Professional Documents
Culture Documents
Antonije R. Djordjevi
Miodrag B. Badar
Vladimir V. Petrovi
Dragan I. Olan
Tapan K. Sarkar
Roger F. Harrington
AWAS for Windows
Version 2.0
Analysis of Wire Antennas and Scatterers
Software and Users Manual
Artech House
2002
9
iii
Abstract
AWAS for Windows is a user-friendly computer program for analysis of wire antennas and
scatterers. It is built on an efficient and accurate technique, which has been developed exclusively
for this program. The technique is based on the two-potential equation for the distribution of wire
currents. This integro-differential equation is solved numerically, using the method of moments
with a polynomial (power-series) approximation for the current distribution. Influence of a real
ground is taken into account using Sommerfelds theory.
AWAS for Windows can analyze wire structures isolated in space, placed above a
perfectly conducting ground plane, or placed above a real ground. Wire structures can have
concentrated or distributed loadings, and they can be driven at one or more ports or excited by a
plane wave of an arbitrary polarization. The program evaluates the current distribution, near and
far fields, port impedance, admittance, and scattering parameters, and so on.
The method inherently requires a small number of unknowns (about 10 per wavelength).
Owing to special techniques used for numerical integration, AWAS for Windows is one of the
worlds fastest programs in its category, and it yields accurate and reliable results, which are in
accordance with experimental data.
The program is user friendly, as it has been designed in accordance with Windows
standards. It allows for easy data input, editing and checking data that define the wire structure,
automatic segmentation, and choosing the approximation functions on wires. It also has ample
graphic capabilities at the output.
The program is delivered in an executable form for any IBM or compatible Win32
platform.
3
+
v
Contents
List of Figures ix
List of Tables xi
Trademarks Used in the Manual xiii
Chapter 1 Introduction 1
1.1 Subject 1
1.2 Intended Users 2
1.3 Program Environment and Distribution 2
1.4 Program Usage 2
1.5 Packaging List 2
1.5 Contents of the Manual 3
1.5 If You Use Previous Versions of AWAS 3
1.7 If You Have Questions or Comments 4
Chapter 2 Hardware and Software Requirements 7
Chapter 3 Installing AWAS for Windows 9
Chapter 4 Getting Started 15
4.1 Starting and Stopping the Program 15
4.2 Getting Help 16
4.3 Quick Tour Through AWAS for Windows 17
4.3.1 Starting the Program 17
4.3.2 Tour Through Help Windows 17
4.3.3 Tour Through Input Data Window 20
4.3.4 Tour Through Graph and Plot Windows 24
4.3.5 Stopping the Program 26
Chapter 5 Tutorial Guide 27
5.1 Introduction 27
5.2 Setup 27
5.3 Basic Principles of Analysis of Wire Structures 29
5.4 Input Data Window 30
5.4.1 Open Input Data Window 30
5.4.2 Structure of Input Data and Input Data Files 30
5.4.3 Load an Input Data Set 31
5.4.4 Parts of Input Data Window 31
5.4.5 Show Window 35
5.4.6 Edit Loaded Data 35
+
vi Analysis of Wire Antennas and Scatterers
5.4.7 Operating Mode and Ground 37
5.4.8 Frequency 39
5.4.9 Far Fields 40
5.4.10 Near Fields 42
5.4.11 Plane Waves 43
5.4.12 Checking 45
5.4.13 Save Input Data 47
5.5 Analysis 48
5.5.1 Start Analysis from Input Data Window 48
5.5.2 While the Program Kernel Is Active 49
5.5.3 Structure of Output Data and Output Data Files 49
5.5.4 Batch Analysis 49
5.6 Output Data Management 50
5.6.1 Save Output Data Files 50
5.6.2 List Output Data 52
5.6.3 Graphic Presentation of Output Data 53
5.7 Recommended Exercise 58
Chapter 6 Theory 59
6.1 Introduction 59
6.2 Two-Potential Equation 61
6.3 Solution of Two-Potential Equation 64
6.4 Evaluation of Antenna and Scatterer Characteristics 67
6.4.1 Current and Charge Distributions 68
6.4.2 Impedance, Admittance, and Scattering Parameters of Antennas 68
6.4.3 Power Balance 72
6.4.4 Near Fields 73
6.4.5 Far Fields 74
6.5 Evaluation of Integrals 76
6.6 Antennas Above Real Ground 77
6.6.1 Introduction 77
6.6.2 Greens Functions 78
6.6.3 Continuity Equation for Wire Segments Above Ground 81
6.6.4 Evaluation of Reflected Near Fields 82
6.6.5 Evaluation of Sommerfelds Integrals 84
Chapter 7 Reference Manual 85
7.1 Introduction 85
7.1.1 Program Features 85
7.1.2 Program Environment 86
7.1.3 Program Distribution 87
7.1.4 Program Installation 88
7.1.5 Program Structure 88
7.1.6 Computation Cycle 90
7.2 Input Data 90
7.2.1 Input Data Structure 90
7.2.1.1 Header 91
7.2.1.2 Operating Mode 91
7.2.1.3 Ground 92
7.2.1.4 Nodes, Segments, Loadings, and Ports 93
Contents vii
7.2.1.5 Polynomial Approximation and Automatic Segmentation 94
7.2.1.6 Operating Frequencies 95
7.2.1.7 Plane Waves 95
7.2.1.8 Near Fields 97
7.2.1.9 Far Fields 98
7.2.2 Input Data Files 99
7.2.2.1 Input Data Set 99
7.2.2.2 File *.GEO 100
7.2.2.3 File *FRQ 103
7.2.2.4 File *.PWE 103
7.2.2.5 File *.NFP 104
7.2.2.6 File *.FFP 105
7.2.3 Number Representation and Restrictions 106
7.2.3.1 Real Numbers and Integers 106
7.2.3.2 Complex Numbers 106
7.2.3.3 Program Restrictions and Range of Input Data 107
7.2.3.4 Input Data Checking 107
7.3 Output Data 107
7.3.1 Output Data Structure 107
7.3.1.1 Polynomial Coefficients and Automatic Segmentation Data 108
7.3.1.2 Port Matrix Parameters and Power Balance 108
7.3.1.3 Near Fields 109
7.3.1.4 Far Fields 109
7.3.2 Output Data Files 109
7.3.2.1 Output Data Set 109
7.3.2.2 File *.OUT 110
7.3.2.3 File *.SPM 118
7.3.2.4 File *.YPM 119
7.3.2.5 File *.ZPM 121
7.3.2.6 File *.CUR 122
7.3.2.7 File *.NFL 125
7.3.2.8 File *.FFL 127
7.4 Main Window 131
7.5 Help Windows 134
7.6 File Management 135
7.6.1 File Loading 135
7.6.2 File Saving 136
7.7 Data Input 137
7.7.1 Input Data Window 137
7.7.2 Editing Header 141
7.7.3 Operating Mode and Ground Input 141
7.7.4 Geometry Input 142
7.7.4.1 Loading dialog box 144
7.7.4.2 Port Dialog Box 144
7.7.5 Geometry Show 145
7.7.6 Frequency Input 146
7.7.7 Plane Wave Input 147
7.7.8 Near Field Input 149
7.7.9 Far Field Input 150
+
viii Analysis of Wire Antennas and Scatterers
7.7.10 Checking Input Data 152
7.8 Computing 152
7.8.1 Individual Analysis 153
7.8.2 Batch Analysis 154
7.9 Listing Output Data File 156
7.10 Plotting 157
7.10.1 Graph Selection Dialog Box 158
7.10.1.1 Plotting Port Scattering, Admittance, and Impedance Parameters 159
7.10.1.2 Plotting Current and Charge Distributions 161
7.10.1.3 Plotting Near Fields 163
7.10.1.4 Plotting Far Fields 165
7.10.2 Plot Window 169
7.10.2.1 Plot Scale and Labels 171
7.10.2.2 Plot Title 173
7.10.2.3 Exporting Plot 173
7.10.2.4 Tabulating Plotted Data 173
7.11 Hardcopy 173
7.12 Configuration 174
7.12.1 Setup Dialog Box 175
7.12.2 Default Files 177
7.13 External Programs 179
Chapter 8 Examples of Analysis 183
8.1 Introduction 183
8.2 Vertical Monopole Antenna 183
8.3 Simple Wire Scatterers 190
8.4 Loop Antenna 194
8.5 Circuit Analysis 196
8.6 Linear Array 197
8.7 Loaded Broadband Monopole Antennas 199
8.8 Yagi-Uda Array 203
8.9 Tower Broadcast MF Antenna 206
8.10 Nuclear Electromagnetic Pulse Simulator Antenna 207
8.11 Wire-Grid Model of Square Plate 209
8.12 Wire Shielding Cage 211
8.13 Vertical Dipole Antenna Above Real Ground 213
8.14 V-Dipole Antenna Above Real Ground 216
References 219
Appendix A Program Limits 221
A.1 Upper Limits in AWAS for Windows and AWAS for DOS 221
A.2 Input Data Restrictions in AWAS for Windows 222
Appendix B Error Messages 225
B.1 Error Messages Issued by the Shell 225
B.2 Error Messages Issued by the Kernel 227
Appendix C Help Windows 231
Index 257
ix
List of Figures
Figure 4.1 Loaded L-monopole antenna with a counterbalance. All dimensions are in
millimeters. 22
Figure 6.1 A wire structure with the coordinate system. 60
Figure 6.2 (a) Antenna port driven by an ideal voltage generator and (b) its representation by
an impressed electric field. 61
Figure 6.3 Local coordinate system for a wire segment. 63
Figure 6.4 (a) A wire segment above a perfectly conducting ground plane and (b) the
segment and its image in the ground plane. 64
Figure 6.5 A wire structure with pulses as weighting functions. 65
Figure 6.6 Equivalent network representation of a multiport antenna in the transmitting
mode. 68
Figure 6.7 (a) Thvenin and (b) Norton equivalent network representations of a multiport
antenna in the receiving mode. 72
Figure 6.8 Coordinate system for evaluation of far fields. 75
Figure 6.9 Coordinate system for the current element above the ground. 78
Figure 7.1 Basic structure of AWAS for Windows. 89
Figure 8.1 Vertical monopole antenna above a ground plane. 184
Figure 8.2 (a) Current and (b) charge distribution of a monopole antenna of height 1 m and
radius 1 mm, at 300 MHz. 186
Figure 8.3 Power gain of a monopole antenna of height 1 m and radius 1 mm, at 300 MHz. 187
Figure 8.4 Axial component of the electric field on the axis of a monopole antenna of height
1 m and radius 1 mm, at 300 MHz. 189
Figure 8.5 Input impedance of a monopole antenna of height 1 m and radius 1 mm, as a
function of frequency. 189
Figure 8.6 Simple wire scatterer excited by a plane wave. All dimensions are in millimeters. 190
Figure 8.7 Monostatic cross section of the scatterer of Figure 8.6, as a function of
observation angle, at 300 MHz. 191
Figure 8.8 Monostatic cross section of the scatterer of Figure 8.6, as a function of
frequency, for broadside incidence. 192
Figure 8.9 Bistatic cross section of the scatterer of Figure 8.6, as a function of observation
angle, for broadside incidence at 300 MHz. 193
Figure 8.10 Scatterer consisting of two crossed wires. All dimensions are in millimeters. 194
Figure 8.11 Square-loop antenna. All dimensions are in millimeters. 195
Figure 8.12 Input impedance of the loop antenna of Figure 8.11, as a function of frequency. 195
Figure 8.13 Schematic representation of an electric circuit. 196
Figure 8.14 Wire model of the circuit of Figure 8.13. All dimensions are in millimeters. 196
+
x Analysis of Wire Antennas and Scatterers
Figure 8.15 Input impedance of the wire structure of Figure 8.14, as a function of frequency. 197
Figure 8.16 Array of five monopole antennas. All dimensions are in millimeters. 198
Figure 8.17 Monopole antenna with concentrated capacitive loadings. All dimensions are in
millimeters. 199
Figure 8.18 Input impedance of the capacitively loaded monopole antenna of Figure 8.17, as
a function of frequency. 200
Figure 8.19 Normalized radiated electric field of the capacitively loaded monopole antenna
of Figure 8.17, as a function of observation angle, at (a) 1 GHz and (b) 2 GHz. 201
Figure 8.20 Monopole antenna with concentrated capacitive and distributed resistive
loadings. All dimensions are in millimeters. 202
Figure 8.21 Input admittance of the capacitively and resistively loaded monopole antenna of
Figure 8.20, as a function of frequency. 203
Figure 8.22 Yagi-Uda array. All dimensions are in millimeters. 204
Figure 8.23 Input admittance of the Yagi-Uda array of Figure 8.22, as a function of
frequency. 204
Figure 8.24 Power gain of the Yagi-Uda array of Figure 8.22, as a function of observation
angle, at 1 GHz, in (a) the zx-plane and (b) the xy-plane. 205
Figure 8.25 Tower monopole antenna with uninsulated guy ropes. All dimensions are in
meters. 206
Figure 8.26 Input impedance of the tower antenna of Figure 8.25, as a function of frequency. 207
Figure 8.27 Simple nuclear electromagnetic pulse simulator antenna. All dimensions are in
meters. 208
Figure 8.28 Input impedance of NEMPS simulator antenna of Figure 8.27, as a function of
frequency. 208
Figure 8.29 Wire-grid model of a square plate. All dimensions are in millimeters. 209
Figure 8.30 Monostatic cross section of the wire-grid structure of Figure 8.29, as a function
of observation angle , at 75 MHz, when the structure is excited by a linearly
polarized plane wave with (a) -component and (b) -component of the electric
field, for =0. 210
Figure 8.31 Wire shielding cage. All dimensions are in millimeters. 212
Figure 8.32 Vertical component of the electric field along the x-axis in the shielding cage of
Figure 8.31 and its vicinity, at 50 MHz. 212
Figure 8.33 Vertical dipole above real ground. 213
Figure 8.34 Input impedance of the vertical dipole of Figure 8.33 above average ground. 214
Figure 8.35 Radiation pattern of the vertical dipole of Figure 8.33 touching moist ground. 215
Figure 8.36 Electric field on the ground surface of a vertical dipole touching moist ground. 215
Figure 8.37 V-dipole above real ground. 216
Figure 8.38 Input admittance of the V-dipole of Figure 8.37, as a function of frequency. 217
+
xi
List of Tables
Table 8.1 Input admittance (Y) and gain in the horizontal direction (g) of two monopole
antennas, of height h=0.25 m and different radii R, at 300 MHz, for various
values of the accuracy parameter. 185
Table 8.2 Input impedance (Z) of the monopole antenna of height-to-radius ratio h/R=1000,
as a function of normalized height (h/). 188
Table A.1 Upper limits in AWAS for Windows and AWAS for DOS. 222
+
+
xiii
Trademarks Used in the Manual
Windows is a registered trademark of Microsoft Inc.
Pentium is a registered trademark of Intel Inc.
+
+
1
Chapter 1
Introduction
1.1 Subject
Wire antennas and scatterers are structures made of wire-like conductors, the radii of which are
much smaller than their lengths and the wavelength at the operating frequency. Examples of such
structures are simple wire dipoles, V-antennas, loops and rhombic antennas used for HF
communications, tower broadcast antennas for MF and LF ranges, Yagi-Uda antennas and log-
periodic dipole arrays used in the HF, VHF, and UHF ranges, and so on. Certain other antennas
can be approximated by wire structures, such as antennas made by printed-circuit techniques.
Quite a few scatterers can be represented as wire structures (e.g., aircraft at lower frequencies),
and many others can be analyzed by approximating their surfaces by wire-grid models.
The computer program AWAS for Windows (Analysis of Wire Antennas and Scatterers)
is capable of analyzing wire antennas operating in the transmitting and receiving modes, as well as
analyzing wire scatterers. It is based on an efficient and accurate technique that was developed for
the first release of this program [1]. The analysis method was designed by substantial
improvement of the method described in [2, 3]. These improvements are summarized in [4]. The
technique is based on the two-potential equation for the distribution of wire currents. This integro-
differential equation is solved numerically, using the method of moments [5] with a polynomial
(power-series) approximation for the current distribution.
The wire structure can be placed in a free space, above a perfectly conducting ground
plane, or above real (imperfect) ground. The influence of the ground is taken into account
following Sommerfelds approach, with numerical integration algorithms developed exclusively
for this program. The wire structure can have concentrated and distributed loadings. If the
structure is an antenna, it can be driven at one or more ports in the transmitting mode, or excited
by a plane wave of an arbitrary polarization in the receiving mode. If the analyzed structure is a
scatterer, it can be excited by a plane wave.
The program first evaluates the current distribution by computing polynomial
coefficients. Based on these coefficients, input characteristics of antennas (port impedance,
admittance, and scattering parameters) are evaluated, as well as near fields and radiation patterns
for transmitting antennas, and induced currents and voltages for receiving antennas. For scatterers,
monostatic and bistatic cross sections are computed.
The analysis method requires only a small number of polynomial coefficients (about 5 to
10 per wavelength). Owing to the particular method used to solve the two-potential equation, as
well as to special techniques used for numerical integration, AWAS for Windows is a very
+
2 Analysis of Wire Antennas and Scatterers
efficient and accurate program, and it can analyze structures up to about 500 wavelengths in total
wire length.
The computer program is user friendly. It allows for easy data input, editing, and
checking. It has an automatic selection of the approximation functions on wires. It also has ample
graphic capabilities at the output. The kernel of the program can be used on a standalone basis, or
it can be incorporated into a user-written program, such as an optimization procedure for antenna
synthesis.
1.2 Intended Users
The program is intended to be used by engineers involved in the research and design of wire
antennas (from the low-frequency range to microwaves), as well as those dealing with radio
engineering, microwaves, radar techniques, electromagnetic compatibility, and communications.
The program can also be used for teaching purposes, primarily for the postgraduate level.
Users are assumed familiar with the basic antenna theory and with the basic commands
and procedures of Windows.
1.3 Program Environment and Distribution
AWAS for Windows is in an executable form, designed to run on IBM 80486, Pentium, and other
100% compatible computers, running Windows NT, 98, or a later version of the operating system.
The software is distributed on one CD. AWAS for Windows should be installed and run in a
Windows environment. See Chapter 2 for further information.
1.4 Program Usage
The program is user friendly and designed in accordance with Windows standards. Using a
dedicated editor, one can supply data defining the wire structure to be analyzed, as well as
frequency, excitation, near field points, and far field directions. All data are checked for
consistency. A graphic presentation of the structure is also supplied. Errors in data entries can be
easily corrected. The input data can be stored to be re-edited for the next run, if desired.
The program output consists of the listing of the coefficients of polynomials
approximating the current distribution, the input characteristics if an antenna is analyzed, the near
electric and magnetic fields produced by the currents and charges of the wire structure, the
radiation pattern if a transmitting antenna is analyzed, the cross section pattern if a scatterer is
analyzed, and so on. Information contained in the output data file can be tailored to users
requirements. In addition, plots of various quantities (current distribution, input characteristics,
near fields, and far fields) are generated on the screen. They can also be printed, if desired.
1.5 Packaging List
The AWAS for Windows package consists of this book (Users Manual) and one CD.
+
Introduction 3
1.6 Contents of the Manual
Chapter 2 provides detailed information on hardware and software requirements. Chapter 3
follows with instructions on how to install AWAS for Windows. Chapter 4 gives instructions on
how to get started with the program and provides a quick tour through the program. Chapter 5 is a
tutorial guide through AWAS for Windows. Chapter 6 summarizes the theoretical background of
the analysis method. For additional details, the reader can consult [25]. This chapter can be
omitted without loss of continuity, as AWAS for Windows can be used without knowledge of the
analysis method. Chapter 7 is a reference guide, providing detailed instructions for using AWAS
for Windows. Chapter 8 gives several examples of program execution, illustrating various wire
structures that can be analyzed and demonstrating the performance of AWAS for Windows.
Appendix A summarizes program limits and input data restrictions. Appendix B lists error
messages. Finally, Appendix C presents Help windows.
1.7 If You Use Previous Versions of AWAS
In 1990, Artech House published a version of AWAS designed for DOS [1], which will be briefly
referred to as AWAS for DOS. In 1995, a Windows version was published, AWAS for Windows
Version 1.0 [6], which was an upgraded and substantially improved version of AWAS for DOS.
Version 1.0 is a 16-bit code and is compatible with Windows 3.0 and later. The present program is
AWAS for Windows Version 2.0. Compared with Version 1.0, one of Version 2.0s main
improvements is its 32-bit code, resulting in substantially increased performance. Version 2.0,
however, is compatible only with Win32 platforms (i.e., with computers running under Windows
NT, 98, and higher versions of the operating system). This section summarizes similarities and
differences among the three versions of AWAS. For additional information, refer to Appendix A.
Version 2.0 includes other major improvements, as follows. Sommerfelds formulation is
included to treat the real ground. The system of linear equations in Version 2.0 is solved more
efficiently than in Version 1.0, utilizing partial sparsity of the system matrix. The shell of Version
2.0 has been redesigned and improved, in particular the graphics output.
AWAS for Windows Version 2.0 has larger limits than the previous two versions, as
summarized in Table A.1 in Appendix A. Thus, Version 2.0 can accept structures with up to 999
nodes, 999 segments, and 99 ports, as compared to 500 nodes, 320 segments, and 16 ports in
Version 1.0. Version 2.0 can handle about five times more unknowns than Version 1.0. Thus,
Version 2.0 can handle wire structures up to about 500 wavelengths in total wire length, as
compared to 100 wavelengths for Version 1.0.
In Version 2.0, the optimal degrees of the polynomial approximation are automatically
determined based on the accuracy parameter defined in the configuration. Version 2.0 has a wider
selection of loading impedances, including series and parallel RLC circuits, transmission-line
stubs, and skin-effect conductor losses. Version 2.0 can handle not only ideal voltage generators as
in the previous versions, but also ideal current generators and matched generators.
If you want to install AWAS for Windows Version 2.0 on the same computer on which
you have already installed a previous version of AWAS, we suggest placing Version 2.0 in a
separate folder (directory). The three versions of AWAS have files with identical names, which
might cause a conflict if the programs shared the same folder.
All three versions of AWAS use similar basic analysis methods for wire antennas and
scatterers. The basic functions of the programs are also alike. Hence, there is a certain degree of
+
4 Analysis of Wire Antennas and Scatterers
compatibility of input data files for the three programs. Input data files used in AWAS for DOS
and AWAS for Windows Version 1.0 can be used in Version 2.0 after a modification by an
enclosed converter program. There is no backward compatibility of input files. The output files are
incompatible.
1.8 If You Have Questions or Comments
If you have any questions or comments, please write or send a fax to:
Artech House, Inc.
685 Canton Street
Norwood, MA 02062-2610, USA
Fax: + 1 781 769 6334, Phone: + 1 781 769 9750
Alternatively, you may write, send a fax, or e-mail directly one of the authors:
Professor Dr. Antonije Djordjevi
School of Electrical Engineering, University of Belgrade
P.O. Box 816, 11001 Belgrade, Yugoslavia
Fax: + 381 11 324 8681, Phone: + 381 11 340 77 11
E-mail: edjordja@etf.bg.ac.yu
Mr. Miodrag Badar
19 Rue George Sand
75016 Paris, France
Fax: + 33 1 42 30 73 16, Phone: + 33 1 45 25 21 63
E-mail: MBazdar@csi.com
Assist. Professor Dr. Vladimir Petrovi
School of Electrical Engineering, University of Belgrade
P.O. Box 816, 11001 Belgrade, Yugoslavia
Fax: + 381 11 324 8681, Phone: + 381 63 866 27 41
E-mail: vpetrovic@telekom.etf.bg.ac.yu
Mr. Dragan Olan
School of Electrical Engineering, University of Belgrade
P.O. Box 816, 11001 Belgrade, Yugoslavia
Fax: + 381 11 324 8681, Phone: + 381 11 321 83 53
E-mail: olcan@etf.bg.ac.yu
Professor Dr. Tapan Sarkar
Syracuse University, Department of Electrical and Computer Engineering
Syracuse, NY 13244-1240, USA
Fax: + 1 315 443 2583, Phone: + 1 315 443 3775
E-mail: tksarkar@MailBox.syr.edu
+
Introduction 5
Professor Dr. Roger F. Harrington
5424 N Strada de Rubino
Tucson, AZ 85715, USA
Phone: + 1 602 299 1086
E-mail: rfh491@cs.com
+
7
Chapter 2
Hardware and Software Requirements
AWAS for Windows can be installed and run on any computer configuration running under
Windows NT, 98, or later operating system. This includes IBM or compatible computers with
Pentium and other compatible processors.
To achieve the optimal performance of the program, it is highly recommended that it be
run on a minimal configuration consisting of the following hardware:
IBM or 100% compatible personal computer with a Pentium processor;
64 MB of available RAM;
100 MB of available hard disk space;
A graphics card and a monitor allowing a resolution of 800600 pixels (with 96 dpi font);
Mouse or trackball;
Printer attached to the computer;
CD drive.
It is assumed that Windows is properly installed, including the support for the monitor,
mouse, and printer. The font Symbol should be available for displaying Greek symbols in dialog
boxes. The fonts Terminal and Courier should be available for making printouts. Please refer to
the Windows documentation for further information.
AWAS for Windows can only be installed on individual computers and not on a network.
9
Chapter 3
Installing AWAS for Windows
AWAS for Windows is distributed on one standard CD, which contains all files necessary to
install and run AWAS for Windows in the Windows environment. The user is assumed familiar
with basic commands of Windows.
To install AWAS for Windows, perform the following steps:
Start Windows.
Insert the AWAS CD into a drive.
Run SETUP.EXE on this CD from the Start menu in the following way:
Open the Start menu by clicking the Start button on the Windows taskbar.
Select Run... in the Start menu. The Run dialog box opens.
Type D:\SETUP if the distribution CD is, for example, in drive D:, as shown here.
Click the OK button.
While the installation program is loading, the Setup preparation dialog box is opened, as
follows.
15
Chapter 4
Getting Started
After AWAS for Windows has been successfully installed following the instructions in Chapter 3,
it is ready to run in the Windows environment.
4.1 Starting and Stopping the Program
To start AWAS for Windows, locate the AWAS for Windows icon and double click it with the
mouse. The Main window opens, as shown here.
Locate the menu bar at the top of the window:
Getting Started 17
Contents opens the Table of Contents.
Using Help opens Windows instructions on how to use Help.
The Table of Contents can also be accessed by clicking the Contents button in a Help
window.
At the bottom of each Help window, there is a list of cross-references (hotspots) labeled
See also. These entries are titles of Help windows that may contain additional useful information
related to the window that you are currently examining.
To close the Help window, you can perform one of the following steps:
Press Alt+F4.
Click the Help window menu bar to open the File menu, and select the Exit option.
Double click the Help window Control-menu box.
4.3 Quick Tour Through AWAS for Windows
Welcome to AWAS for Windows. This section will take a few minutes of your time to show you
the basic features of the program. It is assumed that you have installed AWAS for Windows
together with sample files, following the instructions of Chapter 3, and that you have read Sections
4.1 and 4.2.
It is also assumed that you have not modified the setup data delivered on the distribution
CD. If you have modified these data, see Section 5.2 for further information how to establish the
same setup as delivered on this CD.
4.3.1 Starting the Program
Locate the AWAS for Windows icon. Start the program by double clicking the icon. The Main
window is displayed.
4.3.2 Tour Through Help Windows
Click the Help button in the Main window toolbar. The Help window opens with the Help window
entitled Main window, as follows.
Getting Started 19
Move the mouse cursor to the entry AWAS and click it. The Help window entitled AWAS opens,
as shown next.
+
20 Analysis of Wire Antennas and Scatterers
If you are not familiar with basic capabilities of AWAS for Windows, read the text in this
window. Use the scroll bar (on the right edge of the Help window) or press the PgUp, PgDn, and
arrow keys to move through the window. Locate the entry GENERAL at the bottom of the
window and click it to see additional information on the program.
If you wish, you can further browse Help windows using the Contents, Back, and History
buttons.
Press Alt+F4 to close the Help window and return to the Main window.
4.3.3 Tour Through Input Data Window
Click the Input button in the Main window toolbar. The Input Data window opens simultaneously
with the Show window, as shown. (The actual layout depends on the setting of the size and
positions of these two windows in the previous run of AWAS for Windows.)
Getting Started 21
The Input Data window is used to define all data necessary to specify the geometry of the
wire structure analyzed, operating frequency, excitation(s), and so on.
To fetch data from input files located on the hard disk, open the File menu in the Input
Data window and select the Load option. The Load Geometry dialog box opens with the File
Name menu, which contains a list of files describing the geometry of wire structures, as shown
next.
Getting Started 23
The wire structure consists of a set of straight-line wire segments. Each segment connects
two points in space (nodes), and each node is defined by its Cartesian coordinates ( , , ) x y z . The
data for nodes appear in the left table (the Node table) of the Input Data window. The data for
segments appear in the right table (the Segment table) of the Input Data window. Geometrically, a
segment is defined by the indices of two nodes connected by that segment (labeled Node a and
Node b, respectively) and by its radius. A wire segment can carry loadings (concentrated or
distributed). One or even two antenna ports can be localized on a wire segment. Descriptions for
loadings and ports are embedded into the Segment table.
The Input Data window also shows information on the operating frequencies, the
operating mode of the antenna or scatterer, the ground, plane waves exciting receiving antennas
and scatterers, points at which near fields are to be calculated, and directions at which far
Getting Started 25
In this box, the user can select the quantity to be plotted (by clicking the corresponding
tab), from which data file (in the Data Source area, in the lower part of the box), in which format
(in the Format area), and so on.
Select the tab labeled Z Parameters to plot the input impedance of the antenna that was
just analyzed. These data are located in the file WIRE.ZPM. The name of this file should appear in
the Data Source area in the File Name box.
In the present example, available for plotting is only one impedance parameter, z( , ) 11 , as
a function of frequency. In the Format area, the radio button labeled Real and Imaginary part
should be pushed in, which means that the impedance will be plotted as the real and imaginary
part versus frequency. Click OK at the bottom of the Graph Selection dialog box to produce the
following plot in the Plot window.
27
Chapter 5
Tutorial Guide
5.1 Introduction
During the quick tour in Section 4.3, you must have been eager to learn more about AWAS for
Windows. This chapter will show you the basic principles and commands for using the program.
After completing this chapter, you should be ready to use the program on your own.
It is assumed that you have installed AWAS for Windows together with sample files,
following the instructions in Chapter 3, and that you have read Chapter 4. Start AWAS for
Windows and follow instructions given below. The Main window opens as shown in Section 4.1.
At any time, you may use the Help utilities, as described in Section 4.2.
5.2 Setup
You must be sure that the setup data are the same as delivered on the distribution CD, so that the
program will behave as described in this chapter. To do so, first examine the Setup dialog box.
Open this box by clicking the Setup button in the Main window toolbar. The box should look as
shown below.
Tutorial Guide 29
display the analysis results in textual form, while the other files are used to produce graphs of
output results. In the full form, the file WIRE.OUT can contain an extensive listing of input data,
as well as of all intermediate and output data. To reduce the size of this file, listing unnecessary
items can be disabled, while listing other items can be enabled. This selection is made by clicking
appropriate check boxes. For the present purpose, leave all five boxes checked.
Enabling the last item, automatic show of the wire structure, provides a permanent on-
line display of the wire structure, which is automatically updated as input data are modified. If this
option is disabled, the display of the wire structure is still available per request. Leave this option
enabled. Check again if the Setup dialog box looks as shown above. Correct any discrepancies, if
necessary. Click OK to close the Setup dialog box and return to the Main window.
5.3 Basic Principles of Analysis of Wire Structures
Before continuing, you should become acquainted with the basic principles of analyzing wire
antennas and scatterers in AWAS for Windows. The program can analyze arbitrary structures that
are assembled from straight-line wire segments (of circular cross sections). Each wire segment is
assumed to have its own radius.
Each segment connects two points in space, referred to as nodes. The starting node
(origin) of the segment will be referred to as node a, and the terminating one as node b. Each node
is defined by the three Cartesian coordinates ) , , ( z y x .
Hence, the geometry of the wire structure can be described to the program by first
specifying a set of nodes and then specifying each segment by supplying indices of the starting
and terminating nodes. In addition, the wire radius is required for each segment. Possible
concentrated or distributed loadings (impedances) as well as ports (for antennas) are also required
to describe the wire structure.
The wire structure can be isolated in space, located above a perfectly conducting ground
plane, or above a real ground. In the latter two cases, the surface of the ground coincides with the
Cartesian xy-plane,
An antenna can be driven at its ports (in the transmitting mode) or illuminated by an
incoming plane wave (receiving mode). A scatterer is assumed excited by an incoming plane
wave.
The analysis is performed in the frequency domain, at a set of equispaced frequencies.
For each frequency, the program sets an integro-differential equation (the two-potential equation)
for the distribution of currents along wire segments. This equation is numerically solved using the
method of moments. Details about the analysis method can be found in Chapter 6. However, the
only fact about the technique used in the program you need to know is that the current distribution
along each wire segment is approximated by a polynomial in terms of a local coordinate measured
along the segment.
Based on the numerically evaluated current distribution, the program calculates various
properties of antennas and scatterers, such as antenna port matrix parameters (scattering,
impedance, and admittance parameters), near electric and magnetic fields produced by wire
currents and charges, and far (radiated) fields.
A computation cycle normally consists of specifying input data, performing the
numerical analysis (by the program kernel for numerical analysis), and examining output results in
textual or graphic form.
Input data can be edited in the Input Data window, which can be accessed by clicking the
Input button in the Main window toolbar. These data can be saved in a set of five files referred to
Tutorial Guide 31
5.4.3 Load an Input Data Set
In the Input Data window, open the File menu and select the Load option. The Load Geometry
dialog box opens with the File Name menu containing a list of all GEO files, as shown in Section
4.3.3. These files reside in the working directory (the directory in which AWAS for Windows is
installed). However, you have an option of changing the drive and the directory when you search
for files to be loaded.
Select EXAMPLE.GEO in the File Name menu, and then click the Open button to load
the selected input data set and close the Load Geometry dialog box. You should see the same
Input Data window with the Show window as shown in Section 4.3.3.
Each GEO file contains a header, consisting of three lines, each one 76 characters long.
These lines can contain useful information on the structure. After you have loaded a file, you can
examine the contents of this header by clicking the Header button at the bottom of the Input Data
window. The Header Edit dialog box opens. The header for the file EXAMPLE.GEO looks as
follows.
Click the OK button to close the Header Edit dialog box
The loaded structure is an antenna, with no ground, defined by seven nodes and six
segments (two segments for the monopole antenna and four segments for the counterbalance). The
antenna has one port and one concentrated loading. The analysis of the antenna is requested in the
transmitting mode, at a total of nine frequencies. The radiation pattern of the antenna is requested
in the three Cartesian coordinate planes.
5.4.4 Parts of Input Data Window
The Input Data window has a menu bar:
with the menus File, Edit, Compute, and Help. It has two tables: the Node table:
Tutorial Guide 33
It has the Operating Mode button:
the Ground button:
and 10 buttons at the bottom of the window:
Click the menu bar to open the File menu. The menu has the following options:
Load (Ctrl+L) load an input data set.
New (Ctrl+N) reset input data in program memory.
Save (Ctrl+S) save an input data set with the name of the loaded set.
Save As (Ctrl+A) save an input data set with a new name.
Save Default Files (Ctrl+D or Ctrl+F8) save frequency, plane wave, near field, and far field
data as defaults; same as the SavDef button.
Show (Ctrl+F1) show geometry; same as the Show button.
Exit (Ctrl+X) close the Input Data window (and its pending Show window) and return to
the Main Window.
Click the menu bar to open the Edit menu. This menu is used to facilitate editing data in
the Node table and the Segment table. The Edit menu has the following options:
Copy (Ctrl+C or Ctrl+Ins) place the selected text in the paste buffer.
Cut (Ctrl+X or Shift+Del) cut the selected text and place it in the paste buffer.
Paste (Ctrl+V or Shift+Ins) paste the text from the paste buffer.
Insert Line (Ctrl+F5) insert a line in the active table; same as the InsLine button.
Delete Line (Ctrl+F6) delete a line in the active table; same as the DelLine button.
Undelete Line (Ctrl+F7) undelete the last deleted line in the active table; same as the Undel
button.
Header (Ctrl+F3) edit header; same as the Header button.
Mode (Ctrl+F4) change operating mode and ground; same as the Mode button.
Click the menu bar to open the Compute menu. The menu has the following options:
Check (Ctrl+F9) check input data; same as the Check button.
Run (Ctrl+F10) perform the analysis; same as the Run button.
Click the menu bar to open the Help menu. The menu has the following options:
Help open the Help window for the Input Data window.
Index open the Help window with the Index of available Help windows.
Contents open the Help window with the Table of Contents of available Help windows.
Tutorial Guide 35
5.4.5 Show Window
Now move the mouse cursor to the upper half of the screen, where the Show window is displayed,
and click the Window Title bar. The Show window opens fully. In its lower portion, five
thumbwheels and three buttons are located, as shown here.
The three left-most thumbwheels can be used to rotate the sketch of the structure about the vertical
(RotH), horizontal (RotP), and the perpendicular axes of the screen (RotB). The next thumbwheel
(Step) changes the step of rotation. Try to rotate the sketch using these thumbwheels. The
structure can be returned to the original position by rotating in a backward sequence.
Alternatively, click View in the menu bar, and then select the Reset option.
The scale of the display can be changed using the thumbwheel Scale(%). Increase the
scale to 400%. You can use the mouse and the Up and Down arrow keys to change the scale in
10% steps or PgUp and PgDn keys to change the scale in 100% steps.
Click the Nodes button to display the node indices, the Segments button to display the
segment indices, and the Ports button to display the port index. Now focus the Input Data, and
compare node and segment indices in the two tables and in the Show window.
5.4.6 Edit Loaded Data
We will now try to convert the L-antenna to a T-antenna and remove the loading. We will keep the
counterbalance. To that purpose, we have to add one node and one segment.
First, move the mouse cursor to the left-most field of the first blank line in the Node table
(the line just below the line for node 7). Click there to activate the field for editing. Type in 8 for
the node index. Press Tab or Enter to move to the next field, for the node x-coordinate. To make
the length of the new segment equal to the length of the already existing horizontal segment at the
monopole top, this coordinate should be m 1 . 0 . Type in 1 . 0 , and move to the next field, for the
node y-coordinate. Type in 0 for this coordinate. (Note that you must type in a zero, and not leave
the field blank, as this will result in an error.) Move to the field for the node z-coordinate. This
coordinate should be the same as for node 2 (i.e., 0.1 m). Type in 0.1.
If you leave the last field (using Tab, Enter, scroll bar, mouse, and so on), you will note
that the sketch of the wire structure is updated by displaying the label of node 8 close to the actual
location of this node.
You can use standard Windows procedures to delete a character while editing data in a
field (backspace or Del keys), select one or several characters by dragging the mouse cursor over
them, delete the selection (Ctrl+X), copy the selection (Ctrl+C), paste (Ctrl+V), or move the
cursor (left and right arrows, mouse). In addition, you can use the Edit menu instead of the
keyboard commands.
Editing the two tables can also be facilitated by using the following keys: Ctrl+F5 (to
insert a blank line at the cursor position), Ctrl+F6 (to delete the line at the cursor position), and
Ctrl+F7 (to undelete the last deleted line). Alternatively, you can use the equivalent buttons at the
bottom of the screen, or options in the Edit menu.
Now, move the mouse cursor to the Segment table, to the first blank line (the line just
below the line for segment 6). In this line, type in 7 for the segment index. This segment should
Tutorial Guide 37
You can now become acquainted with the Port dialog box. Click the button for port a for
segment 1, which is labeled by 1. This is the only port of our antenna, located at the foot of
segment 1. The Port dialog box opens, as shown here.
This box defines port 1. The port is driven by an ideal voltage generator, whose complex
electromotive force (emf) is V ) 0 j 1 ( + = V . Converted to polar form, the rms voltage is 1 V, and
the phase is 0. This emf is the excitation of the antenna. The port nominal impedance is
= 50
c
Z .
We have the rms values here, because in the Setup dialog box we have declared complex
numbers as rms values. A voltage of rms value 2 V and phase 90
o
would be V .) 2 j . 0 ( + = V , as
the voltage is entered in rectangular form (by its real and imaginary parts, and not magnitude
and phase). The generator is an ideal voltage generator. Hence, the generator electromotive force
equals the port voltage. For an ideal current generator, the excitation would be the generator
current (in amperes). For ideal generators, the port nominal impedance is used solely for the
evaluation of the antenna scattering (s) parameters. However, a matched generator is a series
connection of an ideal voltage generator and a resistor whose resistance is equal to the port
nominal impedance.
Click OK or Cancel to close the Port dialog box. Note that you can remove a port if you
specify its index to be 0.
Now click the Header button to open the Header Edit dialog box, which appears as shown
in Section 5.4.3. Edit the text in the box to reflect changes in the antenna that you have just made:
the antenna is now a T-monopole, and it is not loaded. Click OK to close the box.
5.4.7 Operating Mode and Ground
A wire structure can be treated in AWAS for Windows as an antenna or as a scatterer. An antenna
has at least one port. A scatterer has no ports. Both structures can have concentrated and
distributed loadings.
.
38 Analysis of Wire Antennas and Scatterers
An antenna can be analyzed in the transmitting mode, when it is driven at its ports. The
program can analyze the antenna simultaneously driven at all ports, for the specified port
excitations. This case is useful for the analysis of radiation pattern of antenna arrays. The program
can also analyze the antenna driven one port at a time (the excitations at nondriven ports being set
to zero).
An antenna can also be analyzed in the receiving mode, when it is excited by an incident
(illuminating) plane wave. Thereby, the program sets to zero the excitations at antenna ports.
A scatterer is always excited by a plane wave. The evaluation of the monostatic or
bistatic scattering is usually the primary objective of the analysis. Wire cages and other similar
shields can also be analyzed in this mode, and the objective is usually to find the electric and
magnetic fields within the enclosure.
The program has the following operating modes available:
Transmitting/receiving antenna (combines the antenna transmitting mode, one port excited at
a time, and the receiving mode);
Transmitting antenna, one port excited at a time;
Transmitting antenna, all ports simultaneously excited;
Receiving antenna;
Scatterer, monostatic scattering;
Scatterer, bistatic scattering.
Click the Operating Mode button in the middle of the Input Data window to open the
Operating Mode dialog box with the Operating Mode menu, as shown next.
Leave the transmitting mode (one port at a time) and click OK or Cancel to close the Operating
Mode dialog box.
The Operating Mode dialog box also opens upon clicking the Mode button at the bottom
of the Input Data window. In this case, after closing this box, the Ground dialog box opens, in
which you can define the ground, as shown below.
Tutorial Guide 39
You can select no ground, a perfectly conducting (PEC) ground plane, or a real ground. In the
latter case, you have to enter the parameters of the ground (its conductivity and relative
permittivity). The real ground can be analyzed in two ways. The first way is to treat it as a PEC
plane when evaluating the current distribution and use the reflection coefficients for plane waves
when evaluating the far fields. The second way is to account for the ground using Sommerfelds
approach. The selection for the ground can also be made by clicking the Ground button in the
middle section of the Input Data window.
A receiving antenna or a scatterer is assumed excited by an incident plane wave. If the
ground is present, the incident wave is reflected from it, and the wire structure is excited by both
the incident and the reflected waves.
5.4.8 Frequency
The wire structure is analyzed at a set of equispaced frequencies between the first frequency,
f
start
, and the last frequency, f
stop
, that is, at
f f i f i n
i f
= + =
start
, ,..., 0 , (5.1)
where n
f
is the number of frequency steps, and the frequency step is evaluated as
f
f f
n
f
=
stop start
. (5.2)
The analysis is performed at a total of ( ) n
f
+1 frequencies. Thus, n
f
= 0 means that the
analysis is to be performed at one frequency (when f f
start stop
= ). Hence, 0 is the minimal
allowed number of frequency steps. If you want to plot results of the analysis against frequency,
the number of frequency steps must be at least 1. The maximal number of steps that can be
handled by the program is 999.
Data for frequency are displayed and edited in the Input Data window Frequency area, as
shown here.
and n
stop start
n
(5.5)
and
=
stop start
n
(5.6)
are angular steps. A pair ( , )
i j
can be considered as a point in the plane whose coordinate axes
are the angles and .
The minimal number of steps is 0 and the maximal is 999. Once the analysis of the wire
structure is done, far fields can be plotted within a group against one angular coordinate as
variable, with the other coordinate constant. If you want to plot far fields against one angle, the
number of steps for that angle must be at least 1.
and
E
) are evaluated in the analysis. These components are presented in the output results multiplied
by the distance r from the antenna and the field phase is reduced to the coordinate origin, so that
the results do not depend on the spherical coordinate r. For antennas in the transmitting mode, the
power gain is also evaluated. For scatterers, the monostatic or bistatic cross sections are evaluated
in addition to the electric field components.
The limits for angles and are 180 and 180 , except when the ground is present. In
the latter case, can vary between 90 and 90 . Note that = 180 and = 180 both
correspond to the negative direction of the z-axis, while = 0 coincides with the positive
direction of the z-axis.
Let us now examine the loaded data. The far field directions are organized in two far field
groups. The first group can be used to plot the antenna radiation pattern in the zx-plane (when
=
start
= 0 ), and in the yz-plane (when =
stop
= 90 ). In both cases, varies in 10 steps.
The second group can be used to plot the radiation pattern in the xy-plane.
Although data in the Far Field dialog box can be edited, we shall skip editing in this
tutorial and only explain the purpose of the buttons at the bottom of the box. The OK button is the
command to close the box and keep the edits. The Cancel button is the command to close the box,
but discard the edits. The Load button opens the Load Far Field dialog box with the File Name
menu with names of available FFP files. A file can be selected and its data loaded into the Far
Field dialog box. The Save button opens the Save Far Field dialog box in which the name of an
FFP file is to be given, and the program then saves far field input data in that file. The Default
button saves the contents of the Far Field dialog box in the DEFAULT.FFP file.
Close the Far Field dialog box by clicking the OK or Cancel buttons.
5.4.10 Near Fields
Near electric and magnetic fields produced by currents and charges of the wire structure (with the
addition of the plane wave fields if the structure is excited by a wave) can be evaluated by AWAS
for Windows for a set of points defined by Cartesian coordinates ( , , ) x y z . These points can be
organized in 1 to 9 near field groups. Data defining these groups can be accessed by clicking the
corresponding button for near field groups in the Waves and Fields area.
Coordinates of points within a group are
x x i x i n
i x
= + =
start
, ,..., 0 , (5.7)
y y j y j n
j y
= + =
start
, ,..., 0 , (5.8)
z z k z k n
k z
= + =
start
, ,..., 0 , (5.9)
where x
start
, y
start
, and z
start
are the minimal values of the coordinates; x
stop
, y
stop
, and z
stop
are the maximal values of the coordinates; n
x
, n
y
, and n
z
are numbers of coordinate steps in
appropriate directions; and
x
x x
n
x
=
stop start
, (5.10)
Tutorial Guide 43
y
y y
n
y
=
stop start
, (5.11)
and
z
z z
n
z
=
stop start
(5.12)
are coordinate steps.
The minimal number of steps is 0, and the maximal is 999. Once the analysis of the wire
structure is done, near fields can be plotted within a group against one coordinate as variable, with
constant two other coordinates. If you want to plot near fields against one coordinate, the number
of steps for that coordinate must be at least 1. At the specified points, the Cartesian components of
the electric field ( E
x
, E
y
, and E
z
) and components of the magnetic field ( H
x
, H
y
, and H
z
)
are evaluated. There is no limit for the coordinates, except that z must be nonnegative if the
ground is present.
Loaded data contain no near field groups (i.e., the near field is not requested in the
analysis). Take a look at the Near Field dialog box by clicking the corresponding button for near
field groups in the Waves and Fields area. Input data for near fields can be edited in this box, as
shown here.
The Near Field dialog box is similar to the Far Field dialog box and it does not require
additional explanation. Close the Near Field dialog box by clicking the OK or Cancel buttons.
5.4.11 Plane Waves
A receiving antenna or a scatterer is excited by an incident plane wave (and by the reflected wave
if the ground is present). In AWAS for Windows, the analysis can be performed for a set of
.
44 Analysis of Wire Antennas and Scatterers
incident plane waves, assuming the structure is illuminated by one wave at a time. These waves
are organized in 1 to 8 plane wave groups. Data defining these groups can be accessed in the Plane
Wave dialog box by clicking the corresponding button for plane wave groups in the Waves and
Fields area. In this box, plane wave input data can be edited, as shown here.
A plane wave is defined by the direction of incidence (i.e., by the spherical angles, and
), and by the (complex) components of the electric field, E
and E
Tutorial Guide 45
5.4.12 Checking
Extensive checking of input data is performed at two points in the program. The first point is at the
Input Data window. Data checking is performed upon the users request (by clicking the Check
button, pressing Ctrl+F9, or selecting the Check option in the Compute menu) or automatically
before data are saved and before the analysis is started. The user is warned about detected errors
and is expected to correct them before data are saved or the analysis is performed.
The second point of an even wider error checking procedure is in the program kernel for
numerical analysis (AWAS_W_2.DLL). If an error is detected, the analysis is canceled, and an
error message is written to the main output data file WIRE.OUT. Some of these errors, which are
the same as those detected in the Input Data window, are detected immediately after starting the
program kernel. This double-checking is required for cases when the input data are created
externally (by other programs) and the analysis is performed by the program kernel of AWAS for
Windows. Other errors can be detected only during the analysis. For example, with increasing
frequency, the total number of polynomial coefficients increases due to automatic segmentation,
and it can exceed the program limit of 3,000 coefficients.
See Appendix B for a complete list of error messages produced by these two checking
procedures.
Additional checking is provided at data input. For example, in numeric entry only
decimal digits, the decimal point, and the minus sign can be entered. The number of frequency
steps cannot exceed 999, and the number of plane wave, near field, and far field groups cannot
exceed the maximal one (8 or 9).
Before a file is saved, it is checked to see if a file with the same name already exists. If it
does, a dialog box opens prompting the user to decide whether to overwrite the old file. If you
want to load a new input data set over existing data in memory or you want to terminate AWAS
for Windows without saving input data, a dialog box opens asking whether or not the requested
action should be performed.
To see how checking works, let us first see if the data we have already edited are correct.
If you have followed all steps in this tutorial chapter, you should see the following Input Data.
Tutorial Guide 47
Change the index of the first node for segment 2 from 2 to 8 and click the Check button.
The message Segments 1 and 2 too close or intersect appears in the Error dialog box, as shown
here.
The problem now is that segment 2 extends from node 8 to node 3, and the end point of segment 1
is in the middle of segment 2. The two segments must not intersect or overlap, but they can have
one common node. The program considers two segments illegally intersected if the shortest
distance between any two points on the axes of the two segments is less than the sum of segment
radii.
Correct the above mistake by typing in 2 instead of 8 for the first node of segment 2 and
click the Check button. The data should now be correct. Click OK.
We are ready to continue.
5.4.13 Save Input Data
An input data set (consisting of five files with the same name and extensions GEO, FRQ, PWE,
NFP, and FFP, respectively) can be saved in the Input Data window by selecting the Save or Save
As option in the File menu. In the first case, the name of the set is the same as that of the loaded
input data set. If no data have been loaded, the user supplies a valid file name. In the second case,
the user is prompted to supply a name for the set. The same options are also available in the Input
menu in the Main window. However, we recommend that the user performs input data saving
from the Input Data window, as all data checking and editing facilities are readily available there.
Input data files with extensions FRQ, PWE, NFP, and FFP can be saved individually by
clicking the Save button in the Frequency area, the Plane Wave dialog box, the Near Field dialog
box, and the Far Field dialog box, respectively. By clicking the Default button in the
corresponding area or dialog box, each of these files can be saved individually as a DEFAULT
file. By clicking the SavDef button, pressing Ctrl+F8, Ctrl+D, or selecting the Save Default Files
option in the Input Data window File menu, all four files can be saved as the DEFAULT files.
If a file with the same name already exists, you will be asked whether to replace it with
the new file or to change the file name. The DEFAULT files are always overwritten without asking
this question.
In this tutorial, we will save the data for the T-monopole antenna (which we have
created) by selecting the Save As option in the File menu. The Save Input Files dialog box opens,
as shown next.
Tutorial Guide 49
Specifying a large wire structure (large in terms of the number of segments and in terms
of the total wire length in wavelengths), a large number of frequency steps, or coordinate steps in
plane wave groups, near field groups, and far field groups may result in long run times. Structures
above a real ground also require long run times, in particular when evaluating the near field at a
large distance. Hence, until you become familiar with the program, avoid such situations.
5.5.2 While the Program Kernel Is Active
While the program kernel is active, the Work window remains opened with the message
Computing... and a rotating wheel indicating that the program is busy, as shown in Section 4.3.3.
The program kernel can be interrupted by clicking the Cancel button. When the analysis
is completed, the Work window closes, the Input Data window and the Show window shrink to
icons, and the Main window becomes visible. Output data are ready for an examination or for
saving.
5.5.3 Structure of Output Data and Output Data Files
The output of the analysis consists of a variety of data, and these data are written to seven files.
The main output data file is WIRE.OUT. In the full format, it contains: all input data; data about
selected polynomial degrees and possible subdivisions of wire segment generated in the automatic
segmentation process; coefficients of polynomial approximation along all wire segments; near and
far fields produced by the structure; and scattering, admittance, and impedance parameters for
antennas. However, based on the settings in the Setup dialog box, some of these groups of data
can be absent. These settings do not affect the remaining six files.
The most basic results are the coefficients of polynomials approximating the current
distribution along the antenna. They are written to the WIRE.CUR file. A user is rarely interested
in knowing them, but these coefficients can be used to evaluate and plot the current or charge
distributions along any segment of the wire structure.
Scattering (s), admittance (y), and impedance (z) parameters for antennas are written to
the files WIRE.SPM, WIRE.YPM, and WIRE.ZPM, respectively. These parameters are not
evaluated for scatterers or for antennas in transmitting mode when all ports are excited
simultaneously.
Near electric and magnetic fields are written to the file WIRE.NFL, while far fields are
written to WIRE.FFL.
The data in WIRE.OUT can be listed from the Main window, and the data in the
remaining six files are used for producing graphic outputs of the program.
These seven output data files are temporary, and they are overwritten when a new
analysis starts. However, they can be saved (copied) for later use.
A set of the seven output data files will be referred to as the output data set.
5.5.4 Batch Analysis
If you want to analyze several antennas or scatterers and you have prepared the corresponding
input data sets, you can let the program run unattended in the batch mode. The program will read
an input data set, pass the information to the program kernel that will perform the analysis, and the
Tutorial Guide 51
Click Save button in the Main window toolbar. The Save Output Data dialog box opens
as shown here, in which you supply the name for the output data set.
Type Tmono for the name of the saved set, and click Save.
The Save Output Files dialog box opens next, as shown here, in which you can select
whether or not each file of the output data set will be saved. In this way, you can save an
incomplete output data set.
In this tutorial, we will save only the files WIRE.ZPM and WIRE.FFL, which have resulted from
the analysis of the T-monopole (in Section 5.5.1). Hence, click the boxes for WIRE.OUT,
WIRE.SPM, WIRE.YPM, WIRE.CUR, and WIRE.NFL to deselect saving these files, and click OK
to save and return to the Main window.
Tutorial Guide 53
5.6.3 Graphic Presentation of Output Data
Analysis results (both the latest ones and saved copies) can be presented on the screen in graphic
form. First, the Graph Selection dialog box should be opened by clicking the Graph button in the
Main window, selecting the Graph Report option in the Output menu, or pressing Ctrl+G. This
box is shown in Section 4.3.4. In this box, you have to select which quantity is to be plotted, from
which source (file), in which format, for which parameters, and so on. Available for plotting are:
Scattering (s) parameters, as a function of frequency, one parameter at a time.
Admittance (y) parameters, as a function of frequency, one parameter at a time.
Impedance (z) parameters, as a function of frequency, one parameter at a time.
Current and charge distributions along wire segments, one segment at a time, at one frequency
and for one excitation.
Near fields, at one frequency and for one excitation, against one Cartesian coordinate with
two other coordinates fixed within one near field group; one Cartesian component of electric
or magnetic field at a time.
Far fields, at one frequency and for one excitation, against one angular coordinate with the
other coordinate fixed within one near field group; or against frequency with two angular
coordinates fixed; one spherical component of electric field at a time, total field, gain for
antennas, and cross section for scatterers.
The selected quantity can be plotted in various formats (real and imaginary parts, magnitude,
phase, and so on).
In this tutorial, we will first plot the input impedance of the analyzed T-monopole
antenna. Click the tab for impedance parameters. The Graph Selection dialog box looks as shown
here.
Tutorial Guide 55
In the Format area, the format of the plot is selected. For the z parameters, the available
options are Real and Imaginary part, Magnitude, and Phase. Select the magnitude by clicking the
corresponding radio button.
Now, all the necessary selections have been made for producing a plot in the Plot
window, which is initiated by clicking the OK button. The following plot is displayed in the Plot
window.
This plot can be edited or saved.
The File menu in the Plot window has the following options:
Copy copy the plot to the clipboard as a bitmap.
Save as Bitmap save the plot as a bitmap file.
Printer Setup configure printer.
Print obtain a hardcopy of the plot on the printer.
Tabulate obtain a table with data comprising the plot; this option opens a menu with the
following options:
File place the table in a file (with the default name extension TAB).
Printer print the table.
Exit close the Plot window.
The Options menu in the Plot window has the following options:
Title edit the plot title.
Scale change the scale of the plot; this option opens a menu with the following options:
Automatic restore the scale determined by the program.
Tutorial Guide 57
Click the OK button to obtain the following Plot window. The plot is normalized with
respect to the maximal value of the gain.
59
Chapter 6
Theory
6.1 Introduction
This chapter outlines the basic theory for the present analysis. The reader may skip this chapter
and proceed directly to Chapter 7, as AWAS for Windows can be used without knowledge of the
specific analysis method.
The method has been developed especially for program AWAS [1] by substantially
improving the analysis described in [2, 3]. These improvements are summarized in [4]. The reader
may consult the References for additional details. For the present version of AWAS,
Sommerfelds theory has been included to take into account the real ground, following the
approach in [3] and developing a set of integration schemes.
To begin with, we consider a structure consisting of a number of wires, sketched in
Figure 6.1, situated in free space. We assume the wires straight and interconnected in an arbitrary
way. A straight piece of wire will be referred to as a segment. (A bent wire can be approximated
by straight-line segments.) The point at which a segment is terminated (at this point the segment is
connected to another segment(s) or it represents a free end of the segment) will be referred to as a
node. The wires can be perfectly conducting, or they can be loaded by concentrated and
distributed loadings. The wire structure can have one or more ports (closely spaced pairs of
terminals), and it can be excited either by voltage or current generators at the ports or by a plane
wave incident on the structure.
The distribution of currents and charges along the wires usually cannot be evaluated
analytically, but only numerically. A variety of techniques exist for the analysis of wire antennas
and scatterers. Practically all of them are based on the solution of integral (or integro-differential)
equations for the current distribution by the method of moments [5]. Most of these techniques are
best summarized in [7, 8].
Junction
Port
Free end
Loading
O
Figure 6.1 A wire structure with the coordinate system.
Once the current distribution is known, the other quantities of interest (antenna port
parameters, radiation pattern, efficiency, gain, near fields, scatterer cross section, and so on) can
be evaluated relatively easily.
The method used in the present program is based on the two-potential equation for the
current distribution. This equation is solved by the method of moments with a polynomial
approximation for the current distribution. The following sections describe the equation and the
method that is used for its solution. They also present other fundamental equations that are used in
the program.
The steady state regime on the wire structure is assumed. The sinusoidal time-domain
quantities (voltages, currents, fields, and so on) are represented in the analysis by their complex-
domain (frequency-domain) counterparts. There are two usual ways of defining these complex
quantities. The first way that relates, for example, the sinusoidal current i t I t ( ) cos( ) = +
m
(where I
m
is the amplitude, the angular frequency, and the initial phase) to its complex
representative I is
i t I
t
( ) Re( ) = 2 e
j
, (6.1)
where Re denotes the real part. The complex number I is referred to as the complex root-mean-
square (rms) or effective value, because I equals the rms value of the current i t ( ) . Another
possibility is
i t I
t
( ) Re( ) = e
j
, (6.2)
in which case I is referred to as the complex amplitude (peak value), because I now equals the
amplitude of the current i t ( ) . The choice of one of the above definitions does not affect any of the
equations in the following sections that are linear relations between complex representatives.
However, it does affect the relations for power, as well as relations between the complex numbers
and the quantities in the time domain that these complex numbers represent.
Theory 61
6.2 Two-Potential Equation
Assume, for the moment, that the wires are made of a perfect electric conductor (PEC). (Loadings
will be introduced later in this section.) In that case, the boundary conditions require that the
tangential component of the total electric field on the wire surface be zero, that is,
( ) E E + =
i
tan
0, (6.3)
where E is the electric field produced by the currents and charges of the wire structure and E
i
is
the impressed electric field that models the excitation of the system. If the wire structure is a
receiving antenna or a scatterer, then E
i
is the field of a plane wave that is incident on the
structure. If the structure is a transmitting antenna, then it is excited at its ports. In our basic
model, we take ideal voltage generators. Each port is modeled by an impressed axial electric field
that is localized in a small region along an unbroken wire, as shown in Figure 6.2. Other kinds of
generators (ideal current generators and matched generators) are treated through circuit-theory
relations (see Section 6.4.2).
emf
(a) (b)
E
i
Figure 6.2 (a) Antenna port driven by an ideal voltage generator and (b) its representation
by an impressed electric field.
The electric field produced by the wire currents and charges can be expressed in terms of
the magnetic vector-potential ( A ) and the electric scalar-potential (V) as
E A = j grad V , (6.4)
where is the angular frequency ( = 2 f , where f is the operating frequency).
Since the wires are perfectly conducting, the currents and charges are located on their
surfaces. The potentials can be expressed in terms of the surface-current density ( J
s
) and surface-
charge density (
s
) as
S g
S
d ) ' (
4
) (
s
0
= r r J r A , (6.5)
S g V
S
d ) ' (
4
1
) (
s
0
= r r r , (6.6)
=
k
g (6.7)
is the free-space Greens function, and
k =
0 0
(6.8)
is the phase coefficient (phase constant or wave number) for a vacuum.
To avoid evaluation of the surface integrals in (6.5) and (6.6), we can do the following.
First, we assume that the wire currents have only an axial component (i.e., the circumferential
component is zero) and that this component does not vary along the wire circumference. This
approximation is valid for thin wires in most practical cases. The largest error in this assumption
occurs near wire junctions and ends, and when wire segments are placed extremely close together.
Second, according to the extended boundary conditions [9], instead of imposing the condition
(6.3) for the conductor surface, we can postulate that the axial component of the total electric field
is zero along the wire axes [3]. Now, instead of (6.4) and (6.5), the potentials along the wire axes
can approximately be evaluated as
s g s I
s
d ) ( ) (
4
) (
a
0
= r u r A , (6.9)
s g s Q V
s
d ) ( ) ( '
4
1
) (
a
0
= r r , (6.10)
where I s ( ) is the wire current (which is related to the surface-current density by I R = 2 J
s
,
where R is the wire radius), Q s ' ( ) is the wire per-unit-length charge density (which is related to
the surface-charge density by Q R ' = 2
s
), s is the local coordinate measured along the wire axis,
u is the unit vector tangential to the axis, while
r r R
a
= +
0
2 2
, (6.11)
where r
0
is the distance between the element ds on the wire axis and the observation point.
Equation (6.11) defines the approximate average distance between the observation point and the
points along the wire circumference (i.e., the average value of the term r r ' ). Equations (6.9)
and (6.10) are exact for the potentials on the axis of a cylindrical wire segment. For other points,
they yield a good approximation except in the immediate vicinity of discontinuities (junctions and
ends) of a wire structure. (In these regions, we have already introduced another error by assuming
the currents to be axially directed.) The largest error occurs for points near the wire surface.
Equations (6.9) to (6.11) have an alternative interpretation as a direct approximation to
(6.5) and (6.6) for the potentials at the wire surface. In that interpretation, the term g r ( )
a
is
referred to as the reduced or thin-wire kernel.
The current intensity I and the charge density Q' are related by the continuity equation,
Theory 63
Q s
I s
s
' ( )
( )
=
j d
d
. (6.12)
From (6.4) and (6.9) to (6.12), we obtain
s r g
s I
k
r g s s I
s
d ) ( grad
ds
) ( d 1
) ( ) ( ) (
4
j
a
2
a
0
(
= u E . (6.13)
We assume the wire structure divided into N segments. These segments often coincide
with geometrical segments. They are bounded by wire junctions, ends, and ports. However, a wire
segment can be subdivided into two or more parts to facilitate the numerical solution for the
current distribution. We assume each segment to have its local axis, s
m
, where m is the index of
the segment (m N = 1, ... , ). The axis origin is at one segment end (corresponding to s
m
= 0 ). The
axis is directed toward the other segment end (referred to as the segment terminal), which
corresponds to s h
m m
= , where h
m
is the segment length (see Figure 6.3). The reference direction
for the current coincides with the direction of the axis.
Combining (6.13) with (6.3) applied to the electric field on the wire axes, we obtain
0
i
1
0
a
2
a
j
4
d ) ( grad
ds
) ( d 1
) ( ) (
=
(
=
E u
u u
p
N
m
h
m
m
m m
m m m p
m
s r g
s I
k
r g s I , (6.14)
where p ( p N = 1,..., ) is the index of the wire segment for which the boundary condition is
imposed. Equation (6.14) is referred to as the vector-scalar-potential equation or the two-
potential equation. It is an integro-differential equation for the current distribution I s
m m
( ) along
the segments of the wire structure.
I
m
m
2R
m
h
m
s
m
u
Figure 6.3 Local coordinate system for a wire segment.
If a wire segment has concentrated or distributed loadings, the axial component of the
electric field along that segment is not zero. For a distributed loading (distributed impedance), the
total axial electric field is proportional to the product of the wire current and the per-unit-length
impedance (Z'). Instead of (6.3), we now have the following boundary condition:
u E E
p p p p
Z s I s + = ( ) ' ( ) ( )
i
. (6.15)
A concentrated (lumped) loading can be regarded as a distributed loading that extends over a short
portion of the segment. This is similar to the way an ideal voltage generator is modeled by an
impressed electric field, which extends over a short portion of the segment. A rigorous analysis [3]
shows that the width of this portion has an influence on the effects of the impedance and the
Theory 65
=
+ =
N
m
m
n N
1
u
) 1 ( . (6.17)
The expansion given by (6.16) should next be substituted into the two-potential equation (6.14).
Assuming the wire segments to be loaded (and the concentrated loadings to be represented in a
way similar to that used for distributed loadings), we obtain
| ( ) ( ) |
( ) .
j
4
j
) ( ' 4
d ) ( grad
1
) (
0
i
0
0
a
1
2
a
1 0
0
+
+
= =
E u
u u u
p
n
i
i
p
p
pi
p
m p
i
m
m
m
i
m
m
m p
N
m
n
i
h
mi
p
m
m
h
s
I
s Z
s r g
h
s
h
i
k
r g
h
s
I
(6.18)
This equation cannot be satisfied exactly at all points along the wire axes. To approximately
satisfy (6.18), following the method of moments, we choose a set of weighting (testing) functions
and evaluate inner products of (6.18) with these functions. We adopt the same kind of weighting
functions as in [10], because significant simplifications in evaluation of the electric field are
obtained. Each weighting function is constant over a subsegment of a wire, and it is zero
elsewhere. Such a function is referred to as a pulse. Some pulses completely lie along one wire
segment, while other pulses extend over two wire segments around the junction of these two
segments, as illustrated in Figure 6.5. There is no pulse in the immediate vicinity of a free end of a
wire.
Figure 6.5 A wire structure with pulses as weighting functions.
When (6.18) is multiplied by a pulse of amplitude 1, which extends between points s
p1
and s
p2
lying on the axis of segment p, and integrated along this pulse, we obtain
{ ( ) ( ) | | }
( ) ) 19 . 6 ( . d
j
4 d
j
) ( '
4
d ) ( ) (
1
d d ) (
0
i
0
0
a a
1
0
2
a
1 0
0
2
1
2
1
1 2
2
1
p
p
s
s
n
i
p
i
p
p p
s
s
pi
m s s
i
m
m
h
m
p m
i
m
m
m p
N
m
n
i
h
s
s
mi
s s
h
s s Z
I
s r g r g
h
s
h
i
k
s s r g
h
s
I
p
p
p
p
p
p p
m
m
m
p
p
+
+
= =
E u
u u
The integrals of the form
( ) , ,...,
a
d 0
0
(6.20)
are evaluated numerically, with special attention taken to achieve accurate results with a low
computation time, as described in Section 6.5. The integrals over s
p
are evaluated using the
trapezoidal rule. If the distance between the end points of the integral is greater than 1/k, the
interval ( , ) s s
p p 1 2
is subdivided appropriately and the repeated trapezoidal rule is used.
Exceptions are the integrals when m and p coincide. By integrating by parts, the evaluation of
these integrals can be reduced to the evaluation of integrals of the form (6.20), but with i running
to ( ) n
m
+1 .
An expression similar to (6.19) is used for a pulse that lies on two wire segments. In that
case, the integral over s
p
is divided into two integrals, each of them extending only over one wire
segment. The trapezoidal rule is implemented for each of these two integrals, thus involving in the
integration the junction point of the wire segments.
In the program, the pulses along a wire segment are chosen so they have equal lengths.
However, the length of a pulse that lies over two wire segments is smaller, and it is restricted to
0.3/k along each segment.
The position of a port can be at the origin or at the terminal of a wire segment (i.e., at
s
p
= + 0 or s h
p p
= ). Thus, the ideal voltage generator driving the antenna at this port lies
within the pulse that extends over the wire segment where the port is located and over the adjacent
wire segment. (A generator cannot be placed at a free end of a segment.) The integral of the
impressed field modeling this generator, given in the last term of (6.19), simply equals the
electromotive force of the generator, multiplied by ) /(j 4
0
. The reference direction for the
electromotive force is the same as for the current of the corresponding wire segment.
The possible locations of a concentrated loading are the same as for a port. Therefore, the
last integral on the left-hand side of (6.19) yields Z I
p p a
( ) 0 for a loading of impedance Z
p a
,
localized at the origin of segment p, and Z I h
p p p b
( ) for a loading of impedance Z
p b
, localized at
the terminal of that segment, divided by j
0
. A distributed loading of total impedance Z
p d
is
assumed uniform over the corresponding segment, so that its per-unit-length impedance is
p p
h Z Z / '
d
= .
Besides equations of the form (6.19), one equation is formulated for each wire junction
and free end, expressing Kirchhoffs current law (KCL) for that node. The total number of all
these equations must be equal to the total number of unknowns, given by (6.17). At a node that is
a junction of M wire segments, there are ( ) M 1 pulses that extend over pairs of segments, so
that there are ( ) M 1 equations of the form (6.19) for these pulses. With the equation expressing
the KCL for this node, this amounts to a total of M equations. Therefore, for each wire segment
( ) n
m
1 pulses are chosen that extend only over that wire segment, where n
m
is the degree of the
polynomial for that segment. The minimal polynomial degree is 1, except for segments having
both ends free. The minimal polynomial degree for such segments is 2.
If the total length of all wires is expressed in wavelengths, extensive numerical
experiments have shown that about 5 to 10 unknowns per wavelength ( 9 ... 4 =
m
n ) are required to
obtain accurate results for the current distribution, where 5 is an absolute minimum. If less than
Theory 67
about 5 unknowns per wavelength are used, the polynomials are inadequate to represent the
current distribution. To be on the safe side, 6 coefficients per wavelength should be taken as the
minimum. If a number of unknowns greater than about 10 per wavelength is used, the run time of
the program increases without much improvement in the results. However, if too large a number
of unknowns is used, the numerical results will start diverging. There are several reasons for this.
First, a large number of polynomial coefficients for a wire segment results in a large number of
weighting pulses, in which case the length of each pulse becomes small. Most methods for the
analysis of wire structures become unstable if this length becomes smaller than the wire diameter.
Second, the pulses become crowded near junctions and ends. However, in these regions the
current distribution is not properly modeled (because the actual current density is not purely axial).
Hence, the boundary conditions (especially the extended boundary conditions) cannot be satisfied
in these regions. Forcing the current distribution to do so leads to instabilities. Third, our model of
the ports assumes, essentially, that the width of the port region is infinitely small (the delta-
function generator). This model inherently introduces an infinite susceptance in parallel with the
antenna ports. Theoretically, any technique for the analysis of wire structures that implements a
delta-function generator should yield diverging results. However, if the pulse that is used for
testing the field in the generator region is not too short, this discontinuity is not pronounced, and
good results are obtained. Fourth, increasing the number of unknowns increases the size of the
resulting system of linear equations in the unknown coefficients, which increases numerical errors
in solving this system.
The present method follows the general behavior of the methods for analysis of wire
antennas and scatterers: the quality of the solution first increases with the increasing number of
unknowns. After reaching a plateau, it starts decreasing. More accurate models of the wire
discontinuities and of the ports are required to obtain more stable solutions. Of course, these
models are computationally less efficient than the usual techniques for the analysis of wire
antennas. A comprehensive treatment of this topic can be found in [3].
The program automatically evaluates optimal polynomial degrees for all wire segments
based on the above guidelines and an accuracy parameter specified in the program (in the Setup
dialog box).
The resulting system of linear equations is solved by using the L-U decomposition, which
gives the coefficients of the polynomial approximation of the current distribution. The equations
based on the KCL are sparse. To speed up the solution, one unknown can be eliminated from each
KCL equation, thus reducing the matrix size. This procedure, however, destabilizes the solution in
some cases (e.g., for electrically small structures). Hence, the elimination is avoided for accuracy
parameters 7 and greater.
Once the polynomial coefficients are known, the required electrical properties of the
antenna or scatterer can be relatively easily computed, as described in Section 6.4.
6.4 Evaluation of Antenna and Scatterer Characteristics
Once the coefficients of the polynomials approximating the current distribution along the wire
segments ( I
mi
) are known, various characteristics of the analyzed antenna or scatterer can be
evaluated relatively easily. These characteristics are the current and charge distributions of
antennas and scatterers, the impedance, admittance, and scattering parameters of antennas, near
and far fields produced by antennas and scatterers, and so on. These characteristics are evaluated
as described in Sections 6.4.16.4.5.
_
68 Analysis of Wire Antennas and Scatterers
6.4.1 Current and Charge Distributions
The current distribution along wire segment m is given by (6.16). There are ( ) n
m
+1
coefficients if the polynomial degree is n
m
. The coefficients I
mi
are complex quantities, and the
current has its real and imaginary parts. Due to the nature of the approximation, the current is a
continuous and differentiable function along the wire segment (except at the segment ends). From
(6.12) and (6.16), the charge distribution along segment m is given by
( ) Q s I
i
h
s
h
s h
m mi
m
m
m
i
m m
i
n
m
' ( ) , = < <
1
1
0 . (6.21)
It is also a continuous and differentiable function along the wire segment.
6.4.2 Impedance, Admittance, and Scattering Parameters of Antennas
If the structure analyzed is an antenna, it must have at least one port. An antenna array is
considered a multiport antenna. The present technique properly takes into account all the coupling
between the array elements. With respect to its ports, the antenna can equivalently be represented
as shown in Figure 6.6. In the transmitting mode, the antenna is excited at its ports by certain
generators, and the output is the radiated (far) fields. With respect to the generators, the antenna
behaves like a passive network.
Passive
network
I
1
V
1
I
2
V
2
I
P
V
P
Radiated
fields
Figure 6.6 Equivalent network representation of a multiport antenna in the transmitting
mode.
This passive network can be characterized by its admittance, impedance, or scattering
parameters. For example, if the antenna has only one port, the first parameter is the antenna input
admittance, the second parameter is its input impedance, and the third parameter is the reflection
coefficient with respect to a reference impedance (the port nominal impedance). Generally, for an
antenna with P ports, the port voltages and currents are related by
Theory 69
V z I z I z I
V z I z I z I
V z I z I z I
P P
P P
P P P PP P
1 11 1 12 2 1
2 21 1 22 2 2
1 1 2 2
= + + +
= + + +
= + + +
... ,
... ,
... ,
M
(6.22)
where V V
P 1
,..., are the complex voltages and I I
P 1
,..., the complex currents at the ports (with
respect to the reference directions shown in Figure 6.6), while z z
PP 11
,..., are complex
coefficients, called the impedance (z) parameters of the network. The set of relations (6.22) can
conveniently be put into matrix form,
[ ] [ ][ ] V z I = , (6.23)
where [ ] V is a column matrix (vector) containing the port voltages,
[ ] [ ... ] V = V V
P 1
T
, (6.24)
[ ] I is a column matrix containing the port currents,
[ ] [ ... ] I = I I
P 1
T
, (6.25)
and [ ] z is a square matrix of the impedance parameters,
[ ] z =
(
(
(
z z
z z
P
P PP
11 1
1
L
M M
L
. (6.26)
In the above equations, the superscript T denotes transpose. Since the antenna is a reciprocal
network, its matrix [ ] z is symmetric.
By solving the system (6.22) in terms of the currents, we obtain
I y V y V y V
I y V y V y V
I y V y V y V
P P
P P
P P P PP P
1 11 1 12 2 1
2 21 1 22 2 2
1 1 2 2
= + + +
= + + +
= + + +
... ,
... ,
... ,
M
(6.27)
where the coefficients y y
PP 11
,..., are the admittance (y) parameters. In matrix form, we have
[ ] [ ][ ] I y V = , (6.28)
where
[ ] y =
(
(
(
y y
y y
P
P PP
11 1
1
L
M M
L
. (6.29)
For antennas, the matrix [ ] y also is symmetric.
_
70 Analysis of Wire Antennas and Scatterers
Instead of the port voltages and currents, at very high frequencies, intensities of waves
traveling at the transmission lines connected to the ports are often considered. The intensity of the
wave traveling toward the network of Figure 6.6 at port i (incident wave) is defined as
i
i i i
i
I Z V
a
c
c
Z 2
+
= , (6.30)
where Z
i c
is the nominal (reference) impedance for port i. (In many practical cases, the nominal
impedances are 50 .) The intensity of the wave traveling from the network (reflected wave) is
i
i i i
i
I Z V
b
c
c
Z 2
= . (6.31)
These intensities are related by
b s a s a s a
b s a s a s a
b s a s a s a
P P
P P
P P P PP P
1 11 1 12 2 1
2 21 1 22 2 2
1 1 2 2
= + + +
= + + +
= + + +
... ,
... ,
... ,
M
(6.32)
where the coefficients s s
PP 11
,..., are referred to as the scattering (s) parameters. This set of
equations can be written in matrix form as
[ ] [ ][ ] b s a = , (6.33)
where
[ ] [ ... ] a = a a
P 1
T
, (6.34)
[ ] [ ... ] b = b b
P 1
T
, (6.35)
and
[ ] s =
(
(
(
s s
s s
P
P PP
11 1
1
L
M M
L
. (6.36)
In our technique, we first evaluate the admittance parameters. For that purpose, we
assume port 1 to be driven by an ideal voltage generator, of electromotive force V
1
1 = V, and all
the other antenna ports are short-circuited. We now have the impressed electric field only at port
1. For this excitation, we evaluate the currents I I
P 1
,..., . Since a port can be located only at the
origin of an antenna segment (at s
m
= + 0 ) or at the terminal (at s h
m m
= ), the port current is
simply evaluated as I
m
( ) 0 or I h
m m
( ) , depending on the location of the port. From (6.27) we have
y I V y I V
P P 11 1 1 1 1
= = / ,..., / , which determines the first column of the matrix [y]. We repeat this
procedure now assuming each other port, in turn, to be driven, and all the other ports to be short-
circuited, and thus evaluate the complete matrix [y]. The matrix [z] is next evaluated as
Theory 71
[ ] [ ] z y =
1
, (6.37)
while the matrix [ ] s is evaluated from the equation
( ) ( ) [ ] [ ] [ ] [ ] [ ] [ ] [ ] s F z Z z Z F
1
= +
c c
1
, (6.38)
where [ ] F and [ ] Z
c
are diagonal matrices,
| |
[ ] diag ... F = Z Z
P c1 c
, (6.39)
| | [ ] diag ... Z
c c1 c
= Z Z
P
. (6.40)
Once the network parameters have been found, the response to the given port excitation
is evaluated. In the current version of AWAS, any generator can be an ideal voltage generator, an
ideal current generator, or a matched generator. The matched generator is a real voltage generator
whose internal impedance is resistive and equal to the port nominal impedance. Further, a
multiport antenna (or antenna array) can be driven with only one excitation present at a time (all
other excitations are zero), or with all excitations simultaneously present at all P ports. The first
option is useful for analysis of combined (multipurpose) antennas or to evaluate the performance
of one array element in the presence of other elements. The second option is typical for evaluating
the performance of complete antenna arrays. In any case, for the given port excitation, the voltages
at the ports are evaluated by combining (6.27) with equations that describe the generators. The
port voltages are then ascribed to the ideal voltage generators driving the antenna in the numerical
model. The current distribution is evaluated by a repeated solution of the method-of-moments
system of linear equations.
If the antenna is operating in the receiving mode, then, with respect to its ports, it can be
represented equivalently by a network containing generators, as shown in Figure 6.7. One possible
matrix equation describing such a network is
[ ] [ ][ ] [ ] V z I V = +
0
, (6.41)
where [z] is the same matrix as for the transmitting mode and [ ] V
0
is the vector of open-circuit
port voltages. Equation (6.41) is essentially the Thvenin equivalent representation of the antenna.
From (6.41) we obtain
[ ] [ ][ ] [ ] I y V I = +
s
, (6.42)
where [y] is the same matrix as for the transmitting mode and [ ] [ ][ ] I y V
s
=
0
is the vector of port
short-circuit currents. Equation (6.42) is the Norton equivalent representation of the antenna.
Finally, the wave intensities are related by
[ ] [ ][ ] [ ] b s a b = +
0
, (6.43)
where [s] is given by (6.38), while
( ) ] [ ] [ ] [ ] [ ] [
0
1
c 0
V Z z F b
+ = (6.44)
is the vector of wave intensities generated in the antenna and launched outwards.
_
72 Analysis of Wire Antennas and Scatterers
I
1
(a) (b)
Passive
network
Passive
network
s1
I 1
V
2
I
s2
I 2
V
I
P
V
P I
P2
1
I
V
1
V
2
V
P
I
P
I
2
o1
V
V
o2
V
oP
Figure 6.7 (a) Thvenin and (b) Norton equivalent network representations of a multiport
antenna in the receiving mode.
In our numerical method, we assume all the ports of the receiving antenna to be short-
circuited. For the given incident uniform plane wave, we evaluate the vector [ ] I
s
. In addition, we
assume the same antenna to operate in the transmitting mode, with one port excited at a time, and
evaluate the matrix [y] of the antenna. Now, according to the above equations, we have the
complete representation of the antenna.
To obtain the voltages and currents at the ports of the receiving antenna, we combine
(6.42) with equations describing the generators (i.e., an ideal voltage generator behaves like a
short circuit, an ideal current generator is an open circuit, and a matched generator is replaced by a
resistance). Another way to evaluate the state at the ports is to treat the receiving antenna like a
loaded scatterer. In this approach, each port should be substituted by an impedance equal to the
load impedance at that port. The analysis will yield the current distribution of this scatterer, and
hence the currents through the loads.
6.4.3 Power Balance
When an antenna operates in the transmitting mode, it is often important to know the average
power fed to the antenna, as well as the average radiated power. The fed power ( P
fed
) can be
determined from the voltages and currents at the antenna ports. One part of this power ( P
loss
) is
dissipated in the wire resistive loadings (if they exist) or in the wire conductors (skin-effect
losses). Another part of the fed power ( P
rad
) is radiated. The power balance thus yields
P P P
fed rad loss
= + . (6.45)
(Losses in the real ground are not included into this balance.)
The antenna efficiency is
= P P
rad fed
/ , (6.46)
Theory 73
and it can be expressed in decibels as 10
10
log dB.
If definition (6.1) is used for the complex representatives, referring to Figure 6.6, we
have for the average power fed to the antenna
P V I
i i
i
P
fed
=
=
Re( )
*
1
, (6.47)
where the asterisk denotes the complex conjugate. If definition (6.2) is used, instead of (6.47), we
have
P V I
i i
i
P
fed
=
=
1
2
1
Re( )
*
. (6.48)
For a given time-domain set of port voltages, both equations give the same results, but the
complex voltages and currents differ in the two cases by a factor of 2 .
If definition (6.1) is used, the power lost in the antenna loadings is given by
| |
P Z I Z I h
Z
h
I s s
m m m m m
h
m
m
m m m
m
N m
loss a b
d
d = + +
=
Re | ( )| | ( )| | ( )| 0
2 2
0
2
1
. (6.49)
The explanation of symbols in this equation is given in Section 6.3. If definition (6.2) is used, a
factor of 1/2 appears in front of the sum in (6.49). The integral in (6.49) can be evaluated
explicitly, because I s
m m
( ) is given by a polynomial, or numerically.
Due to numerical errors, in certain cases (e.g., electrically very short antennas) the
numerical result for the power fed to the antenna can be negative. Theoretically, it must be
positive (but it is extremely small). In addition, for very lossy antennas, the numerical result for
the radiated power can become negative.
6.4.4 Near Fields
The currents and charges induced on the wire structure (regardless of whether it is excited by
generators at its ports or by incident waves) produce their electric and magnetic fields. These
fields can be of interest in several cases, such as studying electromagnetic compatibility problems
(by modeling sources and targets of electromagnetic interference by wire structures, investigating
interference caused by transmitting antennas, or modeling shields by wire structures), evaluating
electromagnetic hazardous areas in the vicinity of transmitting antennas, or solving corona
problems of high-power antennas. The evaluation of the near fields is also a powerful tool for
estimating the quality of the numerical solution for the antenna current distribution by checking
the extent to which the boundary conditions are satisfied along wire segments.
The electric field due to the approximate current distribution can be obtained from (6.14)
and (6.16). It is given by
| ( ) ( ) |
m
i
m
m
N
m
n
i
m
i
m
m
m
h
mi
s r g
h
s
h
i
k
r g
h
s
I
m
m
d ) ( grad
1
) (
4
j
a
1
1 0
2
a
0
0
= =
+
= u E . (6.50)
curl . (6.52)
Using (6.9) and (6.16), we obtain
( )
= =
=
N
m
h
m m
i
m
m
n
i
mi
m
m
s r g
h
s
I
1
0
a
0
d ) ( grad
4
1
u H . (6.53)
The unit vector u
m
is constant along a segment. Therefore, the integrals in (6.53) have the same
form as the second part of the integrals in (6.50).
In (6.50) and (6.53), the reduced kernel g r ( )
a
is used. As mentioned in Section 6.2, this
approximation yields accurate results if the field point is on the axis of a wire segment, or if it is
sufficiently far away from it. The largest error occurs for the radial component of the electric field
and the circumferential component of the magnetic field near the wire surface. For example, (6.50)
gives a nonzero (instead of zero) radial field within the wire. At the surface of the wire, it gives a
radial field that is one half the exact one. In the program, a correction for this effect is included by
appropriately multiplying the numerical results. This correction does not yield reliable results for
field points that are close to segment ends, nor for segments of radii that are not sufficiently
smaller than the wavelength. However, our model does not accurately approximate the current
distribution near the wire ends, and it is also not valid for the segments of large radii. Hence, the
above correction is sufficient for our purposes.
For a plane wave excitation of the wire structure, the fields of the incident wave (and of
the wave reflected from the ground, if it is present) are added to the fields given by (6.50) and
(6.53).
6.4.5 Far Fields
In the case of a transmitting antenna, one of the important quantities is the radiated (far) electric
field. This is the field produced by the antenna at distances that are much greater than the maximal
dimension of the antenna and much greater than the wavelength. Locally, this field has the
properties of a uniform plane wave, traveling radially from the antenna. Therefore, the far electric
and magnetic fields are related by
H
u E
=
r
0
, (6.54)
Theory 75
where
0 0 0
= / is the wave impedance of a vacuum and u
r
is the unit vector in the radial
direction, as sketched in Figure 6.8. With respect to the coordinate system of this figure, in the
general case the vector E has only components E
and E
and H
r
. (6.55)
If definition (6.2) is used, there is a factor of 1/2 on the right-hand side in the above equation.
Hence, both the magnetic field and the Poynting vector in the far zone can be evaluated if the
electric field is known. The antenna power gain with respect to an isotropic radiator (a
hypothetical antenna that radiates uniformly into the whole space) is given by
G
P
r
p
fed
=
P
4
2
, (6.56)
where P
fed
is the power fed to the antenna and r is the distance between the antenna and the field
point. The directive gain is G G
d p
= / , where is the antenna efficiency. The gain can be
expressed in decibels as g G = 10
10
log dB. For an antenna located above a perfectly conducting
ground plane, the gain is 3 dB greater than the gain of the antenna plus its image located in a free
space.
z
x
y
O
O
s
m
h
m
m
I
u
m
m
r
m
r
u
r
u
Figure 6.8 Coordinate system for evaluation of far fields.
For a scatterer, the cross section is evaluated in the following way. Assume the scatterer
illuminated by a uniform plane wave arriving from the direction defined by the spherical angles
i
and
i
The electric field ( E
i
) of the incident wave has, generally, two components, E
i
and
E
i
. The electric field produced by the wire currents and charges (the scattered field, E
s
) is
|
\
|
=
=
j
j
j
j d
m
0
1
0
0
4
exp( )
( ) ( ) exp( )
exp( ) .
kr
r
k
I
s
h
ks s
m m
m
N
r
mi
i
n
m
h
m
m
i
m m r m
m
m
(6.58)
Here, r is the distance between the field point and the coordinate origin (which is located close to
the antenna), and r
m
is the position of the origin of the local axis s
m
of wire segment m, as shown
in Figure 6.8. The resulting integrals in (6.58) can be evaluated explicitly or numerically.
In numerical results of the program, the term [exp( ) / ] jkr r in (6.58) is moved to the
left-hand side to obtain the normalized far electric field, which does not depend on the distance r
between the observation point and the coordinate origin.
6.5 Evaluation of Integrals
A rapid and sufficiently accurate evaluation of the integrals of the form (6.20) and (6.51) is
essential to achieve a good overall performance of the program. In these equations, r
a
has the
common form
r s s t R
m mp mp m a
= + + ( )
2 2 2
, (6.59)
Theory 77
where s
mp
is the coordinate of the orthogonal projection of the field point onto the s
m
-axis, t
mp
is the distance between the field point and the s
m
-axis, and R
m
is the wire radius. The integrals in
(6.20) and (6.51) thus have poles in the complex plane at s t R
mp mp m
+ j
2 2
. These poles can be
close to the interval of integration ( , ) 0 h
m
when the field point is close to the wire segment. In that
case, the integrands in (6.20) and (6.51) have pronounced pseudo-singularities.
If the poles are closer than about 0 2 . h
m
from the interval ( , ) 0 h
m
, the following
procedure for the evaluation of integrals is implemented. First, a substitution of variables
u s s
m mp
= is introduced. Using the binomial formula, the integrals of the form (6.20) are
reduced to the integrals
J u g r u
i
i
s
h s
mp
m mp
' ( ) =
a
d , (6.60)
where
r u t R
mp m a
= + +
2 2 2
. (6.61)
The integrand in (6.60) has an i-fold zero at u = 0 that softens the influence of the pseudo-
singularity at this point. The integral J
1
' can be evaluated explicitly. Next, the pseudo-singularity
is extracted from the integrands: the term [( / ) . ] 1 0 5 r r
a a
is subtracted from the integrand of J
0
' ,
the term ( / ) u r
i
a
from the integrand of J i n
i m
' , ,..., = 2 , and the term [(1 / 0.5r
a a a
r r ) . ] + 0125
is subtracted from the integrand of K. The subtracted terms are integrated explicitly, while the
remainders are well-behaved functions that can be easily integrated numerically. For that purpose,
Gauss-Legndre integration formulas are used. The order of the formula depends on the
polynomial degree n
m
, the electrical length of the integration interval kh
m
, and the distance
between the poles and the interval ( , ) 0 h
m
.
If the poles are not too close to the interval ( , ) 0 h
m
, the integrals J i n
i m
, ,..., = 1 , and K
are evaluated numerically, using Gauss-Legndre formulas. The order of the formula is adaptively
increased if the poles are between 12 . h
m
and 2h
m
from the interval ( , ) 0 h
m
.
6.6 Antennas Above Real Ground
6.6.1 Introduction
In AWAS for Windows Version 2.0, antennas and scatterers above real ground (earth) are
analyzed using the exact, Sommerfelds approach [11, 12]. The ground is assumed to be a
homogeneous isotropic half-space of arbitrary permittivity , conductivity , and permeability
0
.
To include the influence of the ground into the two-potential equation (6.14), the free-space
Greens function should be augmented by the reflected Greens functions for the magnetic
Source point
Field point
Image point
Figure 6.9 Coordinate system for the current element above the ground.
The reflected magnetic vector-potential depends on the orientation of the current element.
For a vertical (z-oriented) and for a horizontal (e.g., x-oriented) current element, these potentials
are given by [12]
) d (
4
d
0 r
s I g
z Azz
z
u A = , (6.63)
( ) ) d (
4
d
0 r
s I g g
z Azx x Axx
x
u u A + = , (6.64)
respectively. Here,
Azz
g ,
Axx
g , and
Azx
g are the reflected Greens functions given by the
Sommerfeld integrals,
=
0
0
0
d
e ) (
0
u
R J g
Azz
Z u
Azz
, (6.65)
Theory 79
=
0
0
0
d
e ) (
0
u
R J g
Axx
Z u
Axx
, (6.66)
= =
0
0
1
b b
d
e ) ( , cos
0
u
R J g g g
Azx
Z u
Azx Azx Azx
, (6.67)
J
0
and J
1
are the Bessel functions of the first kind and order zero and one, respectively,
0
1 0 r
1 0
1 0
1 0
1 0 r
1 0 r
2 , , u
u u
u u
R
u u
u u
R
u u
u u
R
Azx Axx Azz
+
=
+
=
+
= (6.68)
are the reflection coefficients for the magnetic vector-potential,
r
2
1
2
0
, 1 = = u u , (6.69)
and ( )
0 r
/ / j = is the relative complex permittivity of the ground. The principal branch of
the complex square root function is assumed and the branch-cut is adopted on the negative part of
the real axis,
+ < = z z z arg , arg
2
1
arg . (6.70)
This representation for the potential of a horizontal current element (having x- and z-components)
is not unique. Exact fields can also be obtained by using any two, or all three, Cartesian
components [15].
For an arbitrarily oriented current element, the reflected magnetic vector-potential can be
written in a compact form,
) d (
4
) d (
4
d
0 0 r
s I I
A A
u G s G A
= , (6.71)
where
(
(
(
=
(
(
(
=
Azz Azx Azx
Axx
Axx
Azz Azy Azx
Ayz Ayy Ayx
Axz Axy Axx
A
g g g
g
g
g g g
g g g
g g g
sin cos
0 0
0 0
b b
G (6.72)
is a matrix (dyadic) Greens function for the reflected magnetic vector-potential [16] and u is a
unit vector of the current element, (i.e., the unit vector of a straight wire segment corresponding to
m
u in Figure 6.3).
Next, we need to obtain the reflected electric scalar-potential of the current element
(elementary dipole) at an arbitrary point ) , , ( z y x above the ground. This can be done by applying
the Lorentz gauge to the magnetic vector-potential and expressing the result in terms of the dipole
charges. In this derivation, we will use the normalized (electric) coordinates of the field point with
respect to the image point as the coordinate origin, ). ' ( ), ' ( ), ' ( z z k Z y y k Y x x k X + = = =
The Lorentz gauge in normalized coordinates reads
= , (6.75)
where ' s denotes differentiation in the direction of the dipole, with respect to the primed
coordinates. The dipole current is related to the dipole charge as Q I = j . After introducing the
vector Greens function for the magnetic vector-potential,
) d (
4
d
0
s I
=
A
g
A
, (6.76)
we obtain a differential equation for the reflected scalar-potential of a point charge (Q),
A
g
R
Q
s
V
div
4 '
0
0
. (6.77)
This potential can be written in terms of Greens function for the reflected scalar-potential (
V
g ),
V
g
kQ
V
0
4
= . (6.78)
This finally leads to a relation between Greens functions
A
g and
V
g ,
A
g
R
V
s
g
div
'
=
. (6.79)
We will apply (6.79) to a vertical and a horizontal dipole, to obtain the reflected Greens
functions
Vz
g and
Vx
g . The corresponding scalar-potentials are then obtained using (6.78).
For a vertical dipole, from (6.63) and (6.76), we get
z Azz
z
g u g
A
= , and from (6.79) we
have
Z
g
Z
g
Azz Vz
. (6.80)
The simplest solution for
Vz
g (but not the only one) is
Theory 81
Azz Vz
g g = . (6.81)
This can also be written in the form of Sommerfelds integral,
1 0 r
1 0 r
0
0
0
,
d
e ) (
0
u u
u u
R R
u
R J g
Azz Vz Vz
Z u
Vz
+
= = =
, (6.82)
where
Vz
R is the reflection coefficient for the electric scalar-potential of the vertical dipole.
For the horizontal dipole, the derivation of
Vx
g is somewhat more complicated. From
(6.64) and (6.76), we get
z Azx x Axx
x
g g u u g
A
+ = , and from (6.79) we obtain
Z
g
X
g
X
g
Azx Axx Vx
. (6.83)
After transformations we obtain
\
|
0
0
0
0
d
e ) (
0
u
u
R R J
X X
g
Azx Axx
Z u Vx
, (6.84)
whose simplest solution (but, again, not the only one) is in the form of Sommerfelds integral,
=
0
0
0
d
e ) (
0
u
R J g
Vx
Z u
Vx
, (6.85)
where
1
2
1 0 r
0 0
+
=
+ =
u u
u u
R R R
Azx Axx Vx
(6.86)
is the reflection coefficient for the electric scalar-potential of the horizontal dipole.
The above derivation of Greens functions for electric scalar-potential is similar to the
methodology presented in [16]. The obtained formulas for reflection coefficients
Vz
R and
Vx
R are
the same as in [3], but given here in a more compact form.
6.6.3 Continuity Equation for Wire Segments Above Ground
For a straight slanted wire segment (Figure 6.3), the current intensity and the charge density are
related by the continuity equation (6.12), which is also valid for normalized coordinates. ( ' Q in
this case is expressed in Coulombs.) However, knowing this charge density is not sufficient for
evaluation of the reflected electric scalar-potential. This is a consequence of inequality of the
reflection coefficients
Vz
R and
Vx
R . (The reflection coefficients for the vertical and horizontal
dipole can be made equal by a different formulation of potentials [17].) We must decompose the
charge density of the slanted wire segment into the charge densities due to the vertical and
horizontal component of the current and then apply the reflection coefficients separately to the two
charges. To simplify the analysis for arbitrarily oriented wire segments, we also decompose the
horizontal components into the x- and y-components.
The continuity equation applied to the three Cartesian components of the current moment
s d I yields in matrix form
=
(
(
(
=
2
2
2
d
d j
'
'
'
'
z
y
x
z
y
x
u
u
u
s
I
Q
Q
Q
Q , (6.87)
where
x
u ,
y
u , and
z
u are the components of the unit vector of the wire segment (u). From this
equation, it is easy to see that ' ' ' ' Q Q Q Q
z y x
= + + .
Now, for an arbitrarily oriented elementary dipole, the reflected electric scalar-potential
can be written in a compact form as
) d ' (
4
d
0
r
s Q G
k
V
V
= , (6.88)
where
| | | |
Vz Vx Vx Vz Vy Vx V
g g g g g g G = = (6.89)
is the matrix Greens function for the reflected electric scalar-potential.
6.6.4 Evaluation of Reflected Near Fields
The electric and magnetic field can be evaluated from the magnetic vector-potential and the
electric scalar-potential using (6.4) and (6.52). Introducing the normalized coordinates and
Greens functions, and substituting (6.71) and (6.88), the field components of the reflected field
due to elementary sources are obtained,
) d (
4
d ), d (
4
j d ), d ' (
4
d
r 0 r
0
2
r
s G H s G E G E I
k
I s Q
k
H A I E Q
Q
= . (6.90)
The new matrix (dyadic) Greens functions are given by
A R H V R E
G
Q
G G G curl , grad = = . (6.91)
The expressions for
R
grad and
R
curl can be written in matrix form,
(
(
(
(
(
(
=
(
(
(
(
(
(
=
0
0
0
curl , grad
X Y
X Z
Y Z
Z
Y
X
R R
. (6.92)
Applying these operators to the matrix Greens function given by (6.72) and (6.89), we get the
following matrix (dyadic) Greens functions for the reflected electric field
Q
E and magnetic field
H ,
Theory 83
(
(
(
=
(
(
(
=
Ezz Ezx Ezx
Hxz Exx Exx
Hxz Exx Exx
Ezz Ezy Ezx
Eyz Eyy Eyx
Exz Exy Exx
E
g g g
g g g
g g g
g g g
g g g
g g g
Q
sin sin sin
cos cos cos
b b b
b b b
G , (6.93)
(
(
(
=
Hzz Hzy Hzx
Hyz Hyy Hyx
Hxz Hxy Hxx
H
g g g
g g g
g g g
G
(
(
(
(
+
+
=
0 cos sin
cos cos sin cos
sin sin cos sin
b b
b b 2 b
2 1
b 2 b
2 1
b
Hzx Hzx
Hxz Hxx Hxx Hyx Hyx
Hxz Hxx Hyx Hyx Hxx
g g
g g g g g
g g g g g
. (6.94)
The elements of these matrices are given by the following Sommerfelds integrals:
=
0
0
2
1
b
d
e ) (
0
u
R J
g
g
Vx
Z u Vx
Exx
, (6.95)
=
0
0
d e ) (
0
Vx
Z u Vx
Ezx
R J
Z
g
g , (6.96)
=
0
0
2
1
b
d
e ) (
0
u
R J
g
g
Azz
Z u Azz
Hxz
, (6.97)
=
0
0
d e ) (
0
Vz
Z u Vz
Ezz
R J
Z
g
g , (6.98)
=
|
|
\
|
=
0
0
2
2
b
b
d
e ) (
0
u
R J
g
g
Azx
Z u Azx
Hxx
, (6.99)
=
0
0 1
d e ) (
0
Axx
Z u Axx
Hyx
R J
Z
g
g , (6.100)
=
0
0
2
1
b
2
d
e
) (
0
u
R
J g
g
Azx
Z u Azx
Hyx
, (6.101)
=
0
0
2
1
b
d
e ) (
0
u
R J
g
g
Axx
Z u Axx
Hzx
. (6.102)
In the above derivations, the following identities with Bessel functions are used:
\
|
= . (6.103)
6.6.5 Evaluation of Sommerfelds Integrals
Evaluation of Sommerfelds integrals is always a critical task, both from the accuracy and CPU
time point of view. In AWAS for Windows Version 2.0, Sommerfelds integrals are evaluated
using a combined, analytical-numerical method, which will be briefly explained.
Analytical transformations are applied to those integrals that have a singular behavior
when the field point tends to the image point (when both Z and P tend to zero, that is, 0 R ). In
these cases, the corresponding reflection coefficient under the integral is expanded in a short
Taylor series with respect to
1
= t . The first term of the series is integrated analytically,
resulting in a singular function. Numerical integration is applied to the remaining, nonsingular
(convergent), part of the integral.
The general problems concerning the numerical part of integration arise from the several
properties of Sommerfelds integrals. First, Sommerfelds integrals span semi-infinite intervals.
Second, for large horizontal distances, P, the Bessel function is a fast oscillating function of the
variable of integration, . Third, for small vertical distances, Z, the exponential function is a
slowly decaying function. Fourth, for large vertical distances, Z, and 1 < , the exponential
function is a fast oscillating function. Fifth, the reflection coefficients have critical singularities at
1
1
= and
r 2
= , which are branch points of the complex square root function.
The applied numerical method that overcomes these problems is adaptive, tuning its
performance in accordance with the values of the coordinates P and Z, and the relative complex
permittivity
r
. By the change of variables sin = , the first critical singularity is removed. The
similar change of variables is applied for the second singularity if the ground is nonconductive.
Special care is paid to integration in the vicinity of points where the integrand suffers rapid
changes. For the fast-oscillating functions in a limited interval, the order of the integration formula
is adaptively increased and, if necessary, integration is performed on subintervals that are
progressively introduced. For a function that is both fast-oscillating and slowly decaying (which
happens for large P and small Z), integration between the zeros of the integrand is used together
with the iterated Aitken method for speeding up the convergence of slowly convergent series.
Accurate and fast algorithms are used for computation of all functions involved. The combination
of all these strategies leads to an accurate and relatively fast calculation of Sommerfelds integrals.
85
Chapter 7
Reference Manual
7.1 Introduction
Wire antennas and scatterers are structures made of wire-like conductors, whose radii are much
smaller than their lengths and than the wavelength at the operating frequency.
AWAS for Windows is capable of analyzing wire antennas operating in the transmitting
and receiving modes, as well as scatterers. The program solves an integro-differential equation for
the current distribution along the wires, referred to as the two-potential equation. The equation is
based on the boundary conditions for the electric field on the wire axes. It is solved numerically,
using the method of moments, with a polynomial (power-series) approximation for the current
distribution and pulse testing functions. Details about the analysis method can be found in Chapter
6, but AWAS for Windows can be used without knowledge of the analysis method.
The user is assumed familiar with the basic antenna theory and Windows. This chapter
presents in detail the structure of the program, input data, and output data, and gives a full
description of commands and procedures required to use the program. The user should see
Appendix A, which summarizes program limits and input data restrictions, and Appendix B,
which lists error messages issued by the program. Appendix C is a printout of Help windows that
can be used as a shorthand version of this chapter.
7.1.1 Program Features
AWAS for Windows is a user-friendly program, designed following Windows standards, to be
installed and run in a Windows environment.
AWAS for Windows consists of a powerful kernel for numerical analysis of wire
antennas and scatterers and a user-friendly interface (program shell) consisting of a set of
windows for data input, output, and monitoring.
A computation cycle typically consists of defining input data, running the numerical
kernel, and examining results.
The user can supply data defining the structure to be analyzed, as well as frequency,
excitation, near field points, and far field directions, using a dedicated editor. A graphic
presentation of the structure is also provided. All data are checked for consistency. Errors in data
Reference Manual 87
7.1.3 Program Distribution
The software is distributed on one CD. When the program is installed, the following files are
present in the AWAS root directory:
AWAS2.EXE main program and program shell.
AWAS_W_2.DLL numerical kernel.
RUNWIRE.EXE program that activates the numerical kernel; it can also be used on a
standalone basis.
AWAS.HLP text for Help windows.
DFORRT.DLL run-time support for the numerical kernel.
MSVCRT.DLL run-time support for the numerical kernel.
CONFIG.DAT setup data for the shell.
SETUP.DAT setup data for the kernel.
KEY.DAT licensing file.
README.TXT possible last-minute updates.
UNINST.ISU information for uninstalling AWAS for Windows.
DEFAULT.FRQ default file for frequency input data.
DEFAULT.PWE default file for plane wave input data.
DEFAULT.NFP default file for near field input data.
DEFAULT.FFP default file for far field input data.
AWAS_W_2.EXP export descriptor for AWAS_W_2.DLL.
AWAS_W_2.LIB library file for AWAS_W_2.DLL.
The following file groups (input data sets) are located in the subdirectory EXAMPLES:
EXAMPLE input data set for the example used in Chapters 4 and 5.
MONOPOLE input data set for the monopole antenna of Section 8.2.
SCATT input data set for the simple scatterer of Section 8.3.
CROSS input data set for the crossed-wire scatterer of Section 8.3.
LOOP input data set for the loop antenna of Section 8.4.
CIRCUIT input data set for the wire model of a circuit of Section 8.5.
ARRAY5 input data set for the linear antenna array of Section 8.6.
CMONO input data set for the capacitively loaded monopole antenna of Section 8.7.
RCMONO input data set for the resistively and capacitively loaded monopole antenna of
Section 8.7.
YAGI input data set for the Yagi-Uda array of Section 8.8.
RTB input data set for the tower broadcast antenna of Section 8.9.
NEMPS input data set for the nuclear electromagnetic pulse simulator antenna of Section
8.10.
PLATE input data set for the wire-grid model of a square plate of Section 8.11.
CAGE input data set for the wire shielding cage of Section 8.12.
DIPOLE input data set for the dipole above real ground 8.13.
VDIPOLE input data set for the V-dipole above real ground Section 8.14.
Each group consists of five files with the same file name and name extensions: GEO,
FRQ, PWE, NFP, and FFP, respectively.
Located in the subdirectory SOURCE are the following source codes, together with the
corresponding executable and input data files:
Reference Manual 89
The program kernel, which performs the numerical analysis. The kernel is written, compiled,
and linked in Compaq Visual Fortran 6.1.0 environment.
Windows are organized in a hierarchy, starting with the AWAS Main window, and they
can open dialog boxes. Windows and dialog boxes are designed to provide a user-friendly
interface with the program to input all data required for the analysis, examine results of the
analysis by listing or plotting, monitor computations, perform setup, and so on.
The basic structure of the program is shown in Figure 7.1.
DEFAULT
PLANE WAVE
DIALOG BOX
MAIN
WINDOW
BATCH
DIALOG BOX
SETUP DIALOG
BOX
PLANE WAVE
DIALOG BOX
NEAR FIELD
DIALOG BOX
FAR FIELD
DIALOG BOX
DEFAULT
NEAR FIELD
DIALOG BOX
DEFAULT
FAR FIELD
DIALOG BOX
NUMERICAL
KERNEL
DEFAULT
FREQUENCY
DIALOG BOX
DIALOG BOX
GRAPH
SELECTION
WINDOW
WINDOW
WINDOW
WINDOW
WINDOW
PLOT
WORK
SHOW
DATA
INPUT
LIST
Figure 7.1 Basic structure of AWAS for Windows.
All windows have a menu bar from which pull-down menus can be opened. They also
have buttons, hot keys, and a Control menu. The Main window also has a toolbar. Dialog boxes
have menus, buttons, and hot keys.
All of the major program activities start and terminate in the Main window. Subordinate
windows and dialog boxes are, generally, opened by selecting a menu option, pressing a hot key,
or clicking a button.
Reference Manual 93
The information on the ground is located in the input data file describing the geometry of
the structure. See Section 7.2.2.2 for a description of this file. The ground is coded in the input and
output data files in the following way:
0 no ground;
1 PEC ground plane;
1 PEC ground plane with reflection-coefficient formulation for far fields (for real ground);
2 Sommerfelds formulation for real ground.
The ground can be defined in the Input Data window. See Section 7.7.3 for more
information.
7.2.1.4 Nodes, Segments, Loadings, and Ports
AWAS for Windows can analyze structures assembled from straight-line wire segments.
Geometrically, such a structure can be defined by a set of nodes (i.e., points in three-dimensional
space) and a specification of the way the nodes are interconnected by the segments. The wire
structure can have one or more ports (if it is an antenna), and wires can have concentrated or
distributed loadings.
Each node is defined by its Cartesian coordinates ( , , ) x y z . Any two nodes must not
coincide. If a ground plane or real ground is present, the nodes must be above the plane 0 = z or
in that plane (i.e., 0 z ) except for Sommerfelds formulation when 0 > z .
A wire segment is assumed to have a circular cross section. It can be made of a perfect
(lossless) conductor (PEC) or of a real conductor. Each segment is defined by an ordered pair of
two node indices, by its radius, and by possible loadings. A local coordinate axis is associated with
the segment. The segment and its axis are oriented from the first node (origin, labeled a) toward
the second node (end or terminal, labeled b), as shown in Figure 6.3.
For the purpose of the numerical analysis, a polynomial is automatically associated with
each wire segment, which approximates the current distribution along that segment. See Section
7.2.1.5 for more information on polynomials.
Indices of the two nodes for a segment must be different. Two or more segments can
have a common node, thus forming a junction. Two segments must not otherwise intersect.
If the PEC ground plane is used in computations, one end of a wire segment can be
connected to the ground plane, which is simulated by putting z = 0 for that node. However, the
other end must be above the plane (i.e., z > 0 for that node). For Sommerfelds formulation, for
all nodes the condition z > 0 must be fulfilled.
Additional data associated with a segment are its radius and loadings: the concentrated
loading at the segment origin ( Z
a
), the concentrated loading at the segment terminal ( Z
b
), and
the loading uniformly distributed along the segment ( Z
d
). A segment can also carry a port at its
origin, terminal, or both.
Wire radius (R) must satisfy the condition nR h < , where n is the polynomial degree and
h is the segment length. To obtain a good solution, the radius should also be kept much smaller
than the wavelength (at least about 30 times smaller).
Each loading can have various forms. Available options are:
Series RLC circuit (code 1 for R, 2 for L, and 4 for C, code offset 0), with all three elements
present, with one element missing, or with two elements missing;
Parallel RLC circuit (code 1 for R, 2 for L, and 4 for C, code offset 16), with all three elements
present, with one element missing, or with two elements missing;
Short-circuited transmission-line (SCTL) stub (code 32);
Open-circuited transmission-line (OCTL) stub (code 48).
and E
.
Here, is the angle with respect to the z-axis, and is the azimuthal angle with respect to the x-
axis.
Plane waves can be organized in 1 to 8 plane wave groups. Each group has a unique pair
of electric-field components, which define both the intensity and polarization (linear, circular, or
elliptic) of the waves. The directions of incidence vary within the group, and the angles defining
these directions are given by
i
i i n = + =
start
, ,..., 0 , (7.3)
j
j j n = + =
start
, ,..., 0 , (7.4)
where
start
and
start
are the minimal values of the angles,
stop
and
stop
are the maximal
values of the angles, n
and n
stop start
n
(7.5)
and
=
stop start
n
(7.6)
are angular steps. A pair ( , )
i j
can be considered as a point in the plane whose coordinate axes
are the angles and .
The minimal number of angular steps is 0, and the maximal is 999.
For antennas operating in the transmitting mode, the number of plane wave groups is
treated as 0, because the excitation is only at ports. For antennas operating in the receiving mode
or for scatterers, the number of plane wave groups must be at least 1. The electric-field intensity
(modulus) of a plane wave must be greater than zero.
For a scatterer whose monostatic (radar) cross section is evaluated, plane wave groups
are, at the same time, far field groups. In this operating mode, far field points are not specified at
data input. When the far fields are plotted, the information on the plane waves is used to define the
far field directions, as if these were actually input. Far fields can be plotted within a group against
one angular coordinate as the variable, with the other coordinate constant, or against both
coordinates (3D plot). If you want to plot far fields against one angle, the number of steps for that
angle must be at least 1 (amounting to 2 values of the angle). However, if two numbers of steps
are taken to be large, the overall number of plane waves can become huge, resulting in a long run
time and large output data files.
There are no limits for the electric field components. By properly defining these
components, various polarizations of the incident field can be simulated.
A linear polarization is achieved if only one electric field component is present ( E
or
E
), while the other is zero, or if both components are present but have identical phases (or are in
counterphase). For example, V/m j0) 1 ( + =
E and E
in Figure 6.8.
V/m j2) 1 ( + =
and u
.
A circular polarization is achieved if both electric field components are present, have the
same intensities, but are in phase quadrature. For example, mV/m ) 0 j 1 ( + =
E and
mV/m ) 1 j 0 ( + =
E and V/m ) 2 j 2 ( + =
and n
stop start
n
(7.15)
and
)
Reference Manual 99
=
stop start
n
(7.16)
are angular steps. A pair ( , )
i j
can be considered as a point in the plane whose coordinate axes
are the angles and .
The minimal number of angular steps is 0, and the maximal is 999. Once the analysis of
the wire structure is done, far fields can be plotted within a group against one angular coordinate
as the variable, with the other coordinate constant, or against both coordinates (3D plot). If you
want to plot far fields against one angle, the number of steps for that angle must be at least 1
(amounting to 2 values of the angle). However, if two numbers of steps are taken to be large, the
overall number of far field directions can become huge, resulting in a long run time and large
output data files.
For the specified far field directions, spherical components of the electric field ( E
and
E
) are evaluated in the analysis. These components are presented in the output results multiplied
by the distance (r) from the antenna, and the field phase is reduced to the coordinate origin, so that
the results do not depend on the spherical coordinate r. Altogether this amounts to multiplying the
radiated electric field by r kr exp( ) j , where k = 2 / is the phase coefficient. The unit for this
normalized electric field is volt (V). Hence, the results depend only on the spherical angles, and
.
For antennas in the transmitting mode, the power gain (with respect to an isotropic
radiator) is evaluated in addition to the electric field components, and for scatterers, the
monostatic or bistatic cross sections (total, copolar, and cros-spolar) are evaluated.
Limits for angles and are 180 180 and 180 180 , respectively, when
there is no ground present. If the ground is present, the limits for are 90 90 . Note that
= 180 and = 180 both correspond to the negative direction of the z-axis, while = 0
coincides with the positive direction of the z-axis. When the ground is lossy, the far field is
evaluated using the reflection coefficients. Care should be taken when 90 | | as the reflection
coefficient does not properly account for the attenuation of the surface wave at finite distances
from the antenna. The field of the ground wave may be obtained using Sommerfelds formulation
and evaluating the near field at the appropriate distance from the antenna.
See Section 7.2.2.6 for information on files containing data for far field directions. See
Section 7.7.9 for information on input of far field directions.
7.2.2 Input Data Files
7.2.2.1 Input Data Set
Input data that are required for one analysis and that completely describe the wire structure
(including its geometry, excitation, ground, operating frequencies, points at which near and far
fields are to be evaluated, and so on) are passed from the program shell to the program numerical
kernel in a set of five temporary input data files, which form one input data set. These files are:
WIRE.GEO contains data defining the operating mode, ground, geometry of the wire
structure (nodes, segments, loadings, and ports), as well as the header for the data set.
WIRE.FRQ contains data defining the operating frequencies.
WIRE.PWE contains data defining plane waves illuminating the structure.
).
The above example defines two plane wave groups (n
g
= 2). For the first group,
V/m j0) 1 ( + =
E and E
. Further,
start
= 0 ,
stop
= 90 , n
= 90 ,
start stop
= = 0 ,
and n
. Further,
start
= 0 ,
stop
= 90 , n
= 90,
start stop
= = 0 , and n
).
The above example defines two far field groups ( n
g
= 2 ). For the first group,
start
= 180 ,
stop
= 180 , n
= 36 ,
start
= 0 ,
stop
= 90 , and n
= 1. For =
start
, we
have a 360 scan in the zx-plane. Normally, is positive, and for =
start
, the far fields would
be evaluated only in the half-plane where x 0 . The directions for which is negative are in the
other half-plane, where x 0 . Similarly, when =
stop
, a 360 scan in the yz-plane is
simulated, in both cases in steps of 10 . For the second group,
start stop
= = 90 , n
= 0 ,
start
= 180 ,
stop
= 180 , and n
and E
start
,
stop
, and n
. Electric field components are in volts per meter (V/m), and angles are in
degrees. The basic data describing the geometry follow, in the same way as explained in Section
7.3.2.3.
Blocks follow for each operating frequency (300 MHz, 350 MHz, and 400 MHz for the
example presented in this section, but only the block for the first frequency is shown). Each
frequency block contains one or more blocks corresponding to each excitation of the wire
structure. In the present example, the antenna is analyzed for five excitations (one port at a time),
so that five such blocks are listed.
For a transmitting antenna with all ports simultaneously excited, only one excitation
block exists. For an antenna in the transmitting/receiving mode, an excitation block comes first
when the antenna is driven at the ports. It is followed by excitation blocks when the structure is
illuminated by plane waves. Thereby, plane wave groups are ordered as in the file header. Within
)
Reference Manual 125
a group, plane waves are ordered so that variations of the angle are nested into variations of the
angle . For a scatterer, only excitation blocks exist for plane waves.
The block for each excitation contains polynomial coefficients for all wire segments (for
that excitation and the corresponding frequency). The block starts with the number of segments
(five). A block follows for each segment, consisting of two or three lines. The first line contains
the polynomial degree for that segment (three for each segment), segment length ( m 25 . 0 = h in
all cases in the example), and the normalization coefficient 1 / ( ) h required in (6.21) for
evaluation of the charge distribution. The next line (or two lines) contains polynomial coefficients,
ordered according to ascending powers. Each of these coefficients is a complex number, which is
printed as two separate real numbers. One line can contain up to five complex numbers. (In the
above listing, lines are wrapped, so that each line with complex numbers appears as two lines.)
The boundary between blocks for frequencies and blocks for excitation is always one
blank line, while blocks for segments are not separated. The above listing contains one frequency
block, with five excitation blocks, and each has five segment blocks (i.e., a total of 25 segment
blocks).
7.3.2.7 File *.NFL
The file *.NFL (i.e., a file with an arbitrary name and with the default name extension NFL)
contains near fields. This file can be used to produce plots of near electric and magnetic fields or
to export data to other programs.
The file WIRE.NFL is a temporary file created by the program kernel to pass data to the
program shell. A saved copy of this file (with the default name extension NFL) is created by the
shell. The file contains only integer and floating-point numbers, separated by spaces and carriage-
return line-feed pairs.
An example of the file *.NFL is shown below (corresponding to the shielding cage
described in Section 8.12). The structure is analyzed at frequency MHz 50 = f , as a scatterer
(bistatic scattering), illuminated by one plane wave arriving from = 90 and = 0 , with
electric field components j0)V/m 1 ( + =
E and E
and
E
(each printed as two separate real numbers); the second line contains
start
,
stop
, and n
;
and the third line contains
start
,
stop
, and n
= 36,
start
= 0 ,
stop
= 90 , and n
= 0 ,
start
= 180 ,
stop
= 180 , and n
= 36 . Only the header and several lines with far fields are
listed.
2
2.8000E+08 3.2000E+08 8
-1
1
0
-180.00 180.00 36
0.00 90.00 1
90.00 90.00 0
-180.00 180.00 36
7
6
1
0.000000E+00 0.000000E+00 0.000000E+00
and E
. The
basic data describing the geometry follow, in the same way as explained in Section 7.3.2.3.
Blocks follow for each operating frequency. Each frequency block contains one or more
blocks corresponding to each excitation of the wire structure. In the present example, the antenna
is analyzed when it is excited at one port, so that only one such block exists.
For a transmitting antenna with all ports simultaneously excited as well for a
transmitting/receiving antenna, only one excitation block exists. For a transmitting antenna with
one port excited at a time, the number of excitation blocks equals the number of ports. For an
antenna in the receiving mode, no far fields are evaluated. For a scatterer, excitation blocks exist
)
Reference Manual 129
only for plane waves. Plane wave groups are ordered as in the file header. Within a group, plane
waves are ordered so that variations of the angle are nested into variations of the angle .
Further description for scatterers will be given later in this section. For antennas, the
block for each excitation consists of blocks for each far field group, and the block for each group
consists of blocks for each direction. Thereby, far field groups are ordered as in the file header,
and, within a group, directions are ordered so that variations of the angle are nested into
variations of the angle .
The block for a direction consists of two lines. The first line contains spherical
components of the electric field ( E
and E
E and E
i
= 0 . The direction of wave incidence
is defined by
istart istop
= = 90 , n
i
= 0,
istart istop
= = 0 , and n
i
= 0 . One far field group
is defined, with
sstart
= 0 ,
sstop
= 90 , n
s
= 45 ,
sstart sstop
= = 0 , and n
s
= 0 . Only the
header and several lines with far fields are listed.
1
3.0000E+08 3.0000E+08 0
-3
0
1
1.0000E+00 0.0000E+00 0.0000E+00 0.0000E+00
90.00 90.00 0
0.00 0.00 0
0.00 90.00 45
0.00 0.00 0
2
1
0
0.000000E+00 0.000000E+00 -5.000000E-01
0.000000E+00 0.000000E+00 5.000000E-01
1 2
0.0000E+00 0.0000E+00 0.0000E+00 0.0000E+00
-9.9990E+02 -9.9990E+02 -9.9990E+02
-1.5558E-03 -3.7242E-04 0.0000E+00 0.0000E+00
-4.4927E+01 -4.4927E+01 -9.9990E+02
-3.1177E-03 -7.4732E-04 0.0000E+00 0.0000E+00
-3.8888E+01 -3.8888E+01 -9.9990E+02
-4.6921E-03 -1.1272E-03 0.0000E+00 0.0000E+00
-3.5337E+01 -3.5337E+01 -9.9990E+02
-6.2849E-03 -1.5145E-03 0.0000E+00 0.0000E+00
-3.2797E+01 -3.2797E+01 -9.9990E+02
-7.9021E-03 -1.9116E-03 0.0000E+00 0.0000E+00
-3.0806E+01 -3.0806E+01 -9.9990E+02
-9.5495E-03 -2.3209E-03 0.0000E+00 0.0000E+00
-2.9159E+01 -2.9159E+01 -9.9990E+02
-1.1233E-02 -2.7448E-03 0.0000E+00 0.0000E+00
-2.7746E+01 -2.7746E+01 -9.9990E+02
E and E
i
= 0 and with
istart
= 0 ,
istop
= 90 , n
i
= 45 ,
istart istop
= = 0 , and n
i
= 0 . Only the header and several lines with
far fields are listed.
1
3.0000E+08 3.0000E+08 0
3
0
1
1.0000E+00 0.0000E+00 0.0000E+00 0.0000E+00
0.00 90.00 45
0.00 0.00 0
0.00 90.00 45
0.00 0.00 0
2
1
0
0.000000E+00 0.000000E+00 -5.000000E-01
0.000000E+00 0.000000E+00 5.000000E-01
1 2
0.0000E+00 0.0000E+00 0.0000E+00 0.0000E+00
-9.9990E+02 -9.9990E+02 -9.9990E+02
2.2572E-04 5.0484E-04 0.0000E+00 0.0000E+00
-5.4153E+01 -5.4153E+01 -9.9990E+02
9.0241E-04 2.0208E-03 0.0000E+00 0.0000E+00
-4.2108E+01 -4.2108E+01 -9.9990E+02
2.0286E-03 4.5522E-03 0.0000E+00 0.0000E+00
-3.5057E+01 -3.5057E+01 -9.9990E+02
)
Reference Manual 131
3.6013E-03 8.1050E-03 0.0000E+00 0.0000E+00
-3.0050E+01 -3.0050E+01 -9.9990E+02
The structure of the file header is the same as in the previous example. The operating
mode code is now 3 (scatterer, monostatic cross section).
In this case, no far field groups were defined in the input data. However, the program
automatically takes the data for angles (number of groups, start and stop values, and the numbers
of steps) from the plane wave groups, and copies them as data for the far field groups.
The sequence of frequency and excitation blocks is also the same as in the previous
example. However, now there are no blocks for far field groups within an excitation block, and
each excitation block is the same as the block for a far field direction in the previous example.
7.4 Main Window
After AWAS for Windows loads, the Main window opens, as shown next.
All major program activities (data input, computations, the examination of output results
by listing or plotting, as well as the configuration of the program) start and terminate in the Main
window. AWAS for Windows can be stopped only from the Main window.
The Main window has a menu bar and a toolbar. The following menus are available in
the Main window menu bar:
and E
or E
(
(
(
=
(
(
(
z
y
x
r r r
r r r
r r r
z
y
x
zz zy zx
yz yy yx
xz xy xx
'
'
'
, (7.17)
where
zz xx
r r ,..., are the elements of the rotation matrix. The excitation for each array element is
given relative to the excitation of the basic element. The data for the array are written to the file
ARRAY.GEO. As an example, the file ARRAY.DAT contains data for creating an array of the
antennas defined in the file EXAMPLE.GEO.
The program MERGE merges the geometry data for two structures. The program reads
the data from two GEO files (whose names are specified from the keyboard), merges the nodes,
segments, and ports, and writes the result to the file MERGE.GEO. The merged structure has the
combined title of the two original structures. The operating mode and the ground correspond to the
second structure.
The programs SCALE, ARRAY, and MERGE do not work for structures placed above a
real ground.
The program SPAR converts the scattering parameters computed by AWAS into a
standard form recognized by many microwave-engineering programs. The program reads the file
WIRE.SPM and the output goes to the file whose name WIRE and the extension is S1P, S2P, ...,
S9P, S10, ..., depending on the number of ports. The output scattering parameters are given as
magnitude and angle, and the frequency is in GHz. The port nominal impedance must be 50 for
all ports defined in AWAS to produce a meaningful output. The program can serve as an example
of parsing the output files of AWAS with the port matrix parameters.
The program CIRC is an example of parsing the output files of AWAS for the far fields.
The program reads the file WIRE.FFL for an antenna, extracts the right-hand circular (RHC) and
left-hand circular LHC polarization, finds the axial ratio of the polarization ellipse, and tabulates
the results in the file WIRE.CP.
Finally, the program TABCONV2 reads data from a table report, created from the Plot
window, and converts them to a simple ASCII table. The program acts only on a table whose
)
Reference Manual 181
second column contains real numbers. The input to the program is the name of the TAB file and
the name of the output file.
In a more advanced application, the user can create input data, submit them directly to the
kernel of AWAS for Windows (AWAS_W_2.DLL) for analysis, and postprocess the results.
Further, this chain of steps can be nested within an optimization procedure aimed at a computer-
aided synthesis of antennas and scatterers [3].
AWAS_W_2.DLL can easily be executed on a standalone basis or incorporated into a
user-written program. In both cases, the user is responsible for creating the temporary input data
set, described in Section 7.2.2, and for handling the output results present in the temporary output
data set, described in Section 7.3.2. In addition, the files SETUP.DAT and KEY.DAT must be
present in the same directory as AWAS_W_2.DLL and the input and output data files. To run
AWAS_W_2.DLL on a standalone basis, simply activate RUNWIRE.EXE from the Windows
Program Manager, command prompt, or from another program.
The source code of RUNWIRE.EXE can be found in the directory SOURCE, in the file
RUNWIRE.C. The source code was written in C programming language. It is a simple code that
does not include any error handling or allow the program to be interrupted during execution, but it
does show all declarations and variables necessary to activate AWAS_W_2.DLL. This source code
can serve as a template for building more complex Windows applications that incorporate
AWAS_W_2.DLL. In such applications, the status of AWAS_W_2.DLL can be monitored using the
temporary files HEADER.TMP, FREQU.TMP, STATUS.TMP, and MESSAGE.DAT.
>
183
Chapter 8
Examples of Analysis
8.1 Introduction
This chapter presents several examples of the analysis of wire antennas and scatterers using
AWAS for Windows. The examples are aimed at illustrating various capabilities of the program
and demonstrating the numerical performance of the method. Basic input data sets for all
examples are distributed with the program, so that these examples can help the user exercise his
skills in using the program.
The first example is a vertical monopole antenna above a PEC ground plane. Some
features of the numerical convergence are demonstrated in this example. The second example is
the analysis of simple wire scatterers. A square loop is used in the subsequent example to compare
the results of the program with quasi-static results for the loop inductance. A loaded wire structure
that is a realistic model of an electrical circuit is analyzed next, demonstrating the range of validity
of the circuit theory. A five-element array of quarter-wavelength monopole antennas is an
example of multiport wire structures. In the next group of examples (loaded broadband
monopoles, a Yagi-Uda antenna, and a tower antenna with guy ropes), comparisons are made
between the results of AWAS for Windows and available experimental data. The subsequent three
examples (a nuclear electromagnetic pulse simulator antenna, a wire-grid model of a square plate,
and a wire cage) illustrate some special applications of AWAS for Windows. The last two
examples illustrate the application of Sommerfelds theory. For all examples, the accuracy
parameter is 4, unless stated otherwise.
8.2 Vertical Monopole Antenna
We consider a vertical monopole antenna of height h and radius R, placed above a perfectly
conducting ground plane. The antenna is sketched in Figure 8.1. The monopole is defined by two
nodes and one segment. If we place the monopole on the z-axis, the node coordinates are (0,0,0)
and ( ) 0,0,h . The first node lies in the ground plane. We take this node to be the segment origin,
and the antenna port is located at this node. The input data set for this example is MONOPOLE.
= 45,
start stop
0 = = , and n
= 0. The plot was obtained by manually changing the scale for the
ordinate to the range from dB 30 to dB 10 .
Figure 8.3 Power gain of a monopole antenna of height 1 m and radius 1 mm, at 300 MHz.
Next, the same monopole antenna was analyzed at low frequencies, to demonstrate the
behavior of the numerical algorithm for electrically short structures. The results are shown in
Table 8.2 for two cases: when the antenna conductor is assumed perfect (PEC) and when it is
made of copper (whose conductivity is 56 MS/m). The input reactance has the expected nature of
a pure capacitance when h / . < 0 01, where is the wavelength. For the PEC case, the input
resistance is just the radiation resistance. Initially, it is proportional to frequency squared. It
becomes erratic when h / . < 0 001 because it is extremely small, practically zero. These results
are within the numerical noise. For the copper conductor, the input resistance behaves regularly
down to the last entry in Table 8.2. The reason is that the radiation resistance is masked by the
conductor losses.
j6.052 10
3
3 2
10 j6.052 10 815 . 5
10
3
3936 10
4 4
.
j6.063 10
4 3
10 j6.063 10 551 . 6
10
4
5 6
10 j6.063 10 562 . 3
5 3
10 j6.063 10 198 . 2
10
5
6 6
10 j6.063 10 351 . 3
6 3
10 j6.063 10 431 . 1
10
6
7 6
10 j6.063 10 717 . 4
7 3
10 j6.063 10 416 . 1
10
9
10 6
10 j6.063 10 352 . 6
10 3
10 j6.063 10 415 . 1
10
12
13 6
10 j6.063 10 136 . 4
13 3
10 j6.063 10 417 . 1
10
15
16 6
10 j6.063 10 058 . 6
16 3
10 j6.063 10 415 . 1
10
18
19 6
10 j6.063 10 058 . 6
19 3
10 j6.063 10 417 . 1
For the same monopole (i.e., for m 1 = h and mm 1 = R , at MHz 300 = f ), Figure 8.4
shows the resulting axial component of the electric field ( E
z
), produced by the antenna currents
and charges. This near field plot was obtained based on the following specifications for one near
field group: x x
start stop
= = 0, n
x
= 0, y y
start stop
= = 0, n
y
= 0, z
start
= 0, m 5 . 1
stop
= z , and
n
z
= 150. Theoretically, this field must be zero within the antenna (since the antenna is assumed to
be made of a perfect conductor), except at the generator (because in that region we have an
impressed electric field modeling the excitation). However, in a numerical solution the boundary
conditions are satisfied only approximately, so that there is a residual electric field that is a
measure of the error in the solution. Note two large spikes of this field, which extend beyond the
border of the plot. The first spike (around z = 0, and with the extreme value of
( ) V/m 317 Re =
z
E ) is due to the generator field. The integral of the electric field under this peak
balances the generator electromotive force. The second spike (around z = 1 m and with the
extreme value of ( ) V/m 102 Re =
z
E ) is due to the approximate modeling of the antenna current
and charge distributions in the vicinity of the antenna end. The electric field beyond m 1 = z
should, theoretically, be different from zero, because these points are not inside the antenna
conductor.
