod
3|| CS^):
g for 2,J (E^
0Q: phase constant of microwave in free space
v:
depolarization ratio * CB . )
./CE
J
,
mm wooa
may wood
Fn » x )! ;3P.*, « 3P,
j {Ei>
Pj| <P£) :
C-§0
{■OIM .
II
I' *
"
32P)! 3 \
»-f
2
oY
3?
0:
polarization angle (angle of major axis relative to the inciden
polarisation direction)
H:
(=» 0 /9 ) d e p o la riza tio n index
Mj| CMp i
rj|:
M fo r Ej| <2^}
(* a|j+jS||) propagation constant for E,j
A
(=* ctj+j
) propagation constant for
Ejl (Ef): complex reflection coefficient for Ejj (E^)
Tjj CT^): complex transmission .coefficient for E.j (2^)
e (H )«
permittivity (permeability) of free space
£* s relative dielectric constant
ejj * s’ Ejj (3i)
£":
relative loss factor
ejj Cep r s" for
o:
(Ei )
C* s’-je") connlex dielectric constant
£j ( g p : I for Ejj (sp
tan 3;
(» z"/z') loss tangent
xi
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
£:
two-dimensional complex dielectric tensor
:
£a Z (£Za ):
tan
(=*
1 complex diagonal tensor terms
complex off-diagonal tensor
Ctan 6^):
tan <5 :
XZ
tan °zx:
C* tan 5fJ {.tan O )
tan 5 for SfJCSr)
C= £ "/£') coupling "gain” from E, to E»
XZ
XZ
-L
ll
coupling "gain" from Ejj to
xii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1
1.
Introduction
As wood is semi-transparent at microwave frequencies, the loss,
phase shift and polarization of microwaves transmitted through wood
can be measured.
Wood is also strongly anisotropic both in its phys­
ical strength and in its electrical properties.
Such anisotropy leads
to the depolarization of the electromagnetic field which can be meas­
ured to infer the combined effects of the grain angle and summerwood
relative to the direction of the incident electric field.
Since the
dielectric constant E* and the loss tangent, tan 5, differ when the
electric field is parallel and prependicular to the grain direction,
Ej| and Ej^ will have different losses and phase shifts depending on
the thickness.
polarized.
These cause the transmitted field to be elliptically
It splits into two modes for which the resulting instan­
taneous trace of the electric field is the superposition of two
ellipses.
The resultant locus of the electric field is inclined at
a polarization angle 0, with respect to a linearly polarized incident
field, and the sense of rotation depends upon whether Ejj (field par­
allel to the grain) leads or lags E ( (field perpendicular to the
grain) in time phase.
This polarization angle can be measured by
simply rotating the dipole antenna which is used to sense the trans­
mitted fields.
The following defines some of the terminology in reference to
the instrumentation system used:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2
Coherent detection —
the incoming signal is mixed with a
coherently related signal in a nonlinear device.
Coherent detection systems are well known for their extremely
good sensitivity, being of the order of -130 dBm, and their good
linearity over a dynamic range of nearly the same order of magnitude as
the sensitivity. The coherent detection principle basically involves
two channels, i.e., information and reference channels.
Successful
operation depends on sufficient separation between these two channels
prior to the actual detection process.
Homodyne detection — a subclass of coherent detection.
An un­
modulated carrier signal is mixed in a nonlinear device with a mod­
ulated signal having the same carrier frequency, thereby giving a zero
intermediate frequency (IF).
The homodyne system was adopted by the NBS Electronic Calibration
Center as a highly accurate means of measuring relative phase in 1965
[Ellerbruch, 1965].
It was successfully automated using a servo-
control system and gave phase accuracies to a few tenths of a degree
over a wide range of information signal amplitudes.
Concurrently it
was found the system could also be used for highly accurate attenua­
tion measurements, giving an accuracy of approximately 0.01 dB per
dB of attenuation.
This is comparable to the best systems known as
the RF, IF, AF and DC substitution methods.
This system has the
additional advantages of extremely high sensitivity and dynamic range.
The sensitivity is limited by the detector noise (of the order of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3
-130 dBm, depending on the bandwidth). The dynamic operating range
is limited by the sensitivity at the low end, and by the maximum
carrier power in the reference channel at the high end.
Typically
the dynamic range can have nearly the same order of magnitude as the
sensitivity.
Primarily this is because the bandwidth can be made much
smaller than say a super-heterodyne microwave receiver where the band­
width is typically a few MHz.
In our system the bandwidth is deter­
mined by the maximum rate which the field being measured varies as
the probe (or scatterer) antenna is rotated, e.g. of the order of
4 K£z (six sidebands of the spinning frequency).
To accomplish the maximum dynamic range and sensitivity, linear
detection and a bridge circuit (balanced mixers) are used in our sys­
tem.
Linear homodyne detection possesses a transfer function in which
the output signal is directly proportional to the envelope of the
input modulated signal only.
any unwanted carrier
The use of balanced mixers suppresses
due to reflections and leakage in the infor­
mation channel; this significantly extends the high end of the linear­
ly dynamic range.
The balanced mixers have the additional advantage
that noise, which is present in the reference channel, tends to be
cancelled.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4
2.
Theoretical Background of the Problem
Basic theory for waves propagating in anisotropic media can be
found in the literature lLandau, et al. 1960, Tyras 1969, KOng ]975,
Tan, et al, 1979, Kubo, et al. 1979, Cook 1979, Morin, et al. 1979],
The theory for anisotropic media consisting of a mixture of multiphase
dielectrics is generally not available, but some special cases were
reviewed by Tinga [1969].
Here we present the theory in a simplified
form, couched so as to be most easily related to the experimental
parameters which we can measure.
We initially assume that the aniso­
tropy is two-dimensional, i.e., uniaxially aligned normal to the wood
fibers as shown in Fig. 2-1 and that the wood is locally homogeneous.
This is a reasonable assumption since we anticipate E^ generated by the
incident fields E „ or E „ due to the anisotropv in the v-direction
xO
zO
J
will be small and the effects due to E , i.e., t
and t , m=x,z, are
y
my
ym
even smaller.
Anyway, the samples we have chosen in general do not
have anisotropy in the y-direction.
Therefore 3x3 dielectric tensor
can be reduced to a 2x2 dielectric constant.
The problem is akin to that used by Casey [1976] for describing
propagation through epoxy panels impregnated with graphite fibers.
However, our case is more general, since the conductivity normal to
the fiber cannot be assumed to be zero, and the displacement currents
parallel to the fibers cannot be neglected. .Also, we restrict our
attention to normal incidence which greatly simplifies the problem.
This is undoubtedly the most important case relevant to nondestructive
testing.
Then the axes are oriented as shown in Fig. 2-2, with the
x-axis parallel to the grain and the z-axis normal to the grain.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Thus
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
//
\ j / t TWO-DIMENSIONAL
r //
GRAIN
//
//
//
//
/a(y)|
LINEARLY POLARIZED
INCIDENT FIELD
,
t\
L___
ELLIPTICALLY
POLARIZED
REFLECTED FIELD
ELLIPTICALLY
POLARIZED
TRANSMITTED
FIELDS
%
■Xo
■*?i<y)EXo
*»•
/
\
PA(y)e 2o
Figure 2-1
Figure 2-2
cn
6
we need to find the 2X2 dielectric tensor
E =
t
t ^
s
zx
e
zz
XX
XZ
(1)
where each entry is complex, i.e.. t
XX
= £*
XX
je "
XX
, etc.. for an
jtot
assumed e
dependence.
A linearly polarized plane electromagnetic wave incident upon
such a medium will become depolarized upon reflection or transmission.
Consequently, an incident z-polarized plane wave will acquire x- and
z-components as it propagates through wood.
Our measurement technique is equally suited to measure reflection
from wood surfaces or transmission through the wood.
The transmission
method is believed to offer the greatest accuracy as it is unnecessary
to discriminate the desired field from the direct field, so only the
transmission method will be discussed.
The amount of depolarization
is directly related to the grain angle, specific gravity and moisture
content of the wood.
The wave equation is
V x V x e - Sq£E = 0
Assuming TEM plane waves and
3
3
(2)
= 0 (locally homogeneous in x-
and z-directions) then (2) becomes
2
-
-2-f + SJz = o
3E
(3)
2
2
where S = to
is the phase constant of wave in free space, and
0
0 0
the medium is assumed to be magnetically isotropic with y=]iQ.
fi
a
We introduce the complex propagation constants, Fjj and
, for
A
the x- and z-components, respectively.
A
A
Then Re (Fjj) and Im(Fj|) are
A
significantly larger than Re(F^) and Im(F^), respectively.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
This is
especially true if significant moisture is present, since for high. I'd-C.
the complex dielectric constant Sy is significantly greater than e^.
This difference is the result of different dipole polarizations.
Cellulose is considered to be composed of straight chain moleculars
which form crystallites whose long axes are essentially parallel to
the long axes of the wood cells, i.e., the grain of the wood.
The
hydroxyl groups and chemically bound water molecules along the sides
of these chain molecules give rise to permanent dipoles whose rota­
tion or displacement contribute directly to the dielectric constant.
When an electric field is impressed parallel to the long axes of these
cellulose chains, a greater polarization is exhibited by these
hydroxyl and water groups than when the field is perpendicular to the
molecular chains.
Consequently, a larger dielectric constant is ob­
tained when the electric field is parallel to the grain of the wood.
Similar arguments follow for loss tangent, but this is more compli­
cated since the electrical conductivity also needs to be considered
[Yavorski 1952].
As the wave propagates through the wood, the z-component E (Y)
z
of the electric field gives rise to a new field ex (Y) in the’x-direction
which is in quadrature phase with
(Y). Thus, the off-diagonal
(cross-coupling) entries of the dielectric tensor are nonzero.
Sim­
ilarly, we get a field, e (Y) in the z-direction which’in quadrature
z
phase with E^CY). In effect, we have two elliptically polarized modes
traveling in the +Y-direction. Let
ratios of J
(Y) and p,| (Y) denote the signed
CY) J to |E2 (Y)| and |ez (Y)j to Je ^Y) |in dB, respectively,
s.s shown in Fig. 2-2. The signs
P| ^
Pj| ^
will be decided
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
depending upon the nature of ellipticity (left-handed or right-handed)
of the fields.
These can be determined experimentally using our
system.
The electric field E can be expressed in the form
f
A
IIs
E (Y)'
X
i(Y) =
A
Exoe
>
-IiY
- % « > E2oe
=
(4)
-?AT
E (Y)
,2
J
V
.
Exoe
-r,T
+ jpl!<Y)Ex0e
J
where each component is generated by incident fields in both the xand z-directions. Before further discussion, let’s consider two
simple cases:
Cass 1
At Y = 0 we must have p^ (0) = p., (0) = 0 , so that (4) becomes
f
\
xO
E(0) =
(5)
zO
which is the linearly polarized incident field resolved into x- and zcomponents.
Case 2
At large Y, since Re (FjjY) »
E(Y) -»■
Re (F^Y)
-FjY
E e
zO
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(6)
9
corresponding to a single elliptically polarized wave propagating in
+Y-direction.
out.
In this case, the parallel mode is essentially filtered
This is especially true when the moisture content is high.
Returning to the more general cases, (4) can be written as
^ *\
'riY
-jPj_(Y)e
A
A
\o'
•
‘riY
e
"r|]Y
jP|j(*)e
EzC
The transformations (4) and (7) give a clear picture of the
generation of two elliptically polarized modes.
One mode is generated
by E _ and jp„ (Y)E _(e,, in Fig. 2-2) and the other bv E „ and
xO
jjrl!
zO II
3
J zO
jp^(Y)Exg (s^ in Fig. 2-2).
A more general approach to investigate
the two elliptically polarized mode is given in Appendix A.
To correlate the dielectric tensor with the measured electrical
parameters of the wood, the notations to be used are defined as
follows:
o»w= ICE ’) /CE ')J in dS is-the‘substitution loss sufferedbv
max w
max u
the major axis of the polarization ellipse in passing through wood
relative to the major axis in the absence of wood,
of wood in centimeters.
w is the thickness
To localize the loss due to the wood only,
the loss due to interface reflections is corrected as will be ex­
plained later.
A$ in degrees is the substitution phase shift of the major axis
of the polarization ellipse in passing through wood relative to the
major axis in the absence of wood.
Again the effect of interface
reflections on the phase shift will be corrected later.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
10
Y = C(E . ) /(E
) J in dB is the ratio of minor to major
min w
max w
axes of the polarization ellipse and is called depolarization ratio.
It is primarily a function of the degree of wood anisotropy, i.e.,
the contrast of its electric (or dielectric) properties in the direc­
tions perpendicular and parallel to the wood grain.
For an incident
field perpendicular to the grain, ^zQr we can measure (a»w)^,
(A<j>)^ and
Similarly, fcr E^0 we can measure (a*w)|j, (A4>)||
and P|, (Y||).
6 in degrees is the tilt angle of the major axis of the polariza­
tion ellipse after passing through wood, relative to the major axis in
the absence of wood in the transverse plane or x-z plane.
It is
called the polarization angle.
The subscripts W and 0 are with respect to wood and free space
respectively.
Using these notations, let‘■s:.substitute G7T-into the wave equation
giving
-r,Y
.
A
+j2riP|(Y)-jpj|(Y))
y*v
—r .y
e
—r y
‘^ z x ^ l l <» <fn+®0®z2>
•
1 <f> 0 Szz-5B0£zxPl (Y»
-j2r„p* (Yj+jp.V (Y))
xO
= 0
zO
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(8)
11
Note that in this derivation the transmission coefficients at the
front and back interfaces of wood were assumed to be unity.
Since (8) is true for anv arbitrary E „ and E , the general
z xO
zO
3
solution^is that entries of the matrix are all equal to zero.
“ril 1Y
terms e
' can be dropped, since they are not zero.
The
We then have
eight equations to find the eight unknowns in dielectric tensor:
=-(<V ^ /Bo+eiVY)>
- 12W Bo)+£=Upl®
£^z(y) = a / D £) ((2a|13j|-2a1 3i )pi m + 2 6 ip|(Y))
<y ) = (1/D£) ((aj?-8y-a^+6|)p1 (y )+2aip| (y )-pj; (y ))
(9)
ezx(Y) = (1/d£) ((2ct||S|,-2a1 Si )p|1(Y)-2g||P[|(y ))
£zx(Y) = (1/V
((ar 6r ai+si )pii(Y)‘2ai|Pii (Y)+P|l(Y))
£■,«) =
£2zm =-(2aiVBo>-EiAl->
where D£ =
.2
(1-p^ (Y)p|| (Y)).
Note that these formulas are in terms
of the electrical parameters which can be easily measured using our
system.
By orienting the incident field parallel or perpendicular to
A
A
A
the grain, we can measure Pj_ (Y) , pjj (Y), Red^Y), Red^Y), Im(r||Y) and
Im(?jY).
A
Note that while Fjj ^ are constants with Y, p^ y are functions of
Y, e.g., p^ j|(0) = 0 and p^ jj (°°) = constant.
In view of (9), p| jj (Y)
and pj| ||(Y) are involved in the solution for the dielectric tensor.
These terms represent higher order effects which may be negligible in
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
12
some cases.
However, in general, it will be necessary to experimen­
tally measure
tents where
jj(Y) for each species over the range of moisture con­
jj
(?) and pj* jj(Y) are significant.
This can be done by
stacking slabs of wood, all cut from the same board and stabilized at
the same moisture content.
This data table will be stored in the
computer for retrieval as needed when testing in real-time.
Assuming
that the moisture content can be determined with reasonably good
accuracy using the zeroth order solution as obtained by assinning pf ^
and pj| || are negligible, good approximations to p^ jj and p£ ^ are
obtained from the previously stored tables. These can then be used in
(3) and (9) to find the first or higher order solution for £.
A
*
simpler way is to find the normal mode electric field solution of (3)
in Jones matrix formalism and equate this solution with (7)
Pj’
j and pjJ can then be found in terms of dielectric tensor.
ative method can then be used.
p£, p£,
An iter­
We first calculate the first order
tensor assuming p.' and p" are zero, and then use this to find approx­
imate values of p* and p". These are then used to compute the 2nd
order tensor, etc.
Details are given in Appendix A.
In the preceding theory, reflections at the interfaces between
the board and air as well as internal reflections were neglected for
simplicity.
As the reflections are small, an approximate method
involving a simple iterative procedure can be used to find the first
or higher order solutions.
This is done using the zero-order solution
as found from (9) to calculate the reflection coefficients R, and
i
!i
for the field components which are perpendicular and parallel to the
grain, respectively.
Then, the transmitted field components at each
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
13
interface are corrected according to the transmission factor,
A
A
A
= 1-3^, etc.
A
These R and'T in turn are used to correct measured
parameters and (9) is used again to find the first order corrected
tensor.
achieved.
This iteration procedure is repeated until convergence is
The dielectric tensors given in Chapter 4 are the results
after correction.
The contribution from multiple internal reflections
is anticipated to be of the order of 1% or less in most cases, and
hence negligible.
The grain angle, 9, (defined in Fig. 2-2) is one of the most
important physical parameters to be found.
As the axes of the dielec-
trie tensor are defined as being parallel or perpendicular to the
grain, we see that determining the grain angle must be done indepen­
dent of the tensor.
As will be seen from the experimental data in
Chapter 4, finding the grain angle -is straightforward if the moisture
is sufficiently high and the wood is sufficiently thiclc.
In such
cases, it is sufficient to orient the incident field vertically,
rather than parallel and perpendicular to the grain as assumed in the
above paragraph.
Then, for small grain angles and sufficient attenuaA
A
ticn such that Sed^tf) »
dominant as in (6).
Red^W), the perpendicular ellipse is
Thus, the tilt angle of this ellipse, i.e.,
polarization angle 0, is. a direct indicator of grain angle 0.
Coder these same conditions, p^(W), Re(f,W) and Ua(f,w) can also
be measured,
vflhile these data are not sufficient to uniquely deter­
mine the complete dielectric tensor, useful approximations can be
made which will permit reasonably accurate results in some cases. If
Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission
14
this is not possible, it would not he difficult to set up two homodyne systems of slightly different frequencies which will illuminate
the wood with orthogonal fields.
As homodyne systems are completely
coherent (i.e., the signal to he detected only correlates with itself),
a single dipole scatterer can be used for both frequencies.
Finding the moisture content and specific gravity (or density)
is more difficult.
Of the two, moisture content is of considerably
more practical interest.
As water is chiefly contained in elongated
fibers, it will affect both the real and imaginary parts of
£jnr(S]|) more than the corresponding parts of
the less tangent of
ing moisture.
In particular,
exhibits a pronounced increase with increas­
However, the most significant effects appear in the
off-diagonal elements, £XZ and £2 X . In Chanter
4, it will be shown
~
that these elements are very sensitive indicators of moisture.
As a
matter of fact, this is the first time such an effect has been quan­
tified.
It will undoubtedly play a key role in future studies.
To interpret the electrical parameters (or tensor elements) in
terms of moisture content, specific gravity (or density) and grain
angle involves conducting many experiments on different species for
various known moisture contents and specific gravities (or densities).
For each species, these data can be stored in a computer to create
a "standard" table of the electrical parameters (or dielectric ten­
sors) vs physical parameters of the wood.
Then the physical param­
eters of an unknown beard can be found from the measured electrical
parameters using an empirical method.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
15
la principle, this can be done for a given thickness.
However
to remove the effects of thickness and the reflection at the first
interface, it is necessary to normalize the measured data by first
calculating the dielectric tensor, and then iterating to obtain the
true loss/cm (a) and phase shift/cm (8) - The latter was used in the
following.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
16
3.
Instrumentation System
A homodyne system is used to map microwave field distributions
in real-time [King and Yen 1981] . The system is monostatic. A key
feature is the use of an electrically modulated dipole scatterer which
is also mechanically spun to create a double amplitude-modulated backscattered field. This backscattered field is coherently detected by
mixing with a CW reference signal.
A phase-insensitive detector is
used, comprised of two balanced mixers which are fed in quadrature
phase by one of the RF inputs (Quadrature Mixer), followed by a phase
quadrature combiner.
The resulting amplitude and phase are propor­
tional to the square of the RF field component along the instanta­
neous axis of the spinning dipole.
Both can be measured simulta­
neously in real-time as well as the polarization characteristics,
i.e., depolarization ratio and polarization angle.
Thus the field is uniqely described.
Uniqueness, of course,
implies that the amplitude, phase and vector properties (or polariza­
tion) of the tangential fields (either E or H) must be recorded over
a surface.
Ideally, this should be done with a single point probe
which does not disturb the field, making probe corrections minimal
or unnecessary.
While the RF circuitry usually associated with the
probe (e.g., cables, waveguides, mixers, receivers, ..., etc.) can
generally be shielded with absorbers to minimize field disturbances,
the amount of such circuitry should be minimal from the standpoints
of cost, convenience and simplicity. Moreover, the measurement
system should be highly sensitive and linear over a wide dynamic
range.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
17
These needs are largely fulfilled by using a small scattering
dipole probe which is electrically modulated at f
(10 KHz), in con­
junction with a special coherent detection system for detecting the
backscattered field.
The backscattered field is proportional to the
square of the component of the complex phasor field aligned with the
dipole.
Thus since homodyne system is phase sensitive, the phase as
well as the amplitude information is preserved.
The vector informa­
tion can be obtained by spinning the dipole at frequency f
(150 Hz)
in the tangent plane of the surface being scanned.
The use of an electrically modulated scattering dipole in con­
junction with homodyne detection techniques to measure E or H field
distributions is certainly not new.
ically spun dipoles new.
Neither is the use of mechan­
The history, theory and applications of
this technology are thoroughly discussed in King's book (1978).
What
is new here is that both electrical and mechanical modulation are
used simultaneously to uniquely characterize the measured field
in real-time.
If the vector nature of the field is known a priori, the spin­
ning part of the probe modulation is unnecessary.
But more generally,
both modulations are needed when measuring arbitrarily polarized
fields (e.g., fields which are elliptically polarized by reflection,
refraction or diffraction).
For example, the results of field
polarization and the corresponding instantaneous amplitude and phase
patterns obtained by spinning the dipole for linear, elliptical and
circular polarizations are shown in Appendix 3.
The greatest versa­
tility of the system is realized when measuring fields which are the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
18
superposition of two elliptically polarized modes, such as those
created by reflection, refraction or diffraction via anisotropic
media.
The entire instrumentation system used is shown in Fig. 3-1,
heavy dark lines denote the paths of microwave (RF) power flow and
lighter lines denote AF signal paths.
The detector is configured
so that amplitude, phase and polarization properties are available
simultaneously and independently in real-time at a rate of 2f .
s
Another important advantage is that the IF is zero so that all signal
processing can be done at the audio frequency ^ .
The system can be divided into four major parts:
(1) microwave
circuitry, (2) analog signal processing, (3) digital signal processing,
and (4) servo-control system.
Each part is individually described
in the following.
3.1
Microwave Circuitry
Fig. 3-2 shows the details of a C-band (4.4-5 GHz) microwave
homodyne system.
parts:
(a)
In the following, it can be divided into four
microwave power generation, (b) reference channel,
(c) information channel, (d) homodyne detectors.
a.
Microwave power generation
The CW source is a Varian VA-1259J Reflex Klystron which pro­
duced about 2 watts of power at 4.81 GHz (X^ = S.23 cm). A HP-716B
Klystron power supply is used to provide the necessary power to the
klystron.
For the maximum output power, typically, the beam and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
19
*5
5o
“
13
oJ
<o
S?
Ui
o
s
m
z Ui
<s
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
KLYSTRON
POWER
SUPPLY
FREQ WAD
METER VAR. SLo t t ED t
—
ATT- LINE
| 2
e h
REFLEX
KLYSTRON
4.81 GHz
REF
CHANNEL-1
±:
"H" I
MAGIC
TEE
90°
/H Y B R ID
T.
JDE
&
BALANCED
MIXER
I
m
|v!
yVJ.
fiA |
PHASE
IN S E N S IT IV E
COHERENT
DET.
RESISTIVE
LEADS
TRANS.-RECV
ANT.
ISO.
BALANCED
MIXER
AF OSC
PIN DIODE
INF.
\ q u AD.
CHANNEL] MIXERS
QUAD.
COMBINER
}
Vt
&
| E*u(t)|
cos(u>mt —2<#>(♦))
Figure 3-2
to
o
21
reflector voltages are about +750 and -410 volts, respectively.
three-stub tuner
A
and an isolator provide the necessary impedance
matching between the source and the load.
This eliminates "pulling"
of the klystron and stabilizes the frequency.
A frequency meter is
used to check the frequency and a variable attenuator sets the working
power.
b.
Reference channel
A CW reference signal, (same as the local oscillator in a hetero­
dyne system) with power around +13 dBm, is diverted to the balanced
mixers via the RF probe in the slotted line and a 150 cm long
plaxial coax cable.
The power level is properly adjusted to extend
the upper end of the linearly dynamic range as much as possible.
The
plaxial cable is of premium quality, designed so that changes of
attenuation and phase shift during flexing are virtually non-existent.
For linear detection, small variations of the reference signal A due
to the flexing of the coax cable or a change in the position of the
slotted line probe does not affect amplitude measurements.
The depth
of the RF probe controls the signal level in the reference channel.
The level can be set by temporarily connecting a power meter at the
probe output.
The probe is tuned for best matching.
Tuner T2 is
adjusted for a flat slotted line, (minimize the standing waves) such
that the phase of the reference signal is linearly proportional to the
position of the slotted line probe.
This is a simple and effective
way of varying the RF phase for calibration purposes and for choosing
the reference phase for relative RF phase shift measurements.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
22
c.
Information channel 1
Most of the CW source power is split by the magic tee so that
about 1 watt is radiated by the Transmit-Receive open waveguide
antenna at port 3.
at port 2.
The other half is dissinated in the Z -termination
“
0
Tuner T3 is tuned to minimize the power at the isolated
port 4, which amounts to matching the system with the TransmitReceive antenna and open space.
This can be done by temporarily
connecting a power meter at port 4.
The scattering probe is an electrically modulated zig-zag dipole
which is spun at 11,000 rpm (see Fig. 3-3).
This high spinning speed
increases the system's capability to dynamically gather real-time
data.
The dipole is a printed circuit etched on a dielectric sub­
strate disk to provide mechanical rigidity and balance.
This disk
is driven by a hollow dielectric shaft which conceals the leads to
the diode. A p-i-n diode serves as the load impedance Z , switched
at 10 KHz fed from an external oscillator via brushes, slip-rings
and the resistive leads.
This also can be done using a radio telem­
etry method as shown in Appendix C or by some form of magnetic coupl­
ing between rotating and stationary parts [Doebelin 1966].
The p-i-n
diode was selected for its high on-off impedance ratio (e.g.,
Z
= 1.5 f2, Z
— 100 kft) . Hence, it has a large amplitude modulaon
off
*
^
tion index [King 1978].
The electrical switching frequency was
chosen as f - 10 KHz which is above the noise comer of the Schottky
m
mixers (~ 2.5 KHz) used.
This minimizes the 1/f detector noise.
The
resistive leads are oriented perpendicular to the incoming electric
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
23
SCATTERING
DIPOLE
>
*=IO=
>Xwv¥/:
<XvfXf'
'
Xw
i^
x :^ :g |w S :x:::i
iiliili^ilii;:
®
o y>
*»• A
e*S
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ELECTRICALLY
3-3
Figure
MODULATED
AND MECHANICALLY
SPUN
09
24
field. ‘They are essentailly transparent to the RF field, but are
sufficiently conductive (~ 0.6 kft/cm per line) to provide enough
voltage at the diode terminals to switch the p-i-n diode on and off.
They are commercially available filaments of polytetraflouroethylene
rendered semiconducting by uniformly dispersing fine carbon particles
through the plastic while in its semifluid state.
The lines are
drawn in 0.03 inch filaments, and coated with a 0.005 inch nylon
film for strength and insulation.
These leads are bound to the
diode and slip-rings using silver paint.
A zig-zag antenna design is used because the resonant length
of the dipole is significantly less than A^/2, thereby more closely
approximating a point probe.
Tuning is accomplished by symmetrically
trimming away both ends a bit at a time, while the modulated probe
is positioned in the RF field, observing the output of the homodyne
detector.
Resonance is achieved for a length of about 0.37 X^ with
the pitch angle of 45°.
This gives a maximum backscattered signal.
We also tried a resonant dipole which is symmetrically loaded by a
spring-coil on both ends.
When tuned, the resonant length was very
good (about 1.5 cm) , but the backscattered signal was only half of
that of Xg/2 dipole.
Also this kind of dipole is difficult to fabri­
cate and make stable.
Polarization angle 0 measurements require knowing the dipole's
instantaneous rotation angle (9*(t) = ^st). This requires that a
S-reference pulse be generated, say at 6’(t) = 0 and ir.
To do this,
a stationary collar is mounted around the shaft which has a small
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
25
hole in its diameter.
A light emitter and sensor are diametrically
positioned in the collar so that a pulse is sensed when the hole in
the shaft is aligned with the emitter-sensor pair.
This reference
t
pulse is then hard-clipped and used to time the position of a maximum
in the detected backscattered signal.
The entire mechanism in Fig. 3-3 is covered with microwave
absorbing material mounted on a small non-metallic platform.
The
absorber has a vertical slot to permit vertical motion of the probing
mechanism.
An anechoic chamber with dimensions of 40" x 40" x 100"
was built to permit the measurements in a quiet environment, free from
external disturbances and standing waves (see Fig. 3-4).
Reliable measurements require no standing waves in the vicinity
of the spinning scatterer (probe) which is situated behind the board.
Standing waves indicate the presence of reflections from the rear of
the chamber or the presence of a significant field diffracted around
the board.
These reflected, refracted or diffracted waves were
eliminated by using strategically placed microwave absorbers around
the dipole and the specimen being tested as shown in Fig. 3-4.
Also
the proximity effect between the dipole and the specimen is negligible
when the dipole is situated about 1.5 cm away from the wood.
d . Homodyne detection
Appendix D briefly reviews the theory of backscattering from a
modulated dipole.
There it is shown that the relevant backscattered
voltage appearing at the Transmit-Receive antenna port is proportional
to the modulated phasor as follows,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 0 cm
TR A N S M IT-R EC EIV E
HORN
WOOD
.5cm
LIGHT
SENSOR
D.C.
MOTOR
FIN G ER -TYPE
ABSORBER
PYRAMID TYPE
ABSORBER
**■ AUDIO OSC.
PLEXIGLAS
CLAMP
PLEXIGLAS SUPPORT
TROLLEY
MACARTA
GLASS FIBER
PLATFORM
POLE
F IN G ER -TYPE ABSORBER
EXPANDED FOAM
PLEXIGLAS
SUPPORTER
PLEXIGLAS
TRACK
Figure 3-4
to
cr>
i
27
AV = c|i*U(t) [2 (1 .+ m cos 0) t)e 324>(t)
1
■
m
(1)
where |E*U(t) j and <J)(t) are the instantaneous amplitude and phase
respectively of the component of E in the direction of U(t), i.e.,
parallel to the instantaneous axis of the dipole.
This voltage is
split by the magic tee, so that half of the received backscattered
power emerges at port 4 and the other half emerges at port 1 where
it is dissipated in an isolator.
Further, that half from port 4 is
split by a tee such that the waves arriving at the two balanced
mixers (denoted as b in Fig. 3-2) are in phase.
The reference signal power which is diverted via the RF probe
in the slotted line is split by a quadrature tee such that symmetry
is preserved except for a phase shift of tt/2, say in the left branch.
Thus, the reference voltages fed to the left and right mixers are
joit
®L = jeR = jAe
(2)
and the information signal to each mixer is (1) multiplied by exp(jojt),
e.
i
j (wt-2<J) (t))
(3)
Since the mixers operate as linear envelope detectors, the
homodyne w^-outputs from the right and left arm of this quadrature
mixers are
(4a)
and
V = K (E*U(t) I2sin(2cf>(t) )c c s (cj t)
L
L
1
m
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(4b)
28
respectively, where K and K are the mixer conversion factors.
X*
r
These two outputs are then fed to a quadrature combiner which
shifts one of them by tt/2 , say V* in Fig. 3-2, and summed with V ,
I*
R
giving
V = Vo + V* = K|i*U(t) |2cos(w t-2(j)(t))
R
L
IQ
assuming
(5)
= Kr = K is the overall conversion constant.
Contrast (5) with the output of a single mixer, (4a) or (4b),
which contain phase-sensitive amplitude factors cos(2$(t)) or
sin(2(j>(t)). Hence, we term the detection circuitry in Fig. 3-2 as a
"phase-insensitive coherent (homodyne) detector."
The fact that the
amplitude is independent of phase is essential for real-time process­
ing.
The detector has the same form as a well-known circuit for
single-sideband modulation (SSB) of a CW carrier, except that in the
present case it is being used as a detector of one (the upper) side­
band.
The other (lower) sideband results if the quadrature combiner
subtracts rather than adds to the two ^-inputs.
The mixing of a
DSBWC signal with the coherent reference signal has transferred the
phase information at the RF frequency 0) to the modulating frequency
0)
m
.
This is a significant result since we can measure the RF ampli-
tude and phase simultaneously and independently at the audio frecruency, w .
m
The quadrature mixer is readily available in a strip-line pack­
age.
The balanced mixers used are of the 90° balanced variety.
They
provide better phase and amplitude balance at the IF ports than either
the 180° balanced or double-balanced varieties.
The unwanted carrier
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
29
reflected from an object other than the scatterer (e.g., the wood,
supporting trolley, etc.) can be suppressed by the balanced mixers,
and this will increase the dynamic range and give better results for
amplitude and phase measurements.
This quadrature circuitry also has the same form as an image
rejection mixer commonly used in heterodyne detection to give the
upper or lower sidebands at the intermediate frequency (IF).
The IF
of available image rejection mixer is of the order of a few MHz and
up.
For the present application where the IF is zero and f is onlv
m
a few KHz, a specially designed quadrature combiner is required
which can operate at frequencies from 5 to 20 KHz.
— —
2
Referring back to (5), the instantaneous amplitude JE-U(t)j and
RF phase 26(t) can be measured simultaneously and independently as
functions of the spinning frequency f^.
As the dipole spins, the
locus of the electric field will be traced out as shown in Fig. 3-5(a).
(For simplicity, only one ellipse is considered here.
Similar argu­
ments can be applied to two elliptical modes in reference to
Fig. 3-5(a) and Fig. A-3 in Appendix A.)
Assume an incident linearly
polarized wave is depolarized by the anisotropic medium and becomes
ellipticallv polarized, having a polarization angle 0 as shown in
Fig. 3-5(a).
The measured fields are the mean-square values.
2
Fig. 3-5(b) shows how the amplitude E
, depolarization ratio
max
y
and polarization angle 0 are available at 2fg (300 Hz) for an elliptically polarized field.
The RF phase (Fig. 3-5(c)), which also
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
30
rms
a
•mln
® <§+Z ©'
2 !
Figure 3-5
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
31
varies at 2f .. is measured at 10 KHz.
s
Thus the instantaneous ampli-
tude, phase and polarization information contained in (5) is avail­
able in real-time.
Wood is inhomogeneous, lossy and anisotropic.
Consequently,
the amplitude, phase and polarization properties of the field trans­
mitted through the wood are diagnostic indicators of the wood's
physical properties, i.e., moisture content, specific gravity (or
density) and grain angle.
In our system the spinning dipole is
positioned very close to the wood (a ^/6 to X^/3) so as to measure
the integrated field over a local area of the order of
2
The wood
moves horizontally between the Transmit-Receive horn antenna and
the spinning dipole, and the dipole can also be scanned in vertical
steps.
The incident field is vertically polarized, but the polariza­
tion of the backscattered field is a function of the dipole's orien­
tation angle 9(t), and the anisotropic properties of the wood.
The
effects of all of these factors are measurable from the information
provided in (5).
Following detection, the signal processing is done at the audio
frequencies, f and 2f , as described in the next section,
m
s
3.2
Analog Signal Processing
The entire system shown in Fig. 3-1 is being used for non­
destructive testing of wood.
This application is useful to illus­
trate the system’s operation, but the hardware portions which include
microwave, analog and digital circuitry are sufficiently general
for use in most other applications.
Various other circuitry may
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
32
serve as well.
The schematic diagram of this analog circuitry is
shown in Appendix E.
The combiner output is first amplified by a low-noise (noise
figure is 1 dB at 10 KHz), wideband (up to 200 KHz), high sensitivity
(few nv) and high gain (40 dB), pre-amplifier (Ithco-166) followed by
an amplifier with a gain of 20 dB.
The signal is then band-pass
filtered to eliminate all components at 2ntu (n = 1,2,...) and
s
pto (p= 0,2,3,...).
m
The remaining (p=l) component at co is then
m
given by (5), where the filter must have sufficient bandwidth to
accommodate all signficant frequency components at oj ± 2nco (see
m
s
Appendix D). In the present system we found n=6 to be satisfactory
so that the bandwidth is about 4 KHz for 2f ~ 300 Hz.
s
Fig. 3-6(a)
shows a typical waveform of (5) at point a on Fig. 3-1, taken in
free space with no wood present.
As the dipole spins at
i--- .2
square amplitude |E*U(t)j
according to (5).
it sweeps out the instantaneous mean
and twice the instantaneous phase, 2<£(t),
This data is repeated for every 180° of rotation,
i.e., at a rate of 2f^.
Although, in principle, the amplitude and
phase of the field at the dipole are available at every value of
it is sufficient to measure only four parameters for present purposes.
These are the values of amplitude along the major and minor axes of
the polarization ellipse (denoted as the a- and y- branches respec­
tively in Fig. 3-1), the phase along the major axis (the 6-branch)
and the polarization angle (©-branch). If the field is known to be
linearly polarized, the field along the minor axis is zero.
The
circuitry is configured so that the amplitude, phase and polarization
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
33
OSSWC I
MOO. SIS.0.38V
eata
i
EXPOSED MINOR AXIS
LEVEL. T *37.3 dB
•T-BRANCH
-9° WOOD
=
-6
FREE SPACE
460*RF PHASE -
0
2
3
4
5
6
8
9
Figure 3-5
Reproduced with permission of the copyright owner. Further reproduction prohibited w ithout permission.
iO
ii ms
34
properties are available simultaneously and independently in real­
time at a rate of 2f . The functions for each branch in Fig. 3-1
s
3
are explained as follows.
a-branch:
_ _
2
2
The peak amplitude of |e *U(t) I = E
in (5) occurs
1 1
max
when the dipole is aligned along'the major axis.
It is easily ob­
tained using an ideal envelope detector followed by a low-pass filter
designed to eliminate all components of 2to and U) . Variations in
s
m
the a-branch correspond to the varying substitution loss.
In wood,
this loss is primarily a function of moisture content and thickness.
y-branch:
The ratio of the amplitudes at the minor and major axes,
Y2 = (E . /E
)2 in dB
is
min max
'
ellipticallv polarized fields.
2
Emax 1Sava^lat>^-e
of the parameters used to characterize
The amplitude of the major axis, i.e.,
the a-branch, but
determining the amplitude
along the minor axis is considerably more difficult, especially since
Y is typically of the order of -15 to -45 dB!
2
To measure E ^ ^ , the envelope of (5) is again detected using an
ideal diode.
The following low-pass filter eliminates w and all of
m
its higher harmonics, leaving the enveloDe comprised of 2a3 and its
s
•
,
six harmonics.
2
This envolope is then adjusted such that E
at b
max
2
and c in Fig. 3-1 has the same value as E
at k in the a-branch.
max
Two typical waveforms are shown in Fig. 3-6; b is for no wood present
and c is for a 30.5 x 19.0 * 4.3 cm piece of 12% moisture content
Hickory wood with its grain tilted by -9°.
Comparing the values of
2
EITlcLX for b and c , the round-trip substitution loss is 7.2 dB.
mally, however, this loss would be measured in the a-branch.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Nor-
35
To expose the minimum, the waveform in the y-branch is amplified
by another 20 dB and is hard-clipped, as shown in waveforms d and e
in Fig- 3-6-
The minicomputer samples this waveform at a 100 KHz
rate (every 0.5° for f
= 150 Hz), and picks the smallest value.
2
This is repeated ten times, and then averaged and compared to E
from the a-branch.
y is then computed after accounting for the 20 dB
difference in the two branches-
In Fig. 3-6e, y is -37.6 dB.
Obviously, determining y is time-consuming.
For each value of
y at least 5 revolutions or 33 ms are needed for f = 150 Hz.
s
Moreover, determining y becomes more difficult in the presence of
2
noisy 3^,^. This becomes a problem when the wood is too lossy, i.e.,
has high moisture content and/or is very thick.
Mechanical vibra­
tions due to high spinning speed of the dipole also contribute sig2
nificant noise to E . .
min
8-branch:
The HF phase, -2$(t), is measured at f (10 KHz)
m
using an audio phase meter.
To stabilize the phase meter input, (5)
is hard-clipped to preserve the zero-crossings, and compared with
the reference modulating signal from the audio oscillator.
The phase
meter reauires aoproximately 1.5 ms settling time for f = 1 0 KHz.
—
m
Consequently, its output is only accurate during the time when the
phase is slowly changing, i.e., in the vicinity of 6=0 or 7r, corres­
ponding to the phase along the major axis of the polarization ellipse.
In the vicinity of the minor axis, the meter loses lock because the
phase is changing rapidly and the signal is very weak.
To meet the
requirement of the minicomputer analog-to-digital converter which
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
36
requires an input signal 0 to 10 volts, a positive DC bias is added
to the phase meter output.
Two phase waveforms are shown as f and g
in Fig. 3-6, corresponding to free space (linearly polarized) and the
Hickory wood(slightly elliptically polarized)used previously, with
the grain tilted +6°.
The measured round-trip substitution phase
delay of the wood is 460°.
The computer is programmed to read and
store the 3“branch voltage when the dipole is inclined along the
major axis, as determined by measuring the axis tilt from vertical, 0,
as discussed in the next paragraph.
For wood, the phase shift is
primarily a function of the moisture content, specific gravity and
thickness.
The specific gravity of Hickory wood used here is 0.72.
0-branch:
The tilt angle of the major axis, 0, is found by
2
measuring the time difference between E
and the reference pulse
max
s
derived from the light sensor on the motor shaft.
The waveform of
the amplitude of (5) is detected and then integrated to create a zerocrossing at
Hard clipping and an automatic gain control (AGC)
amplifier are used to remove the effect due to amplitude variations.
Resulting waveforms for three cases are compared in Fig. 3-6 corres­
ponding to free space (h), and the Hickory sample with its grain
inclined at +6° (i) and at -9® (j) from horizontal.
The time
displacement of i and j relative to h clearly shows the polarization
rotation (0) of the transmitted field.
Higher moisture content and
larger thickness of the wood makes 0 equal to the grain angle 6. This
waveform and the clipped reference pulses from the light sensor
alternatively turn a Schmitt trigger circuit cn and cff, and the time
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
37
difference is computed by the minicomputer which is explained in the
next section.
3.3
Digital Signal Processing
A 150-foot communication cable is used to connect the microwave
and PDP-11 minicomputer laboratories to do real-time processing and
remote controlling.
The individual analog signals derived from four
branches are sent independently and simultaneously without cross-talk.
Voltage followers are used to improve the signal-to-noise ratio.
Some of the stray 60 Hz signal picked up by the cable can be reduced
using a notch filter.
A massive ground strap is used to reduce stray
noise pickup [Morrison,1967 and Princeton Lab., 1978].
The digital signal processing is done by sending the analog
signals directly to the computer.
To check the system's operation,
a digital storage oscilloscope which can sample the waveforms up to
200 kHz is used.
After recording the data, the oscilloscope is
interfaced with the computer to do the data processing with its
programs given in Appendix F.
Two minicomputers are used to do digital signal processing and
the schematic diagram is shown in Fig. 3-7.
The PDP-11/40 is the
master computer and PDP-11/20 is the slave computer.
They communicate
with each other via the DPll-A parallel link interface.
The PDP-11/20
receives the signal from the microwave laboratory through the DRll-C
general purpose interface and a zero detection circuit.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
A/D
a - BRANCH(MAXIMUM)
/9-BRANCH (R F PHASE)
REF
ST
OFF
©
BRANC
A /D
P D P -ll/4 0
Gate
sig.
Intr.
req. PDP-ll/20 <
o
t
Gate i
s«’g. =
ST
INF
A/D
OFF
Intr. °
req.
