Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Instructor
Dr. Raymond Rumpf
(915) 747‐6958
rcrumpf@utep.edu
EE 5337
Computational Electromagnetics (CEM)
Lecture #1
Introduction to CEM
These notes may contain copyrighted material obtained under fair use rules. Distribution of these materials is strictly prohibited
Lecture 1 Slide 1
Outline
• What is CEM?
• CEM wisdom
• General concepts in CEM
• Classification of methods
• Overview of methods
Lecture 1 Slide 2
1
1/8/2018
What is CEM?
Lecture 1 Slide 3
Computational Electromagnetics
Definition
Computational electromagnetics (CEM) is the procedure we must follow to model and
simulate the behavior of electromagnetic fields in devices or around structures.
Most often, CEM implies using numerical techniques to solve Maxwell’s equations instead
of obtaining analytical solutions.
Why is this needed?
Very often, exact analytical solutions, or even good approximate solutions, are not
available. Using a numerical technique offers the ability to solve virtually any
electromagnetic problem of interest.
r
Zc cosh 1 out Zc ?
2 rin
Lecture 1 Slide 4
2
1/8/2018
Popular Numerical Techniques
• Transfer matrix method
• Scattering matrix method
• Finite‐difference frequency‐domain
• Finite‐difference time‐domain
• Transmission line modeling method
• Beam propagation method
• Method of lines
• Rigorous coupled‐wave analysis
• Plane wave expansion method
• Slice absorption method
• Finite element analysis
• Method of moments
• Boundary element method
• Spectral domain method
• Discontinuous Galerkin method
Lecture 1 Slide 5
CEM Wisdom
Lecture 1 Slide 6
3
1/8/2018
The Key to Computation is Visualization
Is there anything wrong? If so, what is it?
i , j , k 1
Ezi , j 1,k Ezi , j ,k E y E yi , j ,k
xxi , j ,k H xi , j ,k
y z
xyi , j ,k H yi , j ,k xyi 1, j ,k H yi 1, j ,k xyi , j 1,k H yi , j 1,k xyi 1, j 1,k H yi 1, j 1,k
4
xzi , j ,k H zi , j ,k xzi , j ,k 1 H zi , j ,k 1 xzi 1, j ,k 1 H zi 1, j ,k 1 xzi 1, j ,k H zi 1, j ,k
4
E yi 1, j ,k E yi , j ,k Exi , j 1,k E xi , j ,k zxi , j ,k H xi , j ,k zxi 1, j ,k H xi 1, j ,k zxi 1, j ,k 1 H xi 1, j ,k 1 zxi , j ,k 1 H xi , j ,k 1
x y 4
i , j ,k H yi , j ,k zyi , j 1,k H yi , j 1,k zyi , j 1,k 1 H yi , j 1, k 1 zyi, j ,k 1 H yi , j ,k 1
zy
4
Response
i , j ,k i , j ,k
zz H z
H zi , j ,k H zi , j 1,k H y H y
i, j,k i , j , k 1
xxi , j , k Exi , j ,k
y z
xyi , j ,k E yi , j ,k xyi , j 1,k E yi , j 1, k xyi 1, j 1,k E yi 1, j 1, k xyi 1, j , k E yi 1, j ,k
4
xzi , j ,k Ezi , j ,k xzi , j , k 1 Ezi , j ,k 1 xzi 1, j , k 1 Ezi 1, j ,k 1 xzi 1, j , k Ezi 1, j ,k
4
H xi , j ,k H xi , j ,k 1 H zi , j , k H zi 1, j , k yx Ex yx yxi 1, j 1,k Exi 1, j 1, k yxi 1, j ,k Exi 1, j ,k
i , j ,k i , j ,k i , j 1, k i , j 1, k
Ex
z x 4
yyi , j , k E yi , j ,k
yzi , j ,k Ezi , j ,k yzi , j , k 1 Ezi , j ,k 1 yzi , j 1, k 1 Ezi , j 1,k 1 yzi , j 1, k Ezi , j 1,k
4
H yi , j ,k H yi 1, j ,k H xi , j , k H xi , j 1, k zxi , j , k Exi , j ,k zxi 1, j ,k Exi 1, j , k zxi 1, j ,k 1 Exi 1, j , k 1 zxi , j ,k 1 Exi , j ,k 1
x y 4
zyi , j ,k E yi , j ,k zyi , j 1,k E yi , j 1, k zyi , j 1,k 1 E yi , j 1,k 1 zyi , j ,k 1 E yi , j ,k 1
4
zzi , j , k Ezi , j ,k
Lecture 1 Slide 7
Golden Rule #1
All numbers should equal 1.
Why?
(1.234567…) + (0.0123456…) = Lost two digits of accuracy!!
Solution: NORMALIZE EVERYTHING!!!
E 0E x k0 x
0
0
0 or y k0 y
1 m
0
H H z k0 z
0
Lecture 1 Slide 8
4
1/8/2018
Golden Rule #2
Never perform calculations.
Why?
1. Golden Rule #1.
2. Finite floating point precision introduces round‐off errors.
Solution: MINIMIZE NUMBER OF COMPUTATIONS!!!
1. Take problems as far analytically as possible.
2. Avoid unnecessary computations.
r x2 y 2 R x2 y 2
r2 R
g r exp 2 g R exp 2
Lecture 1 Slide 9
Golden Rule #3
Write clean code.
• Well organized • Compact
• Well commented • No junk code
Why?
1. It will run faster and more reliably.
2. Easier to catch mistakes.
3. Easier to troubleshoot.
4. Easier to pick up again at a later date.
5. Easier to modify.
Solution
1. Outline your code before writing it.
2. Delete obsolete code.
3. Comment every step.
4. Use meaningful variable names.
Lecture 1 Slide 10
5
1/8/2018
The CEM Process
There is a rhythm to
computational electromagnetics
and it repeats itself constantly.
Starts with Maxwell’s equations and derives all the necessary
equations to implement the algorithm in MATLAB.
• Equations everywhere! Only a few are needed.
• Implementation does not resemble the formulation.
Organizes the equations derived in the formulation and
considers other numerical details.
• Consider all numerical best practices.
• Should end with a detailed block diagram.
Actually implements the algorithm in computer code.
• Implementation should be simple and minimal.
• The “art” of simulation begins here.
• Practice, practice, practice!
Lecture 1 Slide 11
Don’t Be Lazy
A little extra time making your program more efficient or simulating a
device in a more intelligent manner can save you lots of time, energy,
and aggravation.
Lecture 1 Slide 12
6
1/8/2018
Formulation Wisdom
• Derive equations as far and as simple as possible.
• Build big/complicated matrices from small/simple
matrices.
– This usually requires converting to matrix form early in
the formulation.
• Make your formulation documents very detailed.
• A good understanding of the formulation gives
you the ability to modify your algorithm or to
add/subtract features.
Lecture 1 Slide 13
Implementation Wisdom
• Make a detailed block diagram!
• In the block diagram, include only the equations
you will incorporate into your code.
• Add all other steps to your block diagram.
– Sources
– Building devices
– Extracting information
– Post‐processing data
– Etc.
Lecture 1 Slide 14
7
1/8/2018
Coding Wisdom
• Work hard to write clean, simple, and well
commented code.
– Indent code inside loops, if statements, etc.
– Let linear algebra do the work for you.
– Use lots of comments.
• Match your code to your formulation.
– Try to use the same variables.
• Do not fix your code with incorrect equations.
– Changing signs arbitrarily is a common way to make
things work, but you are hiding a problem and possibly
creating more.
Lecture 1 Slide 15
Simulation Wisdom #1
Simulate devices in multiple steps using models of
increasing complexity.
Avoid the temptation to jump straight to the big, bad, and ugly 3D
simulation in all of its glorious complexity.
Model your device with slowly increasing levels of complexity.
You will get to your final answer much faster this way!
R. C. Rumpf, “Engineering the dispersion and anisotropy in periodic
electromagnetic structures,” Solid State Physics 66, 2015.
Lecture 1 Slide 16
8
1/8/2018
Simulation Wisdom #2
It must be standard practice to ensure your results are
converged.
w 2.0 m w
h 0.6 m nsup
nrib
Effective Refractive Index
a 0.25 m
ncore
nsup 1.0 a h
nrib 1.9
ncore 1.9 nsub
nsub 1.52
Grid Resolution
Lecture 1 Slide 17
Simulation Wisdom #3
Those who simulate the most,
trust the simulations the least.
Never trust your code or your results.
Benchmark. Benchmark. Benchmark.
Lecture 1 Slide 18
9
1/8/2018
Final Word of Wisdom
Do not EVER share your codes!
Only bad things will happen.
The best thing that can happen
is that you become useless.
Instead, offer to simulate devices for
them and make yourself a collaborator.
Lecture 1 Slide 19
General Concepts in
Computational EM
Lecture 1 Slide 20
10
1/8/2018
Physical Vs. Numerical Boundary Conditions
Physical Boundary Conditions
Tangential components
Physical boundary conditions refer to the are continuous
conditions that must be satisfied at the
boundary between two materials. These are
derived from the integral form of Maxwell’s
equations.
Numerical Boundary Conditions
Numerical boundary conditions refer to the what
is done at the edge of a grid or mesh and how
fields outside the grid are estimated.
Lecture 1 Slide 21
Full Vs. Sparse Matrices
Full Matrices Sparse Matrices
A A
Sparse matrices have most of their elements
equal to zero. They are often more than 99%
Full matrices have all non‐zero elements. sparse.
They tend to look banded with the largest It is most memory efficient to store only the non‐
numbers running down the main diagonal. zero elements in memory.
They tend to “banded” matrices with the largest
numbers running down the main diagonal.
Lecture 1 Slide 22
11
1/8/2018
Integral Vs. Differential Equations (1 of 2)
Integral Equations
f x, x dx g x
Integral equations calculate a quantity at a specific point using
information from the entire domain. They are usually written
around boundaries and lead to formulations with full matrices. They
do not require boundary conditions.
Differential Equations
df x
f x g x
dx
Differential equations calculate a quantity at a specific point using
only information from the local vicinity. They are usually written for
points distributed throughout a volume and lead to formulations
with sparse matrices. They require boundary conditions.
Lecture 1 Slide 23
Frequency‐Domain Vs. Time‐Domain
This is what a
frequency‐domain
code calculates.
This is what a time‐
domain code
calculates.
Frequency‐domain solutions are at a single frequency. Time‐domain solutions look
different because there are inherently a broad range of frequencies involved.
Lecture 1 Slide 24
12
1/8/2018
Definition of “Convergence”
Virtually all numerical methods have some sort of “resolution”
parameter that when taken to infinity solves Maxwell’s equations
exactly. In practice, we cannot this arbitrarily far because a computer
will run out of memory and simulations will take prohibitively long to
run.
There are no equations to calculate what “resolution” is needed to
obtain “accurate” results. Instead, the user must look for
convergence. There are, however, some good rules of thumb to
make an initial guess at resolution.
Convergence is the tendency of a calculated parameter to
asymptotically approach some fixed value as the resolution of the
model is increased. A converged solution does not imply an
accurate solution!!!
Lecture 1 Slide 25
Tips About Convergence
• Make checking for convergence a habit that you
always perform.
• When checking a parameter for convergence,
ensure that it is the only thing about the
simulation that is changing.
• Simulations do not get more “accurate” as
resolution is increased. They only get more
“converged.”
Lecture 1 Slide 26
13
1/8/2018
Conservation of Power
When electromagnetic wave is applied to a device, it can be absorbed
(i.e. converted to another form of energy), reflected and/or
transmitted. Without a nuclear reaction, nothing else can happen.
