0% found this document useful (0 votes)
133 views

Manual FDTD Basics

fdtd basic

Uploaded by

ck maiti
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
133 views

Manual FDTD Basics

fdtd basic

Uploaded by

ck maiti
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 14

Photonic Software U a

FDTD Basics

<< Previous Next >>

OptiFDTD is a powerful, highly integrated, user-friendly software that allows computer


aided design and simulation of advanced passive photonic components.

The OptiFDTD software package is based on the nite-di erence time-domain


(FDTD) method. The FDTD method has been established as a powerful engineering
tool for integrated and di ractive optics device simulations. This is due to its unique
combination of features, such as the ability to model light propagation, scattering and
di raction, and re ection and polarization e ects. It can also model material
anisotropy and dispersion without any pre-assumption of eld behavior such as the
slowly varying amplitude approximation. The method allows for the e ective and
powerful simulation and analysis of sub-micron devices with very ne structural
details. A sub-micron scale implies a high degree of light con nement and
correspondingly, the large refractive index di erence of the materials (mostly
semiconductors) to be used in a typical device design.

2D FDTD Equations
The FDTD approach is based on a direct numerical solution of the time-dependent
Maxwell’s curl equations. The rst version of OptiFDTD is in 2D. The photonic device
is laid out in the X-Z plane. The propagation is along Z. The Y-direction is assumed to
be in nite. This assumption removes all the ∂/ ∂y derivatives from Maxwell’s
equations and splits them into two (TE and TM) independent sets of equations.

The 2D computational domain is shown in Figure 1. The space steps in the X and Z
directions are Δx and Δy, respectively. Each mesh point is associated with a speci c type of
material and contains information about its properties such as refractive index, and dispersion
parameters.
Privacy - Terms
/
Figure 1: Numerical representation of the 2D computational domain

TE waves
In the 2D TE case (Hx, Ey, Hz – nonzero components, propagation along Z, transverse
eld variations along X) in lossless media, Maxwell’s equations take the following
form:

where ε = ε0εr is the dielectric permittivity and is the magnetic permeability of


the vacuum. The refractive index is n = √εr.

Each eld is represented by a 2D array — Ey(i,k), Hx(i,k) and Hz(i,k) — corresponding


to the 2D mesh grid given in Figure 1. The indices i and k account for the number of
space steps in the X and Z direction, respectively. In the case of TE, the location of
the elds in the mesh is shown in Figure 2.

Privacy - Terms
/
Figure 2: Location of the TE elds in the computational domain

The TE elds stencil can be explained as follows. The Ey eld locations coincide with
the mesh nodes given in Figure 1. In Figure 2, the solid lines represent the mesh given
in Figure 1. The Ey eld is considered to be the center of the FDTD space cell. The
dashed lines form the FDTD cells. The magnetic elds Hx and Hz are associated with
cell edges. The locations of the electric elds are associated with integer values of the
indices i and k. The Hx eld is associated with integer i and (k + 0.5) indices. The Hz
eld is associated with (i + 0.5) and integer k indices. The numerical analog in
Equation 1 can be derived from the following relation:

Privacy - Terms
/
The total set of numerical Equation 1 takes the form:

The superscript n labels the time steps while the indices i and k label the space steps and
Δx and Δz along the x and z directions, respectively. This is the so-called Yee’s numerical scheme
applied to the 2D TE case. It uses central di erence approximations for the numerical
derivatives in space and time, both having second order accuracy. The sampling in space is on a
sub-wavelength scale. Typically, 10 to 20 steps per wavelength are needed. The sampling in
time is selected to ensure numerical stability of the algorithm. The time step is determined by
the Courant limit:

TM waves
In the 2D TM case (Ex, Hy, Ez — nonzero components, propagation along Z,
transverse eld variations along X) in lossless media, Maxwell’s equations take the
following form:

Privacy - Terms
/
The location of the TM elds in the computational domain follows the same
philosophy and is shown in Figure 3.

Figure 3: Location of the TM elds in the computational domain

Now, the electric eld components Ex and Ez are associated with the cell edges, while
the magnetic eld Hy is located at the cell center. The TM algorithm can be presented
in a way similar to Equation 3.

3D FDTD Equations
In 3D simulations, the simulation domain is a cubic box, the space steps are Dx, Dy, and Dz in x,
y, and z directions respectively.

Each eld components is presented by a 3D array —Ex(i,j,k), Ey(i,j,k), Ez(i,j,k),


Hx(i,j,k), Hy(i,j,k), Hz(i,j,k). The eld components position in Yee’s Cell are shown
in Figure 4. These placements and the notation show that the E and H components are
interleaved at intervals of 1 ⁄ 2Dh in space and 1 ⁄ 2Dt for the purpose of implementing a
leapfrog algorithm.

Privacy - Terms
/
Figure 4: Displacement of the electric and magnetic eld vector components about a
cubic unit cell of the Yee space lattice

In general, the time domain Maxwell’s equations are given in di erential form by

When the second-order nite di erence method, eld notations, and eld
displacements are applied to the above Maxwell’s equation for time domain and
space domain derivatives, the 3D-FDTD formulas can be written as:

Privacy - Terms
/
Privacy - Terms
/
Space Step and Time Step
The fundamental constraint of FDTD method is the step size both for the time and space. Space
and time steps relate to the accuracy, numerical dispersion, and the stability of the FDTD
method. Many references and books have discussed these problems. In general, to keep the
results as accurate as possible, with a low numerical dispersion, the mesh size often quoted is
“10 cells per wavelength”, meaning that the side of each cell should be 1/10λ or less at the
highest frequency (shortest wavelength).

Please note that FDTD is a volumetric computational method, so that if some portion
of the computational space is lled with penetrable material, you must use the
wavelength in the material to determine the maximum cell size.

The following equation is for the suggested mesh size:

where nmax is the maximum refractive index value in the computational domain.

Once the cell size is determined, the maximum size for the time step Δt immediately
follows the Courant-Friedrichs-Levy (CFL) condition.

For 3D FDTD simulation, the CFL condition is:

Privacy - Terms
/
where v is the speed of the light in medium.

OptiFDTD Simulation Procedures


The following is the ow chart for the FDTD simulation in OptiFDTD. It also details the
work ow in OptiFDTD.

Figure 5: FDTD Simulation Flow Chart in OptiFDTD Privacy - Terms


/
Output data
The elds propagated by the FDTD algorithm are the time domain elds. At each
location of the computational domain they have a form similar to that given in Equation
15:

where B is the amplitude of the eld at that particular location, G is the wave pro le,
and φi is the corresponding phase. However, the values of B and φi are not
accessible from the time domain eld values.

In order to get the full amplitude/phase wave information, we need the stationary
complex elds that correspond to the waveform Equation 15. The complex elds are
the source of all useful information, such as output and re ected powers, overlap
integrals with modal elds, etc. Those complex elds are calculated by a run time
Fourier transform performed in the last time period of the simulation. The nal
complex elds can be visualized at speci c Output Planes located properly in the
computational domain.

Figure 6: Output Planes

OptiFDTD uses TF/SF (total eld/scattering eld) technique for the incident plane.
User can specify the incident wave direction. Behind the incident plane, it is the pure
re ection eld region, when the observation detectors are placed in this region, the Privacy - Terms
/
re ection function can be calculated. When the Observation detectors are placed in
the eld transmission region, the transmission function can be calculated.

<< Previous Next >>

Categories

Applications

New Features

References

Training

Videos

OptiFDTD Manuals

Background and Tutorials ▸



Applications

Overview ▸

Technical Support

Installing OptiFDTD

FDTD Basics


Material Models

Boundary Conditions ▾

Input Wave ▾ Privacy - Terms


/
2D FDTD Band Solver ▾
Post-Simulation Data Analysis ▾
Plane Wave Expansion (PWE) Method ▾
Power Transmittance Calculation with VB Scripting ▾
Layout Designer ▾
32-bit vs 64-bit ▾


Lesson 1 - Getting Started

Lesson 2 - Input Wave Setup ▾

Lesson 3 - Photonic Crystal and Photonic Band Gap ▾



Lesson 4 - Multiple Resonant Lorentz Dispersive Material

Lesson 5 - Drude Model for Noble Metal and Surface Plasma ▾

Lesson 6 - Second Order Nonlinearity ▾

Lesson 7 - Four Wave Mixing ▾

Lesson 8 - Plane Wave Simulation ▾

Lesson 9 - FDTD Band Solver ▾

Lesson 10 - Lorentz-Drude Model for Metal and Surface

Plasma ▾
Lesson 11 - Analyzing 1D Photonic Crystals (Bragg Gratings) ▾
Lesson 12 - Analyzing 2D Photonic Crystals ▾
Lesson 13 - Analyzing 3D Photonic Crystals ▾

Lesson 14 - Analyzing 2D Defects in Photonic Crystals ▾


Lesson 15 - Grating Simulation ▾
Lesson 16 - Calculating Power Transmittance and Re ection

using VB Script ▾
Lesson 17 - Parameter Sweep Simulation ▾
Lesson 18 - 64-bit 3D Simulator▾
Privacy - Terms

Heating Absorption ▾
/
Lesson 19
Lesson 19 - Heating Absorption ▾
Lesson 20 - 2D TF/SF Simulation and RCS Detection ▾
Lesson 21 - 3D Surface Plasmon ▾
Lesson 22 - 3D Layout using Non-Uniform Mesh ▾

Simulating the Future of Networks with Optiwave


October 21, 2019
Optiwave Systems is an Ottawa based software company,
boasting a robust variety of photonic design tools, which it
provides to hundreds of leading high-technology institutions.
Founded in 1994 and incorporated in 2005, the company has a
well-established community of over one thousand users in over
seventy countries. Optiwave has come to CENGN in order to
validate…

More Info

Evaluations
Get access to all our software tools instantly! No need to speak
with a sales representative.

Evaluate
Privacy - Terms
/
About Us © 2019 Optiwave Systems Inc. Privacy Policy Terms of Service

Contact Sales: 1-866-576-6784 (toll free) or 1-613-224-4700

Privacy - Terms
/

You might also like