DRII
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Y - BRANCH (MINIMUM)
1MHz
COUNTER
Figure 3-7
IaJ
00
39
The DRll-A parallel link interface bridges the two computers.
The data can be sent byte by byte instead of bit by bit.
This saves
time when many messages are being sent back and forth.
The DRll-C general purpose interface connects the PDP-11
unibus and an external user peripheral.
It provides the logic and
buffer register necessary to program controlled parallel transfers
of 16-bit data between a PDP-11 system and an external device.
The
interrupt control logic permits the interface to gain bus control and
perform program interrupts to specific vector addresses.
The inter­
rupt enable bits are under program control, and the interrupt request
bits are under control of the user’s device.
Signal processings of the a, (3 and y branches are quite straight­
forward as discussed in the last section.
Here only the 0-branch
signal processing is explained.
In Fig. 3-7, the reference and information signals are the two
inputs from the microwave laboratory.
They first pass to the phase
detection circuit where Schmitt trigger circuits (ST), are only
activated by the positive slope zero-crossings, and used to trigger
the zero-crossings of the signals.
Each output of the Schmitt trigger
circuit passes on to drive one D-type flip-flop (DFF) which in turn
activates the interrupt request of the computer.
When the reference
signal first causes the interrupt, it also starts the counter which
is driven by a 1 MHz crystal clock (counter). Then the computer
clears the interrupt and waits for another interrupt.
Once the
counter starts, it just keeps on counting and does net stop.
When
another interrupt is caused by the information signal, the computer
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
40
simply reads the time count.
for another one.
It then clears the interrupt and waits
From this we can compute
the polarization angle 0
using
0 = 180° x (time difference)/(time period)
(1)
where the time difference is that between the reference and informa­
tion signal interrupts and the time period is that between two sub­
sequent reference signal interrupts.
The computer determines the change of polarization angle up to
2 radians and is also programmed to accumulate many cycles.
The
accuracy of the result and the maximum accumulation are limited by
the word length of the computer.
the PDP-11/40.
3.4
All the computations are done on
The data acquisition is implemented on both computers.
Servo-Control System
To automate the measurements of electrical parameters over the
entire board, a mechanism for scanning dimension lumber in both the
horizontal and vertical directions has been implemented with the
electric control circuitry is shown in Fig. 3-8.
The horizontal
scanning speed is 7 cm/sec and is done by moving the sample (sitting
on a trolley) back and forth using a bi-directional DC motor.
Vertical scanning is done by moving the spinning assembly up and
down in 1 cm step using a programmable stepping motor.
When the trol­
ley gets to the end of a scan, a micro-switch is triggered and the
direction is reversed.
Also, the stepping motor moves the spinning
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
41
dipole up or down by one step.
Dimension lumber having lengths of
1-4 ft., widths of 4-12 in. and thickness of 0.5-3 in. can then be
accommodated.
By scanning the assemblies, a raster of data can be generated
and manipulated by the computer.
Tne real-time results can be
recorded by a x-y recorder using the circuit given in Appendix G.
In conclusion, a new sophisticated microwave (homodyne) system,
having good sensitivity and wide dynamic range, has been developed
which can:
1)
measure loss (a*w) , KF phase shift (A<£), depolarization ratio
(y) and polarization angle (0) independently and simultaneously
in real-time with good accuracy.
2)
scan over a planar surface and distinguish the bulk properties
over a local region of the order of a few centimeters for
dimension lumber.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
! O-
TRANSMIT-RECEIVE
MACARTA
HORN
+ 24V
+ 24V
+ 24V
PLEXIGLASS
TRACK # 2
tj S
WIRE
t
MICRO-SWITCH
RELAY
SUPPORT
PLATFORM
PLEXIGLASS SUPPORT
TROLLEY
SPRING
(HORIZONTAL SCANNING)
- E ( dc )3
RELAY
MICRO­
SWITCH
MOTOR
6 0 0 ms
RELAY
-M h"
7400
7400
4,—-6 IO _ e
NE
555
1
100
r.mSn
STM 101
TRANS­
LATOR
CCW MODULE
STEPPING
MOTOR
(VER TIC A L SCANNING)
bms
F ig u r e
3 -8
43
4.
Our objective is to correlate
Experimental Results
the electrical parameters
(a*w, A<p, y and 0) and the physical parameters (w,M.C., S.G. and 0)
of wood.
The notations used for the electrical parameters were
defined in Chapter 2 and those for the physical parameters are
defined here.
w = Thickness in cm
„ _ = Moisture
.
« j. x. = weight of water in wood
M.C
Content
-- —. , -- =
•— ---------weight or dry wood
m
%
_ _ _ _
._. _
.,
, weight of dry wood ,
— ) relative to water
S.G. - Specific Gravity = {— ----volume of green wood
0 = Grain Angle in degrees
The instrumentation operating procedures for these measurements
are given in Appendix H.
Many wood samples of different species and varying physical
parameters were tested to quantify the relationships between the
electrical parameters and the physical parameters.
The following
lists the various species tested and their physical properties:
1)
White
Pine (WP) with S.G. values from 0.31 to 0.38, M.C. of
0, 6,12 and 204%, and w from 1.5 to 5.08 cm.
2)
Basswood (3S) with a S.G. of 0.35, M.C. of 0, 5 and 12% and
w of 1.8 cm.
3)
Aspen
(AS) with S.G- from 0.38 to 0.41, M.C. of 0, 6, 12 and
130% and w from 2.54 to 4.75 cm.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
44
4)
Douglas Fir (DF) with S.G. from 0.38 to 0.46, M.C. of 0, 6, 12,
31, 45, 52, 136 and 150% and w from C.7 to 5.08 cm.
5)
Red Oak (RO) with S.G. from 0.56 to 0.62, M.C. of 0, 6 and 12%
and w from 1.8 to 4.1 cm.
6)
Hickory (HI) with S.G. from 0.72 to 0.73, M.C. of 0, 6, 12,
45 and 48% and w from 2.4 to 4.5 cm.
The grain angle was generally varied from -20° to +20° from
horizontal for all the samples by rotating the beards about the spin
axis of the scattering dipole, as shown in Fig, 4-1.
PIVOT POINT
i-
I-
— WOOD
ABSORBER
ROTATABLE
PLEXIGLAS SUPPORT
if
_ ^ 1 /8 "F IB E R G L A S
PLATFORM
Figure 4A1
i
f
V '
Thi»t results are/ repeatableland the relationships between the
/ ,
„
k
,
/
..........
electrical and physical parameters are similar for all six species.
The piresent instrumentation system gives consistent results for
CL'W,jA<£, y and ©/men the one-way loss is less than 20 dB, although
the sensitivity pan be improved as discussed in Chapter 6.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
45
To isolate the role of the wood on the microwave signal, the
effect due to interface reflections is corrected in all the following
data using an iterative method explained in Chapter 2.
The effect of
multiple internal reflections is very small and hence neglected.
The experimental results are then analyzed and summarized as follows:
4.1
Relationships Between the Electrical and the Physical Parameters
of Wood
a.
Loss (a*w) in dB
1.
As expected, the loss increases linearly with thickness
when the other parameters are held constant (see Fig.
4-2 where only (a*w)| is shown)
2. M.C. was found to be the dominant factor controlling
loss.
This is because water absorbs microwave energy
much more effectively than dry wood, especially when it
is confined in the material.
Thus even a small amount
of water can be detected with fairly good accuracy
through loss measurements.
Fig. 4-3, shows that loss
increases linearly with moisture content, with w as the
parameter (only (a*w)^ is shown). Also the unit
losses
and a.j vs M.C. are shown in Fig. 4-4.
3. Loss increases with specific gravity for both hardwood
and coniferous woods, but with different slopes.
clarity, loss is normalized to unit thickness in
Fig. 4-5.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
For
46
4.
Loss is independent of 0 for
18f<20°,
with small
variations for high moisture content and thicker
specimens (see Fig. 4-6).
5.
Loss has maximum and minimum values at 0 = 90° (II) and
0 = 0° (1) respectively, since Ejj suffers the greatest
loss.
6.
For relatively thin and dry wood, the loss has the
greatest variability.
Presumably this is because the
effects of reflections are comparable with the dissi­
pative loss.
7. -Ihe error in M.C. is smaller at 0 = 90° (II) than that
at 0 = 0° (1) , since E|j suffers the greatest loss and
is therefore the most sensitive to loss.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
MC = 150 %
DOUGLAS FIR
30
m 20
52%
45%
12 %
6%
0%
0
T H IC K N E S S w (c m )
Figure 4-2
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
20
DOUGLAS FIR
DOUGLAS FIR
w = 5 .0 8 cm
CQ
*o
W
-H
3 .7
2.6
2 .4
0.8
,4 .0 .
50
100
150
M O I S T U R E C O N T E N T (%)
Figure 4-3
(uio/gp) nz) puDTc S S 0 1
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
20
15
10
5
0
0
50
100
150
M O I S T U R E C O N T E N T (%)
Figure 4-4
2.8
2.5
} HARDWOODS
^ CONIFEROUS WOODS
2.0
£
CD
/ / !2%
X3
«
■o
c
o
6%
CO
CO
o-J
12%
0.5
0.3
0.3]
WP
10.4 |
AS DF
0.5
|0.6
0.7
RO
SPECIFIC G R A V IT Y
Figure 4-5
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
HI
—
50
DOUGLAS F IR
20
w = 2 .0 2 cm
M C = i5 0 %
LOSS
(ocw) (dB)
»
w = 4.9 cm
MC = 5 2 %
0
w = 2 .5 4 c m
MC = 4 5 %
w =5.03cm
6Q/o ,*2 %
0
r w = 2 .5 4 c m
0%
-20
-10
GRAIN A N G L E
9
(degress)
Figure 4-6
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
51
b.
RF phase shift (Acj>) in degrees
1-
A<p increases linearly with thickness since A<£ is
approximately proportional to -/z' w (see Fig. 4-7
where only A<^ is shown).
2.
For low moisture content, water dipoles are physically
bound to the material, so their mobility is limited.
Water then affects the resultant permittivity less than
for higher M.C. (see Fig. 4-8).
The phase change is
chiefly a function of the wood's real dielectric con­
stant, which a strong function of density.
However,
after exceeding a threshold which is determined by the
material structure or the amount and size of pores or
capillaries, (around 20% M.C. for Douglas Fir) the
material is no longer able to bind more water dipoles.
Free molecules are then formed which affects the result­
ant permittivity to a greater degree, causing A<S> to
increase linearly and with greater slope (only A6^ is
shown). The threshold value is a measure of the
hydroscopic activity.
Thus, the microwave method pro­
vides a means for determining of the characteristics of
water binding forces and for investigating the material
structure.
After removing the effect of thickness, the
phase constants 3j_ and Sjj vs M.C. are shown in Fig. 4-9.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
52
3.
A<j> increases with specific gravity since the dielectric
constant £' increases.
For clarity, A<p is normalized
to unit thickness to give <3^ and 3|| in Fig. 4-10.
4. A<j) is independent of grain angle 0 for (0 [ <20°,
since £' is almost -unchanged (see Fig. 4-11).
5. A<j> has maximum and minimum values at 9 = 90° (II) and
8 = 0° (1) respectively, i.e., Ejj has the greatest phase
shift-
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
DOUGLAS F IR
500
150%
400
Z
300
o>
©
TJ
200
12%
1 0 0
0 %
0
2
4
THICKNESS w (cm)
Figmre 4-7
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
54
I—
z
UI
h
Z
o
o
II I
tr
3
H
cn
i
•"S'
o
O'
fa
CO
Li.
o
cj
10
o
—
—
(UJ0/-6ap)»g
pud
in
1# 1 N V 1 S N O O 3 S V H d
in
IZ
UJ
l-
z
o
CO
o
in
UJ
q:
I
■s*
o
3
H
O'
•H
10
CO
O
2
O CO.
^ ro
lO
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
S3
3
fa
55
70
MC= 12%
} HARDWOODS
A CONIFEROUS WOODS
60
£'
o
w
©
©
o>
©
*o
<2T 4 0
tj
#■»
o
30
12% y 6 %
10
0.3
WP BS
0.6
AS DF
RO
S P E C IF IC G R A VITY
Figure 4-10
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
56
500'
DOUGLAS FIR
A 4> ( degrees)
400
w=4.9cm
’M C = 5 2 ° /c
300 -
200
w=2.54cm
‘MC =52%
6%
lw= 5.08cm
-0
0
f
■Q— 11^ o*
Q/^ 0 %
LS
0j-w = 2.54 cm
0%
100
-10
GRAIN A N G L E Q (degrees)
Figure 4-11
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
57
c.
, Depolarization
ratio
(y) in dB
The measurement ofy is the most difficult.
E . is usually
min
■*
a very weak signal which can easily be smeared by environmental dis­
turbances or noise, resulting in misleading conclusions.
Thus most
of the following data were taken manually due to mechanical vibrations
of the spinning dipole.
over several cycles.
Dynamic operation required averaging E ^ ^
The general conclusions for y in our results
are summarized as follows:
1.
p^ and pjj increase with thickness since the depolariza­
tion path is longer, and p^ and p^ are both weak func­
tions of w for low M.C. (<_ 6%) and thin boards
(w <_2 cm) (see Figs. 4-12 and 4-13, respectively).
2.
p^ increases with M.C.. Presumably E^_.^ is quite small
and comparable to the noise, but there is significant
couding from E
to E . for higher M.C., and E
is
max
man
max
suffering greater loss.
For M.C. <_ 30%,
is generally
less than -30 dB as shown in Fig. 4-14.
3.
p^ increases with M.C., for the same reason as given for
Pj^.
Note that for M.C. <_ 30%, pjj is greater than -30 dB
for most of the specimens, and somewhat greater than
p^ (see Fig. 4-15).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
58
4.
For M.C. <_ 12%,
and py decrease and increase with
specific gravity, respectively.
Presumably the aniso­
tropy increases with specific gravity, so the parallel
polarization ellipse suffers the greatest attenuation
(see Figs. 4-16 and 4-17 respectively).
However, in
some of our results we found that at low M.C. different
fiber structures may change this conclusion 5.
y is maximum around 9 = 45° (Fig. 4-18) , where Ey = E^
at the front interface.
At 9 = 90° (E is 11 to the
grain) y has a relative minimum Cpjj) , while at 9 = 0°
(E is 1 to the grain) y is again minimum (p^) , with
Py > pj^ (see Fig. 4-18 to 4-21).
For large M.C. and
thickness, y is nearly constant with 9 (Figs. 4-20 and
4-21).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
to oc
u.
O
O
ro
O
O
4 -1 3
F ig u r e
THICKNESS
w (cm)
59
O
O
THICKNESS
o
o
o
o
o
o
( 8 P ) Td O l i v a N 0 ! l V Z I d V 1 0 d 3 Q
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
.4-12
£\J
£ 9to9
F ig u r e
tDS? —
w (cm)
(9P)Hd O l i v y N 0 i l V Z i a V 1 0 d 3 Q
60
o rr
Q i=
2
O
u
UJ
tr
i
on
>
ro
cq to GO
CO
» It H
? ?
*
o
cr
CO CO
<£
3
fa
in
O
w o
2
<3*
in
fa
1
<u
n
3
01
•w
fa
CO
II
3y
CO
O
i
(sp)'& oiiva Noiivziaviod3a
I
<u
5-1
3
05
•H
fa
_ to
o
o
ro
O
O
(9P) Td Oliva NOllVZiaVlOdBQ
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
O
p„(dB)
to
T>
MC = 6 %
CL4
z
o
30
_J
O
CL
UJ
Q
RATIO
6%
CC
OH
<
12%
12%
9. - 4 0
5
N
6%
20
DEPOLARIZATION
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
-3 0
-10
AS
6%
w 3 4.3
AS
-20
0.3
0.4
0.5
0.6
0.7
SPECIFIC GRAVITY
Figure 4-16
0.8
0.9
0.3
0.4
0.5
0.6
0.7
0.8
0.9
SPECIFIC GRAVITY
Figure 4-17
cn
H
Y (d B )
RATIO
DEPOLARIZATION
-70 ~
SUGAR PINE
MC= 6%
w= 3.5 cm
-63
-5 6 -4 9 -4 2
n
Pll
n
- 3 5 ' K\
\
-2 8 - \
\
-21
\
N
V
-14
P
S
S
s
s
s
+s
£5*
-7 0
...
0
.1
!
20
1---1---- !--- 1
40
60
1
1 1
8 0 90°
GRAIN ANGLE 8 (degress)
Figure 4-18
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
DOUGLAS FIR
w s5cm
r(dB)
-40
-4 0
RATIO
DOUGLAS FIR
we2.54cm
•30
MC « 0 %
6%
<
•n
o
O
-30
£:
a
52%
MC= 45%
2
O
DEPOLARIZATION
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
-5 0
-50
N
§
-20
_i
o
a.
U
J
ci
-10
-10
o
GRAIN ANGLE 0 (degrees)
-20
-10
-10
O
10
GRAIN ANGLE
6 (degrees)
Figure 4-19
Figure 4-20
64
-40
DOUGLAS FIR
MC = 150%
w = 0 .6 7 5 cm
-3 0
2.02
o - lO
-
10°
0
GRAIN ANGLE
Q
10°
(DEGREES)
Figure 4-21
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
65
d.
Depolarization index (M) in percent
The depolarization index is defined as the ratio of polariza­
tion angle to the grain
angle (M = 0/6) in percent.
It is a measure
of the degree to which the measured polarization angle 0 follows the
grain angle 9.
Some conclusions concerning M are summarized as
follows;
1.
Wood materials having random fiber directions, e.g.,
plywood and fiber board, show no significant.depolariza­
tion.
2. M is too small to be measured reliably for thin
(w <_ 1.5 cm) and dry (M.C. <_ 6%) wood.
5. M increases linearly with w, M.C., and S.G. since
there is either a longer depolarization path or more
anisotropy (Figs. 4-22, 4-23 and 4-24).
Ultimately
M reaches 100% for high moisture content with sufficient
thickness.
M vs M.C. with M normalized to a unit
thickness is shown in Fig. 4-25.
4. M is independent of 9 for small 6 since the change in
anisotropy is small (see Fig. 4-26).
Equivalently M
vs 6 is a constant for j0] <20° and given values of
M.C., S.G. or w (see Figs. 4-27, 4-28 and 4-29).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
66
5-
M has the largest and least value at S = 90° (II) and
9=0°
(1) respectively, since the depolarization is
the greatest when E is aligned with the grain (Ey).
For example, in our preliminary results we found that
My is about twice the value of Mj_ for both Red Oak and
Basswood with 6% M.C.
This suggests that a grain angle
instrument would more effectively use || polarization,
and is a topic which should be considered in future
studies.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
67
PJ*_
h-
Z
UJ
QL
3
h
Figure
Z
o
o
4-23
LU
CO
’Oj
X3QN! N 0 llV Z I3 V 1 0 d 3 a
£
u
4-22
CO
CO
UJ
Figure
o
z
o
X
H
to
X
o
O
o
o
o
CM
o
O
(%) W X3QN’! NOIlVZiaVlOdia
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
60
w= 5 .0 8 cm
MC= 12
Z
o
6 %
cr
<
_ j
o
CL
UJ
Q
n
WP
DF
0.2
0.4
SPECIFIC GRAVITY
0.6
Figure 4-24
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
69
100
DOUGLAS FIR
80
x
20
hi
50
100
150
MOISTURE CONTENT (%)
Figure 4-25
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
70
M (%)
60
DEPOLARIZATION
80
INDEX
D O U G L A S F IR
w = 2 .5 4 c m
M C=45%
<5— >
40
w = 5 .0 S cm
6%
20
— w = 2 . 5 4 cm
o%r
6%
-20
-10
0
I
GRAIN A N G LE S(degrees)
Figuxe 4-26
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
O
©
ANGLE
H l(w c 4 .3 c m )
DF ( 5 .0 8 )
fRO (4.1)
^ 6
IAS (2.6)
-3
GRAIN ANGLE
OF
(2 .5 4 )
BS
<l 8 >
w = 2 .5 4 cm
12
•
POLARIZATION
- J 15°
AS
-3
-6
-•-6
-9
-9
12
Figure 4-27
(12)
D f-H I (12)
DF,HI ( 6 )
Lw
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
O
GRAIN ANGLE
9
■•-12
Figure 4-28
-j
>P
(0
ii
£
o
£
o
00
o
I
e>
2
«»
tO
I - - CVJ
\\--o>
\\!
o
to
n. II
3 3
cvi
o® 3 1 9 N V N O U V Z iy V lO d «■
-M
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure
"S
Cd
n
O
2
4-29
72
73
These limited data have shown trends over limited ranges of the
physical parameters.
Surprisingly, many of these trends are linear.
Overall, the most significant results are that a*v, A<jj and AT are
J
essentially independent of the grain angle for }8 < 20°.
Thus,
knowing a and 3 are sufficient to determine the M.C. and S.G.
uniquely.
Knowledge of y or 0 is not essential.
This is encouraging
because measuring y is a very difficult problem as mentioned before.
Thus an extremely complicated problem involving 64 combinations is
reduced to 4 combinations, i.e., a and 3 vs M.C. and S.G., plus one
straightforward problem (0 vs 9).
In practice then, we should first
find the M.C. and S.G. values by matching measured a and 3 in a
standard table.
In turn these can be used to find M.
the measured 0 we can find the grain angle 9.
Then, from
This simplifies the
testing problem considerably, and will be especially helpful when
dynamically testing in real-time using the minicomputer.
Indeed, in
view of these results, an efficient computer program has been written
which takes advantage of them.
This will be discussed further in
the next chapter.
4.2
Relationships Between the Dielectric and the Physical Parameters
of Wood
Electromagnetic dielectric properties are potentially useful
diagnostic indicators of the physical and mechanical properties of
wood.
The complex dielectric constant is anisotropic (i.e., a
tensor), and a strong function of the specific gravity and moisture
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
74
content,
to important new achievement of this work is that we are
able to compute the complex dielectric tensor from the measured
parameters.
a.
The dielectric tensor
To quantitatively understand the dielectric properties of wood,
the dielectric tensors are computed for various species and M.C.
using the method described in Chapter 2.
First, two known dielectric
materials, i.e. polystyrene foam and plexiglas,
calibrating the instrumentation.
are tested for
Their results were found in good
agreement with published data and are summarized as follows
[Von Hippel 1954]:
1.
