Ray Optics Module

User’s Guide
Ray Optics Module User’s Guide
© 1998–2022 COMSOL
Protected by patents listed on www.comsol.com/patents, or see Help>About COMSOL Multiphysics on the
File menu in the COMSOL Desktop for a less detailed lists of U.S. Patents that may apply. Patents pending.
This Documentation and the Programs described herein are furnished under the COMSOL Software License
Agreement (www.comsol.com/sla) and may be used or copied only under the terms of the license
agreement.
COMSOL, the COMSOL logo, COMSOL Multiphysics, COMSOL Desktop, COMSOL Compiler,
COMSOL Server, and LiveLink are either registered trademarks or trademarks of COMSOL AB. All other
trademarks are the property of their respective owners, and COMSOL AB and its subsidiaries and products
are not affiliated with, endorsed by, sponsored by, or supported by those trademark owners. For a list of such
trademark owners, see www.comsol.com/trademarks.
Version: COMSOL 6.1

Contact Information
Visit the Contact COMSOL page at www.comsol.com/contact to submit general inquiries
or search for an address and phone number. You can also visit the Worldwide Sales Offices
page at www.comsol.com/contact/offices for address and contact information.

If you need to contact Support, an online request form is located on the COMSOL Access
page at www.comsol.com/support/case. Other useful links include:

• Support Center: www.comsol.com/support

• Product Download: www.comsol.com/product-download
• Product Updates: www.comsol.com/support/updates
• COMSOL Blog: www.comsol.com/blogs
• Discussion Forum: www.comsol.com/forum
• Events: www.comsol.com/events
• COMSOL Video Gallery: www.comsol.com/videos
• Support Knowledge Base: www.comsol.com/support/knowledgebase

Part number: CM024201

C o n t e n t s

Chapter 1: Introduction

About the Ray Optics Module 10

The Ray Optics Module Physics Interface Guide . . . . . . . . . . . 10
Common Physics Interface and Feature Settings and Nodes . . . . . . 11
Where Do I Access the Documentation and Application Libraries? . . . . 11

Overview of the User’s Guide 16

Chapter 2: Ray Optics Modeling

Essentials of Ray Tracing 18

The Ray Tracing Algorithm . . . . . . . . . . . . . . . . . . . 18
Basic Requirements of a Geometrical Optics Model . . . . . . . . . 19

Geometry and Meshing 21

Creating a Geometry . . . . . . . . . . . . . . . . . . . . . 21
Domain Selection . . . . . . . . . . . . . . . . . . . . . . 23
Meshing and Discretization Error. . . . . . . . . . . . . . . . . 23
Meshing Guidelines . . . . . . . . . . . . . . . . . . . . . . 25

Domain Properties 27
Refractive Index . . . . . . . . . . . . . . . . . . . . . . . 27
Scattering and Absorption . . . . . . . . . . . . . . . . . . . 28

Boundary Conditions 29
Reflection and Refraction. . . . . . . . . . . . . . . . . . . . 29
Primary and Secondary Ray Releases . . . . . . . . . . . . . . . 30
Wall Conditions . . . . . . . . . . . . . . . . . . . . . . . 32
Special Boundary Conditions . . . . . . . . . . . . . . . . . . 33

CONTENTS |3
 Ray Release Features 34
Grid-Based Release . . . . . . . . . . . . . . . . . . . . . . 34
Release from Domains, Boundaries, Edges, or Points . . . . . . . . . 34
Specialized Release Features . . . . . . . . . . . . . . . . . . 35
Release Features for Multiscale Modeling . . . . . . . . . . . . . . 35

Additional Variables Solved For 36

Modeling Polychromatic Radiation . . . . . . . . . . . . . . . . 36
Intensity, Polarization, and Power . . . . . . . . . . . . . . . . 37
Other Dependent Variables. . . . . . . . . . . . . . . . . . . 39
Order of Initialization of Auxiliary Dependent Variables . . . . . . . . 40

Study Types 41
Ray Tracing Study Step . . . . . . . . . . . . . . . . . . . . 41
Bidirectionally Coupled Ray Tracing Study Step . . . . . . . . . . . 43

Results Analysis and Visualization 45

Ray Trajectories Plot . . . . . . . . . . . . . . . . . . . . . 45
Ray Plot . . . . . . . . . . . . . . . . . . . . . . . . . . 46
Interference Pattern Plot . . . . . . . . . . . . . . . . . . . . 47
Spot Diagram Plot . . . . . . . . . . . . . . . . . . . . . . 48
Optical Aberration Plot and Aberration Evaluation . . . . . . . . . . 51
Plotting Refractive Index and Abbe Number . . . . . . . . . . . . 52

Variables and Nonlocal Couplings 57

Ray Statistics . . . . . . . . . . . . . . . . . . . . . . . . 57
Global Variables . . . . . . . . . . . . . . . . . . . . . . . 58
Variables for Average Ray Position . . . . . . . . . . . . . . . . 59
Using Ray Detectors . . . . . . . . . . . . . . . . . . . . . 59
Nonlocal Couplings . . . . . . . . . . . . . . . . . . . . . . 60

4 | CONTENTS
 Material Libraries 62

Chapter 3: Ray Optics Interfaces

The Geometrical Optics Interface 68

Geometrical Optics Physics Interface Settings . . . . . . . . . . . . 68
List of Geometrical Optics Interface Physics Features . . . . . . . . . 77
Ray Properties . . . . . . . . . . . . . . . . . . . . . . . . 78
Medium Properties . . . . . . . . . . . . . . . . . . . . . . 79
Material Discontinuity . . . . . . . . . . . . . . . . . . . . . 84
Thin Dielectric Film . . . . . . . . . . . . . . . . . . . . . . 89
Wall . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
Mirror . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
Ideal Lens . . . . . . . . . . . . . . . . . . . . . . . . . 95
Axial Symmetry . . . . . . . . . . . . . . . . . . . . . . . 96
Release . . . . . . . . . . . . . . . . . . . . . . . . . . 97
Release from Boundary . . . . . . . . . . . . . . . . . . . 106
Release from Symmetry Axis . . . . . . . . . . . . . . . . . 110
Release from Edge . . . . . . . . . . . . . . . . . . . . . 111
Release from Point . . . . . . . . . . . . . . . . . . . . . 112
Release from Point on Axis . . . . . . . . . . . . . . . . . . 112
Release from Grid . . . . . . . . . . . . . . . . . . . . . 112
Release from Grid on Axis . . . . . . . . . . . . . . . . . . 116
Release from Data File. . . . . . . . . . . . . . . . . . . . 116
Photometric Data Import . . . . . . . . . . . . . . . . . . 119
Gaussian Beam. . . . . . . . . . . . . . . . . . . . . . . 120
Illuminated Surface . . . . . . . . . . . . . . . . . . . . . 124
Solar Radiation . . . . . . . . . . . . . . . . . . . . . . 126
Blackbody Radiation. . . . . . . . . . . . . . . . . . . . . 129
Ray Termination . . . . . . . . . . . . . . . . . . . . . . 130
Scattering Domain . . . . . . . . . . . . . . . . . . . . . 132
Scattering Boundary. . . . . . . . . . . . . . . . . . . . . 134
Release from Electric Field . . . . . . . . . . . . . . . . . . 135
Release from Far-Field Radiation Pattern . . . . . . . . . . . . . 137
Grating . . . . . . . . . . . . . . . . . . . . . . . . . 140
Diffraction Order (Grating) . . . . . . . . . . . . . . . . . . 145

CONTENTS |5
 Cross Grating . . . . . . . . . . . . . . . . . . . . . . . 146
Diffraction Order (Cross Grating) . . . . . . . . . . . . . . . 148
Linear Polarizer . . . . . . . . . . . . . . . . . . . . . . 149
Ideal Depolarizer . . . . . . . . . . . . . . . . . . . . . . 149
Linear Wave Retarder . . . . . . . . . . . . . . . . . . . . 150
Circular Wave Retarder . . . . . . . . . . . . . . . . . . . 151
Mueller Matrix . . . . . . . . . . . . . . . . . . . . . . . 151
Fluence Rate Calculation . . . . . . . . . . . . . . . . . . . 152
Deposited Ray Power (Domain) . . . . . . . . . . . . . . . . 152
Deposited Ray Power (Boundary) . . . . . . . . . . . . . . . 153
Accumulator (Domain) . . . . . . . . . . . . . . . . . . . 153
Accumulator (Boundary) . . . . . . . . . . . . . . . . . . . 154
Nonlocal Accumulator. . . . . . . . . . . . . . . . . . . . 156
Auxiliary Dependent Variable . . . . . . . . . . . . . . . . . 157
Ray Continuity. . . . . . . . . . . . . . . . . . . . . . . 157
Ray Detector . . . . . . . . . . . . . . . . . . . . . . . 158

Theory for the Geometrical Optics Interface 159

Note on Sign Conventions . . . . . . . . . . . . . . . . . . 160
The Electromagnetic Wave Equation . . . . . . . . . . . . . . 161
Introduction to Geometrical Optics . . . . . . . . . . . . . . 165
Optical Dispersion Models . . . . . . . . . . . . . . . . . . 166
Thermo-Optic Dispersion Models . . . . . . . . . . . . . . . 168
Refractive Index at Common Spectral Lines. . . . . . . . . . . . 169
Optical Attenuation Models. . . . . . . . . . . . . . . . . . 169
Initial Conditions: Position . . . . . . . . . . . . . . . . . . 172
Initial Conditions: Direction. . . . . . . . . . . . . . . . . . 173
Initial Conditions: Wavelength or Frequency . . . . . . . . . . . 175
Blackbody Radiation Theory . . . . . . . . . . . . . . . . . 177
Gaussian Beam Theory . . . . . . . . . . . . . . . . . . . 184
Material Discontinuity Theory . . . . . . . . . . . . . . . . . 191
The Fresnel Equations . . . . . . . . . . . . . . . . . . . . 192
Intensity, Wavefront Curvature, and Polarization. . . . . . . . . . 193
Wavefront Curvature Calculation in Graded Media . . . . . . . . . 202
Refraction in Strongly Absorbing Media . . . . . . . . . . . . . 207
Attenuation Within Domains . . . . . . . . . . . . . . . . . 209
Fluence Rate Calculation Theory . . . . . . . . . . . . . . . . 211
Ray Termination Theory . . . . . . . . . . . . . . . . . . . 212

6 | CONTENTS
 Illuminated Surface Theory . . . . . . . . . . . . . . . . . . 214
Theory of Mueller Matrices and Optical Components . . . . . . . . 216
Thin Dielectric Film Theory. . . . . . . . . . . . . . . . . . 219
Diffraction Grating Theory . . . . . . . . . . . . . . . . . . 225
Ideal Lens Theory . . . . . . . . . . . . . . . . . . . . . 229
Surface Scattering Theory . . . . . . . . . . . . . . . . . . 230
Volumetric Scattering Theory . . . . . . . . . . . . . . . . . 232
Interference Pattern Theory . . . . . . . . . . . . . . . . . 239
Monochromatic Aberrations and the Zernike Polynomials . . . . . . 241
Accumulator Theory: Domains . . . . . . . . . . . . . . . . 244
Accumulator Theory: Boundaries . . . . . . . . . . . . . . . 246
References for the Geometrical Optics Interface . . . . . . . . . . 247

Chapter 4: Multiphysics Interfaces and Couplings

The Ray Heating Interface 252

Ray Heat Source . . . . . . . . . . . . . . . . . . . . . . 254

Theory for the Ray Heating Interface 255

Unidirectional and Bidirectional Couplings . . . . . . . . . . . . 255
Coupled Heat Transfer and Ray Tracing Equations . . . . . . . . . 255
Heat Source Calculation . . . . . . . . . . . . . . . . . . . 257

Chapter 5: Glossary

Glossary of Terms 260

Index 265

CONTENTS |7
8 | CONTENTS
 1

Introduction

This guide describes the Ray Optics Module, an optional add-on package for
COMSOL Multiphysics®.

This chapter introduces you to the capabilities of this module. A summary of the
physics interfaces and where you can find documentation and model examples is
also included. The last section is a brief overview with links to each chapter in this
guide.

In this chapter:

• About the Ray Optics Module

• Overview of the User’s Guide

9
 About the Ray Optics Module
These topics are included in this section:

• The Ray Optics Module Physics Interface Guide

• Common Physics Interface and Feature Settings and Nodes
• Where Do I Access the Documentation and Application Libraries?

The Ray Optics Module Physics Interface Guide

The Ray Optics Module extends the functionality of the physics interfaces in
COMSOL Multiphysics with physics interfaces for geometrical optics and ray heating,
where the rays can be polarized, unpolarized, or partially polarized. There are also
tools for computing the intensity, phase, and optical path length of rays. The details of
the physics interfaces and study types for the Ray Optics Module are listed in the table
below. The functionality of the COMSOL Multiphysics base package is given in the
COMSOL Multiphysics Reference Manual.

In the COMSOL Multiphysics Reference Manual:

• Studies and Solvers

• The Physics Interfaces
• For a list of all the core physics interfaces included with a COMSOL
Multiphysics license, see Physics Interface Guide.

10 | CHAPTER 1: INTRODUCTION
PHYSICS INTERFACE ICON TAG SPACE AVAILABLE STUDY TYPE
DIMENSION

Optics

Ray Optics

Geometrical Optics gop 3D, 2D, 2D ray tracing; bidirectionally

axisymmetric coupled ray tracing; time
dependent
Ray Heating — 3D, 2D, 2D ray tracing; bidirectionally
axisymmetric coupled ray tracing; time
dependent

Common Physics Interface and Feature Settings and Nodes

There are several common settings and sections available for the physics interfaces and
feature nodes. Some of these sections also have similar settings or are implemented in
the same way no matter the physics interface or feature being used. There are also some
physics feature nodes that display in COMSOL Multiphysics.

In each module’s documentation, only unique or extra information is included;

standard information and procedures are centralized in the COMSOL Multiphysics
Reference Manual.

In the COMSOL Multiphysics Reference Manual see Table 2-4 for

links to common sections and Table 2-5 to common feature nodes.
You can also search for information: press F1 to open the Help
window or Ctrl+F1 to open the Documentation window.

Where Do I Access the Documentation and Application Libraries?

A number of online resources have more information about COMSOL, including
licensing and technical information. The electronic documentation, topic-based (or

ABOUT THE RAY OPTICS MODULE | 11

 context-based) help, and the Application Libraries are all accessed through the
COMSOL Desktop.

If you are reading the documentation as a PDF file on your computer,

the blue links do not work to open an application or content
referenced in a different guide. However, if you are using the Help
system in COMSOL Multiphysics, these links work to open other
modules, application examples, and documentation sets.

THE DOCUMENTATION AND ONLINE HELP

The COMSOL Multiphysics Reference Manual describes the core physics interfaces
and functionality included with the COMSOL Multiphysics license. This book also has
instructions on how to use COMSOL Multiphysics and how to access the electronic
Documentation and Help content.

Opening Topic-Based Help

The Help window is useful as it is connected to the features in the COMSOL Desktop.
To learn more about a node in the Model Builder, or a window on the Desktop, click
to highlight a node or window, then press F1 to open the Help window, which then
displays information about that feature (or click a node in the Model Builder followed
by the Help button ( ). This is called topic-based (or context) help.

To open the Help window:

• In the Model Builder, Application Builder, or Physics Builder, click a node

or window and then press F1.
• On any toolbar (for example, Home, Definitions, or Geometry), hover the
mouse over a button (for example, Add Physics or Build All) and then
press F1.
• From the File menu, click Help ( ).
• In the upper-right corner of the COMSOL Desktop, click the Help ( )
button.

12 | CHAPTER 1: INTRODUCTION
 To open the Help window:

• In the Model Builder or Physics Builder, click a node or window and then
press F1.
• On the main toolbar, click the Help ( ) button.
• From the main menu, select Help>Help.

Opening the Documentation Window

To open the Documentation window:

• Press Ctrl+F1.
• From the File menu, select Help>Documentation ( ).

To open the Documentation window:

• Press Ctrl+F1.
• On the main toolbar, click the Documentation ( ) button.
• From the main menu, select Help>Documentation.

THE APPLICATION LIBRARIES WINDOW

Each model or application includes documentation with the theoretical background
and step-by-step instructions to create a model or application. The models and
applications are available in COMSOL Multiphysics as MPH-files that you can open
for further investigation. You can use the step-by-step instructions and the actual
models as templates for your own modeling. In most models, SI units are used to
describe the relevant properties, parameters, and dimensions, but other unit systems
are available.

Once the Application Libraries window is opened, you can search by name or browse
under a module folder name. Click to view a summary of the model or application and
its properties, including options to open it or its associated PDF document.

The Application Libraries Window in the COMSOL Multiphysics

Reference Manual.

ABOUT THE RAY OPTICS MODULE | 13

 Opening the Application Libraries Window
To open the Application Libraries window ( ):

From the File menu, select Application Libraries.

To include the latest versions of model examples, from the File>Help

menu, select ( ) Update COMSOL Application Library.

Select Application Libraries from the main File or Windows menus.

To include the latest versions of model examples, from the Help menu,
select ( ) Update COMSOL Application Library.

CONTACTING COMSOL BY EