Solution of optical and microwave problems using HFSS by Banks SE, Gutzeit EM. Kurushin AA (z-lib.org)

Notice that this book is not yet fully corrected and might have some English
grammar mistakes ,but the content is understood .
UDC621.3.049.77.029: 681.3.06
This book is a collection of problems,in which the analysis of a number of microwave
structures of great practical importance. For the first time describes the HFSS Ansoft version 13
software.
HFSS software version 13 is designed for three-dimensional design of microwave devicesand
uses several methods of calculation. At the decision of practical problems, more attention is paid
to the peculiarities of calculation methods and installation of HFSS software options in the course
of constructing three-dimensional models of the waveguide, and microstrip antenna structures.
A number of original heterogeneous structures, filters and contemporary antennas having a linear
and circular polarization analyzed using HFSS. The solution of physical problems associated
with optics, radar, radio physics.
For technical workers, studentsand postgraduate students in the design of microwave
devices and methods for calculating electromagnetic fields in inhomogeneous structures.
atthe need for more detailed information on the proposed CAD, you can take part in seminars
held by the authors at the Training and Counseling Center LLC "Orkada". A preliminary
application for training as well as for the purchase of the program you can send by email.
address:[email protected], telephone +7 (495) 943-5032 and by fax: +7 (495) 943-6032.
The authors express their gratitude to "Orkada" LLC for financial support in the publication
of manuals.
UDC621.3.049.77.029: 681.3.06
Banks SE,
GutzeitEM.
Kurushin AA
Ltd"Orkada" - mock edition
Content
introduction ............................ .............................................. ......................3
1. Modeling nanostructure in an optical frequency range ...... ..7
2. Waveguide array ..................... .. ...................... 18
3. The antenna array of antennas Vivaldi .................................... ... 33
4. The antenna array on the dipole antenna ............................ ... 43
5. Modeling of the frequency-selective surface ................... ... 60
6. Falling plane wave the object and the calculation of RCS ........................ ... 74
7. Calculation EPR object the size of a large electric .......... ...... ... 90
8. Bandpass waveguide filter ........................................... ... .100
9. accounting facilities heating temperature in the HFSS-13 ...........................
116 10.Realizatsiya adjustment mode in HFSS-13 ............................... ... 126
11. ModelingConnector ................................................. ... 132
12. Antenna, mounted on the mast ........................................ ... 141
13. Calculationtemporal process in a microwave integrated circuit .. ............ ... 148
14. Analysis horn antenna in the time domain ........................ .171
15. Design of nanoscale LED modules using electrodynamic simulation
programs ........................... ... 191
16. installationcalculating a distributed configuration on multiple computers
........................................................................... ..220
Conclusion ........................................................................... .239 References
............................................................... ............ .240
About the authors:
Banks Sergey E.- Doctor of Science, Ch. Scien. et al. IRE. RTF graduated from the
Moscow Energy Institute in 1981, graduate in 1986. A specialist in the field of
microwave equipment and antennas, an expert in the field of microwave CAD. The
author of several monographs, textbooks, 150 scientific articles and 20 patents.
Guttsayt Eduard Mihaylovich- Professor of Department. "Light" MEI, RTF
graduated from the Moscow Energy Institute. Author of books and monographs in
the field of microwave electronics and lighting. The initiator of the introduction of
the achievements of microwave technology in nanotechnology.
Kurushin Aleksandr Aleksandrovich- Ph.D., Associate Professor of Department.
AUiRRV MEI. RTF graduated from Moscow Power Engineering Institute in 1979,
graduate in 1985 Ph.D. (1991), thesis "Design of transistor microwave amplifiers
with high dynamic range." Since 1996, he taught in various aspects of microwave
MIEM, MIREA and MEI. The author of 12 textbooks and 70 scientific articles.
introduction
HFSS v. 13 - this is the electromagnetic field calculation program for the
design of microwave structures having multiple calculation algorithms [1]. The
latest version of HFSS software performs calculations using finite element method
in the frequency domain, transient, uses the method of integral equations, as well as
a hybrid approach: the finite element method + method of integral equations.
Each method in HFSS is implemented as a program in which you want to
create a structure under study, set the parameters of materials and calculated
characteristics. After that HFSS generates a mesh for solving the problem of the
finite element method. In HFSS program grid is generated adaptively in dependence
on the characteristics of the structure and characteristics of the field therein.
In HFSS v13 made a big step forward compared to previous versions of the
program, developed by the firm Ansoft. It modifications made mesh generation
algorithms and calculation algorithms. A new fast and stable algorithm generates a
TAU better tetrahedral grids.
Forming a system of equations, providing a mixed order of its units, as well
as decomposition of an arbitrary region solutions,
allow
implement in
HFSS capabilities high-performance calculation (HighPerformance Computing HPC). The program drawing three-dimensional model has
been improved operations such as insertion and transfer of two-dimensional and
three-dimensional models (imprinting), and the interface has been modified to better
use and automation.
HFSS calculates a wide range of external devices and the parameters of the
microwave antennas, which include electrical and magnetic field, currents, Sparameters, near- and far-field and can also calculate the transient and time change
of electromagnetic fields [2-4]
Developers can be assured accuracy HFSS when designing devices that
include passive and active being introduced "chips" and simulate thus active
antennas, multilayer microwave integrated circuits, RF / microwave components
and biomedical devices (Fig. B.1, V. 2).
The new properties are HFSS 13.0:
 New, sustainable method of partitioning into tetrahedra;
 implementationcurved elements;
 Calculationderivatives characteristics change upon variation of the
variables;
 Reading Files ACIS R19. 2 (19 version);
 Improved communicationwith the program ANSYS DesignXplorer;
 Calculationtransitional regime;
 implementationhybrid finite element method and the integral
equation method;
 Integration with a common platform ANSYS;
 Multiprocessor seal partitioning grid;
 Improvedpostprocessing data processing;
 Conclusion of broadband characteristics of the studied structures.
Fig. IN 1.Model of five-waveguide filter with the calculated complex ANSYS in
temperature distribution
It is also importantthat by HFSS -13 program included a number of examples
that can be used as templates, and which show the new features of the program.
HFSS uses as the main tool for solving electrodynamic problems finite element
method. In this method, the entire volume is divided into tetrahedrons, inside which
the field is represented as a volumetric basis functions with unknown coefficients
which are found by solving the system of linear equations.
In HFSS v13 software module added HFSS-IE, which implements the method
of integral equations, which uses a two-dimensional basis functions describing the
currents on the surfaces, including objects with a finite conductivity, which allows
to describe
dielectric and metal objects with losses. This method is often referred to as the
method of moments (MOM). HFSS-IE module is designed for the simulation of
large radiating structures. This method can be done in HFSS:
• calculationScattering of the radar cross section (Radar Cross Section RCS);
• calculationantenna located on a large object such as a vehicle;
• calculationcoupling factor between the remote antennas;
• calculationelectromagnetic compatibility;
• calculation of multi-port microwavestructures and antennas.
Methodintegral equations in the field because of the nature calculates
formulations boundary problem in infinite space [5,6]. Therefore, it does not need
surfaces, simulating the absorption of radiation fields: radiation boundary, perfectly
matched layer, etc. This eliminates the need for the user program to surround
emitting object closed air volume (box) with radiation conditions on surfaces. Also,
it becomes possible to calculate antennas located on the ground surface with finite
conductivity.
This method works in
standard HFSS interface. Will list its features:
• emitting (open) objects are described without the air box;
• there is able to analyze objects, comprising an endless flat structure, including
endless screens, etc .;
• the program allows the use of discrete sources and excitation source in the form
of incident plane waves;
• It supports the calculation in the frequency range, defined as discrete points
and interpolation mode;
• the program calculates the near and far fields;
• for structures with large dimensions electrical HFSS-IE uses compression
techniques discharged matrix accelerating computational process;
• perhaps
separation
model
on composite
using different calculation methods.
parts
analyzed
In HFSS 13 is easy to implement insertion of projects of different types to
each other. To insert a project or HFSS HFSS-IE, just click Insert
HFSS Design or Insert HFSS-IE Design and the new project appears in the project
tree under the name HFSSDesignn or HFSS-IEDesignn, where n - number of added
project in the order of its appearance in the overall project. The project can include
more than one installation at a decision. And you can specify methods for solving
HFSS and HFSS-IE, adding designs, analyzed by different methods. Each
installation solutions include the following information:
 the general data concerning the decision;
 grid partitioning seal parameters, if needed, to partition the mesh was sealed
in regions with high speed field changes;
 analyzing frequency range.
In solving HFSS-IE by tasks you can import calculation performed by finite
element method in the problem that will be solved by the method of integral
equations. To do this, you can import calculated in HFSS far field Far Field Wave
or near-field Near Field Wave. The project can include more than one installation
at a decision. And you can specify methods for solving HFSS and HFSS-IE, adding
structure to be solved in different ways.
Fig. AT 2.The spatial radiation pattern of the antenna array elements in which the
HFSS-13 may be slotted, vibrator, spiral, Pace-antennas, Vivaldi and others.
Radiating elements
The HFSS-13 introduces a new view of the port - port Floquet. It is used for
modeling of periodic structures. The use of this port discuss in the next chapter,
which will be simulated infinite boundary between two dielectric media.
Thank you for fruitful discussions, screening and discussion of the manuscripts
Ph.D., Leading Researcher Mishustina BA and Ph.D. Podkovyrin SI
1. modeling of nanostructuresin the optical frequency range
When designing the optical devices is often the problem arises of optimizing the
characteristics of reflecting surfaces [7,8]. As a promising reflective surface in
research and applied research using a periodic structure consisting of metal cones
(see. Fig. 1.1). This structure is considered promising for the development of optical
amplifiers, frequency converters and other devices.
a)
b)
at)
Fig.1.1. Scheme a) topology and b) an optical amplifier, c) a fragment of the
nanostructure pattern used to create optical amplifiers
presentedFig. 1.1 Implementation of the reflecting surface is not the only
possible. Used as structures with hemispherical ends and elliptical, disk, etc. In all
cases before the electrodynamic simulation system is the problem of calculating the
frequency characteristic of the reflection coefficient of the periodic structure.
Formulation boundary problem as follows: in an infinite periodic structure
XOY plane falls from the upper half plane wave. Required to find the scattered field
structure.
naturalanalysis method of infinite periodic structures is to use periodicity
conditions that reduce the problem of the endless structure to the analysis of one
period.
It should be noted that the analysis of incidence of a plane wave at an infinite
periodic structure is a relatively new object in the application such as the software
HFSS. So, before you go directly to the calculation of the structure shown in Fig.
1.1, it is advisable to numerically using the device of periodic boundary conditions
to solve test problems and compared with the numerical solution of the analytic. In
the interface between two media was chosen as the test structure. For it is known
rigorous analytical solution in the form of coefficients of reflection and transmission
plane waves [6, 7], which shall be compared with the numerical results.
Falling plane wave at the interface between two media.When a plane wave at the
interface of two media, its polarization is not changed. If the vector E lies in the
plane of incidence, such a polarization is called parallel, if the vector E is
perpendicular to the plane of incidence - the perpendicular polarization.
Forsimulation of infinite periodic structures in HFSS program uses periodic
boundary conditions, which are set on opposite sides of the box, covering the period
analyzed structure. Example of the boundary conditions of the type shown in Fig.
1.2. Upper vacuum parallelepiped is filled with a dielectric constant equal to one,
the bottom - with a relative dielectric constant of 2.25. boxing center distance 0X
0Y and is in this example 320 nm.
Endlessthe boundary between two media, strictly speaking, is not periodic
structure. Therefore, application of the apparatus of periodic boundary conditions
for analysis needs to be clarified.
Consider the known [6] representation of the field within the rectangular
channel Floquet. Under the channel portion understood Floquet space bounded by
vertical walls (see. Fig. 1.2) on which periodic boundary conditions are set. For a
rectangular channel
Floquet componentsfieldsE, Hat
z0 are as follows
E(X y,
Z)
 n.m e
mz
.
a
n
W
her
e
i  nxi myn,
m
Lx.y- repetition periods structure axes 0X and 0Y (Fig.1.2)
n 0 2n/ Lx.
n,
m
m02m/ Ly.
k2 2.
2
(1)
amplituden n, m
the wave number of free space.
an, m -
mharmonics,k-
standing
plane wave:
0. 0
are set incident field, which has the form
0ksincos.
 0ksinsin.
- meridional angle of incidence of a plane wave, and- the azimuth angle.
Row(1) it is in the theory of periodic structures field decomposition on the
Floquet harmonics, each of which is a member of (1). Can
see what the Floquet harmonicszero indices
nm0 that
often callthe fundamental harmonic has the field structure, which coincides with the
field of the incident wave. Since we are at the halffield of the reflected wave.
You can record a similar decomposition(1) when z halfspace0. The main
harmonic in this expansion will correspond transmitted (refracted) wave.
higher harmonicsorders necessary for describing the near field, is excited, if
the period has a complicated structure, such as shown in Fig. 1.1 b, c. In the case of
a flat surface type electrodynamic problem known solutions [2,6,7]. Field describes
the reflected and refracted waves. Floquet harmonics are not excited.
Thus, the use of periodic boundary conditions for the analysis of the
interfacetwo media is justified by the fact that the reflected and refracted waves are
described Floquet fundamental harmonics and higher order harmonics are not
excited.
Fig. 1.2. A fragment of the dielectric substrate on which a plane wave.
It is interesting to note that the fieldthe fundamental wave does not depend on
periods Lx, y. Therefore, when solving test problem the size of individual cells can
be selected arbitrarily, as depending on the size of the periods Floquet harmonics of
higher orders, as noted above, are not excited.
Decisionthe problem of a plane wave using HFSS. Consider further the decision of
test tasks of a plane wave incident on the boundary between two media with the help
of HFSS.
Drawing structure. Setting the periodic boundary conditions.
installationFloquet ports.
team Modeler-> Units derive the dialog shown in Fig. 1.3, wherein the
length measuring unit ask - nanometer (nm).
Fig. 1.3. Sizing unit structure
Draw a boxsize of 320 nm x 320 nm and 200 nm command Draw-> Box height
(Fig. 1.4). Dimensions box in the plane XOY insignificant. We chose them equal to
the period of the structure shown in Fig. 1.1. Created box located in the lower half
and filled with a material with a dielectric constant different from the unit.
Fig. 1.4. Boxing drawing - a fragment of the dielectric plane
Define its parameters by pressingon the Add Material button (Fig. 1.5).
el
ev
en
Fig. 1.5.Adding a dielectric constant of 2.25 at the project
Similarly, we define the topthis box, box size of 320 nm x 320
nm and 800 nm in height. This box has a dielectric filling.
Click Next F key to move the objecta selection mode and set at the sides of the
vertical sides of the upper and lower boxes periodic boundary conditions Master and
Slave (Figure 1.6).
Fig. 1.6. Formulation periodic boundary conditions on the side of the boxes
In this way,in the structure will be present 4 pairs Master-Slave type surfaces.
Forsolutions of the boundary problem in HFSS system on horizontal surfaces,
limiting the Floquet channel vertically, you must install the ports, which are called
Floquet ports. To install the Floquet port select the bottom surface of the lower box
and define the command Assign Exitation -> Floquet Port, on which opens a dialog
Fig. 1.7.
el
ev
en
Fig. 1.7. Installation Floquet ports on the lower and upper side
Optionsports are set in the dialog shown in Fig. 1.8. This dialog allows you to
work with multimode Floquet ports. Therefore, it is possible to specify the number
of harmonics, which is used to determine the parameters of the port. We are working
with a two-mode Floquet ports. This mode corresponds to a channel with two
propagating Floquet harmonics. Both of them have an index n = 0, m = 0, which we
will set in this dialog. The difference between the two propagating Floquet
harmonics is their different polarization. In our case, we have a wave with parallel
and perpendicular polarizations. The terminology used in the HFSS these two waves
are designated as TM and TE waves. Therefore, the line number of waves (number
of modes) set figure 2 (two waves) and as a select wave modes TM (H) and TE (E).
Fig. 1.8. Installing Floquet port options for solving the problem
In HFSS Floquet ports can be put only on the isotropic medium with a
dielectric constant equal to unity. Therefore, set the port, we can not directly on the
bottom surface of the dielectric.
12
Us you need to create a certain air gap between the medium and the port.
IfFloquet port between the lower and the dielectric medium is a gap (Fig.
1.12), the calculation error is introduced due to the reflection of the waves from the
new interface. The need to create such a gap can be considered HFSS software flaw.
Forto a secondary reflection excluded, put in place from the bottom port of the
absorbing boundary Radiate, in contact with the underside of the dielectric board.
In this case we lose the ability to calculate the coefficient of transmission through
the structure, because the port at the bottom of said channel Floquet replaced
boundary. Nevertheless, the ability to calculate the reflection coefficient is stored.
Incident plane wave is specified by the commandAssign Exitation ->
Incident Wave -> Plane Wave. dialog appears for this team Fig. 1.9 General Data
tab which establish the coordinates of the point at which the incident wave has zero
phase.
Fig.1.9. The General tab of the incident wave dialogue
The angle of incidence associated with the angle of directionEoThat is givenin
the form of coordinates of the vectorsk (-sin (Teta), 0,
sin (Teta)), Eo
(- cos (Teta), 0, sin (Teta)), are setIncident Wave Source in the dialog
(Fig. 1.10). Teta - the variable that defines an angle of incidence of the wave.
13
Fig. 1.10.Parameterization of the angle of incidence of a plane wave:
propagation vector k and the electric field vector E.
To calculate the reflection characteristicsand refraction of the incident wave at
different angles of incidence, will change the angle of incidence of a plane wave
This is doneusing parametric analysis. Command Optimetric-> Add->
Parametric derive dialog (Fig. 1.11) setting parameters of the parametric assay (Teta
parameter busting).
Fig. 1.11.Target parametric analysis when changing the angle of incidence
As a result of this analysis, we obtain the field characteristics, pattern field
angles of reflection and refraction of electromagnetic waves of various incidence
angles.
Known analytical dependence for the reflection coefficientand passing
(transmitting) at a wave incident on the boundary between two media, i.e., provided
that the thickness of the board tends to infinity. The reflection coefficient of the field
for a wave with polarization parallel (Fig. 1.12) is equal to [7]

cos sin2 
R 
cos sin2 .
(2)
14
- falls
angle of incidence.
field transmission coefficient for a wave with polarization parallel to equal
T 
2 cos
cos sin2
(3).
By virtue of these features of HFSS, do not allow to install (with [14,15] avg.)
Floquet port to the surface with a dielectric constant different from unity, we have
to limit the dielectric layer thickness d and set Floquet port in the surface layer
separated from the bottom surface of the small gap. In this case, the formula (2) and
(3) direct comparison can not be used with the calculated data. They must be
modified with the second boundary of the dielectric layer. Such a modification is
not difficult to perform, using the relations (2) and (3). As a result, we obtain the
following expressions for the coefficients
reflectionand passing through the layer plane wave
Rs, Ts:
2
R
s

R

2ikd
T R
e
.
2 2i kd
(4)
1R e
T2
T 
e
s
2 2ikd
1R e
ikd
.
 sin2.
W k2/ - wave numberfree space. In the formulas (4)
her
e
und R,
er T
understood as the reflection coefficientsand the border
section two media for waves of both polarizations. For example, for a wave
R ,T
parallel polarizationas
R, musttake parameters
(2), (3).
T
Calculations performedon HFSS and characteristics calculated by the
formulas (4) are compared in Fig. 1.12. The curves in Fig. 1.12 plotted for different
15
values of the layer thickness d = 100, 200 and 300 nm. The dots show the data
obtained using HFSS, and solid curves of formulas (4). It can be seen that
computational and theoretical results are in good agreement.
16
Fig. 1.12.Calculated on HFSS reflectance in the range of incidence angles for
different substrate thicknesses
Therefore, we can conclude that for modelingincidence optical range
electromagnetic waves on metal surfaces with a complex structure can be used the
finite element method, implemented in HFSS.
Fig. 1.13.The reflection coefficient of the incident plane wave with
vertical polarization
Making sure that the same numerical calculationswith the theoretical results, it is
possible to perform surface modeling, consisting of the conical metal nanostructures
are used to enhance luminescence in an optical amplifier (Fig. 1.13).
17
silver surfacelocated on them system nanoscale sharp points, is used to
amplify the luminescence mediums adsorbed with the ions of rare earth elements
caused by plasmons. With HFSS-13 can obtain parameters of such systems are
widely used at present in optical fiber amplifiers to a wavelength of 1.54 microns
[4].
18
2. The waveguide array
Simulated waveguide array [1,13].Its radiating aperture is made up of the open
ends of the square waveguides (Fig. 2.1). We believe that the axis 0z axes directed
along the waveguides. The plane z = 0 coincides with the aperture grille.
The sequence of operation is as follows:
 creature
separate
emitter
at form of segment square
waveguide;
 the creation of boxing, located on top of the radiator;
 assignment of boundary conditions PerfectEon the walls of the waveguide;
 the task Floquet ports on the face of boxing;
 job waveguide port;
 install analysis;
 launch in the calculation;
 output characteristics.
Fig. 2.1. The geometry of the antenna array
The geometry of the antenna array shownFig. 2.1. We consider a model in the
form of an infinite grating. Waveguide ports are in
the
z0 and have the form of square waveguide ports (Fig. 2.2) with
main
wave
linear
polarization.
Direction
field vectorthis wave is shown in Fig. 2.1 in the form of arrows.
electric
19
Fig. 2.2. Model HFSS single element antenna array coordinate vectors
infinite array analysis can be reducedAnalysis for one period through the
Floquet theorem mentioned in Chapter 1. Note that the concept of an infinite lattice
directional diagram (DS) can not be correctly identified, since it is introduced only
to the radiation source with finite dimensions. Strictly speaking, the NAM infinite
array describes the delta - function and physical does not make sense. Nevertheless,
there are approximate techniques to use the results of analysis patterns for endless
evaluation finite lattice misfit dislocations. On them will be discussed later. Just
infinite array analysis gives an indication of its task of coordination with a free
space.
Fig.2.2 shows the geometry of one cell of an infinite array. The model consists
of two areas. The lower part is a waveguide, and on top of it is an air box. periodic
boundary conditions are set on the vertical faces of the box. On the upper surface of
the box is given by the Floquet port.
Creating a model of a separate emitter.To create a separatecell array, perform the
following steps:
1. Open the projectand give it a name AGW.
2. team
Draw> Box create
arbitrary
boxing, and then
edit its settingsunder Edit> Properties (Figure 2.3)
3. Selectcreated edit box and its transparency (imaging parameter generated by the
object) Transparency = 0.8.
20
Fig. waveguide 2.3.Parametry
Fig. 2.4.Parametry air box
4. surgery
Draw> Box create the second box, and edit the dimensions
and properties Edit> Properties (Fig. 2.4).
The taskMaster and Slave boundaries.
1. presson the F key, translating HFSS in surface selection mode, select the side of
the upper box and type the command HFSS> Boundaries> Assign> Master. A
dialog Master border.
2. leave the default name like Master1.
3.
Click in the pop up menu U vector, and click New Vector.
appears Measure dialog appears and Create Line.
4. Ask vectorU coordinate system in the location shown on the surface in Fig. 2.2.
Click the lower right corner (start point), and drag the cursor to the left corner
(endpoint) and refine.
5. Click OK, to close the dialog.
6. Select the opposite chamferand call the slave boundary condition HFSS>
Boundaries> Assign Slave ... Appears Slave dialogue with the selected tab
General (Fig. 2.5).
7. SelectMaster1 as the leading boundary Master.
8. drawU Vector vector as shown in Fig. 2.5.
9. SelectReverse direction mode of the vector V.
10. leave unchangedother settings and click OK.
21
Fig.2.5. Dialog reference phase difference between boundaries of Master and Slave
11. Repeat this procedure forMaster2 Slave 2 and boundaries, as shown in Fig. 2.6.
For the vector V on the border Master2 need to install the opposite direction
(Reverse direction).
Fig.2.6. Master and Slave Boundaries mounted on opposite faces of the air box
antenna model
In the dialog shown in Fig. 2.5 there are two installation phase delay between
the Master options boundaries - Slave: corners by scanning Scan Angle and Input
Phase Delay phase delay.
Setting the waveguide port. waveguide segment contains 4 sides on which to set the
boundary conditions PerfectE. This would correspond to that of the metal. Hold
down the Ctrl key and select 4 of the lower box and define them as a perfect electric
wall command Assign Boundarys -> Perfect E.
21
Fig. 2.7. Setting boundaries ideal conductivity on the waveguide walls
that set the waveguide port:
1. Selectthe lower bound of the lower box (Fig. 2.8).
2. press Right-click and from the pop menu
Assign> Excitations> WavePort. It appears WavePort assistant.
3. Installtherein the number of modes equal to 2.
Fig. 2.8. Setting the integral of the square lines of the waveguide port
22
5. In the Integration Line, select New line and draw two perpendicular lines
integrated for each mode, as shown in Fig. 2.5.
6.leave other settings using the Next command to transition
and click OK to the next page. A given waveguide port appears
Excitation in the list.
Note that the choiceport in two modes due to the fact that the square
waveguide is a bimodal waveguide, in which there are two propagating waves of
differing polarization field. integrated lines task shows that we as the waves in the
wave port using two orthogonal linear polarizations.
installationFloquet ports. The top of the set top boxFloquet
port.
Unlike radiation boundary conditions and PML This port allows you to calculate
and display the value of S21 transmission coefficient of the waveguide port in
Floquet port.
perform the following steps to install the Floquet port:
1. Select the top facetop box.
2. Right-clickand select Assign> Excitation> Floquet Port from the pop-up menu.
Assistant appears Floquet port (Fig. 2.9).
Fig. 2.9. The choice of destinations verktorov matching Floquet ports
3. Under Lattice Coordinate System determined directions A and B
Floquet port.
4. ClickNextTaking Phase Delay default, and then Next to go to page setup Floquet
Modes Setup port modes (Fig. 2.10).
Default settings dialog Floquet modes includetwo modes port
Floquet.Floquet modes are determined by two indices
n.m and polarization
23
Polarisation State.
indices
n.m can understandof formula (1) in Chapter 1.
The fundamental mode has zero Floquet indices. This mode (wave) propagatesin all,
no matter how small, lattice periods. If the grating period is large enough, then there
are propagating
waveswith non-zero indices. Each pair of two waves indices n.m comply
differing polarization.
Fig. 2.10. selection Channel characteristics Floquet
The last columnfashion table marked "Attenuation (Attenuation)." This damping fashion along a direction normal to the lattice planes in dB per unit length.
The values for the mirror mode is 0 dB, as they propagate in free space and is
therefore not attenuated. This option is not specified, and is calculated by the
program.
fashion toptypes can be either propagating or evanescent. their propagation
mode independent of scan angle. There are situations in which for small angles of
scanning the higher order mode attenuated, while increasing the angle it becomes
pervasive. If the grating period less than half the wavelength in free space, then only
the two main modes propagate at all angles of scan. In this case, the periods of more
than half a wavelength. Therefore, at high angles of scanning in addition to basic
modes must appear propagating modes with nonzero index.
24
Fig. 2.11.Target scan angles, which range from the number defined in Floquet
modes depends ports
presson Modes Calculator button (Figure 2.12). This dialogue is necessary for
the calculation of the parameters and the correct number of Floquet modes of
selection events, necessary for the correct description of the port. In the dialog box
you need to put the corners, which will be carried out by the beam scanning antenna
array. It was noted above that our grill for large scan angles can operate in
multimode. We calculate the parameters of ten modes (Fig. 2.12), setting 10 in
Number of Modes.
Fig. 2.12. theFashion Floquet ports
25
5. The Post Processing tab, specify the position of the reference plane.
Setting the reference plane is used,when the user is interested in the phase of
the scattering parameter. The default reference plane is in the plane of the wave port.
De-embeding operation allows you to change the position of the reference plane.
This change affects only the phase of the scattering parameters, and does not affect
their modules.
6. In 3D mode, you can Refinement tab Floquet, who are involved in the adaptive
mesh refinement 3D.
Grid partition that is created HFSSon the next steps of adaptation it is actually
a compromise for the simultaneous analysis of all events at the same time.
7. takeother default settings and click the OK, to close the wizard Floquet ports.
Floquet port appears in the project tree in the Excitation section.
We define as the scan anglesproject variables, as follows:
1. In the menu, set the command Project> Project Variables. This will bring up a
window
Properties for this project (Fig. 2.13).
2. Click the buttonAdd. This will cause the Add Property dialog.
Fig. 2.13.Dialogue project variables
3. Install variable name
$ Phi_scan, and its value 0 deg. This
variable will be used as the azimuth angle.
4. Click OK. Nowvariable $ phi_scan
will be a d d e d
Properties in the project window.
5. Click the buttonAddthatshow again Add Property dialog.
26
6. Install the name of the variable describing the elevation name = $ theta_scan
and a value of 0 deg.
7. Click
OK. Dialog Add Property closes
and
added
var iab le Project $ theta_scan.
8. ClickOK,to close the project windowProperties.
It should be noted,that when using the scan angles of the unit cell patterns,
periodicity plane (here the plane of the array) must be parallel to the plane XOY in
the global coordinate system.
Installation on a decision. to do installation on the analysis, follow these steps:
1. Right-clickAnalysis on in the project tree and select
Add Solution Setup. A dialog opens Solution Setup (Figure 2.14).
2. In the tab General, set the frequency solutions Solution frequency =
299.79 Mhz, Maximum Number of Passes = 5 and Maximum Delta S = 0.02.
Fig. 2.14.Dialogue installation solution
3. In the tab Options, select the Do Lamda Refinement mode and set the option
Lambda = Use Default.
4. Set Maximum Refinement Per Pass = 30%,
Minimum Number of Passes = 5 and
Minimum Converged Passes = 2.
5. Selectbasic functions Order Basis function: First Order, and click
OK.
27
Starting the calculationand view the results.Start the calculation of the team HFSS>
Analyze. Once completed, clickResults on the icon and select Solution Data.
Fig.2.15 shows the window with the scattering matrix, which appears after
the calculation. We note the following.
• S-matrix has a dimension of 6 × 6 taking into account a 2-Floquet modes ports.
• fashion Flockin the S-matrix are listed in the order specified in the Setup panel of
the Floquet port. When referring to this panel, we, therefore, have in mind that
FloquetPort1: 1 refers for fashion and Floquet TE00
FloquetPort1: 2 refers to fashion Floquet TM00.
Fig. 2.15.Dialog Data Solution Data solutions
• columnsand rows in a matrix in Fig. 2.15 correspond to the standard definition of
the scattering matrix. The columns correspond to the wave incident on the various
ports, and line - the waves reflected from different ports. For example, the first
column of the matrix corresponds to the excitation wave incident wave port 1. The
elements of that column are equal to the amplitudes of waves reflected by the
different ports in a state where only one wave is excited port 1. Excitation wave
corresponds to the port for transmission grating.
• Columns corresponding Floquet ports describe lattice parametersin the receiving
mode.
28
parametricsweep scan angle. To show the possibilities of the Floquet port, calculate
the dependence of the lattice reflection coefficientport in the wave as a function of
scan angle. When scanning occurs in a E-lattice plane at an angle of 27.5 ° is
observed glare effect. To show this, we need to find the characteristics in E-plane
scan angle when changing from scan = 0 o to scan = 90o.
To perform a parametric analysis:
1. pressOptimetrics on in the project tree and select Add Parametric.
itdisplays Setup Sweep Analsysis dialogue with laying Sweep Definitions.
2. Click the Add button. itdisplays the Add / Edit Sweep dialog.
3. Variable from the pop menu, select the $ theta_scan.
4. SelectLinear step.
5. Install Start = 0 deg, Stop = 90 deg, and Step = 3 deg (Fig. 2.17).
Fig. 2.17. Setting range change
Theta angle in parametric
optimization
Fig. 2.18. selection saving options
and fields partitions grid when
performing parametric optimization
6. Click the buttonAddand OK, to close the Add / Edit Sweep dialog.
DialogSetup Sweep AnalysisIt includesvariable $ theta_scan.
7. Open the General tab, and make sure that the Sim Setup installed Setup1 with
the Include mode.
8. Open a bookmarkOptionsto make sure thatOptions Save Fields and Mesh
marked, and that the option Copy Geometrically Equivalent Meshes (Fig. 2.18) is
not checked. In this case, at each step of the parametric analysis are stored far-field
characteristics.
9. ClickOK. The section appears Optimetrics installation Parametric.
Current couple the fundamental modes sufficient to describe the normal
incidence of the wave, must be supplemented by the higher modes of types to
describe
29
latticeat high angles of the scan. Preparation of events for the parametric
sweep consists of the following steps:
1. re-signin port installations Floquet panel and click the tab, types of waves. To
define a list of events, necessary for modeling at all scan angles, use the calculator
mode.
2. Call events calculatorby clicking the Modes Calculator.
calculator calculation is additional information to create a list of recommended
events forFloquet port. His results need to select the number of modes and do not
affect the calculation of the lattice model.
3. Select10 for the port modes (with a margin). Probably, we will reduce this
number, but that additional research is required.
4. Set the frequency of 299.97 MHz,where simulation is performed. If two or more
sweep frequencies, typically the highest frequency is selected task installation
comprise, in order to detect the appearance of propagating high-order modes.
5. To establish a set oftypes of waves which will be sufficient for each scan
direction in the parametric sweep, the sweep angles of scanning are entered in the
format "start-stop-step". entered angles
- spherical polar angles in the global coordinate system.
Forthis case the angle Phi = 90 °, so that enter this value as the starting value
and the final value of Phi in fields. Theta scan sweep angle changes from 0 ° to 90
° in steps of 0.5 °, so enter these same values in the Theta calculator fields.
6. Click OK,to start the calculation of the calculator and see the recommended list
of events (as in Fig. 2.12).
Fig. 2.20. Characteristics scanning the reflection and transmission
characteristics of the wave types
31
Note the following:
• initial steamfundamental modes TE00 and TM00 modes remain top of the table
Floquet (Fig. 2.12). Attenuation of zero, which means that these modes propagate
without attenuation, ie they are propagating.
• Thenin the table is a second pair of TE01 and TM01 modes. They extend without
attenuation, at least in one direction.
• Thenfollowed by six modes with minimum attenuation of 60 dB / m. These modes
do not apply in any of the selected areas.
Now you need to make the finalSelect the number of events considered in the
Floquet port. This choice is based on the following considerations. Any Floquet
modes propagating in at least one direction should be included in the mode table.
Therefore, it is necessary to include the first four modes, which reached the port
Floquet unimpaired.
Remainingtypes of waves, having a non-zero attenuation, are candidates to be
excluded from the table. From the point of view of efficiency modeling and
interpretation of the results is better to remove them.
In this regard, we note:
since the length ofthe unit cell is 1.25 m, any of the last six events, when
excited in a rectangular aperture to reach the Floquet port with attenuation equal to
1.25 * 60.00 = 75 dB, that is very much weakened. Therefore, in most cases such
Floquet modes in the description of the port can be neglected.
7. Thus, we introduce a number of modesin the Number of Modes = 4. In the list of
events are only the modes TE00, TM00, TE01, and TM01.
8.Click OK, to select the set of events and then run parametric analysis (it will take
a certain time of the calculation).
review the results of the parametric analysis.
Once the simulation is finished, the S-matrix elementsas a function of scan
angle can be seen in Matrix Data tab or on the graph. By analyzing the scattering
matrix for different scanning angles, it can be seen that the connection between the
TE00 wave types and TE01 very small.
To display the module dependencies of reflection coefficientsand transmission TM
as scan angle functions perform the following steps:
1. Right clickthe Results in the project tree and select Create Modal Solution Data
Report> Rectangular Plot. Report dialog box appears.
2. In the Trace tab, section X, select $ theta_scan.
3. ForY select both S Parameter Category, S (Wave Port1: 1) as a characteristic
(Quantity) and Mag module as a Function.
31
4. Click OKTo create a new schedule. The new report is displayed on the screen
and adds the name of Result in the project tree with the first trace shown below the
graph. Add Trace button becomes active in the Reports dialog.
5. Select the name of the characteristics of the project tree. It displays the
properties for the Trace.
6. markName the Specify option to open the Name field and change the name to
Reflection. This will change the name on the characteristic.
7. Add two additional graphicsand change their names: mag (S (WavePort1:
1, FloquetPort1: 2)) corresponds TM00 Transmission; mag (S (WavePort1:
1, FloquetPort1: 4)) corresponds TM01 Transmission.
8. In the project tree, select the XY Plot 1. It shows the Properties
for a schedule.
9. Edit fields Nameon Reflection and Transmission and press Enter.
The results of the parametric analysis are presentedFig. 2.20. Note that the
reflectance tends to unity at an angle of 27.5 °. This effect is known as glare effect,
since in the vicinity of the angle of glare grating efficiency drops sharply, because
all the energy received on the wave port is not radiated into space, it is reflected.
Also note,that type becomes TM01 wave propagating at an angle of
approximately 30 °. This is evidenced by a sharp increase in the gain of the wave in
the wave port TM01 Floquet port at angles greater than 30 °.
In this example, the matching questions are addressed: the reflection
coefficient of the lattice. The next section will be reviewed and radiation
characteristics of the antenna array.
32
3. The array antenna of the Vivaldi antennas
The calculation of the antenna array composedof radiating elements, each of
which is a broadband antenna Vivaldi. Array of antennas Vivaldi created using
periodic boundary conditions and Floquet ports.
The antenna is powered coaxial linewith a waveguide port. On the upper
surface of the array cells is set Floquet port (see. Fig. 3.1). Periodic boundary
conditions are introduced on the sides of the cell. In this opposite and parallel side
borders are declared Master and Slave. The substrate has a dielectric constant εr =
6, and a thickness 1.27 mm. Strip conductors are 2D objects with boundary
conditions PerE. The calculation is performed at a frequency of 4.5 GHz and
interpolation mode in the range of 2 ... 5 GHz.
Fig.3.1. HFSS interface c separate antenna Vivaldi
3.1. Drawing Vivaldi antennas
Establish length units- millimeters and draw a dielectric board as a
parallelepiped measuring 34 x 60 x 1.27 mm (Figure 3.2.) Create command -> Box.
33
Fig. 3.2. Dimensions Vivaldi antenna, which is an element of the antenna array
will make the insulatorsection variables and define the permeability of the box
the box of 4. Parameters can be edited in the dialog Fig. 3.3.
Fig. 3.3. The size of the substrate 6 with a permeability of 1.27 mm thick and
34
Vivaldi antenna operatesa high frequency band, and maximum band width
achieved with an exponential change in the gap from the beginning of the horn to
raster.
Draw aVivaldi antenna conductors using functional relationships. Draw a
line, which is defined as a function performed by the team Geometry -> Curve ->
Analitic Curve. dialog appears Fig. 3.4, into which the functional dependence of
coordinates of points on an edge of the strip conductor by the variable t. Coordinate
Z depends on the variable t is linear. Therefore, the dependence of the coordinates
Y t can be regarded as dependent on the coordinate Z.
Figure 3.4. Setting functions for which the curve is drawn
useto set the shape of the conductor edges exponential function:
X (t) = 0,
Y (t) = 0.25 * exp (0.123 * t),Z
(t) = t.
The width of the gapbetween conductors at the narrowest point at t = 0 is equal
to 0.5 mm. The variable t varies from 0 to 33.3 (Fig. 3.4). The functional dependence
is set in 24 points.
Nowsupplement exponential curves with straight lines and a circle, which
play the role of balancing the power system devices Vivaldi antenna (Fig. 3.7).
35
Fig. 3.5. Topology and size of the balancing cavity
Further draw the
circle,
resonator Vivaldi antenna (Fig. 3.5).
playing
role
balun
Fig. 3.6. circular resonator parameters, which is part of the balun
Now draw a rectangle,which form the slot line resonator and connecting
a coaxial line (Fig. 3.7).
36
Fig. 3.7. Draw a rectangle to create a slot line
after trainingplotting to form a complex which includes a circle, an
exponential line and all surfaces future Vivaldi antenna, we define command Draw> Line capture audio and perform point exponential curve. When we reach the point
where you want to make the transition to the circle, move the drawing mode in
Center Point Arc (Figure 3.8)
Fig.3.8. Transfer from the drawing line segments (Straight) in the circumferential
drawing mode with the center and radius.
Now we need to combine the exponential segmentwith a broken line Unite
team. Next, create a form from a closed plane Modeler team -> Surface -> Cover
Line.
37
Fig. 3.9. The strip conductor antenna with a balun Vivaldi
Further select an exponential segment, and deploy it to 180 °,
just change the sign of the coordinates Y (t)(Fig. 3.10).
Fig. 3.10.Drawing the second strip conductor
Vivaldi antenna can excite different ways. You can, for example, to put a
discrete port between the antenna vibrators.And it is possible to include in the real
length of the coaxial line structure, the input of which is switched the microwave
generator. Draw the inner conductor of the coaxial line as a cylinder 8 mm in length.
Draw the coaxial line runs Draw-> Cilinder team, on which dialogue (Fig. 3.11)
appears.
38
Fig. 3.11. The parameters of the internal conductor of the coaxial line radius
0.375 mm 8 mm long
Further draw the external
shell
coaxial
line,
commandDraw-> Cilinder,and define the settings shown in Fig. 3.12.
also
Fig. 3.12. Draw external outer cylinder, on its upper surface is automatically
set condition Perfect E
Further draw a radiation box that surrounds the antenna Vivaldi, Draw
command -> Box and set the dimensions shown in Fig. 3.13.
Fig. 3.13. Target box sizes, ambient Vivaldi antenna 34 x 36 x 60 mm
39
on the sidesthis box we define the boundary condition Radiate. Define the
port on the end of the coaxial line (Fig. 3.14).
Fig.3.14. Creating a wave port in a
section of coaxial line
Fig.3.15. Cell array antenna, covered by
periodic boundary conditions
Port Floquet used in the simulation of infinite periodic structures (see chap. 1
and 2). Analysis of the endless structure is to analyze the structure of one period
(unit cell). On its opposite sides defined periodicity conditions. On the upper surface
must do one of the possible
'Open' boundary conditions simulating the radiation processin free space. As
surfaces to "open" boundary condition in HFSS used PML, radiation surface and
Floquet port.
dignityFloquet port is that it can be used to describe not only the reflection of
the wave arriving at the antenna input of Vivaldi, but also the transfer of energy in
space. In the infinite periodic structure field in free space has the form Floquet
harmonics. Each of these harmonics is a plane wave propagating in the space at a
certain angle. As noted in Chapter 1 and 2, most Floquet harmonics is evanescent,
and the main harmonic and one or two types of higher harmonics can be propagated.
Thus harmonics of higher orders may under some angles of the scan to be
propagated, and at the other nonpropagating.
UsingFloquet port allows the calculation of the gains of the waveguide port
in the Floquet harmonics, and vice versa. As in
41
case waveguide port, this information is given in the form of the S-matrix,
linking reflectedand the incident waves (Fig. 3.19-3.24).
Fig. 3.16. Floquet port settings in the table specified number of the mode,
statepolarization mode index, damping and sealing mode
Affects Refinements
Furtherwe define the center frequency of 4.5 GHz, and performing the
calculations of the antenna array in the frequency range of 2 to 5 GHz. The results
of calculation of the frequency characteristic | S11 | are shown in Fig. 3.17.
Fig. 3.17. The frequency characteristic of the reflection coefficient of the Vivaldi
antenna
41
Fig. 3.18. near Field sectional Vivaldi antenna and current on the surfaces of the
strip conductors
Fig. 3.18 shows that the antenna is excited symmetrically, which was not I
would be if there was no balun in structure.
Fig.3.19. The first mode channel Floquet
Fig.3.20. The second mode channel Floquet
Note thatin terms of the use of electrodynamics lattice pattern in the form of a
channel can not be expected Floquet Nam this type of antenna, since this model may
strictly be used solely for analyzing an infinite lattice which has NAM zero width,
i.e. in the form of delta functions. However, it is possible to calculate HFSS Nam
Floquet channel by replacing one of the ports on the Floquet radiation surface. This
possibility should be treated with caution, meaning that the resulting Nam is neither
grating Nam or separate component thereof. It gives a qualitative idea of the NAM
lattice when it is driven by one of its inputs. The antenna technique such Nam called
NAM emitter composed of a lattice. However, we must remember that the
calculation method described above allows you to get the mentioned characteristics
are only approximately.
42
4. The antenna array on the dipole antenna
Considerantenna array of four dipole antennas with a metal reflector (Fig. 4.1).
In such a lattice beam position can be controlled by changing the phase of the voltage
applied to its individual elements. Reflective grating panel improves characteristics
by creating a unidirectional radiation in the horizontal plane.
Run analysis of such array can bea unitary construction, i.e. without using
periodic boundary conditions.
Fig. 4.1. The array antenna with reflector
lattice elements are arrangedon Teflon backing thickness
1.6 mm. Excitation carried out via the digital ports. radiation boundary is given on
the sides of the air box. Variable phase shift for postprocessing has phase_shift name.
The center frequency at which the antenna array operates 1.9 GHz.
43
Fig. 4.2. The geometry of one of the four antennas disposed with a step on the
axis Z equal to 103 mm
Draw a dielectric substrate 1.6 mm thick (dielectric constant4), Draw-> Box
team. The obtained dimensions of the substrate can be adjusted in Fig dialog. 4.3.
Fig. 4.3.Conversation piece dielectric board
Move the coordinate systemon the plane on which will draw vibrator first
cursor highlighting one of the planes (pre-pressing F), command Modeler->
Coordinate System -> Create
-> Face CS.
44
Next, draw a vibratorpoint by point how polygon, setting the command Draw
-> Polyline.
Fig. 4.3. Draw one of the vibrator
Next, createnew coordinate system shifted with respect to the X axis of the
system by 18 mm, equal to half the length of the substrate command Modeler->
Coordinat System -> Create -> Relative CS-> Offset. And make a U-turn with a
copy of this part of the vibrator 180 ° command Edit-> Duplicate-> Around Axis.
Fig. 4.4. Selection Y axis about which the vibrator copying scrolls
Now between linesvibrator draw a rectangle (Fig. 4.5)
on which the then established
discrete
port
> Excitation-> Assign-> Lumped Port.
team
HFSS-
45
Fig. 4.5. Formulation of an integrated line on discrete port
Furtherperform up-shift of the dipole antenna to receive antenna 4
constituting an antenna array (Fig. 4.7).
reflector Draw. metallic reflector angular configurationsuch as shown in Fig. 4.6.
The mutual arrangement of the reflector and the dielectric board with vibrators is
shown in Fig. 4.7. The reflector comprises a hole that can be used to output the
exciting lines and connecting them with a power divider outputs, which acts as a
lattice supply circuit.
Fig. 4.6. Regulation reflective plane array antenna with reflector
The total height of the dielectric substrate420 mm, and the dielectric constant
ε = 4. After drawing the lattice can specify settings for calculation and analysis of
the frequency range. When conducting the calculation of the lattice should be
noted that the antenna array is the RF multipole having four inputs. HFSS
calculates the scattering matrix of the multipole in the prescribed frequency range.
However, the values obtained reflection coefficients are not equal, representing
the greatest practical interest reflection coefficients of the lattice in the operating
mode in which simultaneously excites all four inputs.
46
Fig. 4.7. Structure and the size of the antenna array. The total height of
420 mm reflector
To determine the reflection coefficients of the latticeSi in the working mode,
j_corr (i, j = 1, ... 4) necessary to calculate them from the elements of the scattering
matrix S (i, j) according to the formulas which are shown in Fig. 4.8.
Fig. 4.8. The variables considered in which S-parameters of the antenna at its
inputs simultaneous excitation
Fig.4.8 refers to a complex voltage Vi amplitude excitatory lattice elements.
47
It is known that forreducing sidelobe level of the antenna array elements
extreme need to initiate a smaller amplitude than in the center. The amplitudes of
the excitation can be set as a variable (see. Fig. 4.8). If we take V_port1 =
V_port4 = 0.66 and V_port2 = V_port3 = 1, we get the NAM, as shown in Fig.
4.9.
The difference in power phasetwo adjacent antennas is set variable in step
phase_shift postprocessing. To create such a regime in the dialogue, is led away by
a team HFSS> Fields> Edit Sources, which is a variable offset phase. Set phase_shift
factor variable phase shift.
askphase_shift variable can by clicking on HFSSDesign2 project name in the
project tree. The variables of the project can be seen in the Properties window. In
this window, you can change the value phase_shift. We choose phase_shift = 30º,
that would be consistent with the pattern in Fig. 4.9.
Fig. 4.9. Nam separate vibrator array antenna
48
Fig. 4.10. Three-dimensional Nam antenna array
With a large number of antenna array elementsit is no longer possible to
calculate in a single design. To solve this problem you need to use periodic boundary
conditions.
Excitation of each antenna array element is performedusing a multichannel power
divider (MDM). EXAMPLE MDM together with the vibrator shown in Fig. 4.11. In
the case of in-phase excitation ports are excited so that the far field are summarized
in the direction 0y axis. This process can be seen from the field distribution shown
in Fig. 4.11.
Fig. 4.11. nearfield in the plane of the antenna array xOy
49
It should be noted that the joint calculationemitting lattice together with the
power scheme is for HFSS very complex task that requires a lot of computer
resources costs. This situation is due to the fact that the branched microstrip circuit
realizing MDM requires for its analysis of the use very dense mesh. An example of
such a network is shown in Fig. 4.12, which also shows the distribution of currents
in metallic conductors.
Fig. 4.12. Vibrator grating reflector and distribution circuit
Given the largethe complexity of solving the discussed problems in HFSS, is
of interest to use an integrated approach which combines the use of different
software to calculate various pieces of complex structure. When each part is
analyzed using methods in the most adequate physics of its functioning.
CalculationMDM system in MWO. Power distribution circuit will design in
Microwave Office system (MWO) using EMSight routine that calculates the
stripe structures of the method of moments. Such an approach for analyzing
stripline circuits much more efficient method of finite elements, which is used
in HFSS. Wherein the transmitting grating is analyzed using HFSS. Below, we
discuss the possibility of correct use of data obtained from a single system for
the calculations in the other. In this case, the MWO.
The problem of reconciling the calculation data receivedin different projection
systems it is solved by the use of discrete ports. For this model is used for modeling
vibrators shown in Fig. 4.13. Its feature is the use of two discrete P1,2 ports
connected between the conductors of the transmission line which excites the vibrator
and the metal reflector. In this representation array element is calculated as a
microwave HFSS multipole eight outputs (in shestnadtsatipolyusnik
51
the terminology used in Russian and English-language terminology for
vosmipolyusnik - eight port junction).
Fig. 4.13. Modelvibrator with two ports
Using this model allows us to calculate the scattering matrix grilleMDM
without 8x8.
We next consider the main stages of designMDM system in MWO. MDM is
a combination of several elementary power dividers EDM into two channels
connected with each other via transmission lines. MDM structural scheme is shown
in Fig. 4.14.
51
Fig. 4.14. structure MDM
The composition of MDM shown in Fig. 4.14 D1 EDM includes performing
division of power in half, EDM D2, power dividing in a ratio of 3: 7 and exciter
vibrators V. The unequal power division in the EDM D2 provides decaying field
distribution at the edges of the grating, which reduces the side lobe level in its
vertical plane Nam .
exciterThe device acts as a power dipole antenna. 1,2 outputs it is connected to a
vibrator which has two inputs (see. Fig. 4.13).
First stepMDM design includes the design of the EDM and the pathogen.
Their topology is shown in Fig. 4.15 ac. Fig. 4.15 a, b shows the topology D1,2
dividers, and Fig. 4.15 in the exciter topology. They are made on a substrate with a
permeability of 4.6 and 2 mm thick.
52
a
b
at
Fig. 4.15.Topology EDM and exciter
53
Because thedivider D1 has symmetry relative to ports 2 and 3, it reduces to
design harmonization of entry 1. All the outputs of the device are formed as
microstrip lines with characteristic impedance of 50 ohms. The discrete ports are
used as ports. Lateral shoulders divider comprise quarter-wave transformers that
provide output matching transmission lines with input line. Fig. 4.16 shows the
frequency response of the input reflection coefficient divider 1. It is seen that the
best match is achieved in the vicinity of the frequency 1.7 GHz. When removing it
from the reflection coefficient increases, but remaining sufficiently low value not
higher than - 26 dB in the band 1.6 - 1.8 GHz.
Fig. 4.16. frequency
divider D1 Input 1
addiction
module factor
reflection
divider D2It does not have symmetry relative to the inputs 2 and 3, as it
provides an asymmetrical division of power. Therefore, it requires not only the
design matching at the input 1, but also the realization set values of transmission
factors of the input 1 to the inputs 2 and 3. They
mustbe equal to the following values: dB.