Polystyrene foam (published value £r = 1-04) , w = 12 cm.
dielectric tensor
loss tangent
1.0459-jO.00S2
0.00004-j0.000042
tan <5|| = 0.00877
0.000041-j0.000043
1.04296-jO.012
tan <5^ = 0.0115
2.
Plexiglas (published value £
= 2.6), w = 2.9 cm
dielectric tensor
loss tangent
2.322-j0.02926
0.000217-jO.000377
tan S„ = 0.0126
0.000183-jO.000318
2.285246-j0.05036
tan 6^ = 0.022
Thus both materials are very isotropic and have low loss.
The
inconsistency of plexiglas is presumably due to reflections, edge
diffraction and surface waves.
The latter could have a significant
effect.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
75
Dielectric tensors were computed for most of the wood samples
tested, but only five wood species with two M.C. values of 5 and 12%
are listed for comparison.
1.
Hardwoods
6%
a.
Aspen:
12%
(S.G.)
avg
= 0.4
2.11-jO.21
0.0Q137+j0.006
2.4755-jO.64
0.00468+j0.0073
0.00283+j0.0121
1.69-j0.11
0.0151+j0.0235
1.8-jO.183
tan-5^ =0.065
tan <Sjj = 0.258
tan 6^ = 0.104
tan 6|| = 0.1
b.
Red Oak:
(S.G.)
avg
= 0.6
2.71-j0.45
0.00174+j0.0034
3.1-jO.86
0.00165+j0.0032
2.3-jO.212
0.01606+j0.0216 2.56-jO.37
tan 6^ = 0.092
tan 0 j| = 0.277
tan 6|. = 0.166
c.
Hickory:
(S.G.)
avg
0.0425+j0.0057
tan 6^ = 0.145
= 0.72
3.8-jO.87
0.0035+j0.007
4.57-jl.4
0 .01+j0.012
0.03+j0.07
2.95-jO.525
0.07+j0.1
3.58-jO.8
tan 6^ = 0.178
tan 5„ = 0.306
tan
tan 6„ = 0.229
2.
0j_=
0.2235
Coniferous Woods
a.
Douglas Fir (S.G.)
avg
= 0.43
2.35-j0.15
0.0018+j0.0075" "2.73-jO.7
0.00524+j0.00784
0.0027+j0.01
1.9-j0.095
0.0123+j0.0184
2.156-j0.31
tan 0j^= 0.051
tan 6„ = 0.236
tan o, = 0.095
tan 6|j = 0.064
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
76
b.
White Pine (S.G.)
avg
= 0.33
NA
NA
£
1
NA
NA
1.9-jO.091
NA
2.0-j0.14
tan 6^ = 0.042
tan <5^ = 0.07
For comparison, the corresponding data obtained by James and
Hamill [1965] for Douglas Fir are:
6% M.C.
2.0-j0.26
12% M.C.
NA
NA
~
1.8-j0.117
tan 6j| =0.13
tan 6^ =0.065
2.85-jO.627
NA
NA
2.0-j0.24
tan 6|j = 0.22
tan 0j^ = 0.
~
The James and Hamill data were taken at 1, 3 and 8.53 GHz.
These data were interpolated to 4.8 GHz for direct comparison with
our results.
The comparison is in reasonable agreement, considering
that quite different methods and samples were used.
The ratios of £' /£' and £ ’ /£' are larger for most of the
XX
zz
zx xz
3
hardwoods than for coniferous woods, regardless of specific gravity.
This means that hardwoods are more anisotropic than coniferous woods,
and that the depolarization index for hardwoods will therefore be
larger.
These anisotropic characteristics are due to cell wall
orientation rather than microscopic structure differences in wood.
Rafalski [19663 supported this issue by demonstrating that the dif­
ference in the dielectric properties of beech for radial and tangential
directions reduces when wood is compressed across the grain, and that
£' becomes equal in both directions when the specific gravity of the
specimen is increased to 1.45.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
77
Some general conclusions about the computed dielectric tensors
are:
1.
The complex dielectric constant and the loss tangent are the
greatest when E is aligned with the grain, or,
(£' / £” / tan 6 ) > ( £ ' , £" , tan <5 ).
xx
xx
xx
zz
zz
zz
This separation is enhanced by M.C. because the elongated
fibers tend to become more conducting with added water.
2.
The complex
dielectric constant and loss tangent increase sig­
nificantly with M.C. because of the extremely high dielectric
constant and loss of water.
Let y and y' be two different
values of M.C. with y > y'.
Then
(ett'
3.
*“ Vy >
£tt'tanV V
k=x'2
There is more coupling from Ejj to E^ than vice versa, and this
effect is enhanced by moisture.
£
/t
> j £ /t I >
1 zx xx‘v
1 xz zz'y
In general,
]£ / £
! , > ! £ / §
[ ,
1 zx xx'y1
' xz zz'y’
e.g., for sugar pine, with y and y’ equal to 25 and 6%
respectively, the corresponding values are
-32 > -38 > -49 > -50
respectively.
(dB)
Note that the large negative values means off-
diagonal (coupling) terms are much smaller than the diagonal
terms.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
78
4.
Perhaps more important is the fact that the off-diagonal terms
(£' and £' ) increased by a factor of the order of 2 to 8 times
xz
zx
as the M.C. is doubled.from 6 to 12%.
This is considerably
greater than the increase in the diagonal terms, which is
typically less than 1.3 times and is commonly used as a criteria
for detecting M.C..
Clearly, moisture enhances anisotropy of
the wood and the off-diagonal terms are very sensitive functions
of M.C.
To further understand the behavior of the off-diagonal terms
with M.C., Douglas Fir specimens with M.C. values ranging from
0 to 150% were tested:
M.C.(%)
dielectric tensor
loss tangent
tan
0
6
12
52
150
1.9-jO.053
0.00042+j0.0025
o
XX
tan
0.0279
-6
tan
tan
5
zx
5
o
0.00024+j0 .0018
1.68-jO.02
-11.6
0.012
2.35-j0.15
0.0018+j0.0075
0.064
-4.2
0.0027+jO.01
1.9-jO.095
-4.2
0.051
2.73-jO.7
0.00524+jO.00784
0.253
-1.5
0.0l23+j0.0184
2.156-jO.31
-1.5
0.144
Tl.7-j4.8
0.03+j0.049~
0.41
-1.63
0.2+j0.33
6.2-j2.2
-1.65
0.355
"38.3-j25.2
0.96+j0.93“
0.68
-0.97
3.46+j3.36
23.8-jlO.8
-0.97
0.45
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
79
Note that the "loss tangents" for the off-diagonal terms are
negative and greater than 1 except for green wood, in contrast to
the loss tangents of the diagonal terms which are positive and less
Positive values of tan S for the diagonal terms signify a
than 1.
true loss, while the negative values for the off-diagonal terms
imply the "coupling" from the other electric field component.
it signifies there is a "gain" rather than a "loss".
Thus,
Larger "nega­
tive" values for the off-diagonal terms imply stronger cross-coupling
(or depolarization). For green wood, free water presumably dominants
the effect on the microwaves, and the anisotropy (due to the grain
angle) is reduced, giving less depolarization.
To quantify these observations, results for the off-diagonal
terms is tabulated in Table 4-1, where the F ’s denote the ratios
between the dielectric terms at the higher and lower M.C. values
respectively.
Table 4-1
M..C. (%)
F (S' )
xz
0
5
4.3
6
12
3.0
1.05
12
52
5.725
6.25
16.3
18
32
13.9
17.3
10.2
52 -b 150
F(E" )
xz
3
F (£' )
F(£" )
11.25
5.55
4.55
1.84
ZX
ZX
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
80
For comparison, the corresponding data for the diagonal terms
is also tabulated:
Table 4-2
M..C.(%)
F (£' )
F(£" )
F(e’ )
zz
F(E" )
zz
0 -> 6
1.24
2.8
1.13
4.75
6 ■+ 12
1.16
4.67
1.134
3.26
12 -> 52
4.285
6.85
2.875
7.33
52 + 150
3.27
5.35
3.84
4.91
XX
XX
We therefore conclude that the off-diagonal terms of the
dielectric tensor are much more sensitive to the M.C. than the
diagonal terms, in particular, for e’ . The off diagonal terms
ZX
vary exponentially with M.C. as shown in Fig. 4-30 and 4-31
for t
and t
respectivelv and
xz
zx
shown in Fig. 4-32.
tan <5 and tan o
vs M.C. are
xz
zx
They will undoubtedly play a key role in
future M.C. measurements.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
81
in
ro
I
<3ui
CD
U
a
O’
•H
fa
o
O
tri
O
O
Q
XZ-
O
O
p u o «9
0o
c
1
<5*
u.
CO
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CO
O'
fa
82
12.5
CO
10
DOUGLAS FIR
Ton §xz
0,1
Tan 8ZX
x—
-
X
CO
0
50
100
M O IS T U R E C O N T E N T (%)
Figure 4-32
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
150
b.
Complex Dielectric Constant and Loss Tangent
The complex dielectric constant and loss tangent are next
correlated with the physical properties of wood.
(a) Dielectric constant (£') and loss factor (£")
1.
£| |j and
|| increase exponentially for M.C. below the fiber
saturation point (around 25%) and then increase linearly above this
point (see Fig. 4-33 and 4-34).
Below this point the increase is
caused by the high dielectric constant and loss of water compared
with the relatively low dielectric constant and loss of cellulose.
In addition, the polar groups in the cell wall and the cellulose
have increased freedom of rotation with increased M.C. Above this
point, the increase is due to the more free water which
hasan
extremely high dielectric constant.
2.
£_[ [j
£j| || change with S.G. in an exponential way, at
least for hardwoods (see Figs. 4-35 and 4-36 respectively). Also
from these results we are able to distinguish between hardwoods
(AS, RO and HI) and confierous woods (WP and DF) by their
different dielectric properties as shown in Fig. 4-35.
The slopes
are seen to be about the same at low MC but £^ j| are significantly
higher for coniferous woods at the same S.G.
3.
££ || and £j| |, have maximum and minimum values at 0 = 90° (II)
and 0 = 0 °
(1) respectively.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
40
I DOUGLAS FIR
30
*o
g 20
50
100
150
M O ISTU R E C O N T E N T (%
Figure 4-33
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
85
25
DOUGLAS FIR
20
Vl7
-a
c
a
0
50
[00
150
MOISTURE CONTENT (%)
Figure 4-34
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
86
5.0
* } HARDWOODS
a
CONIFEROUS WOODS
4.0
2%
vi
T3
c 3.0
MC=!2%
12%
D
V
6%
H2%>
2.0
0%
-A—
6 % ^
A'-'"'
.
0%
WP
1.0
0.3
AS DF
0.4
RO
0.5
0.6
0.7
S P EC IFIC GRAVITY
Figure 4-35
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
87
1.35
HARDWOODS
12%
CONIFEROUS WOODS
0.8
12%
0.6
6%;
0 .4
0.2
0.1
03
wp
0.4
^S DF
0.5
0.6
RO
SPECIFIC GRAVITY
Figure 4-36
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
0.7 |
88
(b) Loss tangent (tan o)
1.
tan 6^ jj increase with M.C. in a way which is slightly con­
cave downward below the fiber saturation point and linear above this
point, eventually reaching the maximum value for green wood (see
Fig. 4-37).
This maximum is much larger than that of pure water
which has a loss tangent around 0.12 at 4.81 GHz [King 1965],
In
fact, in the majority of wet substances, absorbed water properties
behave very differently from those of pure water.
microwaves are severely scattered inside
Presumably the
the material making the
effective wave path larger than the physical thickness.
Different
fiber structures, porous, granular and powdered, can bind water in
many different ways, depending on the density, temperature, chemical
and physical composition [Tinga 1969].
It is also interesting to compare with the results from James
and Hamill [1965] who worked at frequencies of 1, 3 and 8.53 GHz.
They found that the loss tangent measured at 1 and 3 GHz increased
rapidly at low moisture levels, and then show a concave downward
trend near fiber saturation at arcund 30%.
The loss tangent of green
material was somewhat lower than for material near fiber saturation,
suggesting a maximum value near the fiber saturation point.
However,
the loss tangent measured at 3.53 GHz increased with increasing M.C.
but with a concave upward trend, in contrast to the concave downward
trend observed at 1 and 3 GHz.
The linear slope of our curve for
4.81 GHz is between concave downward and upward.
The loss tangent
increases with frequency because the polar dipoles cannot follow at
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
0.8
DOUGLAS FIR
tan 8
0.6
tan S
0.4
H 0.2
0
50
!00
150
M O ISTU R E C O N T E N T (%)
Figure 4-37
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
90
higher frequencies and thus
£' decreases, but still e" increases
due to the increasing combined conductive and dielectric losses of
free water.
2.
As expected, tan
3.
tan 6 has maximum and minimum values at 8 = 90° (li) and
8=0°
y increases with S.G. (see Fig. 4-38).
(i) respectively.
As expected £' increases with M.C. and S.G. in a manner similar
to that for
y and 8( ||which were shown in Figs. 4-4, 4-5, 4-9 and 4-10.
Our results for the relationships between the dielectric and
physical properties of wood show the same trends as obtained by others
at lower frequencies [Yavorski 1952, Skaar 1948, Venkateswaran, et al.
1964].
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
0.6
i
0.5
a
r
} HARDWOODS
CONIFEROUS WOODS
00
0.4 h
£
MC=!2%
__ .2L— — ‘
0 .3 1
WP
0.41 I
0.5
AS DF
S P E C IFIC G R AVITY
Figure 4-38
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
92
Quantitatively, the diagonal complex dielectric tensor terms
and loss tangent for Douglas Fir at various M.C. are compared against
James and Hamill [1965] in Table 4-3.
Table 4-3
p»
£i
e’l
M.C.
Ours
0
1.68
6
1.9
J&H
Ours
J&H
0.02
1.8
12
2.156 2.0
16
25
tan
Ours
J&H
0.012
0.095 0.12 0.051 0.07
tan 6
£|l
e ii
1•
Ours J&H
Ours J&H Ours
1.9
0.053
2.35 2.1
0.15 0.27 0.087 0.13
J&H
0.0279
0.25 0.144 0.125 2.73 2.85 0.7
0.63 0.253 0.22
2.3
0.4
0.175
3.2
0.85
3.2
0.83
0.26
4.7
0.3
31
4.05
1.346
0.33
45
5.7
2.0
0.35
52
6.2
2.2
0.35
75
*1
11.7
9.8
0.265
4.8
0.41
25.2
0.67
13.0
136
22.2
9.4
0.42
150
23.8
10.8
0.45
38.3
They are in reasonable agreement for M.C. less than 30%, consider­
ing that quite different methods and wood samples were used.
James
and Hamill [1965] had difficulty measuring the loss factor above 25%
M.C. and Tinga [1969] only has data up to 30% M.C.
Density-independent moisture measurement is desirable for many
applications and two interesting methods were recently introduced.
Meyer and Schilz [1980, 1981] concluded that for less dense (S.G. =
0.05 ~ 0.2) and compressible organic substances, the factor
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
93
A = (s’-lj/e" is independent of density and is useful for moisture
measurements.
In our results, we found that this factor is relatively
independent of the specific gravity (or density) for the same species,
and decreases with specific gravity for both coniferous woods and hard­
woods as shown in Figs. 4-39 and 4-40, respectively. This is basically con­
sistent with Meyer and Schilz's conclusion, although our wood speci­
mens are much more dense having S.G. values varying from 0.3 to 0.72.
This makes comparisons more difficult.
The decrease of A with S.G.
in Fig. 4-40 is due to different fiber structures for the different
species.
As expected, this factor decreases with M.C. as shown in
Fig. 4-41, since (£*-l)/£" <= — ^ ^ is playing the role of the loss
tangent reciprocal.
The loss (a) and phase (6) are linearly related, with M.C. as
the parameter as shown in Figs. 4-42 and 4-43 for II and 1 polariza­
tion respectively.
This is also consistent with Schilz and Schielk’s
results [1981].
Kraszewski et al. [1976] described another method wherein the
measured attenuation (A) and phase shift (A<j>) are used along with an
empirical method to calculate four coefficents, say a , a , a ,
X
«
j
and a^.
These are specific for a particular material and its
physical properties, including the thickness.
The M.C. can then
be computed using the equation
Aa -(Acb)a
M c =
(A$)(ai-a2)-A(a3-a4)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
n )
10
DOUGLAS FIR
8
6
4
2
0
0 .3 8
0 .4 0
0 .4 2
0 .4 4
0 .4 6
SPECIFIC GRAVITY
Figure 4-39
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
95
15
HARDWOODS
CONIFEROUS
WOODS
10
MT
I
vu*
II
<
5
t ASf
WP OF
0
0.2
0.4
0.6
SPECIFIC GRAVITY
Figure 4-40
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
0.8
12
10
R0
8
6
4
DF -
2
0
50
100
150
MOISTURE C O N T E N T (%)
Figure 4-41
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
97
20
150%,
DOUGLAS FIR
a,i (d B /c m )
15
10
5
S2%
0
100
200
300
i (d e g re e s /c m )
Figure 4-42
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
DOUGLAS FIR
100
200
^(degrees/cm )
MC = l50°/c
300
Figure 4-43
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
99
4.3
Relationships Between the Reflection Coefficient and the
Physical Parameters of Wood
Measurement of the reflection coefficient can also be a valuable
approach to find the moisture content.
In some cases, such as very
wet (or effectively infinitely thick) material, it is probably the
best approach.
For example, measurements of wet material by Kraszewski
[1973] in the x-band showed attenuation of about 30 d3/cm.
By measure­
ment of the reflection coefficient, an on-line accuracy of better
than 0.3 percent was obtained.
In these measurements, standing wave
detectors were used to form a reflectometer.
Using the dielectric -tensor as calculated from cur transmission
measurements, we can compute the complex reflection coefficients as
well as the transmission coefficients using the method described
in Chapter 2.
The general conclusions for Douglas Fir are summarized
from Figs. 4-44 and 4-45.
1.
The magnitude of the reflection coefficients, i.e.,
j R^ J
and |Rj| | vary periodically vs. w with shorter period
for higher M.C.
The oscillations decrease exponentially
and essentially reach a constant most rapidly for higher
M.C.
2.
IR,. I
and |r . I
increase with M.C.
1 II'max
1 1 ‘max
3.
|Rii I
> |r , I
for given M.C.
1 li'max
' 1 ‘max
The relationship between the complex reflection coefficient,
the thickness and the dielectric constant can be found in many text­
books [e.g., Ramo et al. 1965] and has a mathematical form
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
for a particular polarization, where
cos0 + j sin0
1
a
.
cos9 + j J £^_ sin9
is the normalized load impedance and
0
=
The normalizing
$q w *
impedance is the intrinsic impedance of free space in our case.
The periodic behavior with w is clearly evident.
If we insert z
L
into (2), then
2
, 1
- (£_-l)sin 9 + j sin0 cos9 ' pr
- vr rr~)
(3)
R =
2 + (t -l)sin^9 + jsin0 cos9( i_
/F
-/ y
and
-1
r
£
'max
which predicts that |R lmax
2+(£ -1 )
r
1 f°r very high M.C.
(|
equals
0.86 for 150% M.C. in Fig. 4-45).
These formulas are only approximate in our case since depolariza­
tion due to the anisotropic properties is not accounted for.
this effect is small.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
However,
101
CM
u.
£
o
:n
I
rr
<D
Ui U
D
2 ■cHn
O
X
t-
O.
(S
CM
o
\"V\
r>
v
O
iTa
i
V
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4.4
Effect of Growth-Rings on the Polarization Angle
Two tests were conducted to investigate the effect of growth-
rings on the polarization angle:
Test 1
Two pieces of Douglas Fir were tested, both having the same
M.C. (13%) and w (2.8 cm) but with the growth-rings parallel and
perpendicular to E^ as shown in Fig. 4-46.
The measured depolarization index (M) was 50% for both samples.
Their dielectric tensors and loss tangents were also computed for
comparison:
Sample 1
Sample 2
2.7348-jO.6927
0.00525+j0.008
2.6522-jO.6884
0.00667+j0.008
0.0123+jO.0186
2.1563-j0.31
0.0121+j0.0145
2.1808-j0.296
tan S|| = 0.253
tan 5^ = 0.144
tan 5,, = 0.259
Sj_ = 0.136
Because two samples had slightly different specific gravities
(0.38 and 0.46 for sample 1 and 2, respectively), it was necessary
to correct the data empirically using previous data.
We conclude from these results that the orientation of growthrings for constant orientation of the grain has little or no effect
on the polarization angle 0, or the dielectric tensor and loss tan­
gent.
Th-s, we can assume that the observed polarization properties
are almost entirely due to the macroscopic grain angle 8.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
103
Test 2
Two pieces of Douglas Fir were cut from the same board.
One
is end cut (no. 1) and the other is cross cut (no. 2), as shown in
Fig. 4-47.
In both samples, the growth-rings patterns are horizontal
and parallel to the direction of propagation, while the grains were
1 to the propagation direction in no. 1 and || to the direction of
propagation in no. 2.
Samples 1 and 2 have thickness 3.65 and
3.55 cm, specific gravities 0.41 and 0.36 and moisture content 31 and
75% respectively.
In our results, sample 2 has twice as much M.C. as sample 1,
but shows less than half of the depolarization index M (M = 0.4 and
1 respectively). We therefore conclude that the growth-rings has
much less effect on the polarization angle than the grain angle.
From the results of these two tests, we conclude that grain
angle is the dominant factor in causing depolarization.
Since the
growth-rings are generally visually discernible and the grain is not,
this method of measuring the grain angle should be quite useful.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
104
CO
r in g s
O
CVi
«
growth
o
4-47
\0
i
<o
Figure
u
3
0»
•w
&
o'
z
£L
2
<
CO
o
z
UJ
CL
s
<
co
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1
grain
UJ
-J
105
4.5
Temperature Effects
Basically, temperature is a constant in our measurements.
Others
[Yavorski 1952, Nanassy 1954, Tinga 1969 and Tiuri et al. 19791 have
studied some of the effects of temperature on the physical properties
of wood.
In our preliminary effort, only two cases (frozen and
thawed to room temperature) were investigated.
Several partially frozen or completely
thawed samples(thawing
time = 24 hours) were tested and their results are
summarized as
follows:
1.
Aspen (w = 2.8 cm, S.G. = 0.39 and
frozen
2.
thawed
(Ofw^ (dB)
13.95
18.3
(Ac{>)^ (degrees)
281
420
pj_ (dB)
-25.8
-27.8
Aspen (w = 4.7 cm, S.G. = 0.38 and
frozen
3.
M.C. = 130%)
M.C. = 133%)
thawed
(a-w)1 (dB)
20.7
31.2
(A<J>)^ (degrees)
606
872
pj_ (dB)
-24.0
-16.3
Hickory (w = 2.7 cm, S.G. = 0.72 and M.C. = 43%)
(a*w)^ (dB)
16.
15.5
(A$), (degrees)
521
536
Pj_ (dB)
-30.8
-31.5
M (%)
78
95
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
106
4.
Hickory (w = 4.5 cm, S.G. = 0.72 and M.C. = 45%)
frozen
thawed
(a»w)^ (dB)
27
25.4
(A4>)i (degrees)
895
920.7
Pj_ (dB)
-18.2
-20.4
From these results, we found that the loss, the phase shift,
and the depolarization ratio and index are significantly reduced if the
wood is partially frozen and larger for higher M.C.
Tinga [1969] found that the complex.dielectric constant is
relatively insensitive to temperatures around 20° C.
He was unable
to measure the off-diagonal terms in the dielectric tensor.
In conclusion, our qualitative and quantitative results generally
agree with the theory and the experimental results of others who used
different methods [Yavorski 1952, James and Hamill 1965 and Tinga
1969].
In addition, we have observed some new phenomenae about
microwaves propagating through anisotropic and inhomogeneous medium.
In particular, the ability to measure the polarization parameters and
to compute the complex dielectric tensor are significant advances.
The relationships between the electrical parameters and physical
properties of wood have been found, thereby establishing the measure­
ment system's usefulness.
Fortunately, some of these relationships
turned out to be linear or quite weak.
This greatly simplifies the
problem, and will be especially useful when dynamically testing in
real-time using the minicomputer.
The data processing is the subject
of the next chapter.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
107
5.
Data Processing
The relationships between the electrical parameters (a*w, Acf>,
y and 0) and the physical parameters (w, M.C., S.G. and 9) have been
established as described in the last chapter.
The raw data are
corrected for reflections as explained in Chapter 2, and then nor­
malized to a unit thickness.
This correction procedure requires
knowing [£] which in turn, requires knowing (a*w), A<+> and p for both
II and 1 polarizations.
However, these cannot be measured dynamically
using the current system, so an approximate method must be used.
Since the cross-coupling effect is small and the incident field is
primarily i for small grain angles, [e] is approximately calculated
by setting (a*w)|| = (a»w)^, A^j = A<j>^ and pjj = p^.
Pjj
For very dry wood,
is set to -35 dB and for very wet wood, p^ ^ is set to -25 dB.
A
A
Then, approximate values of R and T are computed.
In this procedure,
the complex dielectric tensor essentially reduces to the complex
dielectric constant for the particular polarization being used
(normally,
.
Due to the difficulty in preparing the samples with M.C.
ranging from 12% to green and different S.G., we were unable to get
enough data points to build a "standard" table.
Therefore inter­
polation, extrapolation or other curve fitting techniques must be
used.
These data are then stored in the disk.
Starting from the smallest values of M.C. and S.G., the cor­
responding values of a, 3 and M are stored as the first line in the
floppy disk.
Successive lines represent incremented S.G. through
its entire range with M.C. fixed.
The S.G. is then reset to the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
108
smallest value and the M.C. is then incremented once.
The next set
of successive lines (page) represent another run through the range
of S.G. and the corresponding parameters with the fixed value of
M.C.
This is repeated for all the data.
M.C. ranges from 0 to 150%
in increments of 3 to 10%, and the S.G. ranges from 0.38 to 0.46 in
increments of 0.02 for Douglas Fir.
When actually testing an unknown board, the measured values
of a and 3 will be matched to the nearest stored data using a least
mean-square
method and the corresponding M.C., S.G. and M are read.
Then M together with the measured 0 determines 6.
The data processing flow chart is shown in Fig. 5-1.
RAW DATA
r ,9
COMPUTE
CORRECTION FOR
REFLECTION AND
NORMALIZATION
MATCHING AND
READ
COMPUTE
M.C., S.G., M
P R IN T
M.C
S.G.
Figure 5-1
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
109
An efficient program (Appendix I) has been written to do the
real-time data acquisition and processing.
This program is not yet
in its final form, and further modifications and corrections may be
needed.
It contains five major subprograms and their functions are
explained below:
1.
MAIN.FOR/L
This program accepts the data and converts them into the raw
parameters (a*w), Acp, y and 0 for processing.
It also prints out
the physical properties, i.e., M.C., S.G. and 0, of the unknown board
in the final step.
2.
P3.MAC/L
This program vises the analog-to-digital devices to sample the
2
2
E
and E . of the polarization ellipse and to calculate the loss
max
min
(a*w) and depolarization ratio y.
3.
P53.MAC/L
This program uses the zero-crossing device and the DRll-C
general purpose interface to find the polarization angle 0, and
samples the KF phase Acf> of the major-axis.
4.
WOOD.FOR/L (ZPOLE.FOR/L and S0LV2.F0R/L)
This program corrects the measured raw data for the effect of
reflections using the method described in Chapter 2.
ZPOLE is used
to determine the quadrant of the RF phase and SCLV2 is a Gaussian
elimination method for computation of the dielectric tensor.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
5.
LKFRA.FOR/L, LKFRB.FOR/L and LKFRC.FOR/L
These programs match the measured data to the stored data.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Ill
6.
Discussion and Future Effort
6.1 Instrumentation System
Discussion
We have developed a microwave homodyne system which can:
1.
measure loss, phase shift and polarization properties in real­
time with good accuracy.
2.
scan two-dimensionally and distinguish the bulk properties over
a local region of the order of a few centimeter for dimension
lumber.
The instrumentation system also has the following advantages:
1.
Low cost —
all microwave components are off-the-shelf items.
All signal processing is done at audio frequencies, f^ and f^.
2.
Zero IF permits narrow bandwidths with sensitivities of the
order of -120 dBm in a 4 KHz bandwidth for Schottky mixers'.
This is comparable to the sensitivity of superheterodyne systems.
3.
A single dipole antenna is used to measure polarization.
This
is an advantage compared to systems which use dual receiving
antennas and a time-shared receiver, e.g., in near-field'
measurement systems used for predicting radiation patterns.
For our system there is no cross-talk and no probe normaliza­
tion (calibration) is needed.
Since the probe is small,
probe corrections are minimal or unncessary.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
112
4.1 The probe functions only as a scatterer, so no RF circuitry
is needed, e.g., cables, waveguides, mixers, receivers, etc.
These can disturb the field being measured.
5.
Coherent detection is used so phase lock loops are not needed,
and the source frequency, U), can be swept.
Future effort
1.
The present system gives consistent results for (a*w) , A<J> and
0 when the one-way loss is no greater than 20 dB (or
.
This can be increased by any one or all of the following
modifications:
(a) Add a
KF amplifier in the information channel prior to
the quadrature mixers.
This should increase the signal-
to-noise ratio which is critical for measuring E . .
non
(b) Redesign the etched zig-zag antenna using a high dielectric
constant substrate (e.g. Epsilon -10 by 3 M Co.) to improve
the scattering efficiency [Bahl et al. 1980].
(c) Increase the dipole's scattering cross-section.
We
anticipate that this can be done by reducing the' loading
effect of the resistive leads which are being used to
feed the 10 KHz modulating voltage to the p-i-n diode.
2.
The chief source of noise in the present system seems to be
mechanical vibration of the shaft which spins the dipole at
about 11,000 r.p.m.
This problem is being remedied by re­
designing a stiffer shaft and using precision high-speed ball
bearings.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
113
3.
Analysis of phase and amplitude errors of the phase-insensitive
detector are given in another paper by King [1981].
It is. suf­
ficient to remark that errors due to quadrature phase or ampli­
tude unbalance between the two mixers can be totally mitigated
by careful design and trim adjustment of the combiner.
These
trim adjustments should be added to the present system.
The only errors which can not be reduced to inconsequential
levels are those inherent to the detection process itself.
They arise when the carrier (w) in the information channel is
comparable to A.
For this reason, the balanced mixers should
be the 180° variety which suppress the unwanted carrier,
typically by a factor of 20 to 30 dB.
The first term in eq. (3)
Chapter 3, represents such an unwanted carrier, but it is
normally quite small and presents no problems.
Much more
troublesome are unwanted reflections from objects in the
foreground of the transmit-receive antenna, or from the antenna
itself.
As noted earlier, tuner T^ is used to minimize this
unwanted carrier.
4.
Separating the power supplies for the individual signal
processing channels and reducing the number of amplification
stages in the circuitry can eliminate interchannel interference
and reduce noise or offset voltages respectively.
A suggested
circuitry is shown in Appendix J.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
114
5.
A parallel analog-to-digital converter device can be used to
simultaneously sample all four signal processing channels to
decrease the response time.
This can also increase the
resolution, especially for E . .
min
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
115
6.2
Measurement Techniques
Discussion
The measurement system is monostatic since the backscattered
field being measured is received by the same antenna from which it
was transmitted.
Although the application described here involves
transmission through a medium, the system can be used equally well
to measure reflected, refracted or diffracted fields, assuming
reciprocity.
It is also possible to measure magnetic fields using spinning
loops which are electrically modulated.
Furthermore, the system can
be readily expanded to accommodate several modulated scatterers
(e.g., an array) simultaneously, simply by electrically modulating
each scatterer at a different w , i.e., u> ., uj _, ... . In this
m
ml
m2
case, each
at the combiner output is selectively filtered and
processed in parallel simultaneously.
Of course, adjacent i^'s
must be far enough apart so that the bandwidths necessary to
accommodate the respective
gj^' s
do not overlap.
Analog amplitude, HF phase, and polarization angle data are
available simultaneously and independently in real-time.
Ultimately,
the data rate is limited by the speed of computer.
Future effort
1.
Our mathematical interpretation of the microwave fields lends
itself to the study of two elliptical modes which exist when propa­
gating through anisotropic media.
The resultant "ellipse" of the
superposition of these two ellipse is very complicated and was found
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
116
to be somewhat asymmetrical especially for high M.C., e.g., neither
2
2
E
nor S . are even-symmetric for subsequent half revolutions,
max
mxn
^
'
2
e * g = t h e subsequent E . are not 180° apart for each half revolumin
tion.
Thus, measuring the resultant amplitude and phase of the field
in all directions of the dipole will be helpful to completely under­
stand the behavior of microwaves propagating through an anisotropic
medium.
The techniques for handling elliptically polarized waves
can be found in many references [e.g., Rumsey et al. 1951, Jensen
1976].
2.
The sense of rotation was found to depend on the orientation of
the grain angle.
Further study is needed to understand this obser­
vation.
3.
The fact that My is greater than
suggests that a grain
angle instrument would more effectively use !! polarization.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
117
6.3
Data Acquisition and Processing
Discussion
A greater rate of data acquisition
higher spinning
can be achieved by using a
frequency, a) , but this also requires increasingto .
s
m
In the present system, due to the settling time of the phase meter,
a substantial increase in
will permit RF phase measurements over
a greater portion of each half of the dipole’s rotation.
mum cos is limited by the speed of the computer.
The maxi­
For multiple
scatterers and hence multiple to 's, time-shared data acquisition
m
will be required.
Ultimately, dedicated microprocessors should be
used for data processing.
y is the most difficult parameter to measure, and it takes the
most computing time because large amounts of data must be averaged
to remove slight system instabilities and noise.
For a high
data rate the scatterer must be spun ata high
speed, e.g., 10
to 20 kPRM.
Mechanical vibrations thenbecome a
problem, particularly when measuring the depolarization ratio,
2
2
y = (Emin
. /E
) .in dB. However, this is not a serious limitation in
max
testing wood since M.C., S.G. and 0 can be determined without using y.
Future Effort
Additional measurements are needed to establish the "standard"
table and minimize the need for interpolation and extrapolation.
It would be especially helpful to acquire test samples having
M.C. from 12% to green and more samples having different S.G. values
for the same species.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
118
Additional tests should be made to show the effects of temper­
ature for various species and M.C., especially on the off-diagonal
terms of the dielectric tensor.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
119
7.
Applications
This research represents an initial effort toward methods of
microwave nondestructive testing of wood.
It is expected that the
measurement systems which are developed and the knowledge gained from
them will lead to other electromagnetic testing methods for other
materials.
Especially, it will be very useful in investigating
inhomogeneous and anisotropic media.
In particular, further research should be directed toward the
development of a microwave vector holographic system for constructing
three-dimensional images of the internal dielectric and conductive
properties of anisotropic media.
In Chapter 6, we have shown that our system should also be
useful for antenna near-field measurements, since the probe is very
small and can be made even smaller using microstrip techniques.
Then the probe will be closer to a "point".
Also it is straightforward to measure the upper and lower
sidebands simultaneously, and their phase difference will be
-4A4> (t). This phase delay can be used to track the position and
velocity of a moving target in Radar and Navigation applications
[Koelle and Depp 1977].
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
120
Appendix A
Extension of Jones Matrix Formalism to Microwaves
Passing Through Anisotropic Media
Assuming e^Uttime dependence, the wave equation for the electric
field in the anisotropic material is
V^E -
V (V -E )
= -
oj^U q £
*e
For a plane wave exhibiting spatial dependence in only the Y-direction,
this becomes
E (yf
rE~
X
Ez (y)
X
=
2
-co y [e]
3Y2
0
_
E
2
(1)
E
L_Y
_
this equation can also be written in terms of only the transverse
field ccmoonents as follows:
3 2 r„y — r E — —w ii_ [o ]c>
3Y
where E =
(2 )
is the two-component transverse electric field column
vector in the a , a
X
2
coordinate system.
1
aV
£'
r~Y->
[e 3 =
XX
Also
~Y
s'
xzi
-ZZL
is the reduced tensor and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
121
g
_ *
£
= £
~
m,n
m,n
g
my vn
*
S ty
m,n = x,z
'
The propagation or differential Jones matrix [N] is a 2x2 matrix,
defined by
E = + [N]i
(3)
where the -(+) corresponds to propagation in the +(-) Y-direction.
The solution can be written in the form [Vernon et al. 1980]
i(Y) = exp (+Y- [N])E (0)
(4)
where
exp (+Y* [N]) =
I i±iL- Yn [N]n
n=0
n'
Fig. A—1 defines the notation used in the following.
Then
(Y) = [M(Y)][TJE1 (0)
(5)
Et (W) = [T’3 [M(W)3 [Tl^tO)
where [M(Y)3 = exp(-Y*[N]) represents the effects of the anisotropic
material on a wave traveling in the positive Y-direction between Y=0
A
and Y=Y.
It is called the Jones matrix.
A
[T] and [T*] are transmis­
sion matrices at the front and back interfaces respectively.
A
Jones [1948] has derived the expression for M(Y) in terms of :he
dielectric elements in reference to (2) and (3).
Thus equating M(Y)
with eq. (7) in Chapter 2, we are able to find p., and p. in terms of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
122
[T]
[T]
1
W O O D
e
'(o)
E f(y)
1
E f(o)
E f(w)
|
1
1
1
1
1
j
I
1
Y =0
Y=W
Y=Y
Figure A-l
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
123
the complex dielectric tensor.
From this we can find n.! . and t>” •.
-I
I,l
^|
|,1
An iterative method can be used to find the dielectric tensor.
First,
we calculate the first order tensor assuming p' and p" are zero.
A
Then using [M], we find approximate values of p' and p".
then used to compute the 2nd order tensor, etc.
having to actually measure
These are
This procedure avoids
y(y) by stacking slabs of the material.
-t
Also, we can find E from (5),
r
/e
XX
sxn0
---------------- =-
2
Et (w) <= E1 (0)
(1V
~
ft
n
£ COS0
xz
^
-£
sin0 +
zx
/ I zz
/
(6)
COS0
d+ I T y
V zz
Equation (6) describes the important fact that the transmitted
field E*" is the superposition of the fields of two ellipses
and
£||, generated by fields E^ (= E1 cos0) and Ey ( = E1 sin0) respec­
tively, as shown in Fig. A-2.
(This figure is drawn for the case
where major and minor axes of the two ellipses are almost in phase.
Figures for other cases can be drawn in a similar way.)
In Fig. A-2,
is the resultant field pattern and the major and minor axes of
t
£± ^£||^ correspond to the second (first) term of Ej^
t
(Ey) and
t
t
Ej| (Ej^) in (6) respectively.
This resultant field is then backscattered by the dipole antenna
through the wood once more to the transmit-receive horn antenna.
The
-t 2
received signal V is proportional to (E ) following the theory of
modulated scatterers [Harrington, 1965], and can be expressed as:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
124
t
sin 0
XX
------- —
,4
U + f
V
i2
V = k 1 (E )
£ ~ ) ‘
XX
/”! X £XZ sin0 cos0
/s2
2-,
.v Xxx
xz
C O S 0 - 2-- — -— --- —
xz
..
,2
d+J T ~>
+ £
V
XX
(7)
_
£ cos 0
/£
£
sin0 cos0
£2.2„
ZZ
_ Y ZZ
ZX
£ sin © + -------- - 2------- — --- -----zx
,, . r?— .4
.. rx— ,2
where the constant k^ is determined by the echo area of the dipole
and the characteristics of the antenna.
The field pattern corresponding to (7) looks like Fig. A-3.
The heavy dark locus is the readings of the instrumentation, since
it reads the mean square of the backscattered signal.
fields Ej. and
Thus, the two
are mutually coupled in a complicated way.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
125
■
<
<o
u
3
Ci
■H
<33
fa
<S>
D
llli
X
X
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Appendix B
Field, Amplitude and Phase Patterns of Polarized Waves
LINEAR
POLARIZATION
ELLIPTICAL
POLARIZATION
CIRCULAR
POLARiZATION
(a.) The locus of the rms field
(b.)
Mean square amolitude IE - H I2VS .9'
lE-S!2
IE-uI2
8
B+ir/Z
S'
!E-ul2
‘" P 12\ C
e b + tt/z
in)2
e
6
s'
9*ir/2
(c.)RF Phase angle A<£ vs. S '
T
ir
v/2
!,9+ir
e -ir/2 I
*■S'
v/Z
,9+ir
/$ e*7r/2
1 /
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
S'
127
Appendix C
Radio Method for Measuring RF Phase
RF TRANSMITTER (o»RF)-^
BATTERY
SHAFT
MOTOR
CONCEALED LEADS
1
REMOTE
RF
RECEIVER
ETCHED DIPOLE
SCANNING
MOTION
PIN DIODE
TO
PHASE
METER
AMPLIFIER
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
128
Appendix D
Backscattering from Modulated Electric Dipoles
The theory and use of modulated scatterers for field measure­
ments has been well documented [Harrington 1962, 1964], so we only
summarize the salient features for electric dipoles to show the
principles as applied to the present system.
Magnetic dipoles for
measuring H qtb treated in a similar manner.
Let us take the transmit-recieve and scattering antennas and
the intervening medium between them as a 2-port network.
Then the
change in the input (port 1) voltage of the transmit-receive antenna
due to the presence of the scattering antenna is
(1)
vi - vi = 4vi -
11
11
222+ZL
where Z is the load impedance connected to the scattering antenna
Xj
terminals (port 2).
= V^/I^
v°
is the self-imnedance of the
transmit-receive antenna and superscript zero implies that the
is the self­
Similarly, Z ^ = ^22^~2
scatterer is removed.
V
impedance of the scatterer.
z
21
assuming reciprocity.
=
0
The mutual impedance is
— ^
V 2
Here,
V J2dT " Z12
(2 )
is the field of source 1 when driven
by Ij, in the absence of the scatterer and
is the current on the
scatterer when it is driven with current I2 - Then, if E^ is uniform
over the length of the dipole,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
129
-E. • u£
Z21 = —
1;- - "
(3)
where u is the unit vector along the dipole axis (Fig. 1), and the
effective length is
4e=r«7j y*id*
(4>
*2
Thus, (1) reduces to
A v i
-
(2u
'
z i i u
(5 ,
i
-nl
)2
-
<5 >
Important conclusions to be drawn from (5) are:
a.
The quantity (Z^-Z^) represents the change in the input im­
pedance of the transmit-receive antenna due to the presence
of the scattering antenna.
While it varies with position
and orientation of the scatterer, it is normally quite small.
But more important, it is not electrically modulated at oj
in
(although it is modulated at 2u ).
s
Thus, this term is com-
pletely eliminated by the narrowband amplifier which follows
the phase-insensitive coherent detector.
b.
The second term is modulated both at
(by electronically
varying the scatterer1s loading imoedance Z_), and at 2oo (by
L
s
varying u through spinning at
, and squaring).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 30
c.
—
-
2
The detected voltage is proportional to the phasor (E^*u(t;) =
|51*u(t) j2
e
where ji., »u(t) j and <J>(t) are the instan­
taneous amplitude and phase respectively of the components of
E^ in the direction of u(t) as the dipole spins at cos- Using
the phase-insensitive coherent detector shown in Fig. 3-2,
|E^«u(t)
and 2<J)(t) are measured in real time, simultaneously
and independently.
d.
Taking 2 as the impedance of a PIN diode which is electroniXj
cally switched between very large and small values at eo , the
m
second term in (5) is modulated double sideband with.carrier
(DSBWC). The spectrum of the backscattered signal is comprised
of the carrier with an evenly distributed cluster of harmonics
of 2w due to the first term in (5), i.e., to ± 2nco . In addition,
s
s
there are evenly, distributed clusters at to ± (pto ± 2nw ) due to
El
S
the second term in (5), where p and n are integers.
but the
Since all
± 2nws components are filtered out after homodyne
detection, it is sufficient to consider only this p=l term.
Thus, the only relevant component in the DSBWC backscattered
voltage appearing at the transmit-receive antenna terminals is
the phasor
AV = c |e , • u(t)
[1' + m cos (to t) ]e
1
1
m
^
(6)
where m is the amplitude modulation index and C is a constant.
Note that this voltage is amplitude modulated at ui„, and
amplitude and phase modulated at to .
s
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
for Real-Time Measurements
A O
Analog
Circuity
Appendix
E
131
c
o
o
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 32
Appendix.F
Digital Signal Processing Calibrated by Digital Storage Oscilloscope
lE C U A .n A C /L
•*LI
PROGRAM T O C O M M U N I C A T E W I T H T H E N I C - 2 0 9 0 S C O P E AND T H E
1 1 / 4 0 V I A T H E H P —I B .