Finally, in Figure 8.5 the input impedance of a monopole of height m 1 = h and radius
mm 1 = R is plotted versus frequency. The antenna was analyzed at 200 frequencies (from
10 MHz to 2,000 MHz, in 10 MHz steps), and the plot was manually scaled to display the range of
impedances from k 1 to k 1 .
)
Examples of Analysis 189
Figure 8.4 Axial component of the electric field on the axis of a monopole antenna of height
1 m and radius 1 mm, at 300 MHz.
Figure 8.5 Input impedance of a monopole antenna of height 1 m and radius 1 mm, as a
function of frequency.
O
Figure 8.6 Simple wire scatterer excited by a plane wave. All dimensions are in millimeters.
First, we evaluate the monostatic cross section as a function of the angle , at the
operating frequency MHz 300 = f . In AWAS for Windows, we select the operating mode for a
scatterer, monostatic cross section, and select no ground. The wire is defined by two nodes, with
coordinates m) 5 . 0 , 0 , 0 ( and m) 5 . 0 , 0 , 0 ( , and one segment. One plane wave group is defined,
with V/m 1 =
E , E
= 0,
start
0 = ,
stop
90 = , n
= 45,
start stop
0 = = , and n
= 0. In
Figure 8.7, the results for the total monostatic cross section are plotted.
)
Examples of Analysis 191
Figure 8.7 Monostatic cross section of the scatterer of Figure 8.6, as a function of
observation angle, at 300 MHz.
In Figure 8.8, the monostatic cross section for = 90 versus frequency is shown. The
results were obtained by supplying input data similar to those given above, with the following
differences. The frequency range was defined by MHz 100
start
= f , MHz 600
stop
= f , and
n
f
= 100, while only one value of was used in the plane wave group.
i
90 = , versus the angle
s
for the scattered wave, at the operating frequency MHz 300 = f .
The result was obtained by selecting the operating mode for a scatterer, bistatic cross section, one
plane wave (arriving from
i
90 = and
i
= 0 , with V/m 1 =
E and E
E and V/m ) 2 / 2 j( =
E and V/m j =
=
j0.146 + 0.045 j0.092 0.169 j0.030 + 0.079 j0.015 - 0.051 j0.013 + 0.039
j0.092 0.169 j0.114 + 0.049 j0.079 0.166 j0.024 + 0.077 j0.015 0.051
j0.030 + 0.079 j0.079 0.166 j0.110 + 0.051 j0.079 0.166 j0.030 + 0.079
j0.015 0.051 j0.024 + 0.077 j0.079 0.166 j0.114 + 0.049 j0.092 0.169
j0.013 + 0.039 j0.015 0.051 j0.030 + 0.079 j0.092 0.169 j0.146 + 0.045
] [s ,
mS
j5.34 22.52 j2.11 + 7.75 j0.45 + 2.18 j0.30 1.17 j0.10 + 0.98
j2.11 + 7.75 j3.25 24.19 j1.80 + 7.12 j0.59 + 1.89 j0.30 1.17
j0.45 + 2.18 j1.80 + 7.12 j3.25 24.33 j1.80 + 7.12 j0.45 + 2.18
j0.30 1.17 j0.59 + 1.89 j1.80 + 7.12 j3.25 24.19 j2.11 + 7.75
j0.10 + 0.98 j0.30 1.17 j0.45 + 2.18 j2.11 + 7.75 j5.34 22.52
] [
(
(
(
(
(
(
= y ,
(
(
(
(
(
(
=
j18.4 + 44.2 j15.6 12.3 j10.2 + 6.5 j7.4 4.4 j5.8 + 3.1
j15.6 12.3 j18.7 + 45.1 j15.8 12.8 j10.2 + 6.7 j7.4 4.4
j10.2 + 6.5 j15.8 12.8 j18.8 + 45.3 j15.8 12.8 j10.2 + 6.5
j7.4 4.4 j10.2 + 6.7 j15.8 12.8 j18.7 + 45.1 j15.6 12.3
j5.8 + 3.1 j7.4 4.4 j10.2 + 6.5 j15.6 12.3 j18.4 + 44.2
] [z .
By comparing the admittance matrix with the results of Table 8.1 (for mm 1 = R ), we
can see that the coupling between the antennas is relatively strong. Therefore, it must be taken into
account in the design of the feeding circuitry for the array.
)
Examples of Analysis 199
6
6
2
.
3
3
2
.
1
1
1
6
.
7
1
4
0
.
9
1
6
3
.
0
9
0
.
5
z
x
C
1
2
C
3
C
4
C
5
C
L
Figure 8.17 Monopole antenna with concentrated capacitive loadings. All dimensions are in
millimeters.
8.7 Loaded Broadband Monopole Antennas
By suitably loading wire antennas (i.e., by placing concentrated or distributed impedances), it is
possible to modify antenna properties. Similar effects can be achieved with wire scatterers. One of
the important objectives of such a loading is to obtain antennas, which have broadband input
characteristics and radiation patterns. Analyses of two such antennas are presented here.
The first antenna is a vertical monopole, sketched in Figure 8.17, which has concentrated
capacitive loadings [3, 21], as well as a compensating series inductance in the excitation region.
This antenna was synthesized by numerical optimization to achieve broadband input admittance in
a frequency range of one octave (1 GHz to 2 GHz). The capacitances are pF 06 . 1
1
= C ,
pF 37 . 6
2
= C , pF 53 . 0
3
= C , pF 91 . 0
4
= C , and
5
C , while the reactance inductance is
nH 62 . 4 = L . We wish to evaluate the input impedance of this antenna in the frequency range
800 MHz to 2400 MHz, in 32 steps, and the radiation pattern in a vertical plane. The input data set
for this example is CMONO.
This antenna is relatively thick, and the choice of polynomial degrees is critical,
especially at the high end of the frequency band. Increasing the polynomial degree for the first
segment, on which the antenna port is located, increases the parasitic capacitance introduced by
the delta-function approximation of the generator. A similar parasitic capacitance is introduced at
each concentrated loading. This capacitance is connected in parallel to the actual loading. The
problem can be partly bypassed by reducing the accuracy parameter and thus using polynomial
approximations of the lowest possible degrees. One far field group is defined, with
start
0 = ,
stop
90 = , n
= 45,
start stop
0 = = , and n
= 0 .
The geometry of the antenna is defined by 7 nodes and 6 segments. However, since the
capacitance
5
C is actually a short circuit, an alternative way of supplying data for this antenna is
to reduce the number of nodes and segments by one (by discarding the node next to the antenna
tip).
)
200 Analysis of Wire Antennas and Scatterers
The results for the antenna input impedance ( X R Z j + = ) obtained by AWAS for
Windows and the experimental results are plotted in Figure 8.18. The accuracy parameter is 2. The
antenna is lossless, and its efficiency is 100%, which can be verified in the file WIRE.OUT.
0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4
-40
-30
-20
-10
0
10
20
30
40
50
60
70
80
90
100
110
R, AWAS
X, AWAS
R, experiment
X, experiment
Z
[
]
f [GHz]
Figure 8.18 Input impedance of the capacitively loaded monopole antenna of Figure 8.17, as
a function of frequency.
In Figure 8.19, the radiation patterns of this antenna are shown for two frequencies (at the
ends of the operating band). The maximum of the radiation pattern moves upward as the
frequency increases, and a scaled version of the antenna can be used in certain HF applications
where the ionospheric wave propagation is involved.
]
f [GHz]
Figure 8.26 Input impedance of the tower antenna of Figure 8.25, as a function of
frequency.
8.10 Nuclear Electromagnetic Pulse Simulator Antenna
Large outdoor simulators of the nuclear electromagnetic pulse (NEMP) can be made in the form of
a wire structure (referred to as the NEMP simulator antenna), which behaves much like a
transmission line. The wire structure is placed above a ground plane. It is driven at one end by a
pulse generator of a very high voltage, while at the other end it is terminated in a resistor. A
simplified version of such a structure [25] is shown in Figure 8.27. The generator is placed at the
lower end of one short vertical segment, and the resistor, of resistance 150 , is assumed to be
distributed along the other short vertical segment. The input data set for this example is NEMPS.
The input impedance of the simulator as a function of frequency is plotted in Figure 8.28.
If the electric or magnetic fields at a set of points within this simulator are evaluated at a number
of frequencies, it is possible to obtain the time-domain waveforms of these fields by performing
the inverse Fourier transform and thus evaluate the performance of the simulator [26].
)
208 Analysis of Wire Antennas and Scatterers
5
2
2
2
0.04
1
0
1
0
1
0
0.5
z
x
y
150
Figure 8.27 Simple nuclear electromagnetic pulse simulator antenna. All dimensions are in
meters.
Figure 8.28 Input impedance of NEMPS simulator antenna of Figure 8.27, as a function of
frequency.
)
Examples of Analysis 209
1000
1
0
0
0
100
x
z
y
E
Figure 8.29 Wire-grid model of a square plate. All dimensions are in millimeters.
8.11 Wire-Grid Model of Square Plate
Figure 8.29 shows a relatively crude wire-grid model of a square metallic plate that is analyzed as
a scatterer at frequency MHz 75 = f . This model consists of 25 nodes and 40 wire segments. The
grid is assumed to be excited by a linearly polarized plane wave. We wish to evaluate the
monostatic cross section in the zx-plane, in two cases. The first case is when the electric field
vector of this wave has only the component E
. Therefore, we select two plane wave groups, with different electric fields, but
identical sets of angles (
start
0 = ,
stop
90 = , n
= 90,
start stop
0 = = , and n
= 0 ). The
input data set for this example is PLATE. The results for the monostatic cross section are shown in
Figure 8.30.
)
210 Analysis of Wire Antennas and Scatterers
(a)
(b)
Figure 8.30 Monostatic cross section of the wire-grid structure of Figure 8.29, as a function
of observation angle , at 75 MHz, when the structure is excited by a linearly polarized plane
wave with (a) -component and (b) -component of the electric field, for =0.
)
Examples of Analysis 211
8.12 Wire Shielding Cage
Figure 8.31 shows a shielding cage made of welded wires. The cage is placed on a PEC ground
plane; its base is a square of side 700 mm; its height is 400 mm. Each eyelet is a square of side
100 mm. The wire radius is 1 mm. The cage consists of 308 segments, and it is defined by 176
nodes. The objective is to evaluate the shielding effectiveness of the cage at frequency
MHz 50 = f . We take the cage to be illuminated by one plane wave, of grazing incidence,
traveling in the direction of the negative x-axis (i.e., arriving from = 90 and = 0 ), and
vertically polarized electric field. Thus, we take one plane wave group, with V/m j0) 1 ( + =
E
and E
= 0 . However, the incident wave is reflected from the ground plane, so that the total
exterior electric field in which the cage is located has an intensity of V/m 2
exterior
= E . We also
define one near field group, with m 5 . 1
start
= x , m 5 . 1
stop
= x , n
x
= 150 , y y
start stop
= = 0 ,
n
y
= 0 , z z
start stop
= = 0 , and n
z
= 0 , and analyze the cage as a scatterer. It is irrelevant if
monostatic or bistatic scattering is selected, as no far fields need to be calculated. The input data
set for this example is CAGE.
On the distribution CD, in the directory SOURCE, is located the file CAGE.FOR, which
contains a Fortran source code that was used to generate input data for the cage geometry, and its
executable version, CAGE.EXE. The program is relatively simple and self-explanatory. This
program can be used as a template for building similar programs for various structures that are
within the users scope.
Figure 8.32 shows the z-component of the electric field along the x-axis. The cage covers
the range from m 35 . 0
1
= x to m 35 . 0
2
= x , and there the resulting electric field is relatively
small. Away from the cage, the intensity of the electric field tends to the intensity of the exterior
field.
The intensity of the electric field in the vicinity of the coordinate origin (which is in the
middle of the floor of the shielded space) is, approximately, V/m 073 . 0
interior
= E . This
information can be obtained by examining the main output data file WIRE.OUT, or by tabulating
plotted data in the Plot window. In the latter case, it may be advantageous to enlarge the plot
before tabulating data. Hence, the shielding effectiveness (for points near the coordinate origin) is
dB 29 dB log 20
interior
exterior
10
=
E
E
E
S . (8.1)
)
212 Analysis of Wire Antennas and Scatterers
400
700
7
0
0
2
z
x
y
Figure 8.31 Wire shielding cage. All dimensions are in millimeters.
Figure 8.32 Vertical component of the electric field along the x-axis in the shielding cage of
Figure 8.31 and its vicinity, at 50 MHz.
)
Examples of Analysis 213
8.13 Vertical Dipole Antenna Above Real Ground
The first example of antennas above a real ground is a vertical half-wave dipole, sketched in
Figure 8.33. The operating frequency is MHz 3 = f . The dipole arm length is 4 / = l and the
radius is = 0.0005 R . The ground conductivity is mS/m 3 = (average ground) and the relative
permittivity is 10
r
= . The input data set for this example is DIPOLE. Figure 8.34 shows the
input impedance of the dipole ( X R Z j + = ) against the distance h between the dipole and the
ground surface in wavelengths. (For 0 = h , the dipole touches the ground.) Results obtained by
other Sommerfeld-based methods [13, 27, 28] are shown for comparison.
x
2R
h
O
, ,
,
0 0
0
Figure 8.33 Vertical dipole above real ground.
)
214 Analysis of Wire Antennas and Scatterers
0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.10
30
40
50
60
70
80
90
100
110
120
130
R, X, AWAS
R, X [27]
R, X [28]
R [13]
Z
[
]
h/
Figure 8.34 Input impedance of the vertical dipole of Figure 8.33 above average ground.
Figure 8.35 shows the radiation pattern (for the space wave) of a dipole touching the
ground for m 75 . 0 = l , mm 21 = R , mS/m 10 = (moist ground), 10
r
= , and MHz 100 = f .
The radiated field intensity is normalized by the factor ( ) r I 2
0 0 0
, where
0
I is the driving-
point current intensity. The radiation pattern shows a zero in the horizontal direction, due to losses
in the conductive ground. This does not mean that there is no radiation in the horizontal direction.
Only the field in the horizontal direction (i.e., the surface wave [11, 12, 29]) decays faster than
r / 1 . This field can be accurately calculated by the near field option of AWAS for Windows. The
results are shown in Figure 8.36, for the same dipole parameters as in Figure 8.35. The x and z
components of the electric field are shown against the distance (along the x-axis) from the point
where the dipole touches the ground. The dipole is fed by a matched voltage generator of 1 V emf
and 50 impedance, resulting in the fed power of 4 mW. For comparison, asymptotic results for
z
E , obtained from Van der Pols diagrams [29] are also shown. For >> x , the field decays
approximately as
2
/ 1 r and it can be accurately calculated as the near field even as far as
000 , 000 , 1 from the antenna.
)
Examples of Analysis 215
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90
0.00
0.25
0.50
0.75
1.00
1.25
1.50
1.75
2.00
2.25
2.50
AWAS
[28]
PEC ground plane
E
n
[
o
]
Figure 8.35 Radiation pattern of the vertical dipole of Figure 8.33 touching moist ground.
0.01 0.1 1 10 100 1000 10000 100000 1000000
1E-13
1E-12
1E-11
1E-10
1E-9
1E-8
1E-7
1E-6
1E-5
1E-4
1E-3
0.01
0.1
1
10
E
x
, E
z
, AWAS
E
z
[29]
E
[
V
/
m
]
x/
Figure 8.36 Electric field on the ground surface of a vertical dipole touching moist ground.
, ,
,
0
0
0
Figure 8.37 V-dipole above real ground.
219
References
[1] A. R. Djordjevi, M. B. Badar, G. M. Vitoevi, T. K. Sarkar, and R. F. Harrington,
Analysis of Wire Antennas and Scatterers, Software and Users Manual, Norwood, MA: Artech
House, 1990.
[2] A. R. Djordjevi, B. D. Popovi, and M. B. Dragovi, A rapid method for analysis of
wire-antenna structures, Archiv fr Elektrotechnik, Vol. 61, 1979, pp. 1723.
[3] B. D. Popovi, M. B. Dragovi, and A. R. Djordjevi, Analysis and Synthesis of Wire
Antennas, Chichester, England: John Wiley & Sons, 1982.
[4] A. R. Djordjevi, M. B. Badar, T. K. Sarkar, and R. F. Harrington, Solution of
Two-Potential Equation for Wire Structures Using Polynomial Expansion and Pulse Testing
Functions, Proc. of Int. Conf. on Antennas and Propagation, York, England, 1991.
[5] R. F. Harrington, Field Computation by Moment Methods, New York: Macmillan,
1968. Reprinted by IEEE Press, New York, 1993.
[6] A. R. Djordjevi, M. B. Badar, T. K. Sarkar, and R. F. Harrington, Awas for
Windows: Analysis of Wire Antennas and Scatterers, Software and Users Manual, Norwood,
MA: Artech House, 1995.
[7] W. A. Imbraile, Applications of the Method of Moments to Thin-Wire Elements and
Arrays, in R. Mittra (Ed.), Numerical and Asymptotic Techniques in Electromagnetics, Berlin:
Springer-Verlag, 1975.
[8] E. K. Miller and F. J. Deadrick, Some Computational Aspects of Thin-Wire
Modeling, in R. Mittra (Ed.), Numerical and Asymptotic Techniques in Electromagnetics, Berlin:
Springer-Verlag, 1975.
[9] B. D. Popovi, Electromagnetic Field Theorems (A Review), Proc. IEE, Pt. A,
1981, pp. 4763.
[10] J. Rockway, J. Logan, D. Tam, and S. Li, The MININEC System: Microcomputer
Analysis of Wire Antennas, Norwood, MA: Artech House, 1988.
[11] A. Sommerfeld, Partial Differential Equations in Physics, New York: Academic
Press, 1949.
[12] A. Baos, Dipole Radiation in the Presence of a Conducting Half-Space, Oxford,
England: Pergamon Press, 1966.
[13] E. K. Miller, A. J. Pogio, G. J. Burke, and E. S. Selden, Analysis of Wire Antennas
in the Presence of a Conducting Half-Space. Part I. The Vertical Antenna in Free Space, Can. J.
Phys., Vol. 50, 1972, pp. 879888.
[14] E. K. Miller, A. J. Pogio, G. J. Burke, and E. S. Selden, Analysis of Wire Antennas
in the Presence of a Conducting Half-Space. Part II. The Horizontal Antenna in Free Space, Can.
J. Phys., Vol. 50, 1972, pp. 26142627.
221
Appendix A
Program Limits
This appendix summarizes the upper limits imposed by AWAS for Windows Version 2.0 that
define the program capabilities and the lower and upper limits for input data. For compatibility
purposes, the upper limits for AWAS for Windows Version 1.0 and AWAS for DOS are also
given.
A.1 Upper Limits in AWAS for Windows and AWAS for
DOS
The upper limits listed in Table A.1 define the program capabilities, as they determine the
complexity of wire structures that can be analyzed by AWAS for Windows. For comparison, the
same limits are listed for the previous versions of AWAS.
225
Appendix B
Error Messages
This appendix summarizes error messages issued by the program shell and program kernel while
checking input data and computing results. Trivial error messages, displayed in dialog boxes, and
requiring an obvious users action (e.g., Cannot open file) are not listed. See also Appendix A,
which lists restrictions imposed by AWAS for Windows.
B.1 Error Messages Issued by the Shell
These messages are issued by the input-data checking procedure in the Input Data window. This
procedure can be activated by the user (e.g., by pressing Ctrl+F9), and it is automatically activated
before input data are saved and before computations are started from the Input Data window. The
error messages issued by the data checking procedure in the Input Data window are displayed in
the Error dialog box, and they are sorted below in the order in which tests are made (** or ***
stand for a number supplied by the program when the message is issued).
Missing node index
All node indices in the first column of the Node table and in the second and third
columns of the Segment table must be supplied. Completely blank lines are ignored.
Node index out of range 1 to 999
All node indices must be between 1 and 999.
Missing node coordinate
All node coordinates must be defined (even if a coordinate is zero). Completely blank
lines are ignored.
Nodes *** and *** coincide
Nodes must differ from each other.
Node *** already defined
Node indices must differ from each other.
Node *** under ground surface
If a ground plane is present, all nodes must be above the plane or at the plane ( z 0). For
a real ground and Sommerfelds formulation, 0 > z .
Node index out of range 1 to ***
)
226 Analysis of Wire Antennas and Scatterers
The program evaluates the number of nodes by counting nonblank lines in the Node
table. All nodes between one and this number must be defined. No node index can exceed
the number of nodes, which is supplied in the above message.
Missing segment index
All segment indices in the first column of the Segment table must be supplied.
Completely blank lines are ignored.
Segment index out of range 1 to 999
All segment indices must be between 1 and 999.
Segment *** already defined
Segment indices must differ from each other.
Node *** undefined
The nodes declared as the origin and terminal nodes of a segment must be defined in the
Node table.
Node *** left unconnected
At least one segment must be connected to a node.
Missing segment radius value
All segment radii must be defined. Completely blank lines are ignored.
Segment *** penetrates into ground
If a ground is present, all segments must be above the plane. A segment is assumed to lie
in the ground if both nodes to which the segment is connected lie in the plane 0 = z or if
both are closer to the ground surface than the segment radius.
Segment *** too short versus diameter
A segment is assumed too short if its length is smaller than its radius.
Segments *** and *** too close or intersect
Segments must not be too close to each other or intersect. Two segments are assumed too
close to each other if the smallest distance between any two points on the axes of the two
segments is smaller than the sum of wire radii. An exception is two segments that have
one common node. Such segments are assumed to be too close to each other if their axes
almost coincide in terms of the wire radii.
Port index out of range 1 to 99
All port indices must be between 1 and 99. (Enter 0 for the port index to discard that
port.)
Missing port nominal impedance value
All port nominal impedances must be defined. Completely blank Port dialog boxes are
ignored.
Missing port excitation value
All port excitations (generator electromotive forces and currents) must be defined.
Completely blank Port dialog boxes are ignored.
Port ** already defined
Port indices must differ from each other. (Enter 0 for the port index to discard a port.)
No ports defined for an antenna
An antenna must have at least one port.
Ports defined for a scatterer
A scatterer must not have ports.
No excitation for a transmitting antenna
At least one electromotive force must be different from zero for a transmitting antenna
excited at all ports simultaneously.
Element modulus out of range 0.000001 to 10000000.
231
Appendix C
Help Windows
Listed in this appendix are AWAS Help windows, including the Table of Contents and the Index.
Hotspots are underlined and window titles are boldface.
Table of Contents for Help Windows
AWAS
GENERAL
WINDOWS
NUMBERS
PROGRAM RESTRICTIONS
RANGE OF PARAMETERS
CHECKING
FILE LOADING
FILE SAVING
HEADER
OPERATING MODE
GROUND
GEOMETRY
NODES
SEGMENTS
LOADINGS
PORTS
FREQUENCY
PLANE WAVES
NEAR FIELDS
FAR FIELDS
INPUT DATA
INPUT FILES
POLYNOMIAL APPROXIMATION
OUTPUT DATA
OUTPUT FILES
PLOTTING OUTPUT
MAIN WINDOW
HELP
INPUT FILE MANAGEMENT
INPUT DATA WINDOW
HEADER EDITING
GEOMETRY INPUT
GEOMETRY SHOW
OPERATING MODE INPUT
GROUND INPUT
FREQUENCY INPUT
PLANE WAVE INPUT
NEAR FIELD INPUT
FAR FIELD INPUT
ANALYSIS
BATCH ANALYSIS
LIST OUTPUT FILE
SAVE OUTPUT FILES
GRAPH SELECTION DIALOG BOX
PLOT WINDOW
HARDCOPY
CONFIGURATION
SETUP DIALOG BOX
INDEX
)
232
AWAS
AWAS for Windows is a versatile program for analyzing wire antennas and scatterers
assembled from arbitrarily located and interconnected straight-line segments. The program
consists of a shell for user-friendly interface and a kernel for numerical analysis.
The wire structure can be in a free space, located above a perfectly conducting plane, or
above a real ground. Antennas can be analyzed in the transmitting or receiving modes, resulting in
port matrix parameters, current distribution, near fields, and far fields. The analysis of scatterers
results in monostatic or bistatic cross sections, in addition to the current distribution and near
fields.
Numerical analysis is performed by solving an integro-differential equation for the
current distribution. This equation is solved using the method of moments with polynomial
approximation for the current. See the Users Manual for further information.
AWAS for Windows can be run on a configuration consisting of a Pentium platform
running Windows NT, 98, and later, preferably with at least 64 MB of memory available to the
program, and a hard disk with at least 100 MB of free space.
The structure to be analyzed is defined by a set of points (nodes) and wire segments.
Segments can have ports (for antennas), and concentrated (lumped) and distributed loadings. Also
required for the analysis are the operating frequency, data defining where near and far fields are to
be evaluated, and data defining plane waves exciting receiving antennas or scatterers.
Results of the analysis are written to files, and they can be displayed on the screen in text
or graphic forms.
See also GENERAL,
PROGRAM RESTRICTIONS.
GENERAL
AWAS for Windows consists of a powerful kernel for numerical analysis of wire
antennas and scatterers and a user-friendly interface consisting of a set of windows for data input,
output, and monitoring.
A computation cycle typically consists of defining input data, running the numerical
kernel, and examining results.
All program activities are initiated from the Main window.
The user is assumed to be familiar with Windows and the basic antenna theory.
See also WINDOWS,
MAIN WINDOW.
WINDOWS
Windows and dialog boxes are designed to give a user-friendly interface to the program:
they allow users to input all data required for the analysis, examine results of the analysis by
listing or plotting, monitor computations, perform setup, and so on.
Windows and dialog boxes are organized in a hierarchical manner, starting with the
AWAS Main window.
All windows have menu bars (from which pull-down menus can be opened), buttons, and
hot keys. The Main window also has a toolbar. The dialog boxes have buttons.
and E
.
Plane waves can be organized in several groups. Each group has a unique pair of electric
field components, but the directions of incidence vary within the group. These angles within a
group are ( , )
i j
, where
i
i i n = + =
start
, ..., 0, ;
j
j j n = + =
start
, ..., 0, ;
start
and
start
are the minimal values of the angles;
stop
and
stop
are the maximal values of the
angles; n
and n
stop start
n
and =
stop start
n
are angular steps.
For a scatterer whose monostatic (radar) cross section is evaluated, plane wave groups
are, at the same time, far field groups. In this operating mode, far field points are not specified at
data input. However, when the far fields are plotted, the information about the plane waves is used
to define the far field directions as if they were actually input.
See also PLANE WAVE INPUT,
OPERATING MODE,
FAR FIELDS.
start
and
start
are the minimal values of the angles,
stop
and
stop
are the maximal values of the
angles, n
and n
stop start
n
and =
stop start
n
are angular steps. A pair ( , )
i j
can be considered as a point in the plane
the coordinate axes of which are the angles and .
Far fields can be plotted within a group against one angular coordinate as variable, with
constant the other coordinate, or against both coordinates (3D plot). The number of steps for the
variable angle must be at least 1.
At these points, spherical components of the electric field ( E
, E
and E
, E
, total), gain for antennas, and cross section for scatterers, as well as the
variable for plotting: the spherical angle ( , ) or frequency (f).
In the Near Field and Far Field tabs, select the corresponding group in the Group area
and the value of the fixed coordinates, as applicable.
In the Currents and Charges, Near Field, and Far Field tabs, select the excitation of the
structure in the Excitation area. For an antenna, the excitation can be at ports or by plane waves.
For a scatterer, the excitation is by plane waves. For the excitation at ports, select the port index if
the antenna is analyzed driven one port at a time. For the excitation by the plane waves, select the
plane wave group and the particular wave within the group. In these three tabs, use the Frequency
area to define the frequency for which the selected target is to be plotted. This area is disabled
when plotting the far fields as a function of frequency.
In the Format area, specify the format of the plot.
When the set of selections is completed, click the OK button to obtain the graph.
See also PLOTTING OUTPUT,
MAIN WINDOW,
PLOT WINDOW.
PLOT WINDOW
Once all selections are issued in the Graph Selection dialog box, a plot appears in the Plot
window. This plot can be edited or saved.
There are several kinds of plots: rectangular, polar (Smith chart and far fields patterns),
and 3D (far fields patterns: rectangular and spherical). The following menus are available in the
Plot window menu bar:
File manage files containing data for plotting; this menu has the following options:
Copy copy the plot to the clipboard as a bitmap.
Save as Bitmap save the plot as a bitmap file, with the default extension BMP.
Printer Setup configure printer.
Print obtain a hardcopy of the plot on the printer,
Tabulate (not available for 3D plots) obtain a table with data comprising the plot; this
option opens a menu with the following options:
File place the table in a file with the default name extension TAB.
Printer print the table.
Exit close the Plot window.
Options (not available for 3D plots) supply a title, or change the scale of the plot; this menu
has the following options:
Title edit the plot title.
Scale (rectangular plots only) change the scale of the plot; this option opens a menu
with the following options:
Automatic restore the scale determined by the program.
Manual manually supply the minimal and maximal values of the coordinates.
Impedance/admittance grid (Smith chart only) toggle the impedance and admittance
grid.
Help obtain help; this menu has the following options:
)
254 Analysis of Wire Antennas and Scatterers
Help open the Help window for the Plot window.
Index open the Help window with the Index of available Help windows.
Contents open the Help window with the Table of Contents of available Help
windows.
Using Help open the Help window with instructions on how to use the Help utility.
See also GRAPH SELECTION DIALOG BOX.
HARDCOPY
A hardcopy of the contents of any window of program AWAS for Windows can be made
using standard Windows utilities. The Plot window and the List window have their own access to
the printer.
See also PLOT WINDOW,
LIST WINDOW.
CONFIGURATION
Configuration includes defining the accuracy parameter, specifying units for data input,
formatting the main output data file, and so on, as well as defining default files for frequency,
plane waves, near fields, and far fields.
Configuration can be accessed from the Main window Config menu. Options in this
menu are to edit the four default files, one by one, as well as to perform the setup of all other
items.
The contents of the default files are edited using the same procedure as in the Input Data
window, except that the edited data are automatically saved as default files by clicking OK.
The other items can be setup in the Setup dialog box.
See also MAIN WINDOW,
SETUP DIALOG BOX,
FREQUENCY INPUT,
PLANE WAVE INPUT,
NEAR FIELD INPUT,
FAR FIELD INPUT.
SETUP DIALOG BOX
The Setup dialog box can be accessed from the Main window Config menu or by clicking
the Setup button. It is used to define the default wire conductivity and accuracy parameter (used in
autosegmentation), specify units for data input, format the main output data file (WIRE.OUT), and
define some modes of program execution.
The default wire conductivity can be defined by editing the corresponding field. It is
always specified in MS/m. Use the slide button to define the accuracy parameter. All other items
are set up by clicking the corresponding radio button or the checkbox.
Available units for coordinates are: m, mm, foot, and inch.
Available units for wire radii are: m, mm, foot, and inch.
Available units for frequency are: Hz, kHz, MHz, and GHz.
Available units for electric field are: V/m, mV/m, and V/m.
Complex numbers can be treated as rms or peak values.
257
Index
About AWAS dialog box, 134
Accuracy parameter, 174, 175
Admittance parameters. See Port admittance
parameters
Analysis, 29, 45, 48, 90, 133, 141, 153, 251
Angular frequency, 60
Antenna array, 68, 197
Automatic segmentation, 48, 86, 94, 108, 117,
229
Automatic show, 29, 146, 174, 175
Autosegmentation. See Automatic segmentation
AWAS, 1, 85, 88, 232
for DOS, 3, 221
for Windows, 1, 3, 221
AWAS_W_2.DLL program, 153
Basis functions, 64
Batch analysis, 30, 49, 110, 137, 155, 156, 251
Batch dialog box, 52, 90, 133, 155, 227
BMP file, 170, 173
Boundary conditions, 61, 63, 67, 73
extended, 62
Broadband antennas, 199
Broadcast antenna, 206
Charge distribution, 68, 108, 122, 158, 162
Check dialog box, 46
Checking input data, 45, 107, 153, 234
Clipboard, 174
Complex amplitude, 28, 60, 97, 107, 177, 255
Complex domain, 60
Complex numbers, 107, 174, 175
Complex peak value. See Complex amplitude
Complex rms value, 28, 37, 60, 97, 107, 177, 255
Computation cycle, 29, 85, 90, 232
Computations. See Analysis
CONFIG.DAT file, 136, 138, 177, 227, 255
Configuration, 134, 174, 254
Continuity equation, 62
Control-menu box, 16, 89
Cross section, 42, 75, 86, 99, 110, 127, 158, 166
bistatic, 76, 92, 110, 130
copolar, 76
cross-polar, 76
monostatic, 76, 92, 110, 130
total, 76
CUR file, 110, 136, 137, 163, 242
Current distribution, 68, 86, 91, 108, 122, 158,
162
Data input, 132, 138, 244
Data Source area, 160, 253
Default conductivity, 174, 175
Default Far Field dialog box, 134, 136, 137, 153,
179, 251
DEFAULT files, 40, 47, 100, 134, 135, 137, 174,
178, 254
Default Frequency dialog box, 134, 136, 137,
148, 178, 249
Default Near Field dialog box, 134, 136, 137,
151, 179, 250
Default Plane Wave dialog box, 134, 136, 137,
150, 179, 250
Default setup, 178
DEFAULT.FFP file, 251
DEFAULT.FRQ file, 249
DEFAULT.NFP file, 250
DEFAULT.PWE file, 250
Delta-function generator, 67
Dialog boxes, 89, 232
Directive gain, 75
Diskette
distribution, 9, 87
Edit
far field input data, 42, 151, 251
frequency data, 39, 147, 249
geometry data, 247
ground, 38, 142, 249
header, 142, 246
loading, 36, 247
near field input data, 43, 150, 250
)
258 Analysis of Wire Antennas and Scatterers
nodes, 35, 247
operating mode, 38, 142, 248
plane wave input data, 44, 148, 250
plot, 171, 173, 253
plot title, 56, 254
port parameters, 37, 247
segments, 35, 247
Efficiency, 73, 75, 91, 109, 199, 202
Electric scalar-potential, 61
Electromotive force. See Port electromotive force
Ends. See Wire ends
Error dialog box, 108, 132, 141, 153, 225
Error messages, 90, 225
Evaluation of integrals, 76
Excitation area, 160
Exit dialog box, 133
Expansion functions, 64
Far Field dialog box, 40, 136, 137, 151, 179, 251
Far field directions, 98
Far field input, 151, 251
Far field input data, 40, 98, 105, 139, 239
Far fields, 29, 42, 68, 74, 76, 86, 98, 107, 109,
118, 127, 158, 166, 223, 239
Fed power, 72, 75, 109
FFL file, 110, 136, 137, 168, 242
FFP file, 30, 100, 105, 110, 118, 120, 121, 122,
125, 127, 135, 136, 152, 240
Field polarization, 44, 76, 96, 194
File loading, 30, 135, 234
File management, 135, 244
File Name box, 136, 138
File Name menu, 21, 31, 132, 136, 138, 139, 234
File Overwrite dialog box, 133, 138, 140
File saving, 30, 136, 235
Find dialog box, 157
Format area, 55, 160, 253
Format WIRE.OUT, 28, 174, 175, 255
Frequency area, 32, 39, 136, 137, 139, 147, 160,
178, 249, 253
Frequency domain, 60
Frequency input, 147, 249
Frequency input data, 39, 103, 238
FRQ file, 30, 100, 103, 135, 136, 148, 240
Gain. See power gain
Gauss-Legndre integration formulas, 77
Generators, 59, 61, 63, 66, 68, 185
GEO file, 30, 100, 133, 135, 136, 140, 240
Geometry, 29, 93, 236
Geometry input, 143, 247
Geometry show, 146, 248
Goto dialog box, 158
Graph Selection dialog box, 24, 30, 53, 56, 133,
136, 159, 253
Greens function, 62
Ground, 64, 86, 92, 100, 142, 235
Ground dialog box, 38, 141, 143
Ground plane, 29, 75, 92
Hardcopy, 174, 254
Hardware requirements, 7, 86
Header, 31, 91, 100, 117, 139, 142, 235
Header Edit dialog box, 31, 141, 142
Help windows, 16, 17, 134, 135, 231, 244
Hotspots, 135, 231
Images, 64
Impedance parameters. See Port impedance
parameters
Impressed field, 61, 66, 70
Individual analysis, 154, 251
Inner product, 65
Input data, 30, 90, 117, 240
Input data files, 30, 47, 99, 135, 136, 240, 244
Input data set, 30, 47, 90, 99, 135, 154, 183
Input Data window, 20, 30, 132, 135, 136, 138,
143, 178, 184, 225, 244, 247
Integers, 107, 233
Isotropic radiator, 75
Junctions. See Wire junctions
Kernel Error Notification dialog box, 227
Kirchhoffs current law, 66
List output, 52, 133, 157, 252
List window, 30, 52, 90, 133, 136, 157, 174, 252
Load dialog box, 136, 234
Load Far Field dialog box, 42, 152
Load Frequency dialog box, 40
Load Geometry dialog box, 21, 31, 50, 132, 139
Load Near Field dialog box, 151
Load Output File dialog box, 157
Load Plane Wave dialog box, 149
Loaded antennas, 199
Loading dialog box, 36, 144, 145
Loadings, 23, 29, 34, 59, 63, 86, 93, 100, 139,
143, 145, 222, 237
Loop antenna, 194
Loss power, 72, 109
L-U decomposition, 67
Magnetic vector-potential, 61
Main window, 89, 131, 135, 136, 242
Menu bar, 16
Method of moments, 1, 29, 59, 60, 64, 65, 85
Monopole antenna, 183, 197, 199, 206
Near Field dialog box, 43, 136, 137, 150, 179,
250
Near field input, 150, 250
Near field input data, 42, 97, 105, 139, 239
Near field points, 97
Near fields, 29, 73, 86, 91, 97, 107, 109, 118,
125, 158, 164, 223, 239
New Input dialog box, 140
Index 259
NFL file, 110, 125, 136, 165, 242
NFP file, 30, 100, 105, 135, 136, 151, 240
Node table, 23, 31, 34, 139, 143
Nodes, 23, 29, 34, 59, 93, 100, 107, 139, 143,
222, 236
Norton equivalent network, 71
Nuclear electromagnetic pulse simulator antenna,
207
Numbers, 233
Operating frequency, 29, 61, 95, 103, 107, 139,
223
Operating mode, 38, 91, 100, 139, 142, 235, 248
Operating mode code, 92, 93
Operating Mode dialog box, 38, 141
OUT file, 110, 133, 136, 137, 242
Output data, 49, 241
Output data files, 49, 50, 110, 242, 252
Output data set, 29, 49, 50, 90, 110, 137, 154,
155
Packaging list, 3
Phase coefficient, 62
Plane Wave dialog box, 44, 136, 137, 148, 179,
250
Plane wave directions, 96
Plane wave field components, 96
Plane wave input, 148, 250
Plane wave input data, 44, 103, 139, 238
Plane waves, 29, 59, 61, 72, 75, 86, 91, 96, 107,
223, 238
Plot data set, 90
Plot labels, 172
Plot output, 53, 133, 158, 242, 253
Plot scale, 171
Plot Scale dialog box, 56, 171, 172
Plot title, 173
Plot Title dialog box, 56, 173
Plot window, 25, 30, 55, 57, 169, 174, 253
Polynomial approximation, 1, 60, 64, 85, 94, 241
Polynomial coefficients, 64, 67, 68, 86, 94, 107,
108, 118, 122, 185, 241
Polynomial degree, 64, 66, 67, 68, 94, 108, 241
Port admittance parameters, 29, 68, 69, 86, 91,
109, 118, 120, 158, 160
Port dialog box, 37, 144, 145
Port electromotive force, 34, 37, 102
Port impedance parameters, 29, 68, 69, 86, 91,
109, 118, 158, 160
Port index, 145
Port matrix parameters, 29, 68, 160
Port nominal impedance, 34, 37, 69, 70, 102,
119, 145, 222
Port scattering parameters, 29, 68, 70, 86, 91,
109, 118, 158, 160
Ports, 23, 29, 34, 59, 63, 66, 68, 86, 91, 93, 100,
107, 139, 143, 145, 222, 237
Power, 107
Power balance, 72, 118
Power gain, 42, 75, 91, 99, 109, 127, 129, 158,
166
Poynting vector, 75
Print Confirm dialog box, 157, 170
Print Setup dialog box, 174
Printer Options dialog box, 174
Printer Setup dialog box, 157, 170
Program
distribution, 2
environment, 2, 7, 86
features, 85
help, 16, 17, 134, 135, 244
installation, 7, 9, 88
kernel, 85, 88
limits, 221
restrictions, 107, 233
shell, 85, 88
starting, 15, 17
stopping, 16, 26, 133
Pseudo-singularity, 77
PWE file, 30, 100, 135, 136, 149, 240
Radiated fields, 68, 74
Radiated power, 72, 109
Radiation pattern, 42, 86, 91
Radius. See Wire radius
Range of parameters, 233
Real ground, 29, 86, 92
Real numbers, 107, 233
Reduced kernel, 74
Save Default dialog box, 148, 150, 151, 152
Save dialog box, 138, 235
Save Far Field dialog box, 42, 152
Save Frequency dialog box, 148
Save Input Files dialog box, 133, 137, 140
Save Near Field dialog box, 151
Save output, 133
Save Output Data dialog box, 51, 133, 137
Save Output Files dialog box, 51, 133, 137
Save Plane Wave dialog box, 149
Save plot, 173
Save Plot dialog box, 170
Save Tabulated Data dialog box, 170, 173
Scale plot, 56
Scattered fields, 76
Scattering parameters. See Port scattering
parameters
Segment axis, 63, 65, 93, 164
Segment origin, 63
Segment table, 23, 32, 34, 139, 143
Segment terminal, 63
)
260 Analysis of Wire Antennas and Scatterers
Segments, 23, 29, 34, 59, 63, 93, 100, 107, 139,
143, 222, 236
Setup, 27, 134, 174, 175, 254, 255
Setup dialog box, 27, 134, 138, 174, 175, 255
SETUP.DAT file, 136, 138, 177, 181, 255
Shielding cage, 211
Shielding effectiveness, 211
Show window, 20, 35, 139, 146, 157, 170, 248
Software requirements, 7, 86
Sommerfeld, 59, 64, 77, 92, 93, 102, 143, 213,
216
SPM file, 110, 118, 120, 121, 122, 125, 127, 136,
137, 161, 242
Steady state, 60
Surface charges, 61
Surface currents, 61
System of linear equations, 67
TAB file, 136, 137, 170, 173
Tabulate plot, 173
Testing functions, 65
Thvenin equivalent network, 71
Thin-wire kernel, 62
Title dialog box, 171
Toolbar, 16, 134
Total number of polynomial coefficients, 65, 95
Tower antenna, 206
Transmission line, 207
Trapezoidal rule, 66
Two-potential equation, 1, 29, 60, 61, 63, 65, 85
Units for data input, 28, 144, 147, 149, 150, 174,
175, 255
Waves and Fields area, 32, 40, 42, 44, 139, 149,
150, 152
Weighting functions, 65
What area, 54, 160
Windows, 89, 232
Wire axis. See Segment axis
Wire ends, 59, 63, 65, 66, 185
Wire junctions, 59, 63, 65, 66, 93, 103, 185
Wire model of electric circuit, 196
Wire radius, 34, 62, 93, 102, 143, 222
Wire scatterer, 190, 209
WIRE.CUR file, 110, 122, 136, 154, 163, 242
WIRE.FFL file, 110, 127, 136, 154, 168, 242
WIRE.FFP file, 100, 154, 227
WIRE.FRQ file, 100, 103, 154, 227
WIRE.GEO file, 100, 154, 227
WIRE.NFL file, 110, 125, 136, 154, 165, 242
WIRE.NFP file, 100, 105, 154, 227
WIRE.OUT file, 110, 133, 136, 153, 154, 157,
227, 242
WIRE.PWE file, 100, 103, 154, 227
WIRE.SPM file, 110, 118, 136, 154, 161, 242
WIRE.YPM file, 110, 120, 136, 154, 161, 242
WIRE.ZPM file, 110, 121, 136, 154, 161, 242
Wire-grid model, 209
Work window, 24, 49, 154
Yagi-Uda antenna, 203
YPM file, 110, 136, 137, 161, 242
ZPM file, 110, 136, 137, 161, 242