A R T 1
Reflectance, R Transmittance, T Absorptance, A
Fraction of power from Fraction of power from the Fraction of power from
the applied wave that is applied wave that is transmitted the applied wave that is
reflected from the device. through the device. absorbed by the device.
Applied Wave
Transmitted Wave
Reflected Wave Absorbed Wave
Lecture 1 Slide 27
How Do You Know if Your Model Works?
In many cases, you may not know.
1. BENCHMARK, BENCHMARK, BENCHMARK
2. Check for power conservation
3. CONVERGENCE, CONVERGENCE, CONVERGENCE
Common Sense – Check your model for simple things like conservation of power,
magnitude of the numbers, consistency with the physics, etc.
Benchmark – You can verify your code is working by modeling a device with a
known response. Does your model predict that response?
Convergence – Your models will have certain parameters that you can adjust to
improve “accuracy” usually at the cost of computer memory and run time. Keep
increasing “accuracy” until your answer does not change much any more.
When modeling a new device, benchmark your model using as similar of a device
as you can find which has a known response. Compare your experimental results to
the model. Do they agree? Reconcile any differences.
Lecture 1 Slide 28
14
1/8/2018
Tips for Matching Simulations to
Experiments
1. Model the geometry of the device as accurately as possible.
2. Incorporate accurate material properties, including dispersion.
3. Ensure the source replicates the experimental source as closely as
possible.
4. Ensure you detect power in your simulation in a way that is
consistent with laboratory experiments.
Lecture 1 Slide 29
Classification of Methods
Lecture 1 Slide 30
15
1/8/2018
Classification by Size Scale
Low Frequency Methods High Frequency Methods
0 a 0 a
Structural dimensions are on the order Structural dimensions much larger than
of the wavelength or smaller. the wavelength.
Polarization and the vector nature of Fields can be accurately treated as
the field is important. scalar quantities.
• Finite‐difference time‐domain • Ray tracing
• Finite‐difference frequency‐domain • Geometric theory of diffraction
• Finite element analysis • Physical optics
• Method of moments • Physical theory of diffraction
• Rigorous coupled‐wave analysis • Shooting and bouncing rays
• Method of lines
• Beam propagation method
• Boundary element method
• Spectral domain method
• Plane wave expansion method
Lecture 1 Slide 31
Classification by Approximations
Rigorous Methods
A method is rigorous if there exists a “resolution” parameter that when taken to
infinity, finds an exact solution to Maxwell’s equations.
• Finite‐difference time‐domain
• Finite‐difference frequency‐domain
• Finite element method
• Rigorous coupled‐wave analysis
• Method of lines
Full Wave Methods
A method is full wave if it accounts for the vector nature of the electromagnetic
field. A full wave method is not necessarily rigorous.
• Method of moments
• Boundary element method
• Beam propagation method
Scalar Methods
A method is scalar if the vector nature of the field is not accounted for.
• Ray tracing
Lecture 1 Slide 32
16
1/8/2018
Comparison of Method Types
Frequency‐Domain Time‐Domain
+ resolves sharp resonances ‐ scales at best NlogN + wideband simulations ‐ longitudinal periodicity
+ handles oblique incidence ‐ can miss sharp resonances + scales near linearly ‐ sharp resonances
+ longitudinal periodicity ‐ active & nonlinear devices + active & nonlinear devices ‐ memory requirements
+ can be very fast + easily locates resonances ‐ oblique incidence
Fully Numerical Semi‐Analytical
+ better convergence ‐ memory requirements + very fast & efficient ‐ convergence issues
+ scales better than SA ‐ long uniform sections + layered devices ‐ scales poorly
+ complex device geometry + less memory ‐ complex device geometry
Real‐Space Fourier‐Space
+ high index contrast ‐ slow for low index contrast + moderate index contrast ‐ field visualization
+ metals + periodic problems ‐ formulation difficult
+ resolving fine details + very fast and efficient ‐ resolving fine details
+ field visualization
Structured Grid Unstructured Grid
+ easy to implement ‐ less efficient + most efficient ‐ difficult to implement
+ rectangular structures ‐ curved surfaces + handles larger structures ‐ spurious solutions
+ easy for divergence free + conforms to curved surfaces
Differential Based Integral Based
+ sparse matrices ‐ volume mesh + surface mesh ‐ full matrices
+ easier to formulate ‐ spurious solutions + Very efficient for many structures ‐ more difficult to formulate
+ easier to implement ‐ more difficult to implement
Lecture 1 Slide 33
Multiphysics Simulations
A multiphysics simulation is one that accounts for multiple
simultaneous physical mechanisms at the same time.
• Electromagnetic
• Thermal
• Fluids
• Motion
• Chemical
• Acoustic
• Optical
Lecture 1 Slide 34
17
1/8/2018
Any Method Can Do Anything
Any method can be
made to do anything.
The real questions are:
• What devices and information is a
particular method best suited for?
• How much of a “force fit” is it for that
method?
Lecture 1 Slide 35
Overview of the Methods
Lecture 1 Slide 36
18
1/8/2018
Transfer Matrix Method (1 of 2)
Transfer matrices are derived that
E x ,trn E x ,2
relate the fields present at the E T3 E
interfaces between the layers. y ,trn x ,2
T3
Ex ,2 Ex ,1 Ex ,trn
E T2 E E
x ,2 x ,1 y ,trn
T2
Ex ,1 E x ,ref Ex ,2
E
E T1 E x ,2
x ,1 y ,ref
T1
Tglobal T3T2 T1
Ex ,1
E Ex ,trn Ex ,ref
x ,1 E Tglobal E
y ,trn y ,ref
Transmission through all the layers
Ex ,ref is described by multiplying all the
E
y ,ref individual transfer matrices.
Lecture 1 Slide 37
Transfer Matrix Method (2 of 2)
This method is good for…
1. Modeling transmission and reflection from layered devices.
2. Modeling layers of anisotropic materials.
Benefits Drawbacks
• Very fast and efficient • Limited number of geometries
• Rigorous
• Near 100% accuracy it can model.
• Unconditionally stable • Only handles linear,
• Robust
homogeneous and infinite
• Simple to implement
• Thickness of layers can be anything slabs.
• Able to exploit longitudinal periodicity • Cannot account for diffraction
• Easily incorporates material dispersion
effects
• Easily accounts for polarization and
angle of incidence • Inefficient for transient analysis
• Excellent for anisotropic layered
materials
Lecture 1 Slide 38
19
1/8/2018
Finite‐Difference Frequency‐Domain (1 of 2)
Space is converted to a grid and Maxwell’s
equations are written for each point using
the finite‐difference method.
y
y
Ez E z x, y 2 , z Ez x, y 2 , z
y y
This large set of equations is written in
matrix form and solved to calculate the
fields.
Ez E y
j H x
y z D Ey e z D Ez e y jμ xx h x
Ex Ez source
j H y
z x D Ez e x D Ex e z jμ yy h y
E y Ex
j H z D Ex e y D Ey e x jμ zz h z Ax b
x y
x A 1b
H z H y
y
z
j Ex D Hy h z D Hz h y jε xx e x
e x
jε e
x e y
H x H z D Hz h x D Hx H
j E y z yy y
z x
H y H x
j Ez
D Hx h y D Hy h x jε zz e z e z
x y
Lecture 1 Slide 39
Finite‐Difference Frequency‐Domain (2 of 2)
This method is good for…
1. Modeling 2D devices with high volumetric complexity.
2. Visualizing the fields.
3. Fast and easily formulation of new numerical techniques.
Benefits Drawbacks
• Accurate and robust • Does not scale well to 3D
• Highly versatile
• Simple to implement
• Difficult to incorporate
• Easily incorporates dispersion nonlinear materials
• Excellent for field visualization • Structured grid is inefficient
• Error mechanisms are well • Difficult to resolve curved
understood
• Good method for metal devices
surfaces
• Excellent for volumetrically • Slow and memory innefficient
complex devices
• Good scaling compared to other
frequency‐domain methods
Lecture 1 Slide 40
20
1/8/2018
Finite‐Difference Time‐Domain (1 of 2)
Fields are evolved by iterating Maxwell’s
equations in small time steps.
Maxwell’s equations are enforced at each point at
Reflection Plane each time step.
TF/SF Planes
Spacer Region
Unit cell of
real device
Spacer Region
Transmission Plane
Lecture 1 Slide 41
Finite‐Difference Time‐Domain (2 of 2)
This method is good for…
1. Modeling big, bad and ugly problems.
2. Modeling devices with nonlinear material properties.
3. Simulating the transient response of devices.
Benefits Drawbacks
• Excellent for large‐scale simulations. Easily
parallelized. • Tedious to incorporate dispersion
• Excellent for transient analysis.
• Accurate, robust, rigorous, and mature
• Typically has a structured grid
• Highly versatile which is less efficient and doesn’t
• Intuitive to implement
• Easily incorporates nonlinear behavior
conform well to curved surfaces
• Excellent for field visualization and learning
electromagnetics
• Difficult to resolve curved
• Error mechanisms are well understood surfaces
• Good method for metal devices
• Excellent for volumetrically complex devices • Slow for small devices
• Scales near linearly • Very inefficient for highly
• Able to simulate broad frequency response in one
simulation resonant devices
• Great for resonance “hunting”
Lecture 1 Slide 42
21
1/8/2018
Transmission‐Line Modeling Method (1 of 2)
Space is interpreted as a giant 3D circuit.
Waves propagating through space are
represented as current and voltage in
extended circuits.
Also called transmission‐line matrix
method (TLM).
Lecture 1 Slide 43
Transmission‐Line Modeling Method (2 of 2)
This method is good for…
1. Modeling big, bad and ugly problems.
2. Hybridizing models with microwave devices.
3. Representing digital waveforms.
Benefits Drawbacks
• Essentially the same benefits at • Essentially the same drawbacks
FDTD and FDFD. as FDTD and FDFD.
• Excellent for large‐scale
simulations. Easily parallelized.
• Excellent for transient analysis.
• No convergence criteria.
• Inherently stable.
• Time‐ and frequency‐domain
implementations exist.
• Excellent fit with network theory
in microwave engineering.
Lecture 1 Slide 44
22
1/8/2018
Beam Propagation Method (1 of 2)
The beam propagation method (BPM) is a simple method to simulate
“forward” propagation through a device. It calculates the field one
plane at a time so it does not need to solve the entire solution space
at once.
A i μ xx ,i D Hx μ zz1,i D Ex μ xx ,i ε yy ,i neff
2
I
1
j z j z i
eiy1 I A i 1 I Ai e y
4 neff 4 neff
Lecture 1 Slide 45
Beam Propagation Method (2 of 2)
This method is good for…
1. Nonlinear optical devices.
2. Devices where reflections and abrupt changes in the
field are negligible (i.e. forward only devices)
Benefits Drawbacks
• Simple to formulate and implement • Not a rigorous method
(FFT‐BPM is easiest)
• Numerically efficient for faster • Limited in the physics it can
simulations handle
• Well established for nonlinear • Typically uses paraxial
materials (unique for frequency‐
domain method). approximation
• Easily incorporates dispersion • Typically neglects backward
• Excellent for field visualization reflections
• Error mechanisms are well
understood • FFT‐BPM is slower, less stable,
• Well suited for waveguide circuit and less versatile than FDM‐
simulation BPM
Lecture 1 Slide 46
23
1/8/2018
Method of Lines (1 of 2)
reflected
source
x
y BCs
The method of lines is a semi‐analytical
method.
Modes are computed in the transverse plane
for each layer and propagated analytically in
BCs
the z-direction.
Boundary conditions are used to matched
the fields at the interfaces between layers.
BCs
Transmission through the entire stack of
layers is then known and transmitted and
reflected fields can be computed.
transmitted
z BCs
Lecture 1 Slide 47
Method of Lines (2 of 2)
This method is good for…
1. Long devices.
2. Long devices with metals.
Benefits Drawbacks
• Excellent for longitudinally • Scales very poorly in the
periodic devices transverse direction
• Rigorous method • Cumbersome method for field
• Excellent for devices with high visualization
index contrast and metals
• Less efficient than RCWA for
• Good for resonant structures dielectric structures.
• Less numerical dispersion than
fully numerical methods • Rarely used in 3D analysis, but
this may change with more
• Easier field visualization than
RCWA modern computers
Lecture 1 Slide 48
24
1/8/2018
Rigorous Coupled‐Wave Analysis (1 of 2)
Field in each layer is
represented as a set of
plane waves at
different angles.
Plane waves describe
propagation through
each layer.
Layers are connected
by the boundary
conditions.
Lecture 1 Slide 49
Rigorous Coupled‐Wave Analysis (2 of 2)
This method is good for…
1. Modeling diffraction from periodic dielectric structures
2. Periodic devices with longitudinal periodicity
Benefits Drawbacks
• Excellent for modeling diffraction from • Scales poorly in transverse
periodic dielectric structures.
• Extremely fast and efficient for all‐dielectric dimensions.
structures with low to moderate index
contrast • Less efficient for high dielectric
• Accurate and robust contrast and metals due to
• Unconditionally stable Gibb’s phenomenon.
• Thickness of layers can be anything without
numerical cost • Poor method for finite
• Excellent for longitudinally periodic structures.
structures.
• Excellent for structures large in the • Slow convergence if fast
longitudinal direction.
• Easily incorporates polarization and angle of
Fourier factorization is not
incidence. used.
Lecture 1 Slide 50
25
1/8/2018
Plane Wave Expansion Method (1 of 2)
The plane wave expansion method
(PWEM) calculates modes that exist in
an infinitely periodic lattice. It
represents the field in Fourier‐space as
the sum of a large set of plane waves at
different angles.
Lecture 1 Slide 51
Plane Wave Expansion Method (2 of 2)
This method is good for…
1. Analyzing unit cells
2. Calculating photonic band diagrams and effective material properties.
Benefits Drawbacks
• Excellent for all‐dielectric unit • Scales poorly.
cells • Weak method for high
• Fast even for 3D dielectric contrast and metals.
• Accurate and robust • Limited to modal analysis.
• Rigorous method • Cannot model scattering.
• Cannot incorporate dispersion.
Lecture 1 Slide 52
26
1/8/2018
Slice Absorption Method (1 of 2)
Virtually any method that converts Maxwell’s equations to a matrix equation can
order the matrix to give it the following block tridiagonal form.
This allows the problem to be solved one slice at a time.
Lecture 1 Slide 53
Slice Absorption Method (2 of 2)
This method is good for…
1. Modeling structures with high volumetric complexity
2. Modeling finite size structures (i.e. not infinitely
periodic)
Benefits Drawbacks
• Excellent for modeling devices • New method and not well
with high volumetric complexity
understood.
• Easily incorporates dispersion
• Easily incorporate polarization
and oblique incidence
• Potential for transverse devices
• Excellent for finite size devices
• Excellent framework to hybridize
different methods.
• Transverse sources
• Stacking in three dimensions.
Lecture 1 Slide 54
27
1/8/2018
Finite Element Method (1 of 2)
Step 1: Describe Structure Step 2: Mesh Structure
This is a VERY
1 1.0
important and
involved step.
r 1.50 2 2.5
Step 3: Build Global Matrix Step 4: Solve Matrix Equation
Incorporate a
Iterate through source.
each element to
populate the
Ax b
global matrix. Calculate field.
Ax 0 x A 1b
Lecture 1 Slide 55
Finite Element Method (2 of 2)
This method is good for…
1. Modeling volumetrically complex structures in the frequency‐
domain.
Benefits Drawbacks
• Very mature method • Tedious to implement
• Excellent representation of • Requires a meshing step
curved surfaces
• Unstructured grid is highly
efficient
• Unconditionally stable
• Scaling improved with domain
decomposition
Lecture 1 Slide 56
28
1/8/2018
Method of Moments (1 of 2)
Lf g
f an v n a Lv g Galerkin Method Integral Equation
n n
n n
a n
n v m , Lv n v m , g
• Converts a linear •Usually uses PEC approximation
v1 , Lv1 v1 , Lv 2
a v , Lg
1 1
equation to a matrix •Usually based on current
a1 v 2 , Lg
v 2 , Lv1 v 2 , Lv 2 equation
j L2
2 e jkr
aN v N , Lg
Ezinc I z z k 2 2 dz
L 2
z 4 r
The Method of Moments
z11 z12 z13 z14 z15 z16 z17 i1 v1
i1 z
i2 i3 21 z22 z23 z24 z25 z26 z27 i2 v2
i4 i5 i6 z31 z32 z33 z34 z35 z36 z37 i3 v3
i7
v1 v2 z41 z42 z43 z44 z45 z46 z47 i4 v4
v3 z51
v4 v5 z52 z53 z54 z55 z56 z57 i5 v5
v6 v7
z61 z62 z63 z64 z65 z66 z67 i6 v6
z z72 z73 z74 z75 z76 z77 i7 v7
71
Lecture 1 Slide 57
Method of Moments (2 of 2)
This method is good for…
1. Modeling metallic devices at radio frequencies
2. Modeling large‐scale metallic structures at radio frequencies
Benefits Drawbacks
• Extremely efficient analysis of • Not a rigorous method
metallic devices
• Poor method for incorporating
• Full wave
• Very fast dispersion and dielectrics
• Excellent scaling using the fast • Long a tedious formulation
multipole method
• Inefficient for volumetrically
• No boundary conditions
complex structures
• Simple implementation
• Mature method with lots of
literature
• Can by hybridized with FEM
Lecture 1 Slide 58
29
1/8/2018
Boundary Element Method (1 of 2)
The boundary element method (BEM) is also called the Method of
Moments, but is applied to 2D elements. The most famous element is
the Rao‐Wilton‐Glisson (RWG) edge element.
S. M. Rao, D. R. Wilton, A. W. Glisson,
“Electromagnetic Scattering by Surfaces of
Arbitrary Shape,” IEEE Trans. Antennas and
Propagation, vol. AP‐30, no. 3, pp. 409‐418,
1982.
Governing equation exists only at the boundary of a device so many
fewer elements are needed.
5000 elements 400 elements
Lecture 1 Slide 59
Boundary Element Method (2 of 2)
This method is good for…
1. Modeling large devices with simple geometries.
2. Modeling scattering from homogeneous blobs.
Benefits Drawbacks
• Highly efficient when surface • Tedious to implement
to volume ratio is low • Requires a meshing step
• Excellent representation of
curved surfaces • Not usually a rigorous method
• Unstructured grid is highly • Inefficient for volumetrically
efficient complex geometries
• Unconditionally stable
• Can be hybridized with FEM
• Domain can extend to infinity
• Simpler meshing than FEM
Lecture 1 Slide 60
30
1/8/2018
Discontinuous Galerkin Method (1 of 2)
The discontinuous Galerkin method (DGM) combines features of the
finite element and finite‐volume framework to solve differential
equations.
Lecture 1 Slide 61
Discontinuous Galerkin Method (2 of 2)
This method is good for…
1. Solving very complex equations.
2. Modeling very electrically large structures.
3. Time‐domain finite‐element method.
Benefits Drawbacks
• Mesh elements can have any • Tedious to implement
arbitrary shape. • Requires a meshing step
• Fields may be collocated
instead of staggered. • Not usually a rigorous method
• Inherently a parallel method. • Inefficient for volumetrically
• Easily extended to higher‐order complex geometries
of accuracy.
• Allows explicit time‐stepping
• Low memory consumption (no
large matrices)
Lecture 1 Slide 62
31