U S IN G 8 - 3 . I T PARALLEL TO TRANSFER
D A T A AND A L S O A L O T O F H A N D 5 H A C K I N G B E TW EE N I C E - 1 1 A AND
THE SCOPE.
T H I S PROGRAM I S O N L Y A S U B R O U T I N E FOR A F O R T R A N PROGRAM
W H I C H I S U S E D TO C A L C U L A T E T H E R A T I O O F T H E M A X I M U M AND
T H E M I N I M U M O F A WAVEFO RM W H I C H I S R ECO RD I N T H E S C O P E '
ALREADY
T IT L E
IE C 1 1 A
GLOBL
IE C
GLOBL
S D 2 B IN
IE C 11A
C IR
SMR
IO R
V SR
CONTROLLER
164100
164102
164104
164106
—
; C O N T R O L AND I N T E R R U P T REG
r S T A T U S A N D M E S S A G E REG
; I N P U T A N D O U T P U T REG
“ V E C T O R S W I T C H REG
G P IB
TALKER
100
TALK 2
L IS T E N
L IS T 2
US
LIS T E R E R
AND T A L K E R
57
40
TKS
TKB
HARDWARE
177560
177562
-------------------------
CRT
TPS
=
177564
TPS
=
177566
MOV
HARDWARE
M A IN
#M ES5.
PC.
JSR
MOV
PC.
2 (R 5 j.
SET
R E G IS T E R S
-------------------------
UP
THE
R E G IS T E R
R E G IS T E R
“ C O N T R O L AND S T A T U S
JOUTPUT BUFFER
JSR
==========
-------------------------
J C O N T R O L AND S T A T U S
; IN P U T BUFFER
==========
:
ADDRESS
I I E C - 1 1 A TALK ADDRESS
; SCOPE TALK ADDRESS
; S C O P E COMMAND L I S T E N A D D R E S S
" I E C - 1 I A L I S T E N ADDRESS
------------------------- K E Y B O A R D
ie c
H ARDWARE R E G I S T E R S
p ro g ram
==========
R4
IG E T
THE
TRACH
P R IN T
ACCEPT
R4
JGET
THE
S T A R T IN G
G P I3
R E G IS T E R
W H IC H
ADR
WANT TO
OF
D A TA
RECALL
ARRAY
==========
T H E F O L L O W I N G S T A T E M E N T S A R E T H E S E T UP COMMANDS FOR T H E
I E C - 1 1 A AN D T H E N I C - 2 0 5 0 S C O P E .
F I R S T ADDRESS THE TA LK ER
THEN THE L I S T E N E R .
S I N C E U S E T H E H P - 1 3 . SO T H E R E I S A L O T
O F H A N D S H A K I N G B E T W E E N TWO D E V I C E S .
E V E R Y T IM E , TH E TALKER
N E E D T O W A I T FO R T H E D A TA A C C E P T E D S I G N A L I N O R D E R TO SEN D
THE NEXT D A TA.
N O R M A L L Y . E A CH D E V I C E H A S ONE T A L K A D D R E S S
AND ON E L I S T E N A D D R E S S .
S U T T H E S C O P E H A S TWO L I S T E N
ADD=E = S
fC N E
"C=
COMM AN D-
AND
ONE
FuR
DAT A : .
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
133
B IS
#40*
B # C IR
5CLEAR
C LR
B IS
B I5
R1
#1*
#100*
B # C IR
3#SMR
JS Y S T E M CONTROL A C T IV E D
J C L EA R IN T E R F A C E S
B IT
BNE
B IC
# 2 0 0 0 0 , S#SMR
.-S
# 1 7 7 4 0 0 ,0# C IR
J W A IT
MOV
B IT
BEG
B IC
MOV
B IT
BEG
B IC B IS
B IT
BEG
B IT
♦ T A L K E R ,8 # I0 R
# 1 0 0 0 0 0 , @ #C IR
.-B
# 1 0 0 0 0 0 *B # C IR
♦L IS T E N ,@ # IO R
# 1 0 0 0 0 0 ,0 # C IR
.-S
# 1 7 7 4 0 0 ,8# C IR
#4,
B #S M R
#400,
8# C IR
.-S
#4000,
B # C IR
JA DDRESS TH E T A L K E R <I E C - 1 1 A >
f W A I T FO R D A T A A C C E P T
BEG
GOOD
NO L I S T E N E R
error
:
MOV
JSR
JM P
TRY
good
:
B IC
MOV
A C T IV A T E *
#ERR,
PC,
LAST
TO
ALL
IN T E R R U P T
JCLEAR
ALL
jgo
standby
to
J W A IT
JANY
FOR
IN T E R R U P T
ANY
L IS T E N E R
STATUS
J P R IN T
THE
THE
ERROR
C O N D IT IO N S
TO
CHANGE
MESSAGE
AND R ETURN
LIS T E N E R
JCLEAR
fS E N D
OUT
RECALL
# 1 0 0 0 0 0 , @ # C IR
.-s
# 1 0 0 0 0 0 ,3 # C IR
TRACK,
3 # IO R
J W A IT
FO R
DATA
#1,
B # C IR
#100,
0*SM R
# 2 0 0 0 0 , s#SMR
,-S
# 1 7 7 4 0 0 ,0# C IR
#T A L K 2 , 0 # IO R
# 1 0 0 0 0 0 ,0# C IR
.-5
# 1 0 0 0 0 0 ,0# C IR
# L I 5 T 2 , 0 # IO R
# 1 0 0 0 0 0 ,0 # C IR
• —S
#4,
0*5M R
#400,
8 # C IR
• -B
#D3UF,
RO
R2
C O N D IT IO N S
E RRO R
S # IO R
.-S
# '0 ,
B # IO R
# 1 0 0 0 0 0 ,3# C IR
.-S
#40,
0# C IR
IE C
A C T IV E D ?
# 1 7 7 4 0 0 , @ #C IR
BEG
MOV
B IT
Bsa
B IS
TO
c o n d it io n
# 'R ,
# 100000 , b # c i r
. —5
# 4 0 0 0 0 , 3 # C IR
.-5
# io o o o o , b # c : r
# 'D ,
S # IO R
# 1 0 0 0 0 0 ,0 # C IR
B IT
J C L E A R TH E DATA A C C EP T B I T
JADDRESS TH E L IS T E N E R A D D R E S S (S C O P E )
J W A I T FO R D A T A A C C E P T
W IT H
B IT
BEG
B IC
MOV
B IT
BEG
B IT
B EG
B IC
MOV
B IT
B IS
B IS
B IT
BNE
B IC
MOV
B IT
BEG
B IC
MOV
B IT
BEG
B IS
B IT
BEG
MOV
CLR
STATUS
FO R R E P L Y
: CLEAR
R4
P R IN T
COM M U N IC A TE
ALL
ALL
IN T E R R U P T
C O N D IT IO N S
COMMAND TO
JC L E A R DATA AC C EPT B I T
J R E C A L L TR A C K S O F D I S K
J W A I T FO R D A T A A C C E P T
JW A IT
FO R T H E
COMMAND
J P O IN T E R A U T O M A T IC A L L Y
J W A I T FOR D A T A A C C E P T
ALL
STATUS
MEMORY
DONE,
J R E S E T DATA A C C EP T B I T
J S E N T OUT D A T A T R A N S F E R
J W A I T FOR D A T A A C C E P T
JRESET
SCOPE
ACCEPT
3Y
COMMAND
ADVANCED
TO C H A N G E
TALKER
JA N D L I S T E N E R
J S Y S T E M C O N TR G L A C T I V A T E D
JUNADDRESS THE L I S T E N E R
J W A I T FOR R E P L Y
JCLEAR A LL IN T E R R U P T C O N D IT IO N S
JCHANGE TH E TALKER
J W A I T FOR D A T A A C C E P T
JCLEAR
DATA
ACCEPT
ERG
B IT
JCHANGE TH E L I S T E N E R
J W A I T FOR D A T A A C C E P T
JGO T O S T A N D B Y S T A T E
J W A I T FOR S T A T E CHANGE
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
134
b a c k
l o o p
b a c
n
:
:
:
:
B IT
# 1 0 0 0 0 0 , S # C IR
;data
BEB
BACK
; no
B IC
MOV
CLR
CLR
B IT
BEQ •
B IC
MOV
SWAB
MOVB
IN C
CMP
BNE
MOV
CMPB
BNE
IN C
# 1 0 0 0 0 0 , S # C IR
5 # I0 R ,
R3
R2
IN D IC
# 1 0 0 0 0 0 , @ #C IR
.-S
IC L E A R TH E DATA ACCEPT B I T
; move data to r e g . o
; C O U N T # O F B Y T E S FOR O N E D A TU M
J IN D IC A T O R FOR T H E S IG N OF DATA
;data ready?
; no
# 1 0 0 0 0 0 , @ #C IR
@ # IO R ,
R3
R3
R3,
CROJ+
R2
R2,
#7
BAC
#DBUF,
RO
(R O ),
# 'P O S IT
IN D IC
; CLEAR TH E DATA ACCEPT B I T
; g et th e d a ta from th e sco pe
; O N L Y T H E H I G H E R B Y T E I S D A TA
;store th e data
; IN C R E A S E T H E COUNTER
; f iv e b y t e s for d a t a , l in e f e e d
CONVERT
THE
MOVB
MOV
#S0.
#DBUF,
PC,
IN D IC
POS
CSP)
CSP)-*-,
CSP)-*# 4 0 3 0 .,
OUT
R1
LO OP
#100,
PC
•
j
m
T
p o s it
:
:
JSR
TST
BE Q
NEG
MOV
TST
CMP
BEG
IN C
BR
B IS
RTS
II
II
II
II
II
= = = ==
po s
o u t
:
:
l a s t
k»
K
P R IN T
D E C IM A L
;
;
;
;
<R4) +
R1
TO
B IN A R Y
NUMBER
THE
COUNTER
JRELEASE TH E L IS T E N E R
; RETU R N TO TH E C A L L I N G
PROGRAM
==========
•
p r in t
=====
S U B R O U T IN E T O P R IN T OUT T H E
T E R M I N A T E A T D A TA I S 0
MESSAGE
P O IN T E D
TSTB
BPL
MOVB
CMPB
BNE
RTS
THE
READY?
II
u
II
II
II
;
:
@#TPS
.-4
J IS
C R 4 ) -*-,
C R 4 ),
P R IN T
PC
ACCEPT
S U B R O U T IN E
9#TPB
#0
LO CA7 IO N
a c c e p t
b a d
:
:
TSTB
BPL
MOVB
B IC
CMPB
BLT
CRT
BY
R4.
’ NO
J S E N D T H E C HAR T O C R T
t IS
T H A T T H E END O F T H E
; no
; R E T U R N TO M A I N PROGRAM
W IL L
MESSAGE?
==========
TO
R E C E IV E
T H E N S T ORE T--iE N U MBER
I S NO T A N U M B E R , T H E N
ON E
IN T O
JUST
N U M E R IC A L
CHAR
FRUri
T E R M IN A L .
L O C A T IO N T R A C K .
I F THE
I G N O R E I T AND NO C H A N G E
CHAR
IN
TRACK.
CMPB
BGT
MOVB
8#TKS
.-4
@#TKB,
#200,
R4,
BAD
R4,
BAD
=4-
MOV
*CRLF,
R4
R4
#50
and
J C H A N G E T H E S I G N OR B LA N K TO Z E R O
; S E T UP P A R A M E T E R S OF S D 2 B I N
' C O N V E R T D E C I M A L # TO B I N A R Y #
; F I N D OUT T H E S I G N OF T H E D A T A
J l FO R N E G A T I V E , 0 FOR P O S I T I V E
JCOMPLEMENT TH E V A LU E
: STORE THE RESULT
:P O P THE STACK
JTA KE 1 0 0 0 DATA
: IN C R E A S E
9#SMR
?
R ESET T H E P O IN T E R
C H EC K F O R M I N U S S I G N
P O S I T I V E NUMB ER
i FO R N E G A T I V E N U MBER
NUMBER
<R0)
-C S P )
S D 2 B IN
ready
J I S THE DA TA READY?
; no
JG E T THE DATA
JC L E A R T H E CHECK B I T
? I S T H A T A N A S C I I NUMBER
; no
#71
TR ACK
: S T O R E T=-i e
R4-
JSENT THE
number
CR A N D
L IN E
FEED
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
135
JSR
RTS
PC,
PC
P R IN T
MEMORY
:
tr a c k
:
.B L K B
.E V E N
•W OR D
in d ic
:
.W O R D
d b u f
7
0
0
m e s s
:
.A S C II/E N T E R
.B Y T E
0
• EVEN
c r l f
:
.B Y T E
.E V E N
e r r
:
THE
#
OF
TRACK
OF M E M E O R Y T O
RECALL:
/
1 2 ,1 5 ,0
.A S C II/
•B Y T E
.E V E N
.E N D
E R RO R
NO L I S T E N E R
IS
A C T IV A T E D /< 1 2 > < 1 5 >
0
IE C
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
136
S C O P E .F O R /L
+LI
CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC
C
c
C
T H I S I S A F O R T R A N PROGRAM U S E D T O C A L C U L A T E T H E R A T I O
C
C
O F T H E M A X I M U M AND M I N I M U M O F A WAVEFORM W H I C H H A S
C
C
B E E N S T O R E D I N T H E D I S K ME MORY O F T H E S C O P E .
C
C
C
CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC
c
c
c
C
C
RESERVE
MEMORY F O R
D IM E N S IO N
C
C
C
C
C
11
C
C
C
1 .
C
C
C
3
C
C
C
4
S U B R O U T IN E
TO G E T T H E D A T A FR OM
T H E W H OLE ARRAY FROM T H E S U S R O U -
T IN E .
CALL
IE C U D A T A )
IN IT IA L IZ E
THE
MAX.
AND
M IN .
VALUE BUFFER
= +1000
= -1 0 0 0
C O M P A R E D A T A W I T H M A X . AND M I N . r
I F DATA GREATER
THAN
M A X , T H E N C H A N GE T H E M A X .
I F T H E DATA LE S S
THAN T H E
M I N . , THEN CHANGE T H E M I N .
DO
IF
2
ID A TA C 4 0 9 B )
C A L L M A C H IN E LANGUAGE
THE SCOPE.
P A S S BA CK
IM IN
IM A X
C
C
C
C
C
DATA
1 I = 1 ,4 0 9 0
( ID A T A < I ) .L T .IM A X )
IM A X = I D A T A ( I )
IF
(ID A T A C I).G T .IM IN )
IM IN = ID A T A (I)
C O N T IN U E
CALCULATE
THE
GO TO
GO TO
2
1
R A T IO
TYPE S ,IM A X
TYPE 7 , IM IN
IF
C I M I N . N E . O ) GO TO 3
TYPE 3
GO T D 4
R A T IO = F L O A T <IM A X ) /F L O A T <I M I N )
P R IN T
OUT
THE
RESULT
T Y P E 5 , R A T IO
TYPE 3
A CCEPT 1 0 , IA N S
IF
<I A N S .E G .1 )
GO TO
C
C
FORMAT
C
3
5
FORMAT( '
FORMAT( '
THE
THE
7
S
9
FORMAT( '
FORMAT( '
FORMAT( '
THE
M IN IM U M
THE
M IN IM U M
DO YO U WANT
10
FORMAT ( I D
11
STATEMENT
R A T IO IS
M AXIM U M
I'.F IO .S )
IS I ',1 1 0 )
IS : ' , I 1 0 )
I S Z E R O , 5 0 THE
R A T IO
IS IN F IN IT E ')
TO S T A R T OVERA G A IN ? 1 - Y E S ,
O -N O ' )
C
0
STOP
END
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
----
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Appendix G
Circuitry for Plotting Two-Dimensional Scans
DC MOTOR (HORIZ. SCAN)
AND
18 k
PO TENTIOM ETER
4 .3 V
O
MOV
o
TO X - IN P U T
STEPPING MOTOR (VERT SCAN)
AND
1
POTENTIOMETER
3 0 il
AAAA
4 .3 V
110 V o
O’
BIAS V ,
R f(IO O k)
v y ^ ------'out
S IG N A L V;
741
TO Y - IN P U T
X°
137
138
Appendix H
Instrumentation Operating Procedure
The operating procedure is described in a numbered sequential
fashion.
The components and apparatus referred to are shown in
Fig. 3-2.
1.
The klystron power supply and the diode modulating audio
oscillator are turned on and allowed to warm up for 50 minutes.
Remember to allow the klystron filament to heat up for 1 minute
before applying the high voltages.
Make sure the cooling blower
is on and is directed towards the klystron.
should be less than 80 ma.
The maximum beam current
Maximum klystron power output is in the
3/4 mode, obtained by properly adjusting the beam and reflector
voltages, typically, +750 V and -410 V respectively.
The klystron
will then be in the optimum operating condition with good frequency
and temperature stability.
2.
After warm up, it is necessary to check the operating
frequency.
The tunable slotted line RF probe is replaced by an
ordinary slotted line probe.
Then using a power meter or a diode
detector which is connected to an oscilloscope, the frequency meter
can be adjusted to observe a dip in cw power as monitored by the
power meter or oscilloscope.
The frequency can be adjusted by
changing the reflector voltage.
3.
Tuner T^ can be adjusted to obtain a minimum standing
wave in the slotted line.
Then the reference RF phase is linearly
proportional to the position of the slotted line probe.
Instead
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
139
of using cw, switch the klystron modulation mode selector to 1 kHz
square wave.
A SKR meter temporarily connected to the slotted line
probe with a diode detector is used to measure the VSWR.
A VSWR
less than 1.05 can be obtained.
4.
With the tunable slotted line RF probe and the audio oscil­
lator disconnected, and a power meter temporarily connected at Port 4
of the magic tee in Fig. 3-2, tuner
output from port 4 almost vanishes.
is adjusted until the power
This step matches the system
with the antenna and open space by nulling unwanted reflections Au.
Au was' found to be less than -15 dBm through all our measurements.
This will insure that the amplitude and phase unbalance are less than
0.5 dB and ±1.5° respectively.
5.
The tunable slotted line probe is then connected to the
reference channel, and the modulating oscillator is turned on.
The
depth of penetration of the slotted line probe controls the reference
signal level A.
To set the optimum depth, a lock-in voltmeter can
be used to obtain the maximum signal-to-noise ratio. This operating
point will lie in the linear region of the mixers.
This can be
checked by introducing a large information signal level b and varying
the phase shifter.
The amplitude of adjacent maximum should all be
identical if the detector is truly linear.
In this optimization
procedure, it is important that the d.c. load impedance of the
detector be very small, of the order of a few ohms.
If this is not
the case, the rectified d.c. current due to the large reference
signal will cause a reversed self bias which tends to shift the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
140
operating point back into the square law region.
At this operating
point, we will have maximum sensitivity and dynamic range.
6.
The modulating voltage at the diode scatterer should be
adjusted to produce a maximum reading just above a "knee" of the
response as observed on the lock-in voltmeter.
This will give the
largest amplitude modulation index.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Appendix I
'Programs for Real-Time Data Processing
M A IM -F O R /L
■*LI
COMMON X ( 1 0 0 > , Y ( 1 0 0 ) , 2 ( 1 0 0 ) , X D ( 1 S , 3 > » Y D ( 1 6 » 3 ) , D M ( 1 S » 3 )
COMMON A N ( I S , 3 ) , P H < 1 6 , 3 ) , A M A X < 2 > , R F ( 2 ) , R A T 0 ( 2 ) , P H A S < 2 )
101
102
212
39
11
12
COMMON L X »L Y » L M , L A N , L P H , J W , W I D T H , L , A T T N , S A T N , T P H , S P H
COMMON S X , S Y , I D 5 K X , I D S K Y , T H T A , DOM ( 1 0 0 )
ACCEPT 1 0 1 ,W ID T H , DSKX f D S K Y , DT A X , DTAY , D TA Z
F 0 R M A T C 7 F 6 .2 )
A C C E P T 1 0 2 , JM
FORM AT ( I D
JJ = 1
S IG N = 1 .
D IM E N S IO N N A M E (5)
T Y P E 11
FORMAT( '
IN P U T THE F I L E
A C C E P T 1 2 , NAME
F O R M A T (IO A )
CALL A S S IG N (2 1 ,N A M E )
DO 2 K = 1 , 1 0
IS U M 1 = 0 .
IS U M 2 = 0 .
NA ME
FO R
D A TA
:
',* >
IS U M 3= 0.
IS U M 4 = 0 .
IS U M 5 = 0 .
DO 1 1 = 1 , 1 0
CALL M A X M IN < II,1 2 , 1 3 , 1 4 , 1 5 )
IS U M 1 = IS U M 1 + I1
IS U M 2 = IS U M 2 -t-I2
IS U M 3 = IS U M 3 + I3
IS U M 4 = IS U M 4 + I4
1
S
5
2
21
91
103
20
32
33
IS U M S = IS U M 5 + I5
C O N T IN U E
AMAX <J J ) = F L O A T < I S U M 1 ) / 2 0 4 8 .
R F < J J >= < F L O A T ( 1 5 U M 5 ) / 1 4 S 4 . 3 - 2 . 5 S 5 ) * 5 6 . S S
R A T O < J J ) = 1 0 . * A L O G 1 0 <F L O A T ( I S U M 1 ) / F L O A T ( I S U M 2 ) * 1 0 1 . 7 5 >
PHAS ( J J ) = F L O A T ( I 5 U M 3 ) / F L O A T < I S U M 4 } » 1 6 0 .
TYPE 5 , A M A X tJ J ),R F < J J ),R A T O C J J ),P H A S < J J >
W R I T E ( 2 1 , S ) A M A X ( J J ) » R F ( J J ) , R A TO < J j ) , PHAS< J J )
F O R M A T (4 F 1 0 .3 )
FORMAT( 4 F 2 0 . 3 )
C O N T IN U E
C L 0 S E (U N IT = 2 1 )
GO TO S 3
T Y F E 91
ACCEPT 1 0 3 , 1 1 1
FORMAT( 'W A IT IN G ,
rO R M A T iII)
J J=2
H IT
ANY
D IG IT
KEY
I F ( I I I . N E . O ) GO TO 2 1 2
A T T N = 1 0 .*A L O G 1 0 (A M A X < 2 )/A M A X < 1 ) )
T P H = (R F (2 )-R F < 1 ))/2 .
DB=RAT0<2>
T H T A =PH A S • 2 ) -P H A S < 1 )
TYPE 52
A CCEPT 1 0 3 , 1 JJ
F O R M A T ( 'W A I T I N G , H I T ANY D I G I T KEY
I F C I J J . N E . O ) GO TO 2 2
TPH=TPH+180.
T Y P E S3
A C C EPT 1 0 3 , IK K
F O R M A T ('W A IT IN G , H I T ANY
I F ( I K K . N E . O ) GO TO 2 2
T P H = T ? H + I3 0 .
TYPE
D IG IT
KEY
(E X C E P T
0)
TO
C O N TIN U E
(E X C E P T
0)
TO
C O N T IN U E
(E X C E P T
0)
TO
C O N T IN U E
34
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
142
34
ACCEPT' 1 0 3 , IJ K
F O R M A T ('W A IT IN G ,
IF (IJ K .N E .O )
H IT
ANY
GO T O
22
DB,
TPH,
D IG IT
22
TPH=TPH+180.
C A L L W O OD (D B ,
15
1
S IG N , T P H , A TTN , EPP)
TY P E 1 5 , W ID T H ,A T T N ,T P H ,D B ,T H T A
F O R M A T (5 F 1 0 .4 ,/)
100
40
30
SO
50
SC
70
200
300
TPH,
KEY
ATTN,
(E X C E P T
ATTN,
0)
TO
W ID T H ,
L=1
X (L )= 3 3 .3 *A L 0 G (E P P /1 .4 3 )
Y (L )= 0 .4 2
L=L+1
C A L L LKFRA
C A L L LKFRB
C A L L LKFRC
D X L = X (L )-X (L -1 )
IF (D X L -D T A X ) 3 0 , 3 0 , 4 0
NN= 1
D Y L = Y (L )-Y (L -1 )
IF (D Y L -D T A Y ) 5 0 , 5 0 , B 0
NN=1
D Z L = 2 (L )-Z (L -1 )
IF (D Z L -D T A Z ) 7 0 , 7 0 , 3 0
NN=1
I F ( N N . N E . l ) GO TO 2 0 0
L=L+1
GO T O 1 0 0
TYPE 3 0 0 , W I D T H , X C L ) , Y ( L ) , Z ( L )
FORMAT( 4 r 7 . 3 , / )
GO T O 2 1
STOP
END
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
C O N T IN U E ',* )
S IG N ,
143
F IL E :
+LI
. GLQBL
P 3 .M A C /L
M A X M IN
CRTC5
=
177564
CRTBUF
=
CRTCS
=
=
170400
AD CS *■ 2
;a/ d
;a/ d
C HO
*
20040
CHI
=
CHO +
; a / d CHANNEL 0 W IT H
; OVERFLOW CONTROL
; a / d channel i
+
! SCREEN
JSCREEN
2
CONTROL STA TU S
D A TA B U F F E R
R E G IS T E R
r
AVCS
ADDS
control
status
convertor
data
r e g is t e r
buffer
9
•
9
400
0
TO + 5 V
range
9
9
CCSR
=
170404
CCSB
CVEC
=
=
CCSR
344
2
*■
JC L OC K
CONTROL
; COUNT
; C LO C K
SET 3UFFER
IN T E R R U P T VECTOR
STATUS
R E G IS T E R
9
«
CSR
QUTLNK
IN L IN K
1S4000
= CSR + 2
= CSR + 4
=
•
r
M A IN
9
PROGRAM
•
9
m a x m in
:
MOV
MOV
MOV
C LR
MOV
CLR
MOV
B IS
#3>
#404,
#STQR,
R1
#1777,
R3
8#CCSB
§#CCSR
RO
#CHO,
#1,
8#ADCS
R2
; U S E RO AS T H E S TO R A G E P O I N T E R
J R l I S T H E MAX
; R 2 t o h o l d t h e CURRENT v a l u e
;R 3 a s t h e c o u n t e r for o n e p e r io d
; ENABLE
O VERF LO W
TO S T A R T
A /D
8#CCSR
9
•
9
BACKMX:
;aga:
;next:
backmn
:
TSTB
BPL
MOV
8#AD C S
.-4
MOV
IN C
CMP
BN E
S # A D D B .- 8 2 £ R 3 ;
R3
R3,
#500.
BAC KMX
; I N C R E A S E T H E C OU N T NUMBER
J L E S S THAN 5 0 0 P O IN T S PER C Y C L E
JN O T Y E T , GO BACK TO G E T MORE
MOV
CMP
BLT
MOV
TST
CMP
BLT
MOV
#S T 0 R + 2 ,R 0
(R O ) ,
Rl
NEXT
; R E S E T T H E P O I N T E R TO D A T A
JC OM PA R E TO M A X IM U M
I S M A L L TH A N M A X , NO CHANGE
< R 0 )»
(R 0 ) +
RO,
AGA
Rl
; U P D A T E T H E M A X IM U M
J A D V A N C E P O I N T E R BY
; i S T H A T T H E END OF
R l,
82CR5)
; RETURN
RO
; r e s e t p o in t e r
; C L R T H E C OU N T
; S T O P T H E CLOCK T O C H A N G E
; C H A N G E TO C H A N N E L 1
J S T A R T T H E C LO C X A G A I N
J I S THE A /D READY Y ET ?
S#ADDB,
MOV
#STO R,
C LR
B IC
R3
#1,
MOV
IN C
TSTB
#CH1,
3 PL
e#ccsR
8#ADCS
; a / d done
“ NO W A I T
(R 0 ) +
#A D R
JSTORE
or
IN T O
NOT?
DATA
3UFFER
2
THE
BUFFER
BUFFER?
; no
8#CCSR
8#ADCS
THE
M A X IM U M
TO
FORTRAN
THE
PRO
A /D
;no
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
144
lo o p
:
MOO
MOO
S#ADDB*
0#ADDB,
IN C
R3
CMP
BLT
MOO
R3,
BOE
MOO
TST
CMP
BNE
MOO
B IC
MOO
#1,
#DATA,
CLR
CLR
TST
BPL
MOOB
IN C
CMP
R2
@#CSR
S#CSR
.-4
@ #IN L IN K r ( R 3 ) +
R2
R2r
#S
LO OP
#DAT A r
R3
(R 3 )+ ,
0 6 (R 5 )
C R 3 )f,
0 1 O (R 5 )
<R3) ,
0 1 2 (R5)
PC
MOV
MOV
MOV
RTS
MEMORY
JSTORE
THE
DATA
JIN C R E A M E N T
#500.
BACKMN
# S T 0 R + 2 iR O
CMP
(R O ) r
S K IP
(R O ) ,
R2
CRO) +
#ADR
RO»
M IN A G A
R2r
0 4 (R 5 )
BNE
MOV
TO
BUFFER
J I S THAT
JNOT Y E T
JRESET
R2
THE
COUNTER
MORE T H A N
THE
ONE
CYCLE?
BUFFER P O IN T E R
TH A T > P R E V IO U S
;is
.B L K W
.WORD
500
0
DATA:
• BLKB
s
.E N D
M A X M IN
P O IN T
;no
; UPDATE
THE
M IN IM U M
J I S T H A T T H E END O F T H E B U F F E R
; N 0 , GO BACK TO T E S T T H E N E X T D A T A
; R E T U R N T H E M I N I M U M TO F O R T R A N PROGRAM
0#CCSR
R3
P O R T IO N
:
s to r
a d r :
(R O )
0 4 (R 5 )
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
145
P53.MAC/L
*L I
T H IS
PROGRAM U S E S
THE
ZERO
C R O S S IN G
D E T E C T IO N
C IR C U IT ,
AND
T H E D R 1 1 - C GENERAL PURP O S E I N T E R F A C E TO F I N D T H E PHASE D I F F
O F T H E S IG N A L AND T H E R E F E R E N C E .
THEN USE TH E D R 1 1 - A
P A R A L L E L L I N K TO S E N D T H E R E S U L T FR OM 1 1 / 2 0 TO 1 1 / 4 0 .
In
T H I S C A S E , D U R IN G TH E T I M E TO C A LC U LA TE TH E PHASE D I F F TH E
1 1 / 4 0 CAN CO M P U TA TE O THE R D A T A .
T IT L E
. GLQBL
P H A S E -M E A S U R IN G
S B IN 2 D
JCALL T H E E XT E R N A L S U B R O U T IN E
TO
5TH E B I N A R Y N O . TO D E C I M A L N O .
V
I
CRTSR
CRTBUF
SCREEN
=
=
HARDWARE
1775S4
CRTSR + 2
KEYBOARD
KEYSR
KEY3UF
KEYVEC
KEYPRI
=
=
=
=
1STATUS R E G IS T E R
; T R A N S M IT T E D BUFFER
HAR D W A R E R E G I S T E R S
177SG0
KEYSR + 2
SO
KEYVEC
2
D R 11-C
DRVEC1
DRVECZ
DRSTAT
DROUT
D R IN
R E G I S T E R S -- -------------------------
=2 0 0
=2 0 4
=1 S 4 0 1 0
=DRSTAT
=DRSTAT
JS TA TU S R E G IS T E R
JKEYBCARD R E C I E V E
BUFFER
; IN T E R R U P T VECTOR
L O C A T IO N
J P R I O R I T Y S E T L O C A T IO N
HAR D W A R E R E G I S T E R S
; IN T E R R U P T VECTOR
J IN T E R R U P T VECTOR
; STATUS R E G IS T E R
; OUTPUT BUFFER
; IN P U T BUFFER
+ 2
+ 4
HARDWA RE C LO C K
CCSR
CCS3
COUNT
XW11P
= 171540
= CCSR - 2
= C C SR ->■ 4
H ARDWARE R E G I S T E R S
JSTATUS R E G IS T E R
; COUNT S E T BUFFER
I COUNT B UFFER
D R 11-A
PARALLEL
L IN K
H ARDWARE
R E G IS T E R S
CS R
GUTLNK
=* 1 6 4 0 0 0
= CSR + 2
; OUTPUT
IN L IN K
= CSR
J IN P U T
+
'C O N T R O L
4
A /D
ADCS
ADBUF
=
=
17S770
ADCS ♦
M A IN
-h-
1
2
C HANNEL HARDWARE R E G IS T E R S
-------------------------
STATUS
R E G IS T E R
BUFFER
SUFFER
-------------------------
2
PROGRAM
RO
Rl
STARTS
HERE
; CLEAR A L L T H E
:FR O M R 0 - R 5
R E G IS T E R S
TO
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
START
CONVERT
146
CLR
CLR
R2
R3
CLR
R4
CLR
R5
START
THE
CLR
@ #A D C S
CLR
B#ADBUF
USE
THE
KEYBOARD
IC H A N N E L
I N T E R R U P T TO
TE R M IN A T E
0,
THE
G A IN
FACTOR
IS
# K E Y IN T ,@ # K E Y V E C
TSTORE
#340,
#100,
T P R IO R IT Y SET TO H IG H E S T
TENABLE T H E I N T E R R U P T
UP
ALL
B#KEYPRI
@#KEY3R
THE
CLR
MOV
@ #D R OU T
CLR
MOV
CLR
THE
THE
IN TE R R U P T
F L IP -F L O P S
TO
S # D R V E C l+ 2
# IN T 2 A , 3#DRVEC2
S#DRVEC2+2
TSET
P R IO R IT Y
THE
CLOCK
TO
RUN A T
THE
TO
RUN
FO R I N T R 1
(L O W EST)
FOR IN T R 2
ZERO
1 MHZ
S#CC5S
0#CCSR
TREPEAT IN T E R R U P T ,
AN D
START
IN T E R R U P T
#140,
#1,
FO R
0#DRVEC1
START
VECTOR
IN T E R R U P T VECTOR
P R I O R I T Y TO Z E R O
IN T E R R U P T VECTOR
# 0 .,
#37,
MOV
B IS
AND
IN T E R R U P T
TSET T H E
TSET TH E
TSET TH E
THE
ENABLE
C O N D IT IO N S
TAL LOW
# IN T 1 A ,
MOM
MOV
W A IT
IN T E R R U P T
THE
1<10V)
PROGRAM
MOV
MOV
START
r
T
'
MOV
SET
T
1
3#DR5TAT
a#DR G 'JT
IN T E R R U P T
; E N A B L E BOTH I N T R 1 AND
IN T R 2
T R E S E T T H E D F F TO R E CO RD T H E
TO T A K E
S IG N A L
PLACE
7
W ATE:
W A IT
IN C
BR
==========
J W A IT
KEYBOARD
REBOOT TH E
k e y in t:
jm p
======
WHEN
IN T Ia :
IN T E P .R U T
SYSTEM
IN T R 1 A
TO
OCCUR
T H IN G S
th e
s y s te m
==========
IN T E R R U P T
COME S
IN ,
# IN T 1 S ,
#3,
IN T R Z A
ANY OTHER
tre s o c t
MOV
WHEN
D O IN G
e #i73ooo
@ # C C U N T ,F IR S T
#1,
3#DR0UT
=====
IN T E R R U P T
= === = = = = = =
W IT H O U T
MOV
B IC
B IS
RTI
FOR
R5
WATE
REME MB ER T H E
COUNT
TREME MBE R T H E C O U N T A T T H I S MOMENT
TCLEAR T H E I N T E R R U P T BY CLEAR THE
J C O R R E SP O N D IN G D F L I P - F L O P
a#DRVECl
@ #D R OU T
TENABLE
i return
THE
to
IN T E R R U PT
c a l l in g
2
program
==========
IN T E R R U P T
COME S
IN ,
READ
LA ST COUNT, THE D IF F E R E N C E
P H A S E D I = = E S E V|C B .
THE
OVER
COUNT,
THE
SUBTRACT
P E R IO D
Or
THE
IT
FR O M T H E
S IG N A L
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
IS
147
r
in t ib
:
MOM
8 # C 0 U N T ,R 0
; REMEMBER
E IC
#1,
JRESET TH E
F IN D
OUT TH E
P E R IO D
IS THE
n e x t
JUST
n e x t i
:
MOV
MOV
JSR
MOV
MOV
JSR
MOV
MOV
JSR
MOV
N E X T l:
a g a in
:
T IM E
F IR S T ,
SECOND,
Rl ,
R l,
RC,
RO
NEXT
RO
R2
NEXTl
R2
FO R
BETWEEN
Rl
R2
RO
R2
R2
MOV
CLR
MOV
MOVB
TSTB
BPL
CM?
BNE
R3,
0#CSR
#ST1,
<R 0 )-“ ,
0#CSR
.-4
RO,
A G A IN
MOV
# IN T 1 A ,
#1,
MOV
B IC
IN C
TSTB
BPL
MOV
B IC
RTI
T H IS
MOMENT
ENABLE
AND A L S O
PURPOSE —
P R IN T
OUT
JSET
THE
C O M P LE M E N T
THE
RESULT
BY C L E A R
THE
THE
UP
THE
PARAMETER
FO R
3 IN 2 D
JD IS P L A Y
THE
RF
PH AS E
S B IN 2 D
R4
P R IN T
ST1
ST2
ST3
RO
0#GU7LNK
.
#BOTTGM
3ADRVEC1
0#DROUT
JC H A N GE BACK T H E I N T E R R U P T V E C T O R
JEN A B LE T H E IN T E R R U P T 1 A G A IN
C A L L IN G
PROGRAM
==========
IN T E R R U P T
P E R IO D
" S T O R E T H E R E S U L T AT L O C A T I O N D I F A D R
JCALL TH E S U B R O U T IN E
J D IS P L A Y THE P E R IO D
JSTORE R E S U L T A T L O C A T IO N P E R IO D
JRETURN TO
IN T R 2 A
:
TWO S I G N A L S
JTAKE
RTI
i ennttzz a
a
AT
J R E C A L L T H E L A S T COUNT
JG E T t h e n e x t c o u n t t o c a l
JTHEN T H E T I M E D IF F E R E N C E
R2,
-(S P )
#D IF A D R , - ( S P )
PC,
4B IN 2 D
RO,
-(S P )
# P E R IO D , - ( S P )
5 B IN 2 D
PC,
R3,
-(S P )
io c c u c r ,- ( 5 ? )
J5R
MOV
MOV
WHEN
COUNT
IN T E R R U P T
J I S RO U N D E R F LO W
JNO
JTAKE T H E COMPLEMENT
T E S T IN G
PC,
#M ESS,
PC,
R2,
RO,
B IS
D IF F E R E N C E
THE
O F T H E R E F E R E N C E . T H E D I F F O V E R T H E P E R I O D OF T H E R E F E R E N C E
PHASE S H I F T OF TH E S I G N A L W IT H R E S P E C T TO THE R E FE R E N C E
MOV
MOV
SUB
SU B
SUB
TST
BGE
NEG
TST
BGT
NEG
:
0#DRGUT
C OMES
IN ,
RECO RD
3 # C G U N T ,S E C O N D
#2,
8#DRCUT
0#ADCS
8#ADCS
.-4
0 # A D B U F ,R 3
# 1 7 5 0 0 0 , R3
THE
COUNT
JCLEAR
JR EA D
" W A IT
THE
THE
FOR
TO
MEMORY C 0 N T 3 2
IN TE R R U P T
REGUEST
RF PHASE
A / D TO C O M P L E T E
====== =RTNt SOUTINE ==========
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
148
:
P R IN T
OUT T H E
CONTENTS
OF
THE
MEMORY W H I C H
IS 'P O IN T E D
BY
»
P R IN T :
TSTB
BPL
@ #C R T S R
.-4
MOVB
TSTB
BNE
RTS
C R 4 )+ ,
(R 4 )
P R IN T
PC
F IR S T :
• WORD
C0NTA2:
.W OR D
second
m ess
:
:
d if a d r
m e s s i
:
:
p e r io d
:
r fp h s e
:
l f c r
:
:
ST2:
ST3:
s t i
bottom
:
;IS
;IF
S#CRTBUF
THE SCREEN READY?
N O T , W A IT
J S E N D O U T ON E C H A R A C T E R
J I S T H A T T H E END O F T H E S T R I N G ?
jn o , t r y to send t h e n e x t char
J R E T U R N T O C A L L I N G PROGRAM
0
0
.W OR D
0
. A S C II/T IM E D IF F
. BLKB
5
.A S C II/
p e r io d
.B LK B
5
IS :
is: /
.A S C II/
RF PHASE
• BLKB
5
.B Y T E 1 2 , 1 5 , 0
.E V E N
.W OR D
.W OR D
.W OR D
.WORD
0
0
0
0
.E N D
START
/
IS :
/
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
149
3 U o * 0 u r ImL •'•OOrOei, 0b2, Pril,
1
S i » N l , p l G N 2 , a ? a 2, t>?)
Dlr'tfioiOn A t a , S ) , S { S > , X l f c )
Ph2,
ATTni,
aTTn2,
wUTH,
CUH?L=.A 2 1 , 2 X 2 , Z A 2 , £ P 5 X , £ P S Z , T l X , T 2 X , T l Z , r 2 Z , 4 Z x 2 , . U Z X 2 , R Z Z 2 ,
2
I K l l i 2, £a,
KJ*57.<96
UO 2 1-1 . 3
UO 2 J - 1 , 3
A 11 , J )*;! ,
CONTlNL£
PZ,ZxIX,Z21X
i\»e
A P K X = «f.
AbKX-O.
APKZ-t!.
AGKZ*<1.
APHl*phl
APH2»pf,2
s
iJlMU3*/2tf . 1
a2*< Ual/24.)
Prtl*APf.i
PH2-AP«2
P I - I . / L I a S X j NI
P2=l,/t2«SiuN2
PI»3.i41S52=S«l
Ul»( A r T N i * 5 . / 4 U T i r )
i>2*< AT IN c « 5 , / ~ J 3 T H )
IFU.cy.S) ^
TU 7
2i-(o 7c./3«o«9c,e.J
£PSX»Cf1rLX(X(l),-X<2) )
cPsz*onPwt(X(7),-xie>)
Z X 1 X * 1 . 0 0 3 5 » C o <«ST (a?3X)««IiiTh
K Z X 2 - 4 1 / l u y u a U X Z x )*«2-t>tPSx«C3lw(ZxlA )«*2)
KIZX2 « t Z l » C C C S { Z X U > » C a l N ( Z X I X > * ( l . - £ P 3 X n
1
(CCLS(Z*iX>«*2. * £rSX*C3ir.(zx:x)««2, >1
ZX2»?(2A2-i.--irLX CJ., 1. )•*.< IZX2
Z Z ! x » i , 0 B a 5 » C 3 rAPT JEPSZ) *
a
/
tCSli* T (fip3X ) •*
I JTH
H 2 Z 2 - Z 1 / ( C C C 5 1 ZZIX >* * 2 * £ P S Z « C 5 I N tZZI*)««2)
K l Z 2 2 - » Z : * C C 0 S ( i Z U ) « C 3 I N C Z Z l A ) « ( i . - £ r S Z > )/ tC S U * r (£ ? 3 Z )•*
I
( C C u S {2 A ;X j « » 2 . * = P S i * C 3 1 i ( i £ I X J * * 2 . ))
ZZ2*><Z a 2*C.i >'LX<;J., 1. ) - * U Z 2
K X ’ ( ZAic-Zl ) / ( Z a 2 * Z U
PZ-tZZc-xl)/(ZZ2+Z1)
C A L L Z r C L £ ( n X , S i J /.Ar5X,Ai>-tX)
C A L L trCLs-l-xZ/HD, 1PPZ,4-,4ZJ
T 1 X * 2 « Z X 2 / ( 2 x 2 a ZI)
TIZ^Z'ZZc/ <i£2*£i )
P T 1 A - O A ? S (T i x )
7
PTlZ=uA-alTlZ)
Ai?Gl«- U M a I«A3( T i X J/Pt AL (TiX ) J
AK( , 3*a TAM A ; ; i A 3 { T l Z ) / P £ A L ( T i t 5 J
Pl - P l -'-TlZ/rl lx
P2»P2*r7lX/>"UZ
Pl"l»K-i + A-»Ol»lSii./PT
P M 2 * P m 2 * A P o o « 1=J./PI
PTlZ*AWCaleJPTii)
PTl X - * i . C u l y ( P T 1 v)
Cl*Cl-''Tl a
U2»0c-PT1Z
bLTAl«P-i*tPl/i5«:.).iO0./WtDTB t T A 2 * ' iA i « ( « j / i 3 c . > « i : : o . / ,'iiDT-'
at! A i» = i T A i - i 3 g , 5 j 8 7 a i
= t T - <:a g £ T A 2 . i ( f 0 . S 0 7 ? i
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
150
S S & - 1 0 1 6 2 . 17413
K** + l
A P A 1 * u 1»WI J?T h / 5 .
a PA2 » J Z * W I D l n / 5 .
AL P * - a 1 « C 1 » A l 0 3 ( 1 S )
ALf*HA2*il2 ««i.03 ( 1
fad >» <B erA l«»2-A LP H A i*«2j/B S U
0 ( 2 > « ( - 2 . « A L P W A 1 *<JET a i J / B S 8
t» ( 3 ) » - 2 . « . A w r ' r i A 2 * f e t T A 2 « P 1 / S 5 < 1
0<?) = iB E rA ic **2 “ ALPnA2««2)/is3ii
a ( a ) ■ ( " B (7 J)*P1
fat5)=-B<2)«P2
o(e>)»B(l)*P2
B ( f a ) » - i, * A u - ’ri42*oErA2/BStf
A(l,l)»l.
A11,4)»?«;
4(2,*)--!.
A {2,3)»P 4
A (3 , 2 )* «Pl
A(3,3>»1.
A(«,l)»-r>l
A U ,4 }*-1 .
A15,S}«1.
4(S,<jJ»?2
A(6,o)*-1.
A (6,7
A ( 7 , o J — ?1
At7 ,7 J » i.
A ( 5 , 3 J «->*l
A ta ,o )= -i.
i\i»g
IS
2d
21
CALL a L L V 2 ( A , a , A , * , !£*<<)
I F t i t . L T . 4 ) ‘ v-u TO =
" K I T E Is, is ) u H i , u S 2 , P * l , F r ! 2 , » P U . i p A 2 , A H j T . - ! r A 0 ><Xf Ai5Kxr
1
A P «L, A & 2 2
F 0 H M A rtlX ,ilF l< !.3 ,/)
* « I T c ( 6 , 2<3) { X t I ) , 1 * 1 , 4 }
wPITt 1 6 ,2 d n H I ) , 1 * 3 , 2 )
FOP>iATliA,r ; « . e , 3 X , •-J' , F K - . 6 , 1 2 X , F 1 . * . o , o a , '-J' ,Flii.ol
*« IT = .lo ,2 i)
F C A n * r {1A,///|
£PP»X( / )
P£TLx<
t'lil
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Z P O L E .F O R /L
*LI
S U B R O U T IN E Z P O L E C Z .R D , A M P ,P H A S E )
COMPLEX Z
X=REAL<Z>
Y = A IM A G C Z )
2
4
5
3
I
8
A M P =C A B S (Z )
IF (X )
1 ,2 ,0
IF (Y ) 3 ,4 :5
P H I= 0 .
GO T O 7
P H I= 1 .5 7 0 7 9 5 3
GO TO 7
P H I= 4 .7 1 2 3 B 8 S
GO T O 7
P H I= A T A N (A B S (Y /X ))
IF
CY) 8 , 8 , 9
P H I= P H I+ 3 .1415927.
GO T O 7
9
P H I= 3 .1 4 1 5 9 2 7 -P H I
GO TO 7
S
P H I= A T A N (A B S (Y /X ))
IF
(Y ) 1 1 ,1 1 ,7
P H I= S .2 8 3 1 8 5 3 -P H I
P H A S E = R D *P H I
RETURN
END
I I
7
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
S0LV2.F0R/L
*LI
100
S U B R O U T IN E S 0 L V 2 ( A , S , X , N , I E R R )
D IM E N S IO N
A (8 ,B ),B < 8 ),X C 8 )
N=8
IE R R = 0
I F ( N . G T . l ) GO T O 1 0 0
I F ( A < 1 , 1 ) . E G . O . ) GO T O 1 0 0 0
X (1 )= B (1 )/A < 1 ,1 )
RETURN
N M 1 = N —1
DO 4 0 0 1 = 1 , NM1
IP 1 = I+ 1
10=1
150
150
IS O
A M = A B S (A (1 , 1 ) )
DO 1 5 0 J = I P 1 , N
A M T = A B S C A (J ,I))
I F ( A M T . L E . A M ) GO T O 1 5 0
AM = A M T
IO = J
C O N T IN U E
I F ( I . E G . 1 0 ) GO T O 1 8 0
DO I S O J = 1 , N
T E M P = A (I, J )
A (I,J )= A (IO ,J )
A < I Q , J ) =TEMP
C O N T IM u E
T E M P = B (I)
S( I)= B C 10)
B < IQ )= T E M P
1000
C O N T IN U E
DO 3 0 0 J = I r 1 , N
R = A (J ,I)/A (I,I)
DO 2 0 0 K = I P 1 , N
A ( J , K ) = A ( J , K ) —R » A ( I , K )
C O N T IN U E
B ( J ) = B CJ ) - R * B ( I )
C O N T IN U E
C O N T IN U E
X iN )= B (N )/A < N ,N )
DO SCO I = N M 1 , 1 , - 1
IP 1 = I+ 1
DO 5 0 0 J = I P 1 , N
B (I)= 5 C I)-A < I,J )*X (J )
C O N T IN U E
X (I)= B (I)/A (I,I)
C O N T IN U E
RETURN
IE R R =1
as
M R IT E C S ,S 3 )
FORMAT( 2 X ,'E Q U A T IO N S
200
300
400
500
600
S IN G U L A R ')
RETURN
E ND
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
L K F R A .F O R /L
*LI
S U B R O U T IN E
LKFRA
nm
D IM E N S IO N S T X ( I O O ) » D T X ( 1 0 0 ) ,S T Y C 1 0 0 ) ,D T Y C 1 0 0 >
COMMON X U O O ) » Y C 1 0 0 ) , Z < 1 0 0 > , X D ( I S , 3 ) , Y D < I S , 3 ) , D M C I S ,
COMMON A N U S , 3 ) , P H < I S , 3 ) , A M A X < 2 ) » R F ( 2 ) » R A T O ( 2 ) * P H A S (
COMMON L X , L Y , L M , L A N , L P H , J M , W I D T H , L r A T T N , S A T N » T P H , S P H
COMMON S X , S Y , I D S K X , I D S K Y , T H T A , D D M ( 1 0 0 )
I F C J W . N E . l ) GO TO 1 1
C ALL A S S IG N < 6 , ' D 1 . D A T ' )
GO TO 1 0 0
I F ( J M . N E . 2 ) GO TO 1 2
CALL A S S IG N C 6 » 'D 2 .D A T ')
GO TO 1 0 0
I F U H . N E . 3 ) GO TO
13
11
12
13
CALL A S S IG N (S » ' D 3 . D A T ' )
GO TO 1 0 0
T F ( J W . N E . 4 ) GO TO
14
CALL A S S I G N ( S r ' D 4 . D A T ' )
14
GO TO 1 0 0
IF < J W . N E .5 )
15
CALL A S S IG N C S r 'D 5 .D A T ')
GO T O 1 0 0
I F < J W . N E . S ) GO T O
IS
IS
17
100
110
56
55
4
33
5
B
555
15
CALL A S S I G N C 6 , ' D S . D A T ' )
• GO T O 1 0 0
I F U W . N E . 7 ) GO T O
17
CALL A S S I G N C S , 'D 7 . D A T ')
GO TO 1 0 0
CALL A S S IG N C S # 'D 8 .D A T ' )
DO 5 5 1 = 1 > I D S K X
DO 5 S J = 1 r I D S K Y
RE A D ( 5 , 1 1 0 ) X D d r J ) r Y D C I r J ) , D M < I , J ) , ANC I , J ) , FHC I , J )
FORMAT C S F 1 0 . 4 )
C O N T IN U E
C O N T IN U E
SX=100.
1
3
GO TO
1
1
=
=
1
S T X C I) = X D C I, LY ) - X C L - l )
D T X ( I ) = ABSCXD < I »L Y ) - X ( L - 1 ) )
I F C D T X C I ) . G E . 3 X ) GO TO 3 3
I F C S T X ( I ) . G T . O . ) GO T O 4
S X = -D T X < I)
GO TO 3 3
SX=DTX( I )
LX=I
1 = 1 + 1
IF ( I
. L T . I D S K X ) GO T O 3
SY=1.
J = 1
J = J
S T Y ( J ) = YDCLX* J ) - Y < L -1 )
D T Y C J )= A B S C Y D C L X ,J )-Y (L -i))
I F C D T Y < J ) . G E . S Y ) GO T O 5 5 5
I F C S T Y C J ) . G T . O . ) GO T O S
3 Y = -D T Y < J )
GO TO 5 5 5
SY=DTY( J )
LY=J
4
=
J
+
1
IF (J
. L T . I D S K Y ) GO TO
D D M ( L > = D M ( L X »L Y )
Z < L > = T H 7 A *{1 ./D D M C L ))
5
RETURN
END
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
15 4
L K F R B .F O R /L
*L I
nm
S U B R O U T IN E LKFRB
D IM E N S IO N S S P H (IO O ) ,D P H (1 0 0 >
COMMON X ( I O O ) > Y < 1 0 0 ) , Z ( 1 0 0 ) , X D ( I S , 3 ) , Y D ( I S , 3 ) , D M ( 1 S ,
COMMON A N ( I S , 3 ) , P H ( 1 6 , 3 ) , A M A X ( 2 ) , R F ( 2 ) , R A T O ( 2 ) , P H A S (
COMMON L X , L Y , L M , L A N , L P H , J W , W I D T H , L , A T T N , S A T N , T P H , S P H
COMMON S X , S Y , I D S K X , I D S K Y , T H T A , D D M ( 1 0 0 )
SPH=100.
J = 0
J = J + 1
S S P H (J ) = P H (L X ,J )-T P H
1
2
D P H ( J ) = A B S ( P H ( L X , J ) —T P H )
I F ( D P H ( J ) . G E . S P H ) GO TO 1
IF (S S P H (J ) . G T. 0 . )
GO T O 2
S P H = -D P H (J )
GO T O 1
SPH=DPH( J )
LPH= J
I F ( J . L T . ID S K Y )
Y (L )= Y D (L X ,L P H )
RETURN
EN D
GO TO
1
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
155
L K F R C .F O R /L
+LI
S U B R O U T IN E
LKFRC
D IM E N S IO N S S A T N (IO O ),D A N < 1 0 0 >
COMMON X C I O O ) » Y ( 1 0 0 ) » Z < 1 0 0 ) » X D U 6 , 3 ) , Y D < 1 6 , 3 > , D M < I S , 3 )
COMMON A N ( I S , 3 ) , P H ( 1 6 , 3 ) , A M A X < 2 ) , R F < 2 ) , R A T O ( 2 ) , P H A S ( 2 )
COMMON L X , L Y , L M , L A N , L P H , J M , W I D T H , L , A T T N , S A T N , T P H , S P H
COMMON S X r S Y , I D S K X , I D S K Y , T H T A , D D M ( 1 0 0 )
SATN=100.
I = 0
1
2
1
= 1+ 1
SSA TN ( I ) = AN( I , L Y ) -A T T N
D A N ( I ) = A B S ( AN f I , L Y ) - A T T N )
I F C D A N ( I K G E . S A T N ) GO T O 1
I F ( S S A T N ( I > . G T . 0 . ) GO T O 2
S A T N = -D A N < I)
GO T O 1
S A T N = D A N < I)
LAN=I
IF ( I
. L T . I D S K X ) GO T O 1
X ( L 5=XD{ LAN, L Y )
RETURN
END
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Appendix J
Modified Analog Circuitry for Real-Time Measurements
T-LM324
66k
2501
W v - n 10k
!50k
100k
T
lots 120k
M 1312 It109 6
F.
-_,3_VJ
>,r
(N r
GjU
[ Q UAli
420k K)0k
I
)k
(Ok
I :>
AA\W*A'Wwr~l>'^
nruLn
{.COMIItl
iNAfleu.W
UAF-41
01 22.8k
^ j-v39
WVj-Vk'
.V*—
J
392k
.00318
30 2k ».9k
|- A / v V tOk
•Ok
301
W
•j-WVO|-(
0 0051.0005.1.
OD
OAF-41
IU
powen
•a IP.)
O I6 t * O is jt
supplies:
Pt ~
(Ok
«0.2k
»fk »0/k
W W -, JW l ' W n
____
|.|X I
T ’^V-
-W v V —
ooooj* KA/vO.OOOv ^
4' ,0k
ink
13741
IM
p6
U tASE
14.18k
994k
9 0tk
X
351
JlOk
^51
|0k<
©_
156
157
Bibliography
1.
I. J. Bahl andG . Bhartia, "Microstrip antenna," Artech House,
1980.
2.
K. F. Casey, "EMP Penetration through Advanced Composite
Skin Panels," Air Force Weapons Lab Interaction Note 315,
Dec. 1976.
3.
R. J. Cook, "The Propagation of Plane Waves through a
Lamella," Rpt. DES 52, National Physical Laboratory, UK,
Aug. 1979.
4.
E. 0. Doebelin, "Measurement System:
Application and Design,"
McGraw-Hill, p. 688, 1966.
5.
J. D. Dyson, "Microwave Phase Measurements," Technical Rpt.,
AFAL-TR-67-17, D. of Illinois, March 1967.
6.
P. A. Ellerbruch, "Evaluation of a Microwave Phase Measurement
System," J. Res. NBS, 69C, p. 55-65, 1965.
7.
R. F. Harrington, "Small Resonant Scatterers and Their Use
for Field Measurements," IRE-MTT, p. 165-174, May 1962.
8.
R. F. Harrington, "Theory of Loaded Scatterers," Proc. IEE,
Vol. Ill, no. 4, p. 617-623, April 1964.
9.
W. L. James and D. W. Hamill, "Dielectric Properties of
Douglas Fir —
Measured at Microwave Frequency," Forest
Products Journal, p. 51-56, Feb. 1965.
10. R. C. Jones, "A New Calculus for the Treatment of Optical
Systems," J. Opt. Soc. Am., Vol. 38, No. 8, p. 671-685,
August 1948.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
158
11.
R. J. King, "Microwave Homodyne System," Peter Peregrimus LTD,
1978.
12.
R.
J. King, "Microwave Electromagnetic Nondestructive Testing
of Wood," 4th Nondestructive Testing of Wood Symposium,
Vancouver, WA, Aug. 28-30, 1978.
13.
R. J. King, "On Airborne Wave Tilt Measurements," Radio Science,
Vol. 12, No. 3, p. 405-414, May-June, 1977.
14.
R.
J. King, "Microwave Nondestructive Testing of Wood,"
Proposal to NSF.
15.
R. J .
King, "Analytic and Experimental Studies of Propagation
of Electromagnetic Surface Waves Across Mixed Paths," Ph.D.
Thesis, U. of Colorado, Boulder, CO, 1965.
16.
R. J. King and Y. H. Yen, "Probing Amplitude, Phase and
Polarization of Microwave Field Distributions in Real-Time,"
1980 North American Radio Science Meeting, Quebec, Canada,
June 2-6.
17.
R. J . King and Y. H. Yen,“Probing Amplitude, Phase and
Polarization of Microwave Field Distributions in Real-Time,"
IEEE Trans, on Microwave Theory and Techniques, Nov. 1981.
18.
R. J. King, "Error Analysis of Phase Insensitive Coherent
(Homodyne) Detectors," IEEE Trans, on Instrumentation and
Measurement (to be published).
19.
A. R. Koelle and S. W. Depp, "Doppler Radar with Cooperative
Target Measures to Zero Velocity and Senses the Direction of
Motion," Proc. IEEE, 65(3), p. 492-493, March 1977.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
159
20-
J. A. Kong, "Theory of Electromagnetic Waves," John Wiley &
Sons, 1975.
21.
A. Kraszewski and S. Kulinski, "An Improved Microwave Method
of Moisture Content Measurement and Control," IEEE-IECI,
Vol. 23, No. 4, p. 364-369, Nov. 1976.
22.
A. Kraszewski, "Microwave Instrumentation for Moisture Content
Measurement," J. of Microwave Power, 8(3/4), 1973.
23.
H. Kubo and R. Nagata, "Determination of Dielectric Tensor
Fields in Weakly Inhomogeneous Anisotropic Media," J. Opt. Soc.
Am., Vol. 69, No. 4, April 1979.
24.
L. D. Landau and E. M. Lifshitz, "Electrodynamics of Continuous
Media," Addison-Wesley, 1960.
25.
C. K. Lau, "An Experimental Study of Electromagnetic Surface
Waves On a Uniform Pin Reactance Structure," M.S. Thesis,
U. of Wisconsin, Madison, April 1977.
26.
R. T. Lin, "Review of the Dielectric Properties of Wood and
Cellulose," Forest Products Journal, Vol. 17, No. 7, pp. 61-66,
July 1967.
27.
W. Meyer and W. M. Schilz, "A New Method for Density Independent
Determination of the Moisture Content of Solids," J. Phys.
D13 (1C), p. 1823ff, Oct. 14, 1980.
28.
w. Meyer and W. M. Schilz, "Feasibility Study of Density —
Independent Moisture Measurement with Microwaves," IEEE-MTT-29,
No. 7, July 1981.
29.
G. Morin and M. Nachman, "Measurement of Anisotropy at Microwave
Frequencies," J. of Microwave Power, 14(3), 1979.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
160
30.
R. Morrison, "Grounding and Shielding Techniques in Instrumenta­
tion," John Wiley & Sons, 1967.
31.
A. J. Nanassy, "Electric Polarization Measurements on Yellow
Birch,” Canadian J. of Physics, Vol. 42, June 1964.
32.
Princeton Applied Research Lab, "Grounding and Shielding in
Electrochemical Instrumentation —
Some Basic Considerations,"
Technical Note 117, 1978.
33.
J. Rafalski, "On the Dielectric Properties of Differently
Compressed Solid Copper Beech Wood," Hciztechnologie, 7(2),
p. 118-122, 1966, East Germany.
34. S. Ramo, J. R. Whinnery and T. Van Duzer, "Fields and Waves
in Communication Electronics," John Wiley & Sons, p. 349, 1965.
35. V. H. Rumsey, G. A. Deschamps, M. L. Kales and J. I. Bohnert,
"Techniques for Handling Ellipticaly Polarized Waves with
Special Reference to Antennas," Proc. IRE, p. 533, May 1951.
36.
W. Schilz and B. Schiek, "Microwave Systems for Industrial
Measurements,” Advances in Electronics and Electron Physics,
Vol. 55, Academic Press, 1981.
37.
C. Skaar, "The Dielectric Properties of Wood at Several Radio
Frequencies," Tech. Pub. 69, N.Y. State College of Forestry,
Syracuse University, 1948.
38. H. S. Tan and A. K, Fung, "A First Order Theory on Wave
Depolarization by a Geometrically Anisotropic Random Medium,"
Radio Science, Vol. 14, No. 3, p. 377-386, May-June 1979.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
161
39.
W. R. Tinga, "Multiphase Dielectric Theory Applied to Cellulose
Mixtures,” Ph.D. Thesis, The University of Alberta, 1969.
40.
M. Tiuri and S. Heikkila, "Microwave Instrumentation for Accurate
Moisture Measurement of Timber," 9th European Microwave Confer­
ence, 1979.
41.
M. Tiuri, K. Jokela and S. Heikkila, "Microwave Instrumentation
for Accurate Moisture and Density Measurement of Timber," J. of
Microwave Power, 15(4), Dec. 1980.
42.
G. Tyras, "Badiation and Propagation of Electromagnetic Waves,"
Academic Press, 1969.
43.
A. Venkateswaran and S. Y. Tiwari, "Dielectric Properties of
Moisture Wood," Tappi, Vol. 47, No. 1, p. 25-28, Jan. 1964.
44.
R. J. Vernon and B. D. Huggins, "Extension of the Jones Matrix
Formalism to Reflection Problems and Magnetic Material," U.
of Wisconsin-Madison, 1980.
45.
A. Von Hippel, ed., "Dielectric Properties and Applications,"
MIT Press, 1954.
46.
C. G. Wang, "Computer Support of Microwave Navigation and
Instrumentation System," M.S. Thesis/ U. of Wisconsin-Madison,
Dec. 1980.
47.
j. M. Yavorski, "A Review of Electric Properties of Wood,"
Tech. Publication No. 73, State U. of New York, College of
Forest, at Syracuse U., Syracuse, N.Y., 1952.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
TITLE OF THESIS
MICROWAVE ELECTROMAGNETIC NONDESTRUCTIVE
TESTING OF WOOD IN R E A L - T I M E ____________________
MAJOR PROFESSOR
R. J. KING___________________________________
MAJOR DEPARTMENT
ELECTRICAL .ENGINEERING_____________________
MINOR(S)
NAME
COMPUTER SCIENCES__________________________________
YOU-HSIN YEN__________________________________________
PLACE AND DATE OF BIRTH
KEELUNG, TAIWAN, R.O.C. OCT. 14 1950
COLLEGES AND UNIVERSITIES: YEARS ATTENDED AND DEGREES _____________
NATIONAL CHIAO—TUNG UNIVERSITY 1968-1972 BSF.E ( T A I W A N )
MARQUETTE UNIVERSITY. 1975-1977 MSEE (MILWAUKEE)___________
MEMBERSHIPS IN LEARNED OR HONORARY SOCIETIES ______________________
SIGMA-XI
PUBLICATIONS
IEEE
TVK. ISHII , Y.H. YEN AND R.J. KIPP.____________
"IMPROVEMENT OF MICROWAVE POWER DISTRIBUTION BY THE USE OF
THE FIRST ORDER PRINCIPLE OF GEOMETRICAL OPTICS FOR_______
SCIENTIFIC MICROWAVE OVEN CAVITY", J. of Microwave Power.
14(3), 1979 ;
R.J. KING AND Y.H. Y E N ,"PROBING AMPLITUDE,
PHASE AND POLARIZATION OF MICROWAVE FIELD DISTRIBUTIONS IN
REAL TIME", 1980 North American Radio Science Meeting,____
Quebec, Canada, June_-2-6 and IEEE —M T T . NOV. 1981___________
DATE NOV.
25, 1981____________
F-5266
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Документ
Категория
Без категории
Просмотров
0
Размер файла
4 906 Кб
Теги
sdewsdweddes
1/--страниц
Пожаловаться на содержимое документа