S211.5 dB S315.34
OnFig. 4.17 shows the frequency dependence of the modules
reflection Sel and transmission coefficients S a S31 divider D2. It is seen,
21
eve
nd
n
54
whatit is sufficiently well aligned for entry 1 and provides a predetermined value
of transmission coefficients.
Fig.4.17. Frequency characteristics divider D2
excitervibrator must perform the following functions. It performs power
division, arriving at input 1 in equal relation between the inputs 2 and 3. In this phase
signals at ports 2 and 3 must be shifted to 1800. In this case, the two-wire line which
feeds the vibrator is excited its fundamental wave, whose currents on the conductors
and phase-shifted by 1800. It is also important to the driver outputs are matched with
the input impedance of the vibrator. Its parameters HFSS chosen in such a way that
the input impedance at the points of accommodation ports had active portion of 60
ohms, and a zero reactive part.
proceedingof these conditions, the pathogen was carried out to optimize the
topology. As a result, frequency characteristics are obtained as shown in Fig. 4.18 a,
b. Fig. 4.18 and shows the frequency dependence of the reflection coefficient at the
input 1, and Fig. 4.18 b frequency
depending on phase coefficients
transfer
S 21
S31. It is seen,
a
what
nd
exciter well-coordinated and provides the complementary division of power.
55
The MWO system it is possible to visualize the current distribution on the
strip conductors. An example of such a distribution on the conductors exciter shown
in Fig. 4.19.
a
b
Fig. 4.18. frequencypathogen characteristics
56
Fig. 4.19.Current distribution on strip conductors of the pathogen
finalMDM design step is to calculate its output parameters. MDM topology
is shown in Fig. 4.20.
Fig. 4.20.MDM topology
As an example, in Fig. 4.21 shows the frequency dependence of the reflection
coefficient for the central MDM entry 1 (see. Fig. 4.14). It can be seen that the
alignment MDM somewhat worse matching EDM, however, in the frequency band
the reflection coefficient is less than - 20 dB.
The next step consists lattice designa calculation of its parameters based MDM
design results and emitting portion of the lattice in the form of four vibrators formed
in HFSS.
This part of the work is convenient to perform,exporting data files from the
HFSS and MWO and treating them with the help of computer systems like Mathcad
or Matlab.
files Data include multiport scattering matrix,
calculatedin HFSS and MWO. grating scattering matrix Sa
is
a 8x8 matrix in accordance with the number of its inputs. Similarly matrix
scattering MDM Sd It has a dimension of 9x9.
57
Fig. 4.20. frequency dependenceMDM reflection coefficient at the input 1
Matrix
SdIt has the following structure:
Seleven
Sd 
S  1
S1 
.
 
S
- input group with the numbers 2 - 9. Thus,
8x8,
S1 - matrix - the line size 1x8,

size 8x1 and
(5)
S - a matrix
S1- Matrix - column
Seleven- the reflection coefficient at the input 1.
For calculation of resultant antenna reflection coefficient R
necessary to analyze cascadingtwo multiport shown in Fig. 4.21.
58
Fig. 4.21. cascadeMDM compound and the lattice
59
For define
relations given in 20]:
factor
reflection
R
We use the
RSele S1(E SSa)1 SS .
ven
(6)
1
where E - the identity matrix of size 8x8.
Also, analysis of cascade-connected multiport, we can
to find the amplitude of the incident wave to the inputs of
the lattice
U n. n1 ... 8.
Their
60
can be used further for the taskConditions Lattice excitation with the influence
MDM. Knowing the amplitude, we can use HFSS to calculate the antenna pattern
(Fig. 4.10).
Using different software describedabove allows efficient use of them to solve
those problems that can be solved with the help of their computer resources with the
least cost. In this case HFSS used to analyze the three-dimensional radiating
structure and for analyzing MWO planar stripline circuit. As a result, the total time
required for obtaining the final result is significantly reduced. In addition, the
increased accuracy and reliability of the solution.
5. Modeling the frequency-selective surface
Among the most interesting problems of applied electrodynamics are the
problems of wave diffraction by periodic structures. Widespread use are twodimensionally periodic planar lattice, which are called frequency-selective surfaces
(CHSP). They are used as spatial filters, polarizing filters, antenna radomes,
radioprotection agents, etc.
Special mention deserves the use ChSPin quasi-feeders reflector antennas
meteorological satellites. Application ChSP to build frequency selective devices can
significantly improve their weight and size.
By devices of this type of strict requirements. For example, the radiometer
ESA MASTER must provide signal separation ranges 294-306, 316-326 and 342349 GHz [13-17]. Thus the insertion loss should not be greater than 0.5 dB, and the
isolation between channels is not less than 20 dB for reliable detection of weak
radiations molecules in the atmosphere. In order to minimize insertion loss used
ChSP aperture, allowing to exclude dielectrics and losses associated with it.
Improving polling characteristics can be achieved using multilayer structures
containing several parallel layers and multielement periodic structures comprising
several reflectors (apertures) in one unit cell of the periodic structure CHSP. Design
of such devices is an urgent task,
Creating a model ChSP.We analyze CHSP using periodic boundary conditions,
allocating unit cell structures as Floquet channel.
The unit cell may CHSPIt is created using boundaries associated with periodic
boundary conditions and two Floquet ports. Result solutions boundary CHSP
analysis task is represented as a matrix of S-linking Floquet modes in the respective
ports.
As an example, consider that in this chapter is a conductive screen having a
hexagonal lattice of circular apertures (holes). The geometry of the lattice shown in
Fig. 5.1. Fig. 5.1 shows the angle of the lattice vectors. The angle between the lattice
vectors is 60 °.
61
Figure 5.1. The geometry of the frequency-selective surface
Consider a normal incident plane waveon the screen with the polarization
vector of the electric field shown in Fig. 5.1 red arrows.
We calculate the modulesand phase of the scattering matrix elements as
functions of frequency. The frequency band of 8 to 20 GHz.
Modelone cell is shown in Fig. 5.2. Side length equal to 1.73 cm circular
aperture diameter is 1.2 cm and height equal to 4 cm cell. It can be seen that the cell
consists of a planar metal rhombus with a hole and boxing rhombic shape, the
surfaces of which are given periodic boundary conditions and Floquet ports. We
begin to create elementary HFSS means the cell. In the first stage will create rhombic
box.
Creature
rhombic
Boxing. For
solutions
tasks
follow
these steps:
1. Open the a new project and name it RhombicArray. Set the length of the unit centimeters (Cm).
61
Fig. 5.2. Rhomboid hexagonal cell CHSP
2. Click Draw> Line,Click at an arbitrary position to set the starting point, and then
click the other three arbitrary points, and then return to the starting point to close the
figure.
3. Right-click to bring up popup menuand then click Done. This polygon is created.
4. In the project tree, expand Sheets> Unassigned> CreatePolyline, selecting the
first operation CreateLine to see the properties of this line.
5. Set the coordinates of the first point Point1 = (0,0,0).
6. Edit coordinates Point2= (0.865, 1.4982, 0). This will shift the segment of the first
broken line, which will give the desired object resolution.
62
7. Repeatthis process with the following three operations in the tree CreateLines
create and edit them the values Point2, Point3 and Point4 (Figure 5.3.) as follows:
Point2 (0.865, 1.4982, 0)
Point3 (2.595, 1.4982, 0)
Point4 (1.73, 0, 0)
Created flat rhombus can be usedto create a diamond-shaped boxing. Before
that, make, Copy command, a copy of the rhombus, which will be useful for creating
metallic diamond.
Fig. 5.3. Rhombus as a section future cell frequency-selective surface
8. To convert a flat 2D object to 3D cell, press
Draw> Sweep> Along Vector.
9.The status bar, enter all zeros in the cell X, Y, Z, to set the starting point.
This places the cursora first point defining the vector, and changes the message
on status bar "Input the second point of the sweep vector (to enter the second point
scan vector)." Tags cell shows the shift amounts dX, dY, and dZ.
10. In the status bar, enter 0, 0 and 4, and press Enter. This draw a line from the start
point to the end point, and displays the dialog Sweep Along Vector (Fig. 5.4).
11. In the dialog box Sweep Along Vector, leave 0 deg as a draft angle, and select
from the drop down menu, Round as the type Draft type. Then click OK.
63
Fig. 5.4. Completion operation sweep along vector
12. This operation converts diamond in the 3D object.
13. Now move the this object down to 2 cm from the operation Edit> Arrange-> Move.
14. opensMeasure data dialog and there is a possibility to enter through the status
bar as well as the cursor, the reference point X, Y, and Z, as a first point offset vector.
15. In the status bar, set the Z = 0, and press Enter.
16. In the status bar, set the dZ = -2.0 and press Enter.
This will move the objectto the desired length.
Creating a flat diamond-shaped with a hole.If you make a copy of a flat rhombus,
the further it can be used.Otherwise, repeat steps 1 - 7.
1. Click copy flat diamond.
2. Command Draw> Circle Draw a circle with the cursor in the center of the
polygon.
3. Open dialogue properties circle and set the radius of 0.6 cm.
4. Select at the same time a circle and a rectangle and click
Modeler> Boolean> Subtract. Subtract dialog opens.
Fig. 5.5. The cell-frequency selective surface
64
5. Move Rectangle Blank in the list, and the circle in the Tool list.
6. Click
OK, to close the dialog and create a circle by subtracting the
aperture of the rectangle.
The taskMaster and Slave boundaries.
ForIn order to simulate CHSP which consists of periodically arranged elements
necessary to set the periodic boundary conditions on the opposite walls.
Define Master borderand Slave to the rhombic object as follows.
1. Select the surface shown in Fig. 5.6 and set command
HFSS> Boundaries> Assign> Master.Appears Master Boundary dialog.
2. leave the default name like Master1.
3. that set the vector U along the edge click New Vector.
appears Conversation Measure dialog and Create Line.
4. draw vectorU vector on the selected surface. Click on the lower left corner as a
starting point, and draw a cursor to the right corner and refine Master1.
65
Fig.5.6.Zadanie Master boundaries (leading) and the Slave (slave)
5. The Master Boundary dialog for vector V Vector, select
Reverse directionand click OK.
6.
Select the oppositeand set aside the command HFSS> Boundaries> Assign
Slave. Slave appears a dialogue with the selected tab General.
7. For this slave select Master1 surface as the leading boundary.
8. drawU Vector vector as shown in Fig. 5.6 and click OK.
9. Repeatprocedure for Master2 Slave boundaries and 2.
Fig. 5.7. Creating a second pair of periodic boundaries
66
Next, askPerfect E boundary condition for a diamond-shaped with a hole. To
do this, choose Aperture surface of wood chronology and define the command
HFSS> Boundaries> Assign> Perfect E. PerfE1 border appears in the Boundary list.
Setting Floquet ports.
ports Floquet must be installed on the upper and lower surfaces of the model.
1.
Select the topsurface of the model and click HFSS> Excitations> Assign>
Floquet Port. Assistant appears Floquet port, showing General page.
2. For the coordinate system Lattice Coordinate System,from the pop-up menu to
the direction A, choose New Vector.
3. Draw the vector, pressinglower angle Z axis start point, and then pressing the
adjacent corner along the axis X. When you double-click to complete construction
of the vector, dialogs Measure Data and Create line disappears and reappears
assistant Floquet Port, showing that a vector is defined. Repeat this procedure for
the vector b.
4. In Modes setup tab, enter Number of Modes = 14. Thus, the table is full Mode,
which will have fourteen lines (Fig. 5.9).
Fig. 5.8. Floquet port in the space-frequency selective surface of the cell
5. Click the buttonModes Calculatorfor informationInstallation modes. mod
calculator appears. Set Frequency = 20 GHz frequency (fig. 5.9).
67
Fig. 5.9. Calculator modes and setting its start
6. leave the remaining default values and click OK, to close the calculator. This
will fill the table with the calculated data.
7. SelectDeembed option, and set the Distance = 2.0 cm.
8. ClickNextfortransition to 3D Refinement page.
9. For 1st Fashion Mode 1, select seal Affects Refinement (Fig. 5.10) and click
Next to go to the Post Processing page.
Fig. 5.10. results operation modes calculator
68
10. ClickOK. The first port Floquet appear in Excitations section in the project
tree.
11. Select the bottom surface of the model, and repeat the process to set up a second
port Floquet. Information first port Floquet copied.
Fig. 5.11. near Field sectional frequency selective surface of the cell
Installation on the test. For Follow these steps:
1. Right-clickAnalysis on in the project tree and select Add Solution Setup. A
dialog Solution Setup.
2. In the General tab, set Solution frequency = 20 GHz. Install
Maximum Number of Passes = 10 and Maximum Delta S = 0.02.
3. In the tab Options, select the Do Lamda Refinement mode, and set
Lambda target = 0.2.
4. Set Maximum Refinement Per Pass= 20%, the Minimum Number of Passes = 6
and Minimum Converged Passes = 2.
5. Select functionFirst Order Basis.and click OK.
69
Fig. 5.12. Selecting types of waves, which is controlled by the process of
convergence
Calculation characteristics in the frequency range. that perform a
calculation in the band:
1. Right clickon Setup1 in the project tree and select
Add Frequency Sweep. Appears Edit Sweep dialog.
2. To Sweep Type, selectcalculating data smoothing option Interpolating frequency
range.
4. Set Max Solutions = 50, and Error Tolerance = 0.5%.
5. Click the buttonadvanced Optionsthatshow dialogue Interpolating Sweep
Advanced Options.
6. Set Minimum Solutions = 5 and Minimum number of Subranges = 1.
7. Forinstallation convergence Interpolation Convergence, select the button Use
Selected Entries Radio, and click Select Entries button. Dialog appears Interpolation
Basis Convergence (Fig. 5.12).
8. The Interpolation Basis Convergence, leave Entry Selection and Mode Selections
as All dialog box. This means that all Floquet modes are listed in the table. Then,
use the vertical scroll bar, find the line FloquetPort2 1 and set to ON in FloquetPort1
column: 1.
9. In line FloquetPort1: 1 and in line Floquet Port2: 1 and set the option
ON.
10. pressOK,to close this dialog box, the OK to close Interpolating Sweep Advanced
Options, and OK, to close the Edit Sweep dialog.
On Fig. 5.13 - 5.14, the calculated field distribution different Floquet
channel modes.
71
Fig. 5.13.Golf 2nd fashion Floquet
Fig. 5.14. Field 4 th mode Floquet
11. click Analyzeto start calculation.
Viewing the results of calculations. Severalfrequency dependencies of the
modules and phases of the reflection coefficients and transmission of the main types
of waves shown in Fig. 5.15, 5.16.
Fig. 5.15. Phase response transfer factor from port 1 to port Floquet 2 by
main wave
Fig.5.15 shows that CHSP has frequency selectivity for the fundamental
mode. From 5.16 it shows that at a frequency of 18 GHz CHSP completely
transparent to the incident wave, because at this frequency it reflectance goes to zero.
It is also interesting to calculate the properties of ChSPfor waves incident at
various angles to a frequency-selective surface. In this case it is necessary to specify
a plane wave, wherein the incidence angle will vary parametrically.
71
Fig. 5.16. The frequency characteristics of the reflection coefficients and moduli
CHSP transmission of fundamental wave
Note that you can also calculate the properties ChSPfor waves incident at various
angles to a frequency-selective surface. In this case it is necessary to specify a plane
wave, wherein the incidence angle will vary parametrically. For frequency selective
surfaces include resonant structure, discussed in Chapter 1.
Materials so that you can setthey will have the frequency dependence of its
parameters. An example of the frequency dependence of the materials shown in Fig.
5.17.
Fig. 5.17. frequency dependencematerial (from Example ViaWizard)
Library HFSS material can be replenished models with polygonal linearly
frequency dependence of the Debye model of a model, models Dvorzheka,
Sakharov, as well as custom models, all parameters which can be entered in tabular
form [1].
71
6. Falling plane wave to the object
and calculation of radar scattering cross section
The radar cross section for the scattering characterizesthe ability of an object
scatters the incident electromagnetic waves at him. In English literature it is
equivalent to the term Radar Cross Section (RCS). This parameter is used to assess
the ability to detect an object (target) by radar [4,5]. Also this parameter for assessing
parameters used radar cross-section (EPR).
Consider determining the parameter RCS. There are two different cases.In the
first transmitting antenna, irradiating the object under study and the receiving
antenna, receives a reflected wave from it in one place. In this case we speak of
monostatic location and purpose of the parameters describe using monostatic RCS.
Inthe second case, the transmitting and receiving antennas located in different
locations (see. Fig. 6.1). In this case we speak of bistatic radar and used, respectively,
the bistatic RCS.
Fig. 6.1. Scheme bistatic radar
parameters RCS and ESR are related as follows:
RCS = σ / λ2.
where λ - wavelength in free space, σ - EPR [m2].
certain
ratio
(6.1) call
(7)
also
RCS,
normalized
71
radar
factor
reflection
because the he is
andimensionless quantity.
EPR is determined by the electric field of the incidentEpad wave on the object
to place the object location and the field strength of the wave scattered by the object
in the direction of the receiving antenna at a location Eotr receiving antenna:
σ = 4πr2 | ENeg|2/ | Epad|2.
(8)
where r - distance from the object to the receiving antenna.
Note that the ratio(6.2) can be used as in the case of bistatic and monostatic
radar the corresponding change in the field intensity of the scattered wave.
from the relation(6.2) that it characterizes EPR extremely energetic properties
of the object, since its definition includes the absolute values of the field. If the
interest is the phase of the scattered wave, in which case a complex PCP, defined
by the complex ESR:
  4rENeg/ Epad
(9).
HFSS-13 counts bistatic normalized bistatic, complex bistaticmonostatic and
PCP. In this example, we will calculate the normalized PCP for bistatic and
monostatic radar positions and goals. IWavePhi changing the angle from 0 to 180
degrees (Fig. 6.2), we would like to make "flying around" goals, irradiating the target
with all sides (Fig. 6.3), and taking the return signal.
Fig. 6.2. plane wave parameters with the settings to calculate the
monopulse RSC
71
Overallprocedure for calculating RCS object by finite element method in the
frequency domain, consists of the following steps:
1. CreatureHFSS project.
2. drawing geometrical model, including:
 installation drawing area (boxing,wherein the object) will be placed;
 object creation;
 object reference material(In this case, the object to which a plane
wave, perfectly conducting).
3. installationsolving the problem of parameters:
 setting the boundary conditions on the box surfaces (we will use the
PML conditions);
 installing a driving source (a plane wave);
 setting criteria for convergence and partitioning options.
4. Calculation of RCS.
5 . Post-processing for rendering RCS.
applicablethe algorithm for calculating the scattering from a simple object - a
perfectly conducting cube in space.
Creating a model (RCS Model).ChoosingDriven Modal type of task.
to install unit of length:
1. ClickModeler> Units. Dialog appears Set Model Units.
2. Select the meter unitSelect units from the menu. Option Rescale to new units
not included.
If you have installedoption Rescale to new units, the geometric grid
automatically scales the distance between the grid lines to the units introduced in
such a way that the difference would meet the established units.
3. Click The OK, to install the meters as the unit of length for this model.
Build a perfect conductor(PEC) cube surrounded by an air box (Figure 6.3).
On Boxing PML surfaces boundary conditions are set.
OnThis cube plane wave. Calculations RCS- Radar Cross Section.
71
Fig. 6. 3. The conductive cube inside the air box
For creation
Cuba, use
team
conducting cube has an edge length0.75 m (Figure 6.4).
Draw> Box.
Perfectly
Fig. 6.4. The parameters of the conductive cube
that set the properties of a cube:
1. Selectdesigned box and click Properties from the pop-up menu. This displays
the Properties dialog box.
2. Give the name of the object.
3. In the material set pec material from the list and click OK.
4. In the Properties dialog, edit the color and select red.
5. Set transparency 0.6.
6. ClickOK, that choose these settings and close the dialog.
71
Creating an air box. Draw a box size1.4 m with a center at the origin, with
transparency 0.9. Call it air_box. Draw a box on the air command Draw> Box, and
edit it in the dialog size, shown in Fig. 6.5.
Fig. 6.5. The dimensions of the air box
Side of the air boxThey will be removed from the test perfectly conducting cube
at a distance greater than the wavelength, given the frequency of 300 MHz, which
we will use.
Setting boundaries PML. To create a PML boundaries:
1. Install
option
choice
surface
and enter
team
Edit> Select> Face.or by pressing F.
2. Select
Edit> Select> By Name.or select the menu. This displays a
dialogue Select by Face.
3. In the list of object names to select air_box.
This l i s t i s c a l l e d
air_box faces.
4. Holding Ctrl key, click on each surface of the box. Air_box All surfaces
should be highlighted (Fig. 6.6).
71
Fig. 6.6. Selection of all sides of the air box
5. In the drawing window, call the right-click pop-up menu and select Assign
Boundary> PML Setup Wizard. Setup Assistant appears (Fig. 6.7).
Fig. 6.7. Fitting Assistant PML surface
6. In the Uniform Layer Thickness (Fig. 6.7), set the layer thickness of 0.4 m.
Parameters PML layers will be corrected automatically in accordance with the new
thickness.
7. Installoption matching angles and edges: Create joining corner and edge objects,
and click Next. This will create a PML objects and dialog appears Material
Parameters (Fig. 6.8).
71
Fig. 6.8. Fitting for calculating parameters PML
8. Install
minimum
frequency Min Frequency = 0.3 GHz,
and minimu m dist anc e radiat i on Min imu m Radiating Distance = 0.3
m.
9 .Click Next to display the final dialogue PML Summary (Fig. 6.9).
Fig. 6.9. The final installation of the assistant PML
10.ClickFinishto closedialogue Fig. 6.9. PML border there appear in the
Boundaries section of the project tree, and PML objects listed in the project tree.
71
partitioning tetrahedra in the air box. PSettingspartitioning air_box boxing will
greatly affect the accuracy of the calculation of the radiation pattern. To set these
parameters:
1. Highlight the box surfaceair_box.
2. Right clickon Mesh Operations in the project tree.
3. Click Assign> On Selection> Length Based. This displays a dialogue
Element Length Based Refinement (Fig. 6.10).
Fig.6.10. Dialog division setting parameters
4. Installthere parameter Maximum length of Elements = 0.2 m.
5. Click OK to closethis dialogue. In the project tree in the Mesh Operations
folder icon appears Length1.
Setting the parameters of the incident wave.
1. Ask teamMenu HFSS> Excitations> Assign> Incident Wave> Plane
Wave. A page Incident Wave Source: General Data (Figure 6.11.).
2. typethe source name in the Name box.
3. Selectsee Vector Input Format as the Spherical.
4. Enter 0, 0,0 for X-, Y-, and Z-coordinates of the Excitation Location
and / or Zero Phase Position (starting point for an incident wave).
81
Fig.6.11. Setting the incident planar wave phase center
5. ClickNext.
6. appearsdialogue Incident Wave Source: Spherical Vector Setup (Figure 6.12.).
a. In the IWaveTheta, enter
Start =0 deg,Stop= 90 deg, and Step = 3deg. For
the monostatic case, the RCS will be calculated only for IWaveTheta angles,
imposed
here.
Therefore way
radar reflection
coefficient will be calculated for thirty angles.
b. Click View Point List (Fig. 6.12) in order to see the set values of the angle θ.
Fig.6.12. Parameters incident plane wave
7. ClickNext. A page Incident Wave Source: Plane Wave Options (Figure 6.13.).
81
8. In the Type of Plane Wave Select Regular / Propagating; all other fields will be
inactive.
Figure 6.13. incident wave parameters
10. ClickFinish. The incident wave that you have determined is added to the list of
Excitations and its fall line can be seen in Fig. 6.14.
Fig. 6.14. The test cube and corners, which will be calculated RCS
Next we describe how to performThe Setup endless areas for monostatic and
bistatic experiment. These configurations can be calculated graphs normalized
bistatic and monostatic RCS.
Create infinite scope to determine the far-field.to calculateradiated far field, it is
necessary to establish a sphere that surrounds the light object.
82
Forthis example, create a setting for the bistatic and monostatic cases.
Whenyou set the spherical surface to analyze the short-range field and far
field, you specify the range of variation and pitch angles change φ and θ (azimuth
angle and elevation angle). They indicate the direction in which you want to evaluate
radiated fields (see. Fig. 6.15). For each value of the angle φ has a corresponding
range of values for θ, and vice versa. This creates a spherical grid. The number of
grid points determined by the step size for φ and θ.
Fig. 6.15.The spherical coordinate system Setting in
monostatic case Monostatic Setup.
1. Click HFSS> Radiation> InsertFar Field Setup> Infinite Sphere.
appearsConversation Far Field Radiation Sphere Setup (Fig. 6.16).
83
Fig. 6.16.Target spheres, which are set points which are calculated in the
far-field characteristics
2. The Infinite Sphere tab, type a name for the scope in the Name box. For
monostatic scope, type the name of the monostatic.
3. Determine the range of variationin the dialogue corners Fig. 6.16. angles do not
change for monostatic case, since the position of the receiving antenna is fixed and
coincides with the position of the transmitting antenna. Therefore, RCS is calculated
only in the direction determined by the incident wave. This direction is determined
by the angles and IWavetheta IWavephi, which have already been set in the
determination of the excitation source (Figure 6.12).
4. Open a bookmarkCoordinate System(Fig.6.16) and orient the sphere in the
global coordinate system (CS). Select the Use global coordinate system. If you want
to orient the scope in accordance with the coordinate system selected by the user,
you can select Use local coordinate system, and select a coordinate system from the
list Choose from existing coordinate systems.
5. Click bookmarkRadiation Surface. Leave the selection of UseBoundary
Radiation Surfaces. If you want to specify a different type of surface, it is necessary
to use the Use Custom Radiation Surface.
6. Click OK. Monostatic sphere is created.
the scope of the task in the bistatic case.
84
1. In the project tree, right click on the Radiation and from the pop-up menu, click
Insert Far Field Setup> Infinite Sphere. Appears Far Field Radiation Sphere Setup
dialog.
2. entername "Bistatic".
3. Set the value of Phi Start, Stop, and Step = 0.
5.
InstallTheta Start = 0, Stop = 180 deg and Step = 1.
6.
ClickOK, to close the dialog.
installation Bistatic Radiate appears in the section of the project tree.
askFurther frequency of 0.3 GHz and run the program for the calculation.
Creating graphics for bistatic
the graph, follow these steps:
RCS. that set the parameters of
1. Right-clickthe Results in the project tree and select Create Far Fields Report>
Rectangular Plot. New Report dialog box will appear - New Traces.
2. leaveSolution option in the section select the solutions both Setup1:
LastAdaptive.
3. To display the graph of the RCS, you chooseone of the geometries in the list
Geometry.To do this, select the schedule Bistatic.
4. Under Category, select Normalized Bistatic RCS.
This choice implies Quantity list to show for NormRCS
Total, Phi, Theta, X, Y,and Z, with a choice of Total Selected.
5. Forthis function, select dB. After this selection box section Y
TraceIt shows db (NormRCSTotal).
85
Fig. 6.17. selection characteristics bistatic radar reflectance
6. Open a bookmarkFamiliesand check that the angle IWaveTheta = 0.
7.Click New Report. Displays a graph in Fig. 6.18.
Fig. 6.18. Schedule for bistatic radar RCS case
axially x this schedule - the angle of observation.
Creating graphics for monostatic RCS.The sequence of creating a schedule
monostatic RCSsimilar to that described above for the bistatic case.
86
1. Right-clickResult on the icon in the project tree and select Create Far Field
Report> Rectangular Plot. This opens the dialog characteristics (Fig. 6.19).
2. In the Context box, select the Geometry section Monostatic.
3. In the Category, select MonostaticRCS, and the list Quantity
select MonostaticRCSTotal.
5. From the Function list, select dB.Then the reflection coefficient will be
dB.
6. SelectX in the characteristics IWaveTheta.
7. pressNew Report. This creates the report and adds it to the project tree.
Fig. 6.19.Dialog for setting parameters generated in the case of monostatic
87
Fig. 6.20. RCS dependence on the angle of incidence of the incident plane wave
(monostatic mode)
On Fig. 6.20 shows a plot of RCS as a function of angle of incidence of a
plane wave.
88
7. ESR calculation object of a large size electrical
To solve the problems,which are modeled by objects having large electrical
dimensions, the method of integral equations. For its designation in HFSS system
uses the acronym EFIE. The significant difference of an integral equation of the
finite element method, which has long been the only method for solving of
electrodynamic problems in HFSS, is that in the sampling method of integral
equations not exposed space, and the body surface, the scattering of electromagnetic
waves. Obviously, with this order of the system of linear equations is considerably
less than in the case of finite element method. Reducing the dimension of the system
allows for real time analysis of the fields generated by the objects with large
electrical dimensions.
comparing the twomethod of solving boundary value problems of
electrodynamics, it should be noted that the finite element method is a universal
method that has no fundamental restrictions on the analyzed structure. The method
of integral equations demonstrates greater efficacy only in certain cases. The fact
that the recording of integral equations is based on the knowledge of the Green's
function. In the simplest form of the Green's function is known only for a limited
set of objects, which include: a homogeneous space, layered structure with endless
layers, infinite baffle, etc. For them, it is advisable to use the method of integral
equations.
However, as soon as it comesan analysis object of arbitrary shape, e.g., a
complex configuration of the dielectric body, the advantages of the integral equation
method becoming less apparent, since formulation of these equations is
complicated. Essentially, we have to look for unknown numerically analytically
Green's function, which requires a significant investment of computer resources. As
a result, efficiency of the two methods becomes comparable.
scattering wave metal bodies arranged in the free space is one of the tasks for
which the method of integral equations is very effective, as for the formulation of
the boundary value problem is sufficient to know only the Green's function of a
homogeneous medium, which is written in a very simple manner.
In this design (RCS of an Ogive) using the method of integral equations
calculated effective scattering surface (EPR) metal airship. The airship is modeled
89
form of
rotation.
axis (see. Fig. 7.1).
three-dimensional
object, which
the body
Surface airship created by rotation 0x curve around the
Fig. 7.1. The model airship
Structure reviewed at
a plane wave incidentangles
frequency 1.18 GHz. it is irradiated
.. Consider a fall wave
XOY plane.The angle θ = 90o, and the angle φ is changed from 0 to 180o from
step 2o.
Let us count the EPR for the case of a monostatic radar,when the direction of
the point of observation coincides with the direction at which the wave is incident
on the object under study. The analysis results in a monostatic EPR can be seen on
the graph, which shows the dependence on the angle of the EPR irradiation target
IWavePhi, varying from 0 to 180º. most
r ef lec t i on
c oef f i c i en t
p r op or t i on a l t o a n ef f ec t i ve sc a t t er i n g s u rf ac e, i s exp ec t ed t o r ea ch
at an angle of 90º.
Considercreating a further sequence of the airship and its pattern analysis
system HFSS method of integral equations. The model is created as follows. First,
in the plane XOY curve is drawn, which is a rotating body the generatrix. Creating
a curve occurs using the equation in analytical form. Then, the curve is rotated
around the X axis to create a 3-D object. Drawing forming the airship performed by
team Draw-> Create Equation Curve. dialog appears for this team Fig. 7.2 in which
the coordinates of a curve-dependent variable variable _t.
91
Fig. 7.2. Draw the curve in space using Equation parametric curve
The equation generator airship is as follows:
F(x)

cos.

(10)
F(X) 5
y
.
1cos
where x, y, z defined in inches (inches), and α - forming an angle selected
for

22.60
The default unit of measurement of length can now be changed
to inches, but they may change later.
1. Click Draw> Equation Based Curve. Dialog appears Equation Based Curve
(Fig. 7.3).
As a parameter using a variable _t, which coincides with the coordinate x: x
= _t. The equation for the y coordinate becomes:
2
t
1 sin 0.3948 cos 0.3948
y
5

1cos 0.3948
Angles are enteredin radians. It should also specify the length of the unit,
which we multiply each an (x and y) by an amount (1in). So, in the Equation Based
Curve dialog box, enter the equation as follows (Figure 7.3.):
91
X (_t)= _t * (1in)
Y (_t) = (sqrt (1 - (_ t * sin (.3948) / 5) ^ 2) -cos (.3948)) / (1-cos
(.3948)) * (1in)) Z (_t) = 0
We define the starting point Start _t = -5, endpoint End _t = 5 and the
number of dots in a line Points = 24 (Fig. 7.3).
Fig. 7.3. Drawing forming an airship
press OK, to see a curved line.
findHistory in the tree section Lines link Equation Curve1 and double-click
CreateEquationCurve. This allows you to see the Properties window. If the line does
not match the expected mean (Fig. 7.4), it can be edited.
The HFSS-IE curved elements are missing. So, all the curved surface
approximated segmented models. With this in mind we recommend entering the
required number of segments in the Number of Segments.
Fig. 7.4. Drawing images of objects
92
2. Select the curve EquationCurve1 and then click Draw> Sweep
> Around Axis. Dialog appears Sweep Around Axis (Fig. 7.5).
3. In Sweep Around Axis dialog box, select the X axis, Angle of sweep = 360 and
Numberof segments = 24. Click OK.
Fig. 7.5. rotation curve parameters to create a three-dimensional model
Material The default for this object, select, for example, copper.
Typically airships (Fig. 7.6) are made of aluminum.
Fig. 7.6. Type of object to be analyzed with the direction of observation
93
4. SelectOgive object and in the Properties window, click in the section Material
copper and choose the Edit ... (Fig. 7.7). This will open the Materials dialog.
Fig. 7.7. Modify the properties of the material
5. The material selection window, select «aluminum» and click OK. Creating the
model is finished. It has no air box, as in the method of not using integral equations.
Installation incident plane wave.Now you need to add the source of the incident
plane wave. We will count the monostatic RCS in the xy plane. Those. radar
reflection coefficient will be considered, in the airship circling the xoy plane (Fig.
7.6).
Since this is the monostatic RCS, you will need to include a range of angles
of incidence. To get an accurate picture, you can select the step change angle is
equal to 3 °. The incident wave is given by the following steps:
1. Deselect model and right-click
AssignExcitation> Incident Wave> Plane Incident Wavefrom the pop-up
menu. A dialog opens Incident Wave Source: General Data (Figure 7.8.).
2. In this dialog, select the Spherical option to cover the format of the airship and
leave the position of the phase center at the point (0,0,0).
94
Fig. 7.8. Target point plane wave radiation
3. Click Next and set the window Spherical Vector Setup (Fig. 7.9)
meaningcorners Iwave Phi:
Start = 0, Stop = 180, and Step = 3.
4 .For IWave Theta select Start = 90, Stop = 90 Step = 0.
5. Then, set the value Eo Vector Phi = 1 and E Theta = 0.
Fig. 7.9. Setting the parameters of the scope of the far field
6. Click Next and then Finish.
Excitation set. The incident wave (Incident Plane) appears in the section of
Excitations in the project tree. If you select it, the window appears with the image
of the airship "fan" in the form of calculated angles.
Excitation set. The incident wave (Incident Plane) appears in the section of
Excitations in the project tree. If you select it, then in the window
95
airship pattern appears with the "fan out" in the form of angles, which deviates the
incident plane wave.
This setting on the decision to address the incident waves at θ = 90 ° and φ = 0
180 ° for 61 points.
...
Polarization is determinedan angle φ,
directed
So,
what we
can
calculate
monostatic radar
reflectance (monostatic RCS).
NowPerform a solution for that:
1. pressRight-click on Analysis in IEDesign section in the Project Manager
window and select Add Solution Setup, to open a dialogue Solution Setup (Fig.
7.10).
2. The General tab to change the frequency Frequency = 1.18 GHZ. The other
parameters, leave the default.
Fig.7.10. Frequency setting calculation
3. Review the data in tab
Options - seal partitioning into cells
lambda refinement = 0.25 (= λ0/4).the calculation of the installation completed.
To run the simulation:
1. pressRight-click the Setup, and then click Analyze.
2. Save your problem and enter the name of the project. Simulations were
performed for two iterations.
Output current on a surface. that see the current to the airship surface
(Figure 7.11.):
1. Select the surface of the model and set the output current distribution
Fields command> J> Mag_J. This opens the Create Field Plot dialog.
96
Fig. 7.11.
The currents on the surface of the airship
2. Click Done.
You You should see an induced current that is induced
in the metal when subjected to an incident plane wave.
conclusion schedule RCS coefficient.
To graphically display the RCS, you must first enter the setting far-field - the
same way as is done in HFSS. Create far field installatio
= 0 ° and one angle θ = 90 °.
weWe will graphically depict the dependence of monostatic RCS when changing
the incidence angle phi of the plane wave. Therefore each observation point
corresponds to only one corner (Fig. 7.12).
Fig. 7.12.Monostatic radar graph of reflectance when changing the incidence
angle of Phi
To display the schedule dependency monostatic RCS depending on the azimuthal
angle phi incident plane wave:
97
1. Right-click on Radiation in the project tree and select Insert Far Field Setup>
Infinite Sphere ... Appears Far Field Radiation Sphere Setup dialog.
2. enterStart and Stop for Phi equal to 0, and a step 10º (Fig. 7.13).
3. enter Startand Stop for Theta = 90º and Step size = 10º.
3. Click OK. This will close the dialog and create the scope InfiniteSphere1 in
the Radiation section of the project tree.
Fig. 7.13.Schedule setting far field calculation data. In fact, one corner and
one angle
5. To create a graph, click on Results in the project tree and set the command
Create Far Fields Report> Rectangular.
So,This section is intended radar reflection coefficient for the object whose size
is much larger than the wavelength. The calculations use the method of moments,
implemented in the HFSS-IE program.
98
8. The bandpass waveguide filter
In this section we will look at the analysis using HFSS bandpass filter,
which is widely used in practice, the comb structure. Further, the band pass filter
model, we will use HFSS to illustrate additional features, such as accounting for the
temperature dependencies and Tune tuning mode.
microwave synthesisthe filter can be performed using the first stage of
designing popular program Microwave Office [6]. Utility Synthesis Filter This
program allows you to calculate generalized parameters that make up the filter
transmission lines, as well as their sizes are realizing the predetermined frequency
characteristics. Note that the calculation of the geometric dimensions is possible
only for transmission lines planar type: stripline, microstrip, etc. Similarly, analysis
of the frequency response of the filter in MWO system can be carried out only for
stripline constructions. At the same time, in practice, often used comb filters based
on transmission lines with the cylindrical and rectangular conductors, which have a
much higher quality factor than the strip line. The analysis of such structures can
HFSS means.
We perform an analysis of three-dimensional comb bandpass filter (Fig. 8.1) with
1 GHz bandwidth. The filter consists of eight pins, each of which is a cavity. At the
inlet of the first pin connected to a coaxial transmission line with a characteristic
impedance of 50 ohms.
Fig. 8.1. Type of filter in HFSS-13 interface
99
pectinatebandpass filter has an order related to the number of resonators. It
depends on the number of resonators of the filter quality: transmission passband,
and the recession is the transmission coefficient bandwidth.
The length of the individual resonators is not more than a quarter wavelength, and
often much less due to the effect of shortening that occurs due to the capacitance
between the terminal pin and the wall of the housing. The distance between the pins
determines the coupling coefficient between the resonators (Fig. 8.2).
Fig. 8.2. The dimensions of the comb filter
101
atsolving this example uses many HFSS program features, including the
calculation of the filter characteristics depending on the ambient temperature.
Driven Terminal feature of the method lies in the fact that the ports are set up as the
sources relative to the outer casing.
Create a new project.To create a new project, click the command File> New New.
From the Project menu, click Insert HFSS Design Design. To install the solution
method, click HFSS> Solution Type. Dialog appears Solution Type (Fig. 8.3), in
which select the Driven Terminal.
Fig. 8.3. The choice of method of calculation of the bandpass filter
MethodDriven TerminalIt calculates S parameters for ports that are formed
multiconductor transmission lines (multi-conductor transmission line ports).
Elements of S-matrix are then determined by the voltage and currents that
distinguishes its standard power determine the scattering matrix through the wave
power. Note, however, that for the transmission lines with the TEM - waves (coaxial
symmetric stripline, etc.), both determination of the scattering matrix give
equivalent results.
Creating a three-dimensional model. Set the model unit. To do this, follow these
steps:
1. Click Modeler> Units. A dialog opens Set Model Units (Fig. 8.4).
2. Select:in (inches).
3. Click OK.
101
Fig. 8.4. Installation length Units - inch (34.5 mm)
In this example, when drawing will apply in the line of the coordinate input
state, i.e., in the form of numbers (Fig. 8.5). To draw a three-dimensional
substrate:
1. Click Draw> Box.
2. Use input field coordinates, a we coordinate input
X: -1.0, Y: -1.7, Z: -0.3125,
pressing each time Enter.
Fig. 8.5. Indication status line
Fig. 8.6. The dimensions of the filter housing in inches (in)
3 . Using the coordinate input field, we introduce the opposite corner
Boxing
dX: 2.0, dY: 3.4, dZ: 0.0,
pressing Enter.
4.Using the coordinate input field, we introduce the substrate height
dX: 0.0, Y: 0.0, Z: 0.625,
pressing Enter. The filter housing
will have the form shown in Fig. 8.6.
Creation of the outer cylinderan input coaxial line.To create an outer conductor
of the coaxial line:
102
1. Click Modeler> Grid Plane> YZ, to select the active plane, which will be
plotted base of the cylinder.
2. Click Draw> Cylinder.
3. Using the coordinate entry field, enter the central position
1.0, Y: -0.9, Z: 0.0 and press Enter.
X:
Fig. 8.7. coordinate input field at the bottom of the interface HFSS
4 .Using the coordinate entry field, enter the cylinder radius (Figure 8.7.)
DX: 0.0, dY: 0.14, dZ: 0.0 and press Enter.
Fig. 8.8.The filter housing and the first cylinder of the coaxial line
5 .Using the coordinate entry field, enter the cylinder height
dX: 0.75, dY: 0.0, dZ: 0.0 and press Enter.
In the Properties window, select the name
filled the vacuum (Fig. 8.9). Click OK.
feed1and leave the cylinder
Fig. 8.9. Selection vacuum to fill the space
Creature
internal
conductor
coaxial
draw inner conductor of the coaxial line filter at the inlet:
lines.
that
1. Click Draw> Cylinder.
2. Using the coordinate entry field, enter the center position
103
X: 1.0, Y: -0.9,
Z: 0.0
and press the Enter key.
3. Using the coordinate entry field, enter the cylinder radius
dX: 0.0, dY: 0.06, dZ: 0.0
and press Enter.
4.Using the coordinate entry field, enter the cylinder height
dX: 0.75, dY: 0.0, dZ: 0.0 press Enter.
Define the name of the cylinder (Fig. 8.10) as the feedpin1 material and
fill it as the PEC.
Fig. 8.10. internalconductor of the coaxial input
This inner core has continued into the filter housing. To draw a line
inside the coaxial line (pin):
1. Click Draw> Cylinder.
2. Using the coordinate entry field, enter the center position
X: 1.0,
Y: -0.9,
Z: 0.0
and press Enter.
3.Using the coordinate entry field, enter the cylinder radius
dX: 0.0, dY: 0.06 dZ: 0.0
and press Enter.
4.Using the coordinate input singing, the value of the height of the cylinder
dX: -0.15, dY:0.0
dZ: 0.0 press Enter.
Specify a namethis feedprobe1 cylinder.
Fig.
8.11. Pin connection with filter
104
Creating resonators
Resonators are plotted as a metal box. Drawing parallelepiped (fig. 8.12) is given
point, and then dimensions. To create the cavity with the name l1:
1. Click Draw> Box.
2. Using input field coordinates, enter the position of boxing
X:0.85, Y: -0.9625, Z: -0.03, press Enter.
3.Using the coordinate entry field, enter the opposite corner Boxing
dX: -1.7, dY: 0.125, dZ: 0.06, press Enter.
Fig. 8.12.Drawing a first line filter
To create a filter with the name of the pin l2 (Figure 8.13.):
1. Click Draw> Box.
2. In the field enter the coordinates, enter the boxing position
X: -1.0,
Y: -0.75, Z: -0.03, press the Enter key.
3. Next, enter the coordinates of the opposite corner Boxing
dX: 1.818, dY: 0.125, dZ: 0.06, press Enter.
105
Fig. 8.13.Draw a second line filter
To create a third pin of the filter named l3 (Figure 8.14.):
1. Click Draw> Box.
2. Using the coordinate entry field, type a parallelepiped: X: 1.0,
0.48,
Z: -0.03, press Enter.
3. Using input field coordinates, enter the size of the X: dX: 1.818, dY: 0.125, dZ: 0.06, press Enter.
Y: -
Fig. 8.14. Draw a third line filter
To create a fourth resonator l4 (Figure 8.15.):
1. SelectDraw> Box team.
2. Using input field coordinates, enter the boxing position: X: 1.0,
Y: -0.2, Z: -0.03 and press Enter.
3. Using input field coordinates, enter the opposite corner Boxing: dX: 1.818,
dY: 0.125, dZ: 0.06, press Enter.
106
Specify the name of the l4 element in the Properties window and click OK.
Fig. 8.15.Created fourth resonator
filter
Fig. 8.16.A dedicated section of the port
Creating a wave port
highlight chamferport (Figure 8.16.), as follows:
1. Click the command Edit> Select> Faces (or press F).
2. Select the external chamfer coaxial line at X = 1.75in.
To set the excitation wave in the port:
1. Click HFSS> Excitations> Assign> Wave Port. dialog appears
Reference Conductors for Terminals (Fig. 8.17).
2. Ask as the name of the port p1.
3. Select the option to name the terminal Use port object name.
4. leaveoption Use as Reference (Fig. 8.17) unchecked.
107
Fig. 8.17. Description port
Fig. 8.18.Load icon in the project
tree and dialogue terminal
5.Click OK. In the project tree under Excitations sign appears p1 load. The method
Terminal solution to each port is given the reference impedance. Therefore, at the
same time c port appears and the load icon in the project tree (Fig. 8.18).
Next, we execute a rotation around the center of the housing with copying for the
created of the filter. To select objects to replicate:
1. press"O".At the command Edit> Select> By Name Select Object dialog box
appears (Fig. 8.19).
2. Select it with the names of objects: feed1, feedpin1, feedprobe1, l1, l2, l3, l4, hold
down Ctrl + Left.
108
Fig. 8.19.Dialog selection of objects
Fig. 8.20.Conversation not
duplicated and rotation of the
filter
3. ClickOK.To create the rest of the pin filter, follow up with a turn at 180º for that
click Edit> Duplicate> Around Axis, and then in Figure dialog. 8.20 choose:
Axis:Z
Angle: 180
Total Number: 2
and click OK.
Fig. 8.21.Duplicate half elements by rotation about an axis
bandpass filter model takes the form shown in Fig. 8.21.
View boundary conditions.To check how the boundaries are set, click HFSS>
Boundary Display (Solver View).
After
a preliminary decision, a dialog Solver View of Boundaries (Fig. 8.22). In this
dialog box Visibility option, you can mark boundaries for which you wish to see.
Note the following:
109
- Ground plane background (Perfect Conductor) is shown if the name of the
noted outer (Fig. 8.22).
Fig. 8.22.Display Options borders
- All Guides, which are assigned to the boundary conditions Perfect
Conductors will be shown as a border smetal.
Click View> Visibility, visibility to remove any parts of the project. You can
also change the type of boundary conditions and ports. Visibility of objects can be
set in the dialog box that appears on the command View-> Active View Visibility
(Fig. 8.23).
Fig. 8.23. selection the visibility of the objects included in the project
Viewing visibility of boundary conditions and facilities allowing verify the
correctness of the design creation. Click also HFSS-> Validate Check, to run the
validator checks geometric construction problem. If after checking the Validator no
comments, you can make settings for analysis.
111
Settings for analysis.
Click HFSS> Analysis Setup> Add Solution
Setup. Dialog appears Solution Setups (Fig. 8.24). In the General tab, make the
settings shown in Fig. 8.24.
Fig.8.24.Dobavlenie calculation
in the frequency range
Fig. 8.25.Target calculation in
the frequency range by means
of fast scan
To calculate the frequency range refine HFSS> Analysis Setup> Add Sweep.
Appears Edit Sweep dialogue in which we define the parameters indicated in Fig.
8.25.
Calculationcharacteristics of the bandpass filter.To start the process
of solution, click HFSS> Analyze. To see the process of dialogue and solutions
Solution Data:
1. Click HFSS> Results> Solution Data. Solution Data dialog appears.
2. Open the tab Profile, to see the solution file.
3. Open theBookmark Convergence, to see details of the convergence solutions.
Convergence can be seen in the tabular form (Table Table), or in a graphical
representation (Plot).
4. The Matrix Data tab, you can see the matrix. And to see a change Matrix Data
matrix data during calculation in select Setup1 Last Adaptive.
5. ClickBookmark Mesh Statistics, to see the data on the grid subdivision.
111
CalculationIt is formed on a center frequency 1.5 GHz and a frequency range of
0.6 to 2.4 GHz. C-scan type Fast calculation is executed in the frequency points
where the filter characteristics are changed at a high speed.
Output characteristics of the graph.To derive the frequency dependence of
the filter S-parameters of the frequency:
1. Click HFSS> Results> Create Terminal Solution DataReport>
Rectangular Plot. A dialog Reports.
2. Ask following options:
Solution: Setup1: Sweep1
Domain: Sweep
Quantity:St (feed1_T1, feed1T1); St (feed1_T1, feed1T2)
Function: dB
3. Click New Report button.
4. Click Close.
Fig. 8.27. frequency characteristicwaveguide comb filter and adding markers on
the graph
Thenyou can add another characteristic of St (feed1_T2, feed1_T1) in
dB,Add Trace by pressing the button and then the Done button.
112
Fig. 8.28.Chastotnye filter characteristic
For the convenience of viewing the schedule can be changed
Do the following:
scale. F o r
1. Twiceclick on the Y-axis. Dialog appears Y-axis properties.
2. Click tab Scaling, which sets out options: Autoscale: Remove
this option
Min: -1.0
max:0.0
3. Click OK. In this case, you can more accurately see the change in the
gain of the filter passband.
Viewing field inside the filter.This filter need to install a source that has an
internal resistance. Such a source is called Terminate Port. To install the port load:
1.Click HFSS> Fields> Edit Sources. Appears Edit Sources dialog in which
concentrated on all the project data sources. In this dialog box, select:
Select source: p2:
T1Terminated:
Checked Resistance:
50
4. Click OK.
reactance:0
113
Fig. 8.29. Editing sources included in the project
Conclusion calculated field in the section. Select the plane
XoY,activating Global XY Plane.
1. Using the Model Tree, expand Planes folder.
2. Select Global plane: XY.
3. Click HFSS> Fields> Fields> E> Mag_E. Create Field Plot opens a dialog
in which to choose:
Solution:Setup1: LastAdaptive, Quantity:
Mag_E,
In Volume: All.
4. Click the Done button.
IfWe need to change the form of the field graphics, for example to emphasize
the weak fields, click HFSS> Fields> Modify Plot Attributes. Opens the Select
Plot Folder dialog.
Select E Field, and then the following options in the dialog E-Field1
Click the Scale tab.
Select Use Limits, and set: Min: 5
Max: 1500
Scale:Log.
Click
Close.Fieldon the surface becomes more understandable form
(Fig. 8.30) which can be estimated and resonance phenomena in the filter.
Fig. 8.30. Type field in the middle of the filter section
114
So, in this section was carried out drawing and calculation of threedimensional filter to the pins. To perform further optimization of the filter, you
must specify the variables to select the method and optimization objective
function.
Objects that are used as components in Ansoft complex designs, which are
arranged in the interface circuits (Workbench Project Schematic) may be regarded
as system. The complex Ansoft programs that serve to solve the circuit and system
tasks, is: Rmxprt 14.0, Designer 6.1, and Simplorer 9.0. Such Ansoft products as
HFSS 13.0, Maxwell
14.0 and Q3D Extractor are used to solve the electrodynamic structures. If you call
ANSYS DesignXplorer (Fig. 10.31), the variables of any project are also exported
to the project and are available from the interface Workbench.
Fig. 8.31. Adding modes and optimization solutions
All products
Ansoft united a common interface Workbench
team,management, etc. One of the opportunities available in HFSS v.13
- calculation of the temperature. These features will be described in the following
chapter.
12. The antenna is mounted on the mast
This example is solved by the method of integral equations, which is implemented
in HFSS_IE block. This method makes it possible to analyze the structure of large
electrical size.
Consider as an example of an antenna mounted on the metal mast. It consists
of three dipole antennas (ris.12.1). The mast is located on the infinite ground plane.
Fig. 12.1. The antenna on an aluminum leg
The mast hasheight 3.1 m. vibrators are modeled as a 2-D objects with
boundary PerfE. Excitation - discrete ports. Mounting structure modeled
polystyrene holders. The mast stands on a conductive ground, so the project is set
active option infinite ground InfGndPlane1 board under boundary conditions (Fig.
12.1).
Create a new project by clicking on the
. Choosing a unit of length, see the
dialog (Figure 12.2.), Called Modeler team -> Units.
Fig. 12.2.Unit of length for drawing objects
Options method for solving HFSS-IE c a n set in Fig dialogue. 12.3
which is caused HFSS-> Options command.
Fig. 12.3.method of installation options HFSS-IE
Draw a mast with a height of 270 cm, a section of 20 cm x 20 cm (12.4 in Fig.)
And the fill material from the library: aluminum.
Fig. 12.4. Dialogue parallelepiped - mast antenna
Next, draw the pin,creating a first octagon, and then converted into the threedimensional volume, holding it on the axis Z (Fig. 12.5) command Draw-> Sweep> Along Vector.
Fig.12.5. Dialog, wherein the edited-size antenna mast
Further fulfill
drawing
polystyrene
dimensions first shown in Fig. 12.6.
holders
(Fig.
12.6).
Fig. 12.6.polystyrene holders
the creation of threepolystyrene holders execute the command Edit -> Copy.
Further fulfill
drawing
vibrators
of Draw rectangles team -> Rectangle.
antennas (Fig.
12.7)
at form
Fig.12.7. Dialogue vibrator antenna
Vibrators consist of two arms, between which is installed a discrete port.
Platform for the port also draw a rectangle as a team Draw-> Rectangle. Geometric
port settings are shown in Fig dialogue. 12.8.
Fig.12.8. site parameters for the discrete port dipole antenna
The antenna on the mast is above the infinite ground. This fact is reflected in the
fact that the project is put into an earthen Board of infinite size (Figure 12.9.) Team
HFSS IE -> Boundary-> Assign-Infinite Ground Plane.
Fig.12.9. The antenna
on the mast
Fig. 12.10. The task endless
plane
earthen
Groundcard is always set in XOY plane. In the dialog box Fig. 12.10 inputting
coordinates on Z and roughness of the surface from which depend ground board
loss. The material can be set by pressing the Select Material button.
on the job is created modeling Add Solution command on which the figure is
displayed dialogue. 12.11. Carrying out tasks for the calculation of the antenna by
HFSS-IE.
Fig. 12.11. Specifying installation calculation and decomposition of the grid in
the method
HFSS-IE
Calculationantenna characteristics is performed at the same frequency of 0.9
GHz, so establish the adaptation process at this frequency (Fig. 12.11).
Fig. 12.12. Characteristics antenna match
Start the solution by clicking Analyze. After calculation command HFSS-> Fields
-> Edit Sources establish excitation on all three ports (Fig 12.12.), And viewing the
current flowing through the metal coating of the mast and the directional pattern
(Fig 12.13, 12.14 fig..).
Radiation
Pattern 0
-thirty
1
thirty
12.80
9.60
-60
60
6.40
3.20
-90
90
-120
120
-150
150
-180
Fig. 12.13. The
Fig. 12.14. NAM-section of the antenna on the
currents on the
mast
surface of mast
In conclusion, the method of integral equations allows to perform modeling of
structures both with dimensions much larger than the wavelength, and with
dimensions much smaller than the wavelength. As a practical example may be
mentioned calculation wire antenna (Fig. 12.15) 100 m in height, operating in the
wavelength range 100 kHz.
Fig. 12.15. Wire antenna with stretch marks
The sizethis antenna is much smaller than the wavelength. The input impedance
of the antenna at a superlong range is very small. Increase it, you can use the
selection of the number of braces, the choice of the angle at which the backstay are
relative to the ground. The input impedance also significantly depends on the
conductivity of earth surface. The main task of designing such an antenna can be
regarded as an increase in efficiency (Fig. 12.16), which depends on the frequency.
Fig. 12.16. The frequency dependence of the efficiency of the umbrella antenna
height of 110 m and 110 m in length braces.
In such an antenna structure can occur very high voltages. To calculate the voltage
between the antennas can be decomposed extensions distance from individual
antennas to ground fragments to form individual elements. The program calculates
the field strength and voltage in these fragments, using, for example, Exell program
can be integrated to find the total voltage and the voltage between the antenna and
the ground.
HFSS canand work in tandem with other programs: Mathlab, MathCAD,
AutoCAD.
13. Calculation of the temporal process
in a microwave integrated circuit
It is known that methodscircuit theory and calculation can be performed in the
frequency domain and the time domain. The HFSS-13 can perform a calculation in
the time domain in order to see a short pulse propagation through the device, in
addition to receiving the S -parameters.
A program for calculating the transient HFSS Transient - expects to dynamically
change the electromagnetic field using a discrete Galerkin (DGTD - Discontinuous
Galerkin Time Domain). This method uses a tetrahedral mesh partitioning, and the
fundamental finite element method, which made HFSS standard of accuracy in
electromagnetic simulation. You can now research by asking a short radio pulse,
penetrating the ground, electrostatic
discharges and electromagnetic discharges lightning. It is also possible to get a
picture of the field, which varies in time (TDR) at an arbitrary given time signal at
the input.
Consider the topology (Fig. 13.1) Alinks_BGA project, located in the Help
folder of the installation folder HFSS (and not in the Example folder). In the
simulation process in the project is calculated S-parameters in the frequency range,
and there is a short pulse propagation.
Geometry has four signal lines that are connected to the ports and the ground
bus. We perform modeling of geometry already created (Fig. 13.1). Add excitement
and to install solution before analysis and calculation of output data.
Fig. 13.1. structure of distributionlines
Click HFSS> Solution Type
and set the calculation method Transient
(Fig. 13.2). This method of calculating process in time domain.
Ris.13.2. Methods for solving the problem in HFSS
WorkaroundTransientIt hasNetwork Analysis option. If you choose Network
Analysis option when installing on a calculation appears Input tab that allows you
to, for example:
• Modeling of the excitation pulses in the form of such structures as a broadband
antenna, light fibers, electrostatic surge;
• field visualizationwhen exposed to a short pulse;
• analysis of time-dependent reflektomerov.
Verify boundary conditions on the installation of the finished geometry. The
analyzed structure is a fragment of the microwave chip. He covered the surface
air_box, which assign boundary conditions Radiate (Fig. 13.3). To the left in Fig.
13.3 shows the boundary conditions that can be set in the method of calculation of
the transition process.
Fig.13.3.Slozhnaya form surface Radiate
RadiationBoundary- the boundary of the radiation (Figure 13.3.). To
accurately model the volume form on the sides of which are set radiation limits must
be sufficiently large.
Another boundary condition in the Boundaries section, ReferencePlanes set on
the excavation fee ( "the base") to the ports. Since the signal and earth lines are
connected with these planes, each of the signal current must be a closed path.
We distinguish immediatelytwo planes (Fig. 13.4)
BONDWIRE_REFPLANE_1 and SOLDERBALL_REFPLANE_1 and they
we define conditionsPerfect E named Reference Plane.
Fig. 13.4.Target reference plane (counterpoise)
These planes will be used in describing the excitation sources.
Fig.13.3, you can see what the boundary conditions can be applied in HFSS
Transient mode. The boundaries are not available for use in this mode 151
This frequency-dependentborder (laminate impedance, the impedance of the display
screening impedance), which can not be implemented directly in the time domain.
The task structure of the excitation sources.
1. Select the first of the planes yet undescribed LINK_12_BW in the project tree
and set it on a digital port (Figure 13.5.) Command Assign Axitation -> Lumped
port.
Fig. 13.5.team task to create a discrete port project
2. dialog appears Fig. 13.6, in which the column "Use as reference", a check mark
to select conductor SOLDERBALLREFPLANE_1 as a reference plane for the
terminal.
151
Fig. 13.6.Definition of port options in Driven Terminal mode
In the same way, we create all4 ports from the reference conductor
SOLDERBALLREFEPLANE_1.
3. Open the Properties dialog for the properties of the port and its Transient tab (Fig.
13.7). Not every port must be "Active" (if you specify all ports active, it will take a
great time to get all the S-matrix).
Fig. 13.7. selection the port status in the mode of calculation of the transition
process
Ifyou want to reduce the calculation time, make only a few ports active. Passive
ports will work as terminals. Then you get only a partial S-matrix.
4. Open bookmark Post Processing, and leave the Do Not Renormalize option.
152
Fig. 13.8.Bookmark postprocessing data
If you chooseanother option, Renormalize All Terminals' Spectral Domain Date,
the data are normalized to the thermal data of spectral analysis.
5. Highlight the port area (in the Lumped Port) and right-click on this icon in the
Excitation port section and select "Auto-Assign Terminals". According to the
operation sequence of each port is automatically generated terminal (load).
Note that there mayand the other excitation method HFSS Transient. However,
some problems are not solved yet in this method v13: individual cell of the periodic
structure (phased array antennas, periodic chastotnoselektivnye surface) or a
magnetic pattern with an offset (ferrite circulator, ferrite phase shifters). These
examples are better calculated in the frequency domain using HFSS.
Installations for the calculation of the transition process.
first,
we define list of facets, which will remain calculated fields of the transition
process, as follows:
1 .Switch to the selection mode of facets F and move the cursor to the upper chamfer
for both signal and ground plane (Fig. 13.12), hold down the Ctrl key.
153
Fig. 13.12. Selecting several surfaces
2. Create a list of facets Modeler command> List> Create> Face List. In the List
list appears FaceList1. Change the name on PlotFields.
Then prepare to install solution for that:
1. Right-click Analysis> Add Solution Setup.
2. The tab General (Fig. 13.13), we define the maximum number of grid
compaction steps equal to 6.
Fig.13.13. The General tab
Frequency, on which the partition of the grid will be automatically selected on
the basis of time-dependent.
3. Select the tab Input Signal (Fig. 13.14).
154
4. Ask Broadband pulse from DC to 10 GHz, and request the calculation
S-parameters every10 MHz.
Note that the calculation in the time domain can be done by askingminimum and
maximum frequencies. If you specify a lower cutoff frequency equal to zero, the
waveform will change to include the frequency to DC.
Also note,that in addition to the sweep frequency can be determined TDR pulse.
This - pulse whose spectrum extends from DC to some highest frequency of which
depends on the rise time.
Finally, note that you specify here one time dependence for all active excitations.
In the project with more than one common transition process ( "non Network
Analysis"), you can set different time according to different excitations, and run the
simulation at the same time with all the excitations. The method of calculation of the
transition process, all active excitation have the same time dependence, you get a
simulation result when a single active drive.
Fig. 13.14. Input tab
5. Duration Select the tab (Fig. 13.15).
155
Fig. 13. 15. Target damping level transient
Modeling is executed until the transient fields are attenuated to a level Target
Residual (Fig. 13.15). Furthermore, the maximum time is determined by simulation
20x (model size) / (Speed of light).
In this transmission line 1 nsec duration signal can move several times from the
source to the load and back end. Therefore, it is reasonable to limit the time of
calculating the value At most = 1.25 ns (Fig. 13.16), which is equal to the
aforementioned 1 nsec plus input duration.
Fig. 13.16. Page parameter setting time process
Open bookmark Saved Fields (Fig. 13.17), and set the mode to save the field in
the planes, setting option PlotFields.
Keep field every 4 ps. During the 4 ps signal deepened by 0.6 mm in the insulator,
so that such frequency samples is to provide smooth animation of the model.
156
Fig. 13.17. Saving installation on the field during the transition process
In Radiated Fields tab (fig. 13.17) will choose the frequency at which field will
be stored. Modeling requires about 400 MB of RAM per excitation. If you set all
ports active and have the opportunity to perform a distributed simulation, select this
calculation. In a distributed simulation with at least eight processors and the RAM
enough, each of the eight excitation give its own process, and they will all be
addressed simultaneously. So:
1. Save the project and run the simulation.
HFSSwill first perform an adaptive calculation in the time domain.
After that,He will perform simulations of eight, one for each excitation.
2. Afterhow the calculation is running, you can right-click the Results> Create
Terminal Solution Data Report> Rectangular Plot, to follow in the
calculation process for the input and output signals to the various ports as
you complete the simulation. Depending derive by selecting the desired
characteristics in Fig dialogue. 13.17.
157
Fig. 13.17. The output characteristics in the time domain
SelectInput Input (LNK_12_bwire_T1), as well as signals Output
(TXCLK_12_bwire_T1, LNK_12_bwire_T1) and Output (TXCLK_12_sball_T1,
LNK_12_bwire_T1). Then create a graph in Fig. 13.18.
Fig. 13.18. Schedule time process
Fig. 13.19 on Y axis output parameter Residual (difference) in dB. This - the
value of any of the process, which is used as a stopping criterion. It is installed in
the Duration tab. By default, when
158
peak field decreases to 0.001 with respect to
maximum, Modeling deemed
to be fulfilled.
ControlResidualschedule duringcalculation provides information which may be
required during simulation.
Windowthe
state
provides additional information; thisbased on the maximum simulation time.
Fig. 13.19. Schedule of the transition process
The difference is not reduced to a value of -60 dB, so we stopped modeling before.
postprocessing matrix data
In the Report, in the Context section, change the option from Time Solution to
the Spectral. This enables a schedule S-parameters in the frequency domain obtained
by the simulation of the transient. Even when the simulation is still running, transient
solver already give S-parameters in the frequency domain on the basis of information
about the transition process, which is adapted to the present time. Schedule will be
updated very often. Modeling can be slow because each modification requires a
transformation from the time domain to the frequency.
159
Fig. 13.20. Output frequency dependences S parameters obtained by FFT in the
spectral region
Details of S-parameters outputted duringsimulation are shown in Fig. 13.21. Sparameters plotted on this graph - a transmission and reflection coefficient for a
particular signal line.
Fig. 13.21. The frequency response of the signal line
To display field propagating along the line:
1. In the model tree, select the list PlotFields planes.
2. In the Project Manager, right-click on the Field Overlays> Plot
Fields> E_t> Mag E_t.
161
3. Select
Done. At this command will be output in a plane tangential nye
components of the electric field.
The field can be animated in the plane of the resonator in the last moment of time,
ie, animate from 0.6 ns to 1.25 ns with a scale for adjusting the pitch. EXAMPLE
animation field shown in Fig. 13.23.
Fig. 13.22. Output fields and animation it
Fig. 13.23. Type running current in accordance with a time process
161
Canalso see the animation field distribution and resonance phenomena in the last
time. Animation from 0.6 ns to 1.25 ns can be seen with adjustments to field scale.
Fig. 13.24. Type traveling field along the structure
Graphics output in the time domain
Studyin the time domain is to calculate the response (reflection) in the time
domain (Time-Domain Reflectometry TDR) at excitation signal structure as the
shock exposure (transfer function). In this calculation you need to choose the method
of interpolation. After calculation in the Report dialog, you need to choose Time
from the Domain list. You must also specify the input signal, or a step or impulse.
Fig. 13.25. Selection signal type when specifying temporal process
at Time The choice of the area, you can choose from several Categories and
related Quantities, to make the schedule, such as | S11 |. When you compose the
graph in the time domain, each parameter in the frequency
162
the
at first
converted
in
temporary
region
before
than the calculation is performed according to the formula. For example,
if you want to display dependency
Seleven/ (1 - Seleven)
(eleven)
and output this function in the time domain,
transform IFFT:
it uses inverse Fourier
IFFT (S11 * input) / (1 - IFFT (Seleven* input))
(12)
It should be noted, that this expression does not match the words
IFFT (S11 * input)/ (1 - Seleven)* input).
(13)
These two expressions are not equivalent.
Ifyou choose Time Domain Impedance as a category Category, you can choose
the amount of TDRZ. It determines how
TDRZ (t)= Zref * (1 + IFFT (S11 * input)) / (1 - IFFT (Seleven* input))
(14)
where "input" represents an input signal (step or impulse), and "IFFT (.)"
denotes the inverse Fourier transform.
This equation is a ratio of the instantaneous voltage in the time domain v (t) to the
instantaneous value of the current i (t). Voltage and current are determined (in
frequency domain) in terms of incident and reflected waves a and b, and are
V  Z0(ab)
I
1
Z0
(ab) 
Z0(1Sii) a
1
(1Sii) a
Z0
(15)
(16)
Suppose that the incident waveis input in the form of steps, and thus, when we
take the inverse Fourier transform (IFFT) for V and I, we obtain v (t) and i (t) in the
time domain. Let's create this relation as a function of time, and then get TDRZ
function (t). Default Zo = 50 ohms.
To create a schedule in the time domain:
1. For the project with the existing frequency range, perform the following
steps 2 - 4 to save the properties.
163
2. In the dialog Report, in the Domain list, click Time.
This allows you to select the TDR Options button, and for the calculated load
data include Terminal TDR Impedance in the list of Category.
3. Click the Options button to the TDR. Appears TDR Options dialog.
4. Select signal input, Step (a jump) or Impulse (Dirac pulse).
Step size is a jump signal change, whereas there is a short Impulse excitation.
Impulse is very narrow rectangular pulse with a rise time of zero time and fall with
the magnitude of a time step of 1 and a height equal to 1 / (time step).
Selecting Step makes field Rise Time and Impulse inactive.
5. Ifyou have selected Step, enter the pulse rise time in boxing Rise Time.
Timerise should correspond to frequencies. With a bandwidth from DC to
Fmax, better time resolution that can be achieved is equal to 1 / (2Fmax). Rise time
1 / (2Fmax) is a very short rise time, which can be solved. However, the rise time of
0 s gives the same information, so 0 is the default value of this panel.
6. entertotal time on the chart in the text box Maximum Plot Time.
Defaultmaximum time TDR Options dialog box associated with the frequency
range Δf: it is equal to the time 1 / 2Δf, during which IFFT outputs information. This
length of time is often a very long relative to the time delay, which corresponds to
the length of the test on your device, so you may want to reduce this value.
Alternatively, you can edit the time-axis characteristics TDR after it was created.
7. Installnumber of time points for outputting a graph for a box Delta Time. By
default, it may be equal to the number of points in the frequency range.
Time difference based on the frequency bandwidth: a frequency band from DC
to Fmax minimum time resolution that can be obtained is equal to 1 / (2Fmax). IFFT
algorithm provides frequency interval between the values of 1 / (2Fmax), but you
can smoothly interpolate between
164
points, settingthe very best solution is, for example, 1 / (10Fmax), due to the
additional computation time.
8. Optional,TDR in box, change the type of box and width.
Windowfunction algorithm FFT Fast Fourier transform signal to obtain a nonzero value must be removed from Fmax. Each window function between the ability
to resolve signals of comparable frequency and the ability to resolve signals of
different strength and frequency.
List of window types include:
window function
Rectangular
use cases
Functionlow dynamic range giving good resolution for signals with
similar amplitude. Not suitable where the signals are very complex
amplitude. w (n) = 1.
Bartlett
Function high dynamic range, low-resolution, suitable for a wide
range of applications.
Function high dynamic range, low resolution, designed for a wide
range of applications.
upgraded
function
dynamic
range
established for Examples a narrow strip.
Function
modified
dynamic
range,
c a l c ul a t e d f or e xa m pl e with a narrow passband.
The choice of the Kaiser function also allows you to define the area of
related parameter of the Kaiser. The larger the parameter of the Kaiser,
the wider the window. It controls the tradeoff between the width of the
central lobeand the area of petals from the sides.
This approach is applied parabolic shape to a window with the
data in the frequency domain.
Blackman
Hamming
Hanning (Default)
Kaiser
Welch
9. You canuse the Save as Default to set the current values as a default, and the Use
Defaults button to use previously saved options. Note that when you select a feature,
first shows the values corresponding
selected
char ac t eri s ti cs.
10. Click OK.
To display characteristic Terminal TDR Impedance (which means either calculate
S-parameters for waveport1 port in the frequency range, instead of calculating the
delay time with respect to specific impedance), do the following:
a. In the Category, click Terminal TDR Impedance.
b. The Quantity list, click the value for the application of the schedule.
165
The default impedance (Zo) for the magnitude of TDRZ is 50 ohm, unless you
specify otherwise, when you set the renormalization for the impedance of the
terminals, which are created in the model. If you want to change the impedance
value, it can be edited in the Report dialogue (as shown below), or you can create a
variable Zo × (1 + Sii) / (1-Sii) selected Zo. Edit value Zo can be Report dialogue,
for which:
1. Under Category, select Options Terminal TDR Impedance, Port
and Function.
Fig. 13.26. The timing of the x-axis
2. Edit the value to be set in the Value field. In this example, the value for Zo varies
from the default value of 75 ohms when administered 'Zo = 75ohm' column in the
Y-field.
Fig. 13.27. Selecting functions for Y axis
c. Under Function, select the mag for the module output.
3. Click Done. Characteristic appears in the preview window.
IfS11 = 0 at DC, a step time interval goes to zero, and tends to move TDRZ Zref.
If S11 is different from zero in the DC, the step response the time interval is set to a
nonzero value, and TDRZ set at a value different from Zref. Pulse time interval will
always fall to zero, as can be seen from the characteristics of the derivative. Pulse
transition process TDRZ will always be normalized to Zref.
The chart below shows the difference between the short rise time of nonzero and
zero rise time to the transmission line segment 94 ohms. Note that characteristic with
zero rise time starts with the correct line impedance, while the impedance at the other
time points perenormalizuetsya. In addition, one
166
characteristic- shifted copy of another (Figure 15.25.). The reason for this is that the
voltage and current of a load connected to 50 ohm, so that the time interval is stored,
v = Zref * i. When the pulse increases, TDRZ characteristic begins to change from
a steady state, because there is reflected back to the source, the impedance of which
differs from the characteristic impedance of the transmission line.
Fig. 13.28. Transitional process TDRZ impedance changes
note the following points to keep in mind when using
TDR:spatial resolution
c
x
2B
(17)
where c - velocity of light in the medium, and B - the signal bandwidth. Since
TDR is usually based on the frequency range, which starts with a constant
current, the spatial resolution becomes
x
c
2F max
(18)
where Fmax - the highest frequency in the frequency range. For example, if Fmax
= 15 GHz, and the medium has a permittivity εr = 4, then the step will be equal
to the spatial (1.5E8 m / s) / (3E10 1 / s) = 5 mm.
The spatial resolution equal to c / (2Fmax) corresponds to time resolution:
t 1 /
(2Fmax)
(19)
167
Let N - number of points in IFFT. N equals the number of time samples, and it is
also equal to twice the number of frequency samples. The density of the time
samples in sweep frequency is:
where T - total time.
Increased frequency sample density leads to an increase in the total time T. In
practice, this often leads to large transient decay time. Therefore, TDR Options
dialog can be set at a maximum time value.
InterfaceTDR Options also allows selection minimal Δt, than given by equation
(3) above. When you select a smaller .DELTA.t, you increase the Fmax
"complement zero", that is, adding zero to S11 is the estimated bandwidth. In
practice, this leads to a smoother signal TDR.
HFSS allowsset the rise time of the input signal. The rise time must be equal to
at least 1 / (2Fmax). The input signal with a large rise time is less dense at higher
frequencies and will result in less "ringing" in the TDR response.
Hamming or Hanning filteralso reduce the high frequency content and results in a
more smooth characteristic TDR. With these filters can be chosen width. Width
100% often - a good choice.
To calculate the temporal process is necessary to obtain broadband SPICE
model. This is the model that reflect the behavior of electrodynamic structures, but
represent a model with nodes. It executes the program Full-Wave SPICE, which
allows you to perform a simulation of frequency-dependent models of SPICE,
which are generated in HSPICE format, PSpice or Spectre RF for accurate
simulation in the time domain. It provides broadband SPICE models at the touch of
a button. Program Full-Wave SPICE - is an additional module for HFSS and Ansoft
Designer. It calculates accurate broadband SPICE model. This capability allows
developers of electronic components take into account the effects of microwave.
Program Full-Wave SPICE enables the user to implement:



MethodTranslation of wave types to nodes Modes-to-nodes technology;
MethodFast sweep ALPS fast-sweep technology;
algorithmscreating a SPICE circuit models.
Calculationtemporal process often requires a lot of computer resources - memory
and time. Therefore it is recommended to use a distributed simulation in solving
these problems.
distributed simulationIt requires a license for the calculation tasks that require
large computational resources. This license is called the High Perfect Computing
168
(HPC). Distributed Simulation mode set in the Options tab of an Analysis, which is
located in the dialog Tools> Options> General Options.
169
Fig. 13.29. Task allocation calculation options
For example, in Figure dialog. 13.29 The user has access to a computer with
eight or more processors is caused by command \\ large_many_proc_machine. In the
dialog box Fig. 13.29 computer are listed eight times, according to the number of
processors.
Supposethat the computer has 16 processor. This case corresponds to the plant
shown in Fig. 13.30. This number can be defined in tab Solver dialogue HFSS
Options.
Fig. 13.30. Options HFSS: number of processors in a distributed mode
170
In this example, for those processes that can not be executed in a distributed
fashion, the user requests the use of all 16 processors. But such transient HFSS better
finds in multiprocessor mode. For distributed modeling, two processors in the
section Number of Processors (Fig. 13.30) can be chosen. Detailed implementation
techniques of distributed simulation will be discussed in Section 17.
14. Analysis of the horn antenna in the
time domain
In this example we will use the method of calculation of the transition process
to get the S-parameters of a horn antenna in a wide frequency band. The broadband
antenna may be used to transmit short pulses, for example as part of the radar system
of the earth subsurface location.
weWe will use existing examples in the library model containing the antenna
geometry, create the rest of the model and choose the method of calculation of the
transition process.
This section explains how to prepare the model and the analysis of the
transition process in the horn antenna in the time domain.
launch HFSS and discovery model.
1. Run HFSS 13.
2. teamFile> Opendownload filebroadbandhorn.hfssfrom the Help folder in the
installation directory HFSS 13. This arrangement differs from the arrangement in
the example folder. You'll add borders, excitement and set the surface on which are
displayed field.
171
Fig. 14.1. Viewhorn antenna
172
Fig. 14.2. dimensions horn antenna and port construction
3. Set menu HFSS> Solution Type, to select the type of solutions
Transient Network Analysis(Fig. 14.3).
Fig. 14.3.Dialog select the type of solutions
5. Set the properties of the materialmouthpiece as copper and objects pin. The
geometry also includes an air box, the sides of which are covered with layers of
PML.
6. Create an air box with initial vertex (X, Y, Z) = (- 100, -250, 180) and the size (dX, dY, dZ) = (550, 500, 360). The properties of this box are
shown in the dialog shown in Fig. 14.4.
172
Fig. 14.4.Parameters base box horn antenna
chamferthis box are arranged at a distance greater than 100 mm from a speaker.
Since we are going to use the PML, and the minimum frequency is 0.7 GHZ, it is
even more than you need.
Fill the air, boxing, give it a name, color, and permeability.
On the sides of the box we define the boundary conditions PML.
7. highlightAll of the air box bevel.
8. Select Boundaries> PML Setup Wizard. Wizard appears perfectly absorbing
surfaces (Fig. 14.5). Master itself corrects the material parameters to obtain a good
absorption in a layer with a predetermined thickness.
Fig. 14.5.Create perfectly matched layerThickness = thickness with 100 mm in
the assistant PML
9. Click Next.
173
10. Set the minimum frequencyand the minimum distance to the plane Minimum
Radiating Distance radiation, as shown in Fig. 16.6: 0.7 GHz and 100 mm.
These settings need to HFSS generated the correct parameters PML layer.
Fig. 14.6.Setting the frequency and distance of PML assistant
11. ClickNextand then Finish.
Setting planesto output the dynamically changing fields.You need
to decide in advance,where we want to see the field in complete three-dimensional
model at each time step change. Create two perpendicular deployed rectangle on
which we want to keep the field (Fig. 16.7,16.8).
1. Create rectangle in XY plane with the point (X, Y, Z) = (- 100, -250, 0)
the size (Xsize, Ysize) = (550, 500).
174
Fig. 14.7. The horizontal plane
intersectinghorn antenna
Fig. 14.8. Plane,which displays the
field
2. Create rectangle in XZ plane with the point (X, Y, Z) = (- 100, 0, -180)
and the size (Xsize, Zsize) = (550, 360).
3. Derive theserectangles in the form of frames. Now we need to define this list as
a "list" in order to later save the field for them.
4. Select both rectangles.
5. Askteam Modeler> List> Create> Object List (or press Face List if you
selected them in chamfers selection mode).
Fig. 14.9. selectionnamed "Objectlist1" on "PlotFields" in the
properties window
Properties
6. Specify the name of the list, for example,"PlotFields", as shown below, through
the model tree and the properties window.
7. Save the model Save.
175
sources excitation.
Plane"Source" is the wave port.
1. ClickTools> Options> HFSS Options,to bring HFSS Options dialog. Of
General tab, in the Assignment Options section, check the option Auto-assign
terminals on ports is not checked and click OK.
Fig. 14.10.The General tab dialog HFSS Options
2. In the project tree, select a plane named "source".
3 View> Fit Selection. Note
what this an object is an
the coaxial cable, externalwhose outer conductor is part of a circle.
section of
4.AskWave Portforthis plane.
176
Fig. 14.11. Selecting an object for the reference (ground) plane
5. GivePort p1 name and click Next.
6. AskActive port option. Click Next.
Fig. 14.12. wave port assignment active
7. Select "Do Not Renormalize" (Fig. 14.13). The impedance of the port is expected
to be exactly 50 ohm, and is considered as a valid impedance which will be correctly
matched to the impedance models.
Fig. 14.13. The last step is to install a waveguide port
8. Click Finish.
177
The draft with the method Transient Network Analysis with multiple ports, to
complete the S-matrix to provide a signal on one of the ports. This port is called
active. Each simulation with one active port provides one column of the S-matrix.
When you define a port as a passive, this means that the signal is not supplied to it,
and it is loaded by a matched load.
The project with the calculation of transient (non-network analysis), all active
ports will be simultaneously included (ON), while remaining passive ports will only
act as terminators (OFF).
Terminal Specification.since the program HFSS Transient uses
Terminal-Driven ports, we need to set the resistance of a port.
1. P1 Click the right mouse button in the project tree and select Auto Assign
Terminals.
Fig. 14.14. Selection option
automatically detecting terminal
Fig. 14.15. Specification conductors
related to the port, which will be a
counterweight
178
Fig. 14.16. The components of the wave port
2. Select Reference horn as the reference plane, so that the pin will be
terminal.
Calculation and output the results.
This section describes the
installation of the decision of the interim process and view the results.
1. In the project tree, select Analysis> Add Solution Setup.
2. Gridthe transition simulation is created on the basis of simulation in the
frequency domain. In this simulation, the program stops at an appropriate frequency
to perform an adaptive calculation. In this case, use mixed orders of the elements
and the iterative solver. Accept the defaults in the tab General (Fig. 14.17).
Fig.14.17. The settings in the General tab
3. The Input Signal tab (Fig. 14.18), set the bandwidth from 700 MHz to
1.6 GHz. This band corresponds to the modulated Gaussian pulse in the time
domain, which can be seen in the panel.
179
Fig. 14.18. Time process to the input port
4. The Duration tab, you must specify a time when the boundaries of the analysis
of the transition process. For this example, select Auto Terminate.
Fig. 14.19. Building of the accuracy of calculation of the temporal
process
ParameterTarget Residual(Difference until convergence) represents the
minimum field in the model at a given time relative to its largest value. As soon fall
below 0,001 field from the highest value, the simulation may be stopped.
Additionally, you can specify a maximum (At most) and minimum (At least) time
intervals in which to simulate the transition process. Embedded minimum and
maximum time intervals are suitable in most cases. They take into account the size
of the model and the type of signal.
5. The tab Saved Fields (Fig. 14.20), select the option Object List or Face.
181
Set interval that will be stored field value equal to 30 ps. It is - "sufficiently
small" part of the broadband pulse duration to obtain further smooth animation field.
Fig. 14.20.Filing Options tab field Save Fields
6. In the Radiated Fields set saving options and fields in the time and frequency
domains at a frequency of 1.2 GHz, as shown in Fig. 14.21.
181
Fig. 14.21. Select the frequency at which the radiated field is saved
7. Stepinstallation solution Solution Setup is executed. Click OK.
Modeling.
1. Start modeling.
The whole areaProgress solutions (see. Fig. 14.22) corresponds to the
maximum simulation time in the Duration dialog tab Solution Setup, or if it is not
specified, the value of 20 * (diagonal pattern) / (speed of light).
FIELD red arrow indicatessimulation time relative to the maximum time during
which the simulation is performed.
182
Fig. 14.22. Observation of the calculation process, showing the
simulation relative to the total calculation time
In the simulation, you can see the progress of the calculation in a different
way.
2.pressRight-click on Results in the project tree and select Create Terminal
Solution Data Report> Rectangular Plot.
Fig. 14.23. Selection of characteristics in the time domain
Output Input and Output on the same graph, and a Residual dB20 on another.
183
rice14.24. Voltage Input and Output (excitation and reflection) as a function of
time
InputOutput and show excitement and reflection on the port as a function of
time.
Characteristics Residual (ris.14.25) indicates a measure for the level max field
in the model. Simulation is complete when Residual spada- is below 0.001 with
respect to its peak at any time.
Fig. 14.25. ParameterResidual logarithmic scale as a function of time
184
Visualization structure characteristics.
1. To display the characteristics of the S-parameters as a function of frequency,
right-click again on the Results of the project tree and select Create Terminal
Solution Data Report> Rectangular Plot.
2. changechoice in the Solution window on Spectral as shown in Fig. 14.26.
Fig. 14.26. Preparation concluded parameters S-parameters as a function of
frequency, i.e. a spectrum time process
3. Apply module| S11 |. The resulting graph is shown in Fig. 14.27. This graph
changes when the calculation is still ongoing. This allows you to see this chart in
the early steps of calculation.
Fig.14.27. The frequency characteristic | S11 | broadband horn antenna
Traveling field of the horn antenna.
185
To create a traveling field, select the list called "plotfields" in the model tree, ie,
a list containing the two rectangles that you have created to display the field.
In the project tree, right click Field Overlays> Plot Fields> E_t> Mag_E_t and
select an arbitrary non-zero time on the next panel (Fig. 14.28). You can always
adjust the time by changing his choice of Modify.
Fig.14.28. Creating a traveling field at a given time t = 3.044 ps
Now thereopportunity to observe traveling field. Moreover, unlike the method
FEM, here we see the field, which varies in time according to the time process the
input signal (Fig. 14.29).
Fig. 14.29. Distribution of the field in two perpendicular planes
186
3.changethe scale field to make a visible field in a range of 0 to 10 V / m.
4. Animation graphics wave propagation can be done by right-clicking on the name
of the graph and selecting Animate. The Setup Animation window, you can select
the number of steps. If necessary, HFSS perform interpolation between the stored
field solutions.
Fig. 14.30. animation field setting window
visualization of the field radiation.
To calculate the field in the time domain:
1. Right-click on the Radiation and choose Insert Far Field Setup> Infinite Sphere.
Fig. 14.31. Execution of the far-field installations
2. In the project tree, right-click on the Radiation and choose Insert Far-Field
Setup> Infinite Sphere. Set the angle range, as shown in Fig. 14.32. This range in
the XZ plane, and range from zenith to horizon in the direction of the direct radiation
of the antenna.
187
Fig.
14.32. Specifying only one point in the field of far field
3. Now
at tree
project,
click
right
Results> CreateFar- Fields Report> Rectangular Plot.
button
mouse on
4. Select rEz characteristic as a function of time.
Fig. 14.33. The output characteristics of the field intensity in the far field
5.
188
Fig. 14.34. Bookmark families output characteristics
clarification: The component Ez isdominant. Therefore, it is selected in Fig.
14.33. r product of Ez useful because it does not depend on the distance from the
source to the observation point in the far field. RE parameter has the dimensions
[B], since the field has a dimension E [V / m], and the distance r [m].
OnFig. 14.35 shows the time dependence of the shape of the far field. This
graph shows how the distorted signal emitted by the antenna. This information can
be used to improve the antenna design, such as placing a reasonably resistive ribbons
in some of its points. It - also useful information at the stage of postprocessing data
calculation.
Fig. 14.35. Changefar-field radiation, co-polarized, in the center of the main
beam.
IfYou need, you can also create a frequency-dependent far field when changing
solutions to the Spectral and change the angle Theta as the primary variable, as
shown in Fig dialogue. 14.36.
189
Fig. 14.36. The output characteristics of the far field in the spectral region
The Families tab (Fig. 14.36), we can see that the previously set frequency 1.2
GHz, in Radiated-Fields bookmarking solution Solution Setup, and thus, you can
select Spectral.
15. Design of nanoscale svetodiodiodnyh modules
oneof the actual problems arising at the interface between the optical and
microwave - ranges, is to provide an effective light-emission of electromagnetic
waves in media with complex frequency dependence of their properties. These
media, in which the analysis of electromagnetic waves is rather complicated
problem can be attributed electrodeless discharge plasma metamaterials with
negative permittivity and permeability, as well as semiconductor heterostructure.
This section covers the possibility of increasing the efficiency of the LED modules
on the basis of resonator quantum systems with sources of electromagnetic radiation
[8].
The development of microwave technologywith increasing frequency are always
constrained by the technological possibilities. Therefore, for example, in the
submillimeter wavelength range could not be used as an effective resonator and
slowing down the system as in the centimeter range.
butcurrently in connection with the successes in the development of
nanotechnology, such opportunities have emerged, and now you can use the rich
experience in the development of microwave systems in devices of optical range,
including lasers and LEDs. Furthermore, the results of simulation of streams of
electrons and photons allow modeling active and passive elements in a single
process. Thus as sources of electromagnetic radiation can be set point sources in the
form of short current segments.
Developers should consider HFSSsuch opportunities. The library materials
can include nanoscale isotropic and anisotropic materials. However it required to
check the accuracy of work programs of the calculation of structures with extremely
190
small dimensions which exist vozdeyst- electromagnetic light wave and microwave
wavelengths. Especially important is the fact that it is necessary to correctly describe
the sources of light waves. In the electromagnetic simulation programs as radiation
sources can use voltage or current source with governmental Properly speaking
dimensions much smaller than the wavelength.
Consider light emitting diodes (LEDs)a cavity resonator (ER), which are
considered promising [8], as have significant advantages of the compared with
conventional LEDs. For example, the intensity of the spontaneous emission of the
LED using a high-Q resonator increases the order of magnitude due to the narrowing
of the luminescence spectrum. Besides,
improving spectral purity and enhanced orientation and thermal stability of the
radiation, as stated in [8].
191
15.1. Modeling
structure
and
sourcesin the optical wavelength range
excitation
The operating principle of LEDs based on the use of semiconductor
heterostructures, which are composed of layers of n- and p-conductivity Stu. Due to
different concentration of charge carriers in the individual layers and the presence
of this complex structure of the areas in which there is a rapid accumulation of
charges, there are conditions for the transition of charge carriers from one energy
level to another, and there are photon emission, providing illumination in a certain
wavelength range. These plots concentration of charge carriers are called quantum
wells, threads and dots depending on how they are localized photons [9- 11]. Simple
models of these quantum portions in accordance with [10] are shown in Figure 15.1.
a)
b)
at)
Figure 15. 1. The active regions of heterostructures in which the light sources
are presented in the form of quantum wells (a) and filaments (b), and points
(a)
In [8] are real pictures and LED structure with the PR, which are used in fiber
optic communication systems dia- pazona infrared wavelengths. As most short-LED
emitting at a wavelength of 650 nm is represented by the structure depicted here
ris.15.2 as well.
a)
b)
Ris.15.2. Structure (a)spectra and (b) the LEDs based on GaInP / AlInGaP
with Bragg mirrors and MQW of layers of AlAs / AlGaAs.
192
The resonator formed multilayer Bragg mirrors. The region of the active LED
comprises a plurality of quantum wells (MQW). The emission spectra of the LEDs
are illustrated in ris.15.2 b, where for comparison the intensity and Shih Rine shows
also a conventional CD spectra at different currents.
From ris.15.2 should be used, that the introduction ofOR provides an increase
in the intensity of luminescence in a given direction. Wherein the lighting govo ryat
the possibility of obtaining a highly concentrated light intensity curve, and
electrodynamics terminology used in this case the concept of narrow directivity
pattern.
atdesigning LEDs with ER in [8] is recommended to select the smallest
resonators are long excited at a fundamental oscillations (VC) and that have the
highest Q-factor of its own, i.e., minimum absorption (loss) in the cavity. However,
the actual CD with the PRs, including distributed Bragg mirrors and running on the
VC, not quite meet these recommendations.
To solve the problems of optical range using electrodynamic simulation program
and simulation of discrete point sources of light waves, as well as distributed sources
can be done, you can use the current segments along the transmission line.
Absorption losses in the metal and can be modeled as it is done in the microwave
range, i.e. a skin effect, a metal pattern with negative permittivity, using the surface
impedance (resistance to transverse dimension of the square). So, after the
formalization of the problem and solving the problem of modeling the structure, it is
possible to consider the problem of designing and optimization of passive structures
of the LED modules.
15.2. Falling plane wave to a frequency-selective surface
An important task in practical light range, calculation can be regarded as a
light wave falling on the surface with a complex structure. Surface with an
arbitrary structure can be attributed to the special case of tea the frequencyselective surface. Therefore, we pose the problem to calculate the
electrodynamic characteristics of the surface to various shapes, materials and
the losses inherent in the surface.
If we assume that the spectrum of the light wave emitted heterostructure
swarm, has the form shown in Fig. 15.2, b, knowing the coefficient of reflexion
of the frequency-selective surface reflection spectrum can be obtained
193
zhennoywave. In order to solve the problem of electromagnetic wave is
incident on a surface, it is necessary to model the structure of infinite size and
apply periodic boundary conditions, which are implemented in the HFSS
software. Using periodic boundary conditions greatly reduces the amount of
tasks and allows on a modern computer with RAM 4 GB to solve the problem
with the necessary accuracy.
Consider an infinitemulti-pin structure consisting of a cylindrical rod shown
in Fig. 15.3 and graphic HFSS tools. For the simulation of infinite structures
need to use special boundary conditions, the so-called associated periodic
boundary conditions. Calculations showed that the wave incident on the pins
(Fig. 15.4), stimulates the in-phase field Me- forward pins.
HFSS program allows to solve the problem of incidence of the
electromagnetic wave on an endless structure with an arbitrary shape and a
loss into account. Above and below the structure covered by periodic boundary
conditions, we choose a point at which the calculated near field plane before
and behind the plane (i.e., the reflected and transmitted waves). Of interest to
determine the frequency characteristics of the selective surface with different
structures, different surface roughness as pins, pictures, shapes, sizes
structure, its periodic or random repetition character. Frequency
characteristic can be determined by performing simulation of incidence of a
plane wave on the surface, as well as exciting the structure by means of point
light sources.
One embodiment of a quarter-wave devices PR and RT in the form of quantum
discs (CD) is shown in Fig. 15.3, where a perspective and projection of the whip
shows multicavity system (MSHS) with semiconducting heterostructure and
phosphor.
On ris.15.3 b shows the electric powerMSHS inlet line and CD gallium nitridebased blue luminescence instigating lyu- yellow phosphors were used. The length l
of pins corresponds to a quarter of the wavelength of blue color, i.e., about 115 nm.
The spatial period (pitch) frequency-selective structure is 3-4 times less than the
length l. Point sources in the form of a CD By arranging lozheny staggered through
one pin in the common-mode electric fields push-mode oscillations (π-type). This
VC, as is known in the art of microwave, is the most stable.
CD Dimensions (Fig. 15.3 in) taken[11], where it was noted that the structure
of CT GaN AlN in a matrix formed on a sapphire substrate by molecular beam
epitaxy.
194
a)
b)
at)
Fig.15.3. LED module based on quarter-wave multi-pin RS (a, b) with
quantum dots in a quantum CD discs (c).
On ris.15.3 bdiagram also shows the electric field in and near MSHS
quantum disks. The electromagnetic field from penetrating into the semi-MSHS
Vodnikova structure decreases exponentially and it is important to select the
dimensions of the pins with intervals therebetween so as to obtain the maximum
electric field in the quantum nai- disks to provide more effective interaction with
the pin systems heterostructure.
Sequence solve the problem in HFSS system
can summarized as follows.
teamDraw-> Cilindercreate a metal cylinder with a radius of 50 nm and 200
nm in height. Then we use the command Edit-> Duplicate Copy for copying and
reproduction coordinate pins X and Y-coordinate on the pitch distance (Fig. 15.4).
Fig. 15.4. Radiusand the height of the pins - Parameterisable variables
variables with initial values of 50 nm and 200 nm
195
We define a planewave incident on the structure of the command Source
-> Plane Wave.on ris.15.5- 15.7 shows the fields falling fragments, razhonnoy The
relative and total waves.
askboundary conditions surrounding this piece box. In order to set this piece
as an infinitely repeated, must be set on opposite sides of the periodic boundary
Master-Slave conditions. On the lower plane Ground ask Perfect E, and on top of
the plane - the ideal condition for absorption Radiate.
Define the calculation frequency, e.g., 400 THz.
Ris.15.5. Leaning
component (Incident)
Ris.15.6. The reflected
component (Scattering)
Fig. 15.7.The total
field (Total)
Onthis frequency, after the calculation, it is possible to deduce the incident
reflected component of the total field and a sectional frequency-selective surface
(Fig. 15.5-15.7).
Now add command analysis frequency bandAdd Sweep.In the resulting
dialog, this command Fig. 15.8 will make optical THz frequency range 100 - 1000
THz to 20 THz step.
Fig. 15.8.Setting the frequency range.
Run the command calculation HFSS -> Analyzes All. As a characteristic of
the reflection wave from the frequency-selective surface
196
can bringvalue RCS (Radar Cross Section), which depends on both the frequency
and the angle of incidence of a plane wave.
Another way to determine the frequency of the surface properties is to set the
discrete radiation sources in staggered manner, which will simulate quantum dots.
On ris.15.9 and 15.10 shows the calculated electric field vectors and sources of
electromagnetic radiation.
Fig. 15.9. FieldE sectional pins modeled surface
Fig. 15.10. Detail of the surface with the pins, on which are deposited
discrete sources (quantum dots)
ForIn order to control the excitation of the quantum dots, defined as discrete
sources of electromagnetic radiation (. Fig 15.10), we define command HFSS ->
Fields -> Edit Source Modules and set equal to the excitation source 1, and missing
springs - equal
0. Excitation structure using discrete sources enabling a graph parameter | S11 |,
which characterizes the frequency range reflection coefficient of the surface (Fig
15.11.).
197
Fig. 15.11. The frequency characteristic of the frequency-selective reflection
from the rough surface in the optical range
frequencyet al. characteristics depend on the shape of pins. If the pins have
square cross section, then it is concentrated on the edges of the electric field,
which leads to increased losses and the corresponding change in the resonance
characteristics.
15.3. Characteristics
light
at multiresonator structures
radiation
Also interdigitated (open almost all sides) nanoscale structures of interest
resonator systems of different shape, open at one end only, i.e. quarter-OP. Closed
OR have a higher Q factor and therefore provides a more effective radiation.
On ris.15.12 shows a multicavity system with quarter cylindrical resonators
(CHTSR) excited at lower H111 kinds oscillations. Resonators are cylindrical
recesses in the copper plate.
SystemIt consists of a chain of coupled resonators (CSR). Quantum dots (QDs)
are disposed in the cavity through one-phase electromagnetic fields.
198
Fig. 15.12.Resonant-cavity surface Related
Distances between CSR should be considerably greater than the distances
between the resonators to avoid communication between the rows and to
provide excitation VC H111 with orientation along the electric power line CSR.
We choose size in nm
frequencies in THz. Draw a base portion
fragment surface that will create a roughened surface (see Fig. Ris.15.13),
Draw-> Box command.
Fig. 15.13. Detail of the surface with a chain of coupled resonators
Furtherdraw a cylinder height h = 100nm and subtract from the parallelepiped
constituting the fragment surface.
excite
this
surface
(Ris.15.14) or discrete sources.
can
falling
flat
wave
199
Fig. 15.14. Field in the cross section of the resonance holes when falling on the
flat surface of the wave
To solve the problem, the fragment will cover the surface of the periodic
boundary conditions (Fig. 15.15), since, as was done in chapters 2-4.
Fig. 15.15. Detail of the reflective structure with cylindrical cavities open (infinite
structure connecting the periodic boundary conditions of a separate fragment
consisting of four resonators)
notice, thatthe results obtained in HFSS, confirmed that the system with
cylindrical resonators has a higher Q than the interdigital structure. Using a
cylindrical coordinate system is
201
thoroughly discussed in the following sections. Here when considering fields near
rough surfaces back to the multi-pin system.
To display the near field along the line, you must first draw the line you want to
add Insert Near Field Setup command
in the project tree in the Radiation section. This line is shown in ris.15.16 for
interdigital structures. Also shown are near-field vectors in the upper sections of the
resonant rough surfaces.
Fig. 15.16. Fields sectional resonators near rough surfaces
To perform a parametric analysis of the rough surface with round pins, ask team
Optimetric. As parametrically variable height pin select variables from 60 nm to 220
nm. On ris.15.17 shows the resonance characteristics obtained for different heights
of cylindrical pins.
Fig. 15.17. The frequency characteristics of the structure (reflection type), for
different heights of the cylindrical pins
Based on experimental calculations confirmed that the maximum electric
field strength is achieved in the upper section of the pins
201
andMS.The diagram of the electric field and the electric vectors are shown in
ris.15.18.
a) RS, consisting of round
bars.
Fig.
15.18.
(a)
b) MS, consisting of rectangular bars.
multi-pin
structure
and rectangular (b) bars.
from
round
For the structure shown in Fig. 15.18 and round pins, we present results of
calculation of the operating bandwidth and quality factors, which are obtained
in HFSS for materials with different conductivities, and roughness.
Fig. 15.19. Conversation job boundary condition to the finite conductivity of the
material, which can be entered in the form of additional parameters roughness
(Surface Roughness) surface material.
Copper
Copper
om
Molybdenum
fr
Tungsten
= 5.8 ∙ 107S / m
rough surface
= 1.76 ∙ 107S / m
= 1.85 ∙ 107S / m
100 THz
200
250
220
Q = 10
5
4
4.5
It should be noted that the valuesQ-factors listed in the table are too small.
Therefore, instead of the more open structure multi-pin system consider other
resonator with a high Q factor. Here we pay attention to the possibility of using
inferior and VC ensure dimensioned resonators to VC fundamental resonance
spectrum corresponds to the maximum emission of the active medium, in which as
in the present case it is assumed the use of quantum wires and dots.
15.4. Modeling stabilizing resonance structures with
resonators
The technique of stabilizing the microwave resonators are often used, which
affect the open cavity so that the self-resonant frequency becomes more stable by
increasing the unloaded Q of the resonator entire system [8]. This phenomenon can
be used to create a complex surface, operating in the optical wavelength range.
As a first example, consider the use of a stabilizing resonator of the resonator
system consisting of two quarter-cylindrical cavities associated with prismatic
stabilizing a half-wave resonator. This system is a triad of blue, green and red CT
shown in ris.15.20. The cylindrical resonators are excited in antiphase oscillations
lower N111 types that are supported and stabilized views N103 oscillations in a
stabilizing cavity created on the basis of a rectangular waveguide with the wavetype lower H10.
Fig. 15.20. The system in the form of a triad consisting of a quartercylindrical cavities with prismatic stabilizing resonators.
This system uses the coupling slots in the magnetic field, disposed at a distance
of half a waveguide wavelength in a prismatic cavity.
The active element is supposed to use the quantum dot (QD) as a nanoscale
diameter disc t, disposed between the quarter-cylindrical cavities in high electric
field (Fig. 15.21).
The dimensions of the stabilizing cavity for the blue, green and red wavelengths
(λ1 = 460 nm, λ2 = 525 nm and λ3 = 635 nm) are defined by where b <(usually).
For example, at 250 nm obtain d1 = 304 nm, d2 = 362 nm and d3 = 482 nm.
Then, based on the expressions and,
We find a1 = 302 nm, 341 nm = a2 and a3 = 396 nm; then choose b1 = 150 nm
b2= 170 nm and b3 = 200 nm.
Length stabilizing resonators with N103 based on the VC obtain equal expression
L1 = 1062 nm, L2 = 1235 nm and 1596 nm L3 =.
Analysis emitters in HFSS system also makes it possible to obtain
pulsingpicture of electromagnetic radiation in the dynamic mode. The results of
analysis of radiation from a single quantum dot OR discussed in HFSS system
illustrated in ris.15.20, which shows the field at a fixed time.
Ris.15.21. Radiation from CT picture at a fixed time
A ris.15.22 recorded on radiation in a time when the communication in the
slots maximum electric field, and near minimum CT due to the distance of a
quarter wavelength.
a)
b)
Ris.15.22. Radiationof CT (a) at the time point when cracks due maximum
electric field (b)
It should be noted thatof a known technique of microwave high-Q resonators
stabilizing than prismatic N10r on VC. These include, for example, are cylindrical
resonators based on VC N01r, which will be discussed in the next section.
15.5. quantum wiresand the point of the resonators
15.5.1. resonator systemwith quantum wires on the lamellae
Consider the example of a high-most CP based on the unique features of waves
of type H01 in a circular waveguide, wherein propagation losses in the waveguide
walls are extremely small. Of microwave technology it is also known that, as used
in coaxial magnetrons CP on the types N011 vibrations give rise to its own Qmultiresonator system is almost an order of magnitude [11,12]. In this connection,
it is proposed to use the ideas implemented in obraschonno- coaxial and face
stabilized magnetrons (OCM and TCM), to create nanoscale emitters.
One of these options applies to the emitter of the circular action, which can be
represented as an analog of the RCM. Thus, in the emitter circuit ris.15.23 shows a
system converts the lamellar type, π-excited on oscillations by an azimuthally
symmetric type N011 in the inner cylindrical cavity via mc disposed through one
resonator.
Fig. 15.23. circular action emitter (analog OKM)
Onthe ends of the lamellae (L) is also disposed through one A quantum wires
(KH) is no longer than the width L. A width dimensions and length is less than a
quarter of the wavelength radiated oscillations and CP length with blue, green and
red in this case CN can be taken approximately equal to 290, 370 and 620 nm with
identical diameters resonators 900 nm.
Another embodiment provides SDM axial radiation in the form of CP as an
analog oscillation H011 SCI (Fig. 15.24).
Ris.15.24.Emitter quantum asterisk (SCI analogue)
Unlike the circular radiator steps A and mc are located on the end surface CP
views H011 oscillations. At the ends of L (the even or odd) of the same color are
arranged CN. And such radial KN on the face can be perceived as a "quantum of
stars." [18,19] .Rassmatrivaemaya design is convenient for the application of the
phosphor coating consisting of a phosphor grains (RL), whose dimensions are
comparable with the width L. Yellow phosphors in combination with blue KH
provide white light. In this case, it is advisable to use a reflector to produce the
desired light distribution of the radiator.
radiators,presented at ris.15.23 and 15.24, were analyzed in HFSS system. The
calculated resonance characteristics support the possibility of increasing the
unloaded Q when administered Wed.
Fig.15.25 and 15.26 are illustrations produced in the system
HFSS when analyzing the second resonator quantum asterisks on the lamellae.
a)
b)
Fig. 15.25. Resonator system with end fins in perspective (a) and plan view (b).
Ris.15.26. RadiationQuantum of stars at a fixed time.
9. Consideration of temperature conditions in HFSS
accountingthe effect of temperature on the parameters of microwave systems
is a complex task that allows different formulations. The first formulation is
registered depending on the ambient temperature, which sets the temperature of the
microwave device and affects the parameters of the materials forming it, as well as
the dimensions of its elements. In this formulation, the problem is linear, as
feedback, that is, the effect of microwave power to the temperature of an object is
not considered.
In the second formulation to be calculated taking into account the temperature
of the object of heating due to impact of a sufficiently strong RF signal. This
approach involves the calculation parameter SAR (Specific Absorbtion Ratio). In
calculating this characteristic is first determined by the loss in the material, then the
relative magnitude of power is absorbed therein. Consequently, the temperature will
depend on the capacity of the generator.
Finally, the thirdformulation requires taking into account not only the
temperature changes due to exposure to the object of the microwave signal, but also
change its settings caused the microwave heating.
AT this section we consider
some
capabilities
HFSS,
related with temperature control and its changes.
Description depending on materials
from temperature.
Consider
opportunit y accounting dependency on temperature parameters of materials.
Each material,included in the project, it can be described by their dependency
on temperature characteristics. This relationship can be applied to any material
properties:. Permittivity, conductivity, etc. Dependency may be given in a tabular
form, or may be given as formulas.
It is also necessary to set the temperatureeach object. HFSS calculates the
temperature on the characteristics of materials. Using these values, calculated RF
system performance. Such heating temperature of the accounting objects in HFSS.
teamEdit-> View Edit Materialcall the dialog in Fig.9.1, in which:
1. noteThermal Modifier mode, see View / Edit Modifier for.
Fig. 9.1. Turning to the properties of the temperature characteristics of the
description
materials
In this case, a table of material properties Properties of the Material will be
expanded to include a column Thermal Modifier.
2. Select Edit ... from the pop-up menu, or None. dialog appears
Edit Thermal Modifier.
3. Select the Expression button to display the text field Parameters Modifier or
Quadratic button to show the table for Basic Coefficient Set and Advanced
Coefficient Set.
In the case of describing the dependence of expression Expression, we can write
an equation for the temperature dependence in the field Parameters Modifier.
markoption Use temperature dependent data set in the Modifier text field. Now it
is possible on the temperature dependence of the characteristics entered in tabular
form set Add / Import Dataset data.
IfQuadratic option is selected, a Basic Coefficient tab, you can edit a field and to
TempRef units and golf C1 and C2 to the following equation:
P (Temp)= Pref [1+ C1 (Temp - TempRef) + C2 (Temp - TempRef) ^ 2]
where Pref is determined by the relative permittivity
(eg). In this formula P-parameter.
• Ifselected Quadratic, the tab Advanced Coefficient Set (Fig. 9.2) can be set lower
and upper limit temperature (TL and TU, respectively).
Fig. 9.2. Selection of the upper and lower temperature boundary changes
Default,They are calculated automatically. Removing Auto Calculate TML
option, TMU¸ can introduce new bottom value (TML) and the upper temperature
boundary changes (TMU).
Setting the temperature of objects. To set the temperature of the object:
1. Use the HFSS and HFSS-IE> Set Object Temperature,
to display the Temperature of Objects dialog box (Fig. 9.3).
This dialog is reduced to objects of the project table. The first column lists the
object name, then the material, then switch-on the temperature dependence of the
properties option, and then the temperature unit.
Fig. 9.3. Dialog inclusion of temperature dependences of individual objects in
the thermal analysis of structure
2. To edit a property, select the option Include Temperature Dependence. This
makes it possible to select objects and set them on the temperature. The column
headings for the Object Name and Material columns include arrows for sorting
direction. You can invert the sort direction by clicking on the header of each column.
Ifthe list is large, you can use the field Select by name. Enter the name of the object
and press Select. The selected objects are displayed. You can make several choices.
3. To set the temperature to the selected object or objects, enter a variable or a
specific amount, or select an existing name in the text field.
4. Select the unit of the column Unit.
5. Click Set, to apply this value to the selected object, or click Set Default to set this
value by default. If you press Set Default, the column for the selected objects will
show the magnitude of temperature.
6. To edit a material object, when the dialog is closed, you can click on a column
material, and see a list of the material in the pop up menu, and click Edit ....
Byteam HFSS-> Set Object Temperature can be set to the temperature of an
object.
after the calculationtemperature distribution, it can be derived as a field team
Fields-> Plot Fields-> Other ... -> Temperature (Figure 9.4.).
Fig. 9.4. The output of the temperature field distribution in a horn antenna
body
The dialogue shows the temperatureFig. 9.5. It should be noted however, that a
fully-fledged calculation was based on the temperature dependence of linear
expansion of materials, etc. performed in ANSYS complex.
Fig. 9.5.Dialog display the temperature distribution maps
121
The ANSYS interface project tree are options and operations that are designed
to communicate the temperature characteristics of the object (Fig. 9.6).
Fig.9.6. The project tree in the program ANSYS Mechanical
Opportunity feedback temperature
Mechanicaland HFSS.
between ANSYS
Thisoption appears in the Advanced tab of the Solution Setup dialog. Open
this tab if you want to use a bidirectional thermal connection between HFSS
and Maxwell and ANSYS interface. You must make this choice before starting
on the solution (Fig. 9.7). Then you will have the opportunity to bidirectional
transmission of the temperature distribution data of the ANSYS Workbench
interface to HFSS and Maxwell.
121
Fig. 9.7. Optional inclusion of temperature parameters
Beforehow to solve the set on HFSS and Maxwell, for this type of object, you have
to put the initial value for the temperature dependence of the object (Fig. 9.8).
Fig. 9.8. Setting the initial temperature for each object, taking into
account the temperature dependence
Ifyou mark the option Enable Thermal feedback from ANSYS Mechanical, will
create a subdirectory with the extension .THM in the Solution folder included in the
project when the import will be carried out through the data interface of HFSS and
Maxwell. In this directory, the file is written a new centroid.xml (Fig. 9.9) to be
used in the future withdrawal
122
the temperature distribution. ANSYS Workbench 12 program consistently
exporting temperature information in a file called mechanical.ths in the same Project
directory itself.
Fig. 9.9. The location of the files with the temperature data
In the interface of ANSYS Workbench has 12 Export Results command.
Calculation of temperature distributionusing the connection with the
program ANSYS Thermal.
after the decisionProject on HFSS 13 or Maxwell, with the installation of "Enable
Thermal Feedback from ANSYS Mechanical" option in the Advanced tab in the
setup analysis, and after the related thermal analysis in ANSYS Workbench 12, you
can get the temperature distribution of thermal solutions. ANSYS Workbench 12
writes the feedback files directly in the Project Solution HFSS directory, or Maxwell
C: \ Ansoft \ HFSS13.0 \ Help \ hfss. chm :: / savinganewproject. htm> directory.
In this case:
1. Open the project in HFSS and Maxwell.
2. Start the analysis of the project, automatically using feedback from ANSYS
Workbench 12.
NotWe need to make any additional changes to the settings in HFSS solutions or
Maxwell. Just run the analysis Analyze command. HFSS or Maxwel know that the
decision need to use information about the changes in temperature. HFSS
or
Maxwell will give altered the results based on the new temperature distribution,
which can be imported through ANSYS Workbench 12 created earlier ANSYS
Termal Link. To do this, follow these steps:
1. Close the project HFSS and Maxwell.
123
2. In ANSYS Workbench interface 12, command clear
Cleanpreviously imported data.
3. Import the new results from the HFSS and Maxwell.
4. Run thermal simulations again (Fig. 9.10).
Fig. 9.10. Calculationtemperature characteristics via Re-solving After ANSYS
Thermal Link Feedback
The Edit Thermal Modifier tab can be set lower and upper bounds, which will be
true quadratic dependence of the temperature change defined above.
Afterthe end of the calculation, you need to select the object (Figure 9.11.), in
which we want to lead the field and temperature values, and then ask the team
HFSS -> Set Object Themperature.
124
Fig. 9.11. The output of the temperature field in the plane of the analyzed objects
In the Field there is a temperature profile at the site (Fig. 9.12).
Fig. 9.12. icon in the project tree, which shows the preservation temperature
distribution
Thus, Ansoft systemcomprising program electrodynamic analysis and analysis
program mechanical structures in the temperature mode, will execute an electroanalysis in the conditions of geometry changes with temperature. To solve these
problems need a special license.
125
10. Implementation of the adjustment mode in HFSS
In earlier versions of HFSS had several options to optimize the structure to be
analyzed. Frequently used busting operation when changing one or two parameters,
while the others remain unchanged. To solve this problem HFSS program for each
of the searched parameter completely solve electrodynamic problem. In the version
of HFSS-13 have an opportunity to solve this problem quickly and efficiently Tune
operation, so that after the decision can be changed by hand engine to the setting,
and immediately see how changing the characteristics of the analyzed device.
In order to realize the Tune operation, you first need to have variables in the
project. Then Solution Setup dialog tab appears Derivative, in which it should be
noted that the program should find derivatives, ie, increment of the function
depending on the variation of the argument. After the calculation of such an option,
derivatives used by Tune Report team for instant display and study of small design
variations, without having to re solutions. "Nominal" solution, plus derivatives
contain all the information needed to learn how to change characteristics with small
changes of design parameters.
Consider the example of F-shaped antenna Pace (Fig. 10.1). It consists of broken
out in the form of letter F vibrator which is shorted to ground the board at the right
end and has an open line to the other.
Fig. 10.1. Side viewF on the planar shaped antenna to adapt to the position of
the port
126
Fig. 10.2. View from aboveon the F-planar antenna
Foodfrom a discrete source is supplied to the antenna through the cylindrical
wire with 1 mm radius. Suppose a variable displacement feed point F-shaped
antenna (equal to 4.8 mm in Fig. 10.2). The position of its feed point determines the
input impedance at the central frequency of the working range.
To HFSS calculated derivatives to a variable:
1. Open Derivative Solution Setup tab in the dialog (Fig. 10.3).
2. Forvariable feed_pos, which will adapt to, select
Use.
Fig. 10.3. Setting options for the calculation of the derivative feed_pos
In this mode derivatives feed_pos parameter will be calculated, and then the
calculation can be menu command Tune Report Result visible change S-parameters
for varying the position of the feedpoint of the antenna. This feature does not apply
to the frequencies and near fields.
127
Creating a model F-shaped antenna.The global coordinate system construct the
antenna base. Command Create -> Box draw a box, the parameters of which are
shown in Fig. 10.4.
Fig.10.4. The base antenna height of 5 mm.
Furtherdraw the top of the planar antenna operation Draw-> Box, to the
dimensions shown in Figure dialog. 10.5.
Fig. 10.5. Pace-top antenna thickness 1 mm
Cylinder coaxial
line,
position
which
change
when
it changes variable
feed_pos (Ris.10.5) created
surgery
Draw> Cilinder. and its dimensions can be adjusted in Fig dialogue. 10.6.
Positionthe center can be changed by introducing the option feed_pos.
Fig. 10.6. cylinder parameters
128
coverage model antenna box, which draw the command Draw Box (Fig. 10.7)
and on the surface of which set the boundary condition Radiate.
Fig. 10.7. The size of the radiation box that covers the antenna
The lower part of the coaxial line supply port, releasing the cylinder (Fig. 10.8).
Fig. 10.8. Installing a port in a section of coaxial line
Now we define a reference for calculation in the frequency range of 1.5 GHz - 2
GHz (Figure 10.9.).
Fig. 10.9. Setting the frequency range in interpolation mode
129
Press Start. Afterwe define the calculation output characteristics partial MagS
(1,1, feed_pos) on a two-dimensional graph (Fig. 10.10), and then under derivative
Derivative (Fig. 10.11) select a variable feed_pos.
Fig. 10.10. Output
Options
characteristics
Fig. 10.11. characteristic selection
TuneS (1,1, feed_pos)
Nowyou need to open a dialogue operation Tune Report Report Tuning (Fig.
10.12).
Fig. 10.12. Observation of the frequency response of the displacement process of
the antenna when the position of the feed point
By changing the position of the slider in the window Report Tuning, you can
change the position of the feedpoint of the antenna, and thus see how this changes
the frequency characteristic.
131
Fig.10.13. The output characteristics of axis y
After can perform adjustment for the predetermined supply bias point position
calculated far-field characteristics. The spatial radiation pattern F-shaped antenna
is shown in Fig. 10.14.
Fig. 10.14. NAM planar antenna
So,In this section, we met with such an operation, as the adjustment. In addition,
the HFSS geometry optimization can be performed for a given objective function,
which includes the characteristics of the antenna.
The HFSS-13, gradient search techniques realized minimum of the objective
function Quasi Newton, simleks method, and the random search techniques
Pattern Seach and Genetic Algorithm (genetic algorithm).
131
11. Connector Modeling
In this example, the calculated characteristics associated lines complex shape.
Two extreme line connector (Reams. 11.1) are short-circuited with the ground plane.
Connector This project is located in the Examples folder and is executed by
Driven Terminal. The structure has 4 ports, twisted wires are connected. Ground fee
the same for all ports, and are at the bottom, and are at the top of the structure.
Fig. 11.1. The design of linked 4 curved lines
TOonnektor (Fig. 11.1) has discrete ports at each end of the two inner ends of the
curved guides (pin). The two outer conductors are earthed on both sides. The last
fastening parallelepipeds (substrate) are filled FR4 material. Boundary Radiate
radiation is applied to the surrounding boxing.
Calculationperformed an adaptation of the mesh partition of the center frequency
5 GHz. We choose calculation method Driven Terminal (Fig. 11.2) and unit of
measurement mm.
132
Fig. 11.2. selection method of calculation
AllGuides in the connector designed as three-dimensional objects made of copper.
The finished project in the Connector Examples folder set the port and boundary
conditions. Ports and terminals arranged in FR4 substrates, which are arranged
above and below the associated lines (Fig. 11.3).
Fig. 11.3.Setting ports and terminals
Ports are applied to the rectangular plane and the top of the rib set terminals (Fig.
11.4). Driven Terminal feature of the method is that in one section can specify
multiple ports with a common earth bus.
133
Fig. 11.4. ports1 and 2 and the terminals on the left and the right - short
circuit
will establishports 1 and 2 on the middle pad (Fig. 11.4). Ports 3 and 4 are
connected to the same lines that begin port 1 and 2, but on the other hand coupled
lines (Fig. 11.5).
Fig. 11.5. Setting the port between the conductor and the ground plate
deriveConversation determining ground plate (Fig. 11.6) Assign Automatic
command.
134
Fig. 11.6. In the dialog box indicates that the conductor gnd
becomingthe reference plane for all wireline
It is
Selecting gnd as the "Use as Reference" shows that the board gnd
established one for each wire, for all ports.
The calculation in the frequency band perform interpolation method so that
the calculation is performed from the DC to DC upper frequency of 5 GHz.
Fig. 11.7. Type of connector Connector
135
Fig.
11.8. Installation on a decision
To calculate the frequency range ask Add Sweep command Interpolating mode
select and define the frequency from 0 GHz to 5 GHz in steps of 0.01 (Fig. 11.9).
Fig. 11.9.Target analysis frequency band
136
Afterstart the decision-making process can be seen by pressing the right mouse
button on Setup1 and selecting Profile to open a dialogue Solution. Details of the
decision can be seen in the section Convergence, Matrix Data, and Mesh Statistics.
S matrix can be seen as well as the characteristics of ports (Fig. 11.10).
Fig.
11.10. S-calculated matrix connector
To see a graph of S parameters can be displayed in the Cartesian coordinate
system, adding the schedule section Results (Fig. 11.11).
Fig. 11.11. Frequency response parameters S
Of frequency characteristics of Fig. 11.12 you can draw the following
conclusion: the line
have
a good
harmonization, but
link between separate
linesconsiderable. So, between a number of lines reaching | S23 | = 0.8.
137
Fig. 11.12. The electric field in the cross section of connector pins
In concluding this section we note that the Driven Terminal method is used when
there are several lines, and associated with one bearing an earthen foundation.
Fig. 11.13. Calculated E field in a section of transmission lines in the excitation
of the 1st port
In order to observe and study the distribution of the field in the excitation of a
port, or a certain type of waves in this port, you need to set the Edit Source command
and bring the rice dialogue. 11.14, in which the opt-ins and phase sources operating
in the project.
Fig. 11.14.Ustanovka excitation sources on ports
138
Discrete portsimplemented in the HFSS-IE.The HFSS-IE implemented
discrete ports which HFSS-IE method different from those ports to HFSS. Discrete
port in HFSS-IE has a voltage of 1 V between the terminal and the reference
terminal to ground plane, while the port logic determines in HFSS electric field at
the site. Discrete ports (fig. 11.15) are similar to traditional wave ports, but may be
placed inside the structure, and should have a user-specified complex impedance.
Discrete port S-parameters calculated in the port section.
This port is an internalto the
field of solutions. Rectangle
that is placed port regards
to the signal line, and on the
other hand comes to earth
board
This port is an internalto the
field of solutions. On one
side of a rectangle with
regard to the signal line, and
on the opposite - from the
ground plate
this portinternal to the
field of solutions. Port is
set in a circle around the
ball
Ris.11.15.Implementation of ports in HFSS-IE
Discreteport can be set on a rectangle, going from the edge of the line to the
ground plate (Fig. 11.15). On all edges of the discrete ports that are not in contact
with metal or other boundaries must be set boundary condition Rerfect H. These
settings depend on the selected mode: modal or terminal. It is important to note the
following:
• complex impedance port must be non-zero, and the resistance must be nonnegative.
• allowedonly one mode, or load only one (terminal), if the decision is carried
out in the terminal mode.
• for signal ports integral line must be determined.
• each load It must be connected to the border with perfect or infinite
conductivity ground board.
139
Consider,that the port is set up correctly only if the maximum distance from
the terminal to ground the board less than one-twentieth of the wavelength. If this
condition is violated, a warning appears.
Modern digital devices operate at microwave frequencies. Characteristics of
transmission lines between the processors and digital blocks affect the performance
of computers. Not less important is the creation of supercomputers with tens,
hundreds of computers connected in a single supercomputer. Therefore, one of the
most pressing contemporary challenges consists in modeling and design of
communication lines in digital systems and devices.
15.5.2. The resonator system quantum rings
You can imagineor more simple devices emitters using "quantum rings." One of
these emitters is shown schematically in Fig. 15.27.
Ris.15.27. emitterwith the quantum ring.
140
This radiator to some extent, is an analogue of cells shown in Fig. 15.19. It uses
several quarter-cylindrical resonators (e.g. four) circumferentially disposed. In this
example, these resonators are excited species phase oscillations N111 connected
through a communication slot (mc) with a stabilizing cavity based on a half-wave
of the circular waveguide with N01r view (the figure shows lines of force oscillation
type N012). Instead KH provided in a cell introduced blue quantum ring (CCM)
arranged in its matrix (M) along the electric force lines. Thus, the radiating element
in the present emitter is like CN, coiled into a ring.
On ris.15.28 shownpattern of the electromagnetic field in the resonator system
and in the near zone of the emitter in a fixed time, and the frequency characteristic
of the resonator system without semiconductor structures, which were obtained in
HFSS.
Fig. 15.28. Frequency response and the pattern of the electromagnetic field at the
operating frequency of 650 THz
From ris.15.28 seen thatin the frequency range from 500 to 1000 THz several
resonance characteristics obtained, one of which corresponds to the resonance N012
VC in CP at a frequency of 650 THz. Proof of this fact is a picture of two variations
of the field along the axis of the superlattice.
Note that the VC N012 is not a lower, and therefore resonances are observed at
lower frequencies. However, separation of the VC is quite sufficient in order to
avoid any "parasitic" oscillations, lowering the radiation efficiency at the operating
frequency.
must be pay attention to the fact that the resonant characteristics of the PC with
the CP on the VC N012 is the most narrow, indicating the high quality of the PC
at the operating frequency.
141
On ris.15.29 shownfield pattern at the time point when cracks due electric field
reaches the maximum value, and near the CC and the zero field, as is kk at a distance
of a quarter wavelength from mc.
On ris.15.29 alsoshows the distribution of the azimuthal component of the electric
field in the radial direction, i.e. along the line shown in the figure. The azimuthal
component of the electric field on the VC N01r is the only and the curve
corresponding to an operating frequency of 650 THz, which indicates where the
maximum electric field and where to put the spacecraft.
Ris.15.29.The electric field distribution in the radial direction and an illustration
of the radiation at the time point when cracks due maximum electric field.
Note that the crossemitter dimensions considered limited diameters quarterwave resonators, which should not exceed 0,6 λ, so that they allow the occurrence
of only the lower VC N111, and the diameter of the CP, which should not be
substantially greater than 1,2 λ, so that no set higher VC.
This fact is in contrast with the need to ensure adequate technological sizes for
effective manufacture of semiconductor structures on the basis of the CT in a matrix
InGaN GaN [19]. Therefore, it is desirable to increase the size of the matrix to RT.
For this purpose, the emitter is provided an apparatus schematicallyshown in
ris.15.30, which uses a coaxial CP excited on oscillations as H011 and disposed
within multiresonator system consisting of a cylindrical quarter-wave resonator
(CR) with views N111 oscillations. Fig. 15.30 8 shows resonators antiphase
oscillations in communication with through one CP CR. Above these resonators is
circular matrix
211
(MM) CT. Calculated dimensions resonators for blue and red CT according to the
markings on ris.15.30 shown in Table.
Table. Dimensions (in nm) of the radiator shown in ris.15.30.
DCR
Dcf.
Colour
Wavel
d
h1
h2
ength
Blue
460
340
680
100
555
370
Red
640
480
960
140
750
755
Dm
1360
1920
Fig. 15.30. Emitter stabilizing coaxial resonator.
Forprovide the most efficient electromagnetic coupling between the CP and
the CR their dimensions should be chosen so that the lengths of waveguides of
which are obtained resonators have the same critical and, accordingly, the
waveguide wavelength, i.e., h = 1 3 / h and 4λv λv = 2/2.
Device considered emitter convenient in that the diameter
Dm matrix can be increasedby increasing the number of the Czech Republic. With
increasing diameter dav is correspondingly necessary to increase the diameter d of
the rod to maintain the resonance frequency.
Summarizing consideration proposed emitters should be noted that the
luminous flux of such nanoscale devices is very small, and these
211
radiators should be collected in an array. Each emitter can be supplemented with the
reflector. Such transducers can be grouped, for example in full-color (blue, green
and red) triad. The resulting light emitting device can be used as the light source
with adjustable chroma chromatic or white radiation and also be used in displays
with high resolution.
luminous efficiencyradiator with blue KN combined with a yellow phosphor is
largely dependent on its own Q-multiresonator system, tuned to the frequency of the
blue. When used without the color triads of phosphors luminous efficiency of each
color element also greatly depends on the quality factor and adjustment system with
stabilizing multiresonator resonator to the wavelength emitted by the quantum
asterisk or quantum ring.
15.6. Analysis of the structure composed of eight
resonators in the optical wavelength range
Consider the analysis stepsresonator structure in the optical range of wavelengths
in the example system HFSS emitter shown in ris.15.30.
At the first stage, the drawing and creation of the resonator structure. We choose
drawing unit "nm" and "THz" frequencies.
For blue light? c = 460 nm
D = 1.5 microns, Dc = 750 nm, d = 100 nm, Dch = 380 nm, lch = 490 nm and lc =
325 nm;
forred λk = 640 nm
D = 2.1 microns, Dc= 1050, d = 150 nm Dch = 525 nm Dlinych = 675 nm and lc =
450 nm. The length of the coupling slots 400 nm.
resonatorstructure and size with said quantum ringlet executed graphically
HFSS, ris.15.31 shown in a perspective and plan.
212
Ris.15.31.The structure in which slots are arranged through one external
resonator, i.e., linked to the central coaxial resonator 1,3, 5, and 7, the outer
cylindrical resonators
Creaturethese figures carried out using the following commands. Draw a
baseline resonator structure command Draw-> Box (ris.15.32).
Ris.15.32. dimensions base metal box
Furtherin this boxing need to cut one central cylinder radius of 525 nm and an 8cylinder radius of 262.5 nm. 8 resonators are arranged so that the side is located
between the central coupling slot (ris.15.33 and 15.34).
213
Fig. 15.33.Tsentralny cylinder
Ris.15.34. small cylinders
Command Edit-> Duplicate-> Around Axis duplicate the cylinders around an axis
and is multiplied by 8.
Next draw the inner rod located along the axis Z (ris.15.35)
Ris.15.35. dimensions internal metal rod.
Furtherat 400 nm draw the ring, subtracting one from another cylinder (ris.15.36
and 15.37).
Ris.15.36. dimensions the
outer cylinder.
Ris.15.37. dimensions the inner
cylinder.
As a result of subtraction we obtain planar ring 20 nm wide. We establish the
Perfect E boundary conditions on the quantum ring. To insert the ring 8 ports that
will excite
214
consistent
current
at
ring,
thinned
planar
ring
8
rectangles,size of the first of which corresponds to Fig. 15.38.
at
Fig. 15.38.rectangle coordinates in the gap of the quantum ring, which set a
discrete port.
By creating this rectangle, ask him a discrete port command HFSS-> Exitation> Assign Lumped Port, and setting impedance line down the middle of the rectangle.
Further fulfill this rectangle overlaps with the preservation of boundary conditions
with respect to the axis Z.
Ris.15.39.Dialogue discrete port supplied in the gap of the quantum
ring.
Created a structure shown in Fig. 15. 31, will cover the box, created by a team
Draw-> Box, which is in the nanometer size are shown in Fig. 15.40.
215
Ris.15.40. Dimensions Radiate boxing.
On This will install a box radiation boundary conditions, and on the side wall periodic boundary Master-Slave conditions on the opposite wall of the radiation
box (Figure 15.41.).
Ris.15.41.Formulation periodic boundary conditions Master-Slave on opposite
sides of box covering one resonator cell.
Perform a calculation HFSS- team> Analysis setup-> Add Solution Setup. The
center frequency of 467 THz respectively. wavelength 640 nm (Fig. 15.42).
216
Ris.15.42.Setting the center
frequency analysis
Ris.15.43.Target bandwidth
(mode interpolation)
Notethat if the quantum dots are arranged uniformly in the quantum ringlet, they
will be excited in antiphase in accordance with the fields in cylindrical resonators.
radiation pattern in this case is shown in ris.15.44 and 15.45.
Ris.15.44. Radiationelectromagnetic wave optical range of quantum ringlet in
a system with an internal coaxial resonator stabilizing
217
Ris.15.45.Near field in the cross section of the resonator when
installing antiphase excitation of the quantum dots.
On ris.15.46 shows the frequency characteristic, and ris.15.47 - a picture of the
electromagnetic field when excited CT located in CC through one resonator.
XY Plot
1
ansoft LLC
0.00
HFSSDesign1
Curve Info
dB (S (1,1))
Setup1: Sw eep1
-2.50
-5.00
-7.50
dB (S
(1,1))
-10.00
-12.50
-15.00
-17.50
-20.00
-22.50
200.00
300.00
400.00
500.00
600.00
Freq [thz]
700.00
800.00
900.00
1000.00
Fig. 15.46. The frequency response of Selevenfor each port, the introduction of QC
(8 ports)
218
Ris.15.47. Painting electromagnetic field when excited CT located in
the cavity through one QC
Analyzing the dynamic movement pattern of the electromagnetic field, Thoroe
Ko is calculated by changing the phase of the excitation source, it is possible to see
how the interactions of electromagnetic fields radiated from the individual
resonators, and the overall picture of the field types of waves that resonate in
separate cylindrical cavities resonators.
Conclusions.Options considered SDMbased OR quantum dots, rings or sprockets
and in no way limit the variety of these devices. Not difficult to imagine, and other
design options based on SDM quarter Multiresonator systems with other types of
vibrations, including a variety of high-Q resonators stabilizing.
Currently, the representation of the emitters in the form of quantum zvozdocheck and rings may seem unreal fantastic and Modeling such sources still requires
a multifaceted experimental confirmation. However, the progress of the theoretical
analysis of promising nano-materials and nano success shows that it will be realized
in the near future.
numerous publications which show the use of
HFSS show that it can be successfully used to optimize the structure of the LED
module. It is safe to say that the search for optimal solutions when creating the
SDM on the basis of electro- dynamic systems with quantum dots will
significantly improve the parameters of lighting devices with nanoscale The
radiation sources.
219
16. The configuration for distributed based on multiple
computers
HFSS -13 has the ability to solve problems of enormous size, the number of
tetrahedra of the partition can reach many hundreds of thousands.
In the last 10 years have seen tremendous progress in enhancing computer-tools
to solve large-scale problems - from the creation of large computer servers,
to
the distributed multiprocessor clusters and desktop workstations. In parallel went
the improvement of mathematical algorithms and programs to implement them. Fig.
16.1 It shows the progress of milestones in terms of physical problems in the
mechanical analysis, and solution of electrodynamic problems Ansoft field. It is
now believed that the use of HFSS in the design process by 15% accelerates the
development of new products from idea to implementation in hardware.
Fig. 16.1. growth story system performance characteristics HFSS
221
One way to increase productivity is to parallel simulation. Fig. 16.1. It shows
that in 1998-1999 it was possible to solve a system of equations 10M cells. In 2004,
HFSS Ansoft solved the structure of a cell 100M. Today ANSYS may analyze the
structure of 1 billion cells, and the level of simulation accuracy is very high.
Currently, the key is a new method for the implementation of which is necessary
to have a special license for High Performance Computing (HPC).
This method allows for the calculation of 2 sets:
1) Domain Decomposition,
2) Multi-processing.
HPC Solution method reduces to the calculation of multiprocessing using
distributed calculation options that can combine up to 100 multiprocessor cores.
Note that description in this section, the term
"Computer" mayrelate to a personal, as well as a super-computer. Yadry and
processors can characterize any of these machines.
Starting with HFSS v.12, Ansoft company develops HPC method using the
method of the task decomposition. Method MP (multi-processing) has been
implemented in HFSS-7. Option distributed DSO solutions implemented since
HFSS-10. These options are used today to reduce simulation time from days to
hours, allowing the engineer to perform multiple simulations within one working
day. For this purpose have been developed and proposed special methods
multiprocessor and multicomputer calculation HPC and implemented additional
improvements in HFSS v.12 (enhanced partitioning into mesh and mixed orders of
the elements of the partition) to substantially increase productivity.
Let us clarify the terms,that came in designing, together with the strengthening
of the power of the numerical simulation.
1. HFSS HPC -A new option,which appeared in HFSS 12
HPC = High Performance Computing
2. Enables Domain Decomposition, decomposition method -a new method
for solving in HFSS 12. DDM = Domain Decomposition Method. A new
method of DMP in HFSS. DDM - is a method of high-paralleling,
where:
– this method of combining memory and nuclei number of
computers via a network;
221
– DDM It realizes a significant increase in memory for large and very
large-scale problems;
– DDM providesa quick solution, using a large number of
processors.
3. Multi-Processing - Multiprocessorparallelization method separation, pull
a memory, introduced in HFSS 7.
MP = Multi-Processing
– "Traditional" multiprocessing SMP.
4. The number of licenses "HPC licenses" depends for modeling the number of
cores that are required for the calculation.
5. Work Package mode (Packs) or "floating (pool)" mode.
– Packsallowsa predetermined number of nuclei for the
program;
– "Pool" licenses allow you to use any number of cores managed
by multi-servers.
• SMP = Shared Memory Parallel.
• DMP = Distributed Memory Parallel.
• DSO = Distributed Solve Option - mNogo-parametric (geometry,
frequency) parallel simulation.
• RSM = Remote Simulation Manager -A new control programdistributed
analysis, appeared in HFSS v12, designed to work in computer network
Sample calculations which are made using HPC license
Example 1. The antenna is located at the military vehicle Humvee.
•
•
•
•
•
The problem is solved for electromagnetic compatibility 1.8GHz frequency
and
- where the wavelength at the operating
frequency, m) and uses:
basis a second order function (2nd order);
730,000 tetrahedra;
14M unknown in the matrix (it is even a little, to use the method
DDM);
Perform calculations with the help of three related groups of computers;
Shared memory = 115GB.
222
Fig. 16.2. Calculation Example antenna structures on a moving object
The calculation resultsthe spatial pattern of the pokaza- in Figs. 16.2.
Example 2. The antenna mounted on the body fighter F35. Analyzed slot antenna (Fig. 16.3) is in the bottom of the wing of the
aircraft fuselage
• The F-35 Joint Strike Fighter: UHF blade antenna @ 350 MHz
•
The size of the problem= 1400λ3
Fig. 16.3. Military fighter F-16 with the antenna at
the bottom of the fuselage
Ris.16.4. Slot antenna
at solving this problem, a method of DDM.
• DDMsubregion split into partitioning net in small sub-regions "sub-domains"
so that the field in which is calculated in parallel on multiple computers. For
this is done:
- Selection of unit "master" of the iterative solver for the general solution
223
- Automatic partition of solutions on a subdomain
• User
determines set
N available
computers
for
use by DDM:
– n = 1 is the node "master", a single nucleus;
– n = 2 to N is nodes subdomains solutions compared to direct solution.
Fig. 16.5. partition of based on tetrahedra, region and subregion
The calculation results are the same,HFSS directly in solving program, and using
the method of DDM. However, the calculation time significantly different.
224
Fig. 16.6. The cross-section of
the radiation pattern antenna
on the aircraft body
Fig. 16.8. frequency
characteristic
Fig. 16.7.The spatial radiation
pattern
Fig. 16.9. time
calculationdepending on the
number of areas partitioning
Example 3: Two-mirror parabolic antenna(Fig.
16.10) createdKossegrena scheme. The size of the problem ~ 17,500 λ3.
Time solutions on 8-core computer 22049 seconds. The solution is
accelerated in 17-24 times using HPC license.
225
Fig. 16.10. Two-mirror parabolic antenna
The table in Fig. 16.11. given machine resources expended for modeling twomirror of a parabolic antenna, which depend on the number of nuclei. And accelerate
the process of solving the problem with the inclusion of additional cores. We see
that acceleration comes to shevti compared with a basic configuration with 8 cores.
Note that the acceleration calculation as compared with the work can not be
performed on a single processor.
Fig. 16.11. Computer resources required for solving the two-mirror antenna.
Accelerating the Speed-up compared to the base computer with 8 cores
For
solutions like
modeling RSM
of tasks must
establish
managerremote
(Remote Simulation
Manager).
This
226
program is installed asa separate program and automatically creates a remote or
related to the simulation of network communication mode "links".
• Method new distributed analysis work on remote computers
• activates solver capabilities in DDM mode, DSO and solutions mode on
remote computers.
Installation configuration distributed calculation.
To create a new configuration based on a distributed multiple computers:
1. Click or HFSS HFSS-IE> Tools> General Options to display the dialog
General Options and select the tab Analysis Options.
Fig. 16.12. Dialogue HFSS options
2. Selectbookmark Analysis Options, and in the section Analysis Machine Options,
select the Distributed option. These changes will be displayed in the tab Analysis
Machine Options and show the panel for viewing and editing computers
configuration (Fig. 16.13).
227
Fig. 16.13. Options analysis computer in the network
3. Click Edit Distributed Machine Configurations button.
Fig. 16.14. Dialogue distributed calculation configurations
This displays a dialogue Distibuted Machines Configurations (Fig. 16.14). This
list shows all the existing configurations of the machine and the selected
configuration, as well as information, they are locked or not. Here you can add a new
configuration by clicking Add, Edit to edit the configuration command to delete the
selected configuration Delete command, or duplicate the existing configuration
Clone command, usually editing the name and configuration content.
4. To create a new configuration, click on the figure the dialogue. 16.14
buttonAdd.
This opens a dialog Distributed Analysis Machines (Fig. 16.15). Clicking Edit or
Delete also opens Distributed Analysis Machines dialogue, but also includes having
the selected configuration.
228
Fig. 16.15. Dialogue list of machines used in the calculation of the distribution
Here You can edit the configuration name, type, adding, tes- tirovanie and
blocking machines in the list.
5. Askconfiguration name. It can not be empty and can not have previously
given or reserved word.
6. For each machineTo manually add a list, under Remote Machine details, set the
IP address, DNS name or UNC name. You can also use the Import Machines from
File ... to use a text file to simplify the process.
Fig. 16.16. Section, which are included in the list of machines
button controls allow the machine to add to the list Add machine to list command
(Fig. above) or remove the Remove command (Fig. below) of the machine from
the list.
229
Each machine in the current list has the Enabled box. You can block or connect
any machine in the list according to your desire. On top of the table, the dialogue
takes into account the total number of cars, and this number can be enabled or
disabled.
Fig. 16.17. Blocking and the inclusion of machines in the calculation of the
distribution
In general, the HFSS and HFSS-IE use machines in distributed analysis machines
listed in the order in which they appear. If you have selected distribution method of
calculation, and you run some calculations from the same interface, the HFSS and
HFSS-IE choose machines that perform the least amount of computers in the order
in which the computer is on the list. For example, if the list contains 4 cars, and you
run a simulation that requires a single machine, the HFSS chooses the first machine
in the list. If another simulation run, while the previous is done, and this simulation
requires two machines, the HFSS chooses the machines 2 and 3 from the list. If the
first modeling then completed, and we run another simulation, requiring three
machines, the HFSS chooses the machine in the order 1, 4, and 2.
Displayed list always shows the order in which you entered them regardless of the
load on the machines. To control the order in the list, select one or more machines,
and use the Move up or Move down buttons. Moving up and down are allowed when
you select one or more adjacent machines is. Also, when you select one or more
host names, you will see a text field below the power supply control, showing the
name of the first car, how many times it appears, it allowed and blocked.
231
Fig. 16.18. The movement of vehicles on the list to change their
switching sequences
7. TestMachines- When several users on the network using a distributed remote
solution.you need to check the status of these machines before the simulation can be
run to make sure that no other processes are not running Ansoft. To do this, you can
select one or more of the machines in the dialog (Fig. 16.18), and click Test Machine
button. Opens Test Machines dialog. After the test is given a report on the status of
each machine.
Testing can be performed on a group of machines that are included in the
calculation of the distribution. Enable or disable the machine can be in the list
dialogue Fig. 16.18.
8. Click OK, to save your changes and close the dialog Distributed Analysis
Machines. Only machines marked as Enabled appear in the list of machines
distributed in the Analysis tab.
independentlyfrom the machine on which the analysis is actually performed, the
number of processors and installation Desired RAM Limit, and the default process
priority settings are now read from the machine from which you run the analysis.
If configuration is set, you can select it on the instrumental panel, you choosing
icons:
Fig. 16.19. HFSS buttons on the panel, allowing switches between the local mode,
remote and distributed work
231
Note that this option is only active if there are several machines for distributed
analysis. distributed analysis mode is recommended in the event that not enough
resources of a single computer.
HFSS on the same computer works as follows. Grid - essentially a large matrix,
which must be inverted and solved. HFSS is trying to do it all in RAM. As soon as
the available RAM is exhausted, the HFSS starts to divide the matrix into pieces, in
which some elements of the matrix are written to the hard disk, and other parts are
stored in RAM. However, there is a minimum size of the matrix that is to be stored
in the RAM, and must remain in the memory space to be able to invert the matrix
and solve. If the matrix is so large (ie, mesh partitioning very dense), that the
minimum size no longer fits in RAM, HFSS stop.
the user hastwo several options.
It can 1) increase the computer's RAM, or 2) can be usediterative solver, or 3) to
use the license, and therefore HPC method to solve the problem. We recommend
using the iterative solver. Using Iterative Solver option can dramatically reduce
RAM requirements.
Terms of Use iterative solver
1. The iterative solver works more efficiently when the structure does not contain
many excitation sources. (For example, number of excitations - less than twice the
number of processors).
2. If you want totake advantage of the iterative solver, and your analysis includes
calculation of many discrete frequencies, adaptive decision should better go to the
highest frequency in the frequency band.
232
Fig.16.20. Section Solution Options, comprising selecting the order of the basis
function and the parameters of the iterative solver
3. Relative ResidualIt provides a stopping criterion. The difference in
estimatesiterative solver for solving a matrix equation. Value acts on the
performance of the iterative solver as follows:
• Defaultit is equal to 1E-4. This gives accurate S-parameters and fields which are
indistinguishable from those obtained in the usual calculation. Ansoft recommends
setting this value Residal.
• In largervalue Residual, e.g., 1E-1E-3 or 2, the iterative process stops with a
smaller number of iterations, and the solution will converge less. S-parameters will
not differ much from the results obtained by direct solution, for example, the
difference is in the third or the second digit. Fields and antenna pattern - visually the
same.
• Residual = 0.1 size can be used for fast adaptive grid refinement early in the
adaptive process, but the S-parameters will differ significantly.
• valueResidual = 1 should never be used. In the box there is no opportunity to enter
the value Residual than 0.1.
Selecting the order of the basis functions.The HFSS-13 can select a
basic function using HFSS field interpolation values between the tetrahedra nodes.
To do this, see Options dialog Solution Setup
233
selectedOption Order of Basis. It can be a Zero Order, First Order (the default),
Second Order, or Mixed order. Options Order of Basis affect the mesh seal criterion
Lambda Refinement and selected as follows.
Option zero order (Zero Order) is useful when the model is divided into a grid with
more than 100,000 tetrahedra, but the size of the model is small compared with the
wavelength. Higher orders increase the number of unknowns for each tetrahedron,
and are used when greater accuracy is required for calculating the field. If you choose
Zero Order Solution Basis, all tetrahedra in the model must have a length of the ribs
is less than 1/20 of the wavelength.
Installation options when running on the same computer
Fig. 16.21. Options computer calculation to work independently
For multiprocessing, select Enable multiprocessing using HPC licenses. This
option allows the use of HPC license, including multiprocessing.
the RAM requirements increase linearly with an increase in size of the model.
When using the iterative solver is also the case for linear increase RAM
requirements. If you are using a basic solver, the RAM requirements increase
linearly.
Settingssolutions with domains always use licensing HPC.
Even at
Tom, what domains distributed
and
can use
234
multiprocessingtreatment, they do not use a distributed solution or a multi-license all this is included in the licensing of HPC, including domains.
HPC license is based on a calculation of the total number of cores in the
simulation. For example, performing a distributed solution using 10 nodes with
multiprocessing consisting of 2 cores per node, the total number of nuclei 2x10, or
twenty nuclei.
License HPC restricts the type and number of licenses to be verified for a given
number of nuclei. For HPC, one type of license is checked for each kernel is used.
Thus, the modeling of twenty cores would require twenty HPC licenses. For license
type HPC Pack, one pack includes working with eight cores, and each additional
package includes four times more cores. Thus, modeling with twenty two cores
require HPC Pack licenses covering operation with a 8x4, or 32 nuclei.
5. Formultiprocessing, check Enable multiprocessing using HPC licenses option.
This option allows the use of HPC license, including multiprocessing, even on tasks
that do not include the area of domains. In this case, HPC licenses serve as MP
license.
6. Select one of the following priorities from a list of Default Process
Priority:
• Critical (highest) Priority (Not recommended)
• Above Normal Priority (not recommended)
• Normal Priority (Normala priority)
• Below Normal Priority (below normal)
• Idle (lowest) Priority (slow priority)
Can set these values using the additional programming in Visual
Basic.
235
Fig.16.22. Dialogues properties working dialogue Ansoft
To transfer dataand management solution to the problem using dialogue
Submit HPC Job (Fig. 16.22).
To submit a job using the Submit HPC Job dialog box should be set to the main
cluster host name. Selecting the command Tools> Windows HPC> Select Head
Node ... prompted (Fig. 16.23), the choice of the cluster host name that you want to
use.
Fig. 16.23. Selecting a host name
You can enterthe main name of the cluster node in the dialog box, and you can
click on the button [...] to browse the network for the cluster that you want to use.
236
Fig. 16.24. Defining multiple calculation settings dialog
After setting the node headerhostname, select Tools> Windows HPC> Submit
HPC Job ...
You canselect the installation section Analyze Setups (Fig. 16.24).
Here you can choose from the following options:
• All setups in the project;
• All setups in a specified design:youYou select a project from the opening
Gosia list;
• Specify setups:You can addone or more settings By using Add Setups
dialog ...
Ifyou specify multiple settings, called the task sequentially in the
order displayed in the edit box.
BlockResource RequirementsIt manages the computing resources used.
settingsDistributed Analysismanage the use of resources for part of the analysis,
which may be distributed on various
237
cores. Settings Non-Distributed Analysis controlled initial part of the analysis that
works before starting any distributed analysis
- this initial part can not be allocated to separate components, and should be
performed on a single core.
There are two settings Distributed Analysis. The number of "machines" - a
number of groups of computers that run in parallel; computers can run on individual
nodes. The set number of cores on one computer controls the degree of parallel
processing within each computer; these parallel execution of data flows must be on
the same site. In each case HFSS distributed mechanism represents one large
domain model part of the frequency sweep analysis or analysis part Optimetrics
installation.
There is only one setting for not distributed analysis, because this part of the
analysis should take place on a single node. Number of cores controls the degree of
parallelism used for this part of the analysis.
mark Log Analysis Progress, to record the review process in the file
<Projname.progress>. This file
It contains messages
about
errors,prevention and other. Information.
Clicking Next takes the current settings, and improves the dialog box Submit
HPC Job: Properties. Clicking Cancel closes the dialog box.
238
conclusion
HFSS v13 software developed by the American company Ansoft, the
development of which is in the learning process of the subject of this book - not the
only, which has been widely used in Russia in the scientific and studied vatelskih
and educational institutions. The greatest distribution was received by the so-called
"troika»: HFSS, CST, FEKO.
Oftensoftware selection is dictated by tradition, or speed training on the
software product. The HFSS, starting with versions developed Hewlett Packard
companies and the Agilent, was the most popular in Russia among software
electromagnetic modeling of three-dimensional structures, and contributes to the
learning process in many Russian universities.
The structure of the radio now firmly divided into tsif- rovuyu part and the radio
- radio receivers and transmitters, and often radio-technical part - is the antenna and
the very first stages following them. Digital processing occurs and recaptures all of
the bol- Shui radio. Moreover, in the management of a large antenna array performs
the role of the digital part. Therefore, we can say that for the development of radio
engineering HFSS - only part of the way of preparation.
Selection software actually boils down to finding a compromise between the
speed of calculation, the available computing resources and a volume of (the size in
terms of wavelength). It all boils down to the problem of optimizing the design
process variables present-microwave devices, the development of the relation
between analytical and numerical methods, solving urgent problems facing both the
organizers of the scientific work, and in front of the performers from researchers to
engineers.
The process of development of modern software on the bench of undergraduate,
driven to understand the intricacies of the process proektiro- Bani, is a new and
salutary step of education and training qualifications radiospe- ists.
We hope that the problems of the analysis method usedthe tutorial - step through
the construction and analysis of the results - will help developers master the
microwave devices HFSS program and widely use it in their work.
239
Literature
1. HFSS- High Frequency Structure Simulation. Manuals, Ansoft, 2010.
www.ansoft.com - website of Ansoft-ANSYS - HFSS software developer.
2. Computer-aided design of microwave devices. Ed. VV Nikolsky / AM Radio
and Communications, 1982. - 272 p.
3. Roger Harrington, Time Harmonic Electromagnetic Fields, McGraw-Hill, New
York, NY in 1961.
4. K.A. Balanis. Antenna Theory: Analysis and Design, Wiley & Sons, 2 nd
edition, 1997, 942 pp.
5. Baskakov From.AND.Fundamentals of Electrodynamics, Moscow, Moscow,
Sov. Radio, 247
p.
6. Andrew Peterson, Scott Ray, Raj Mittra, Computational Methods for
Electromagnetics, IEEE Press, New York, NY in 1998.
7. LER Peterson et al. "Analysis of Periodic Structures via a Time-Domain
Finite-Element Formulation with a Floquet ABC" IEEE Trans, AP, March 2006,
pp 933- 944.
8. Schubert F. Light Emitting Diodes. Second edition. - Cambridge University
Press, 2006. Schubert FE LED / Per. from English. ed. AE Yunovich - M .:
FIZMATLIT, 2008.- 496 p.
9. Zhores Alferov The history and future of semiconductor heterostructures. FTP,
1998, v. 32, №1, pp 3-18.
10. Zvezdin AK Optical microcavities,waveguides, photonic crystals. Nature. 2004.
№10.
11. Alexandrov IA, KS Zhuravlev, VG Mansurov, Nikitin Yu Nonradiative
recombination in the quantum dots GaN / AlN // Abstracts 6th VC "Nitrides of
gallium, indium and aluminum - structures and devices". June 18-20, 2008. St.
Petersburg. Physico-Technical Institute. AF Joffe RAN.- S.210,211.
12. Guttsayt E. M. technology and devices at microwave frequencies. Radio and
communication, 1994. 224 p.
241
13. Banks SE, Kurushin AA Calculation of antennas and microwave structures with
the help of
HFSS Ansoft - M, ZAO "NPP" Rodnik ", 2009, 256 p.
14. www.microwavestudio.com - company-developers program
Microwave
StudioCST.15.www.edem3d.ru (program
EDEM).
16. Siteofficial dictribyutera modern software company OAO
"Rodnik"www.rodnik.ru.
17. Distance Learning Sitedesign and application of microwave
techniqueshttp://ipso.ioso.ru/distance .
18. Shlifer EDElectronic devices microwave M-type, coaxial and reverse coaxial
magnetrons // Results of science and technology. Ser. Electronics. T.17.- M .:
VINITI, 1985. - s.169-209.
19. VS Sizov, AA Gutkin, AV Sakharov, VV Lundin, Brunkov PN Tsatsulnikov
AF phase decompositionand nonradiative recombination of the carriers in
the active regions of the light-emitting devices based on InGaN quantum
dots in a matrix of GaN or AlGaN // FTPP. 2009 t.43, vyp.6, s.836-840.
20. DM SazonovAntennas and microwave devices. M .: Higher. wk. 1988. 432 c.
241