Simulation of Electromagnetic Fields:

The Finite-Difference Time-Domain (FDTD)

Method and Its Applications
Veysel Demir, Ph.D.
demir@ceet.niu.edu

Department of Electrical Engineering, Northern Illinois University, DeKalb, IL 60115

1/34

Veysel Demir
Bachelor of Science, Electrical and Electronics Engineering, Middle East Technical
University, Ankara, Turkey, 1997.
System Analyst and Programmer, Pamukbank, Software Development Department,
Istanbul, Turkey, July 1997 August 2000.
Master of Science, Electrical Engineering, Syracuse University, Syracuse, NY, 2002.
Doctor of Philosophy, Electrical Engineering, Syracuse University, Syracuse, NY, 2004.
Research Assistant, Sonnet Software, Inc. Liverpool, NY, August 2000 July 2004.
Visiting research scholar, University of Mississippi, Electrical Engineering Department,
University, MS, July 2004 Present.
Assistant Professor, Department of Electrical Engineering, Northern Illinois
University, DeKalb,IL, August 2007 present

2/34

Computational Electromagnetics
Maxwells equations can be given in differential or integral form

Finite-difference
time-domain
(FDTD)

Transmission line
matrix (TLM)

Finite element
method (FEM)

Finite-difference
frequency-domain
(FDFD)

Differential equation methods

Method of Moments
(MoM)

Fast multipole
method (FMM)

Integral equation
methods
3/34

Computational Electromagnetics
Maxwells equations can be given in time domain or frequency domain
Time-domain methods
Finite-difference
time-domain
(FDTD)

Transmission line
matrix (TLM)

Finite element
method (FEM)

Method of Moments
(MoM)

Finite-difference
frequency-domain
(FDFD)

Fast multipole
method (FMM)

Frequency domain methods

4/34

Commercial software packages

Commercial software packages
Finite element method (FEM)

Method of Moments (MoM)

ADS Momentum

HFSS

Transmission line matrix (TLM)

Finite-difference time-domain (FDTD)

CST Microstripes

5/34

The Finite-Difference Time-Domain Method

6/34

FDTD Books

7/34

Yearly FDTD Publications

The most popular method in computational electromagnetics

Source: Allen Taflove, A Perspective on the 40-Year History of FDTD Computational Electrodynamics,
Applied Computational Electromagnetics Society (ACES) Conference, Miami, Florida, March 15, 2006.
Can be found at http://www.ece.northwestern.edu/ecefaculty/Allen1.html

8/34

Maxwells Equations
The basic set of equations describing the electromagnetic world
Shows that lightis anelectromagnetic wave.

Gausss law
Gausss law for magnetism

Faradays law

Amperes law

Constitutive relations

D= v
B= 0
B
E=
t
D
H= J+
t
D = E ,and B = H
9/34

FDTD Overview Finite Differences

Represent the derivatives in Maxwells curl equations by finite differences
We use the second-order accurate central difference formula

df ( x)
f ( x + x) f ( x x)
= f ( x)
dx
2 x

10/34

FDTD Overview Cells

A three-dimensional problem space is composed of cells

11/34

FDTD Overview The Yee Cell

The FDTD (Finite Difference Time Domain) algorithm was first established
by Yee as a three dimensional solution of Maxwell's curl equations.

K. Yee, IEEE Transactions on Antennas and Propagation, May 1966.

12/34

FDTD Overview Material grid

A three-dimensional problem space is composed of cells

13/34

FDTD Overview Updating Equations

Three scalar equations can be obtained from one vector curl equation.

= H
t

H x Ey
x
=

t
z
H y Ez
y
=

t
x
H z Ex
z
=

t
y

Ez
y
Ex
z
Ey

Ex H z
x
=

t
y
Ey H x
y
=

t
z
Ez H y
z
=

t
x

Hy
z
Hz
x
Hx
y

H
= E
t

x
14/34

FDTD Overview Updating Equations

Represent derivatives by finite-differences

Ex H z H y
x
=

t
y
z

Exn + 1 (i, j , k ) Exn (i, j , k )

x (i, j , k )
=
t
n + 0.5
n + 0.5
H zn + 0.5 (i, j , k ) H zn + 0.5 (i, j 1, k ) H y (i, j , k ) H y (i, j, k 1)

y
z

15/34

FDTD Overview Updating Equations

Represent derivatives by finite-differences

H x E y Ez
x
=

t
z
y

H xn + 0.5 (i, j , k ) H xn 0.5 (i, j, k )

x (i, j , k )
=
t
E yn (i, j , k + 1) E yn + 0.5 (i, j , k ) Ezn (i, j + 1, k ) Ezn (i, j , k )

z
y
16/34

FDTD Overview Updating Equations

Express the future components in terms of the past components

H zn + 0.5 (i, j , k ) H zn + 0.5 (i, j 1, k )

Exn + 1 (i, j , k ) = Exn (i, j , k )

x (i, j , k ) H yn + 0.5 (i, j , k ) H yn + 0.5 (i, j , k 1)

E yn (i, j , k + 1) E yn + 0.5 (i, j, k )

H xn + 0.5 (i, j , k ) = H xn 0.5 (i, j , k ) +

n
n
x (i, j , k ) Ez (i, j + 1, k ) Ez (i, j , k )

17/34

FDTD Overview Leap-frog Algorithm

Exercise 1D
18/34

Absorbing Boundary Conditions

The three-dimensional problem space is truncated by absorbing
boundaries
Most popular absorbing boundary is Perfectly Matched layers (PML)

Exercise 2D PML
19/34

Active and Passive Lumped Elements

Active and passive lumped elements can be modeled in FDTD

E
H=
+ E+ J
t

Voltage source

Current source
20/34

Active and Passive Lumped Elements

Resistor

Capacitor

Inductor

Diode
21/34

Active and Passive Lumped Elements

A diode circuit

22/34

Transformation from Time-Domain to Frequency-Domain

Results can be obtained for frequency domain using Fourier Transform
A low-pass filter

S11

S22

Exercise 2D object
23/34

Near-Field to Far-field Transformations

An inverted-F antenna

24

24/34

Modeling fine geometries

It is possible to model fine structures using appropriate formulations

A wire loop antenna

25/34

Incident plane wave

26/34

Scattering Problems

( H inc + H scat )

= ( Einc + Escat )
t

H inc

= 0 Einc
t

A dielectric sphere
Exercise 3D PML
27/34

Scattering from a Dielectric Sphere

28/34

Earth / Ionosphere Models in Geophysics

Snapshots of FDTD-Computed Global Propagation of ELF
Electromagnetic Pulse Generated by Vertical Lightning Strike off South
America Coast

Source: Allen Taflove, A Perspective on the 40-Year History of FDTD Computational Electrodynamics,
Applied Computational Electromagnetics Society (ACES) Conference, Miami, Florida, March 15, 2006.
Can be found at http://www.ece.northwestern.edu/ecefaculty/Allen1.html

29/34

Wireless Personal Communications Devices

Source: Allen Taflove, A Perspective on the 40-Year History of FDTD Computational Electrodynamics,
Applied Computational Electromagnetics Society (ACES) Conference, Miami, Florida, March 15, 2006.
Can be found at http://www.ece.northwestern.edu/ecefaculty/Allen1.html

30/34

Phantom Head Validation at 1.8 GHz

Source: Allen Taflove, A Perspective on the 40-Year History of FDTD Computational Electrodynamics,
Applied Computational Electromagnetics Society (ACES) Conference, Miami, Florida, March 15, 2006.
Can be found at http://www.ece.northwestern.edu/ecefaculty/Allen1.html

31/34

Ultrawideband Microwave Detection of Early-Stage Breast Cancer

Source: Allen Taflove, A Perspective on the 40-Year History of FDTD Computational Electrodynamics,
Applied Computational Electromagnetics Society (ACES) Conference, Miami, Florida, March 15, 2006.
Can be found at http://www.ece.northwestern.edu/ecefaculty/Allen1.html

32/34

Focusing Plasmonic Lens

Source: Allen Taflove, A Perspective on the 40-Year History of FDTD Computational Electrodynamics,
Applied Computational Electromagnetics Society (ACES) Conference, Miami, Florida, March 15, 2006.
Can be found at http://www.ece.northwestern.edu/ecefaculty/Allen1.html

33/34

Thank You

Exercise 2D PEC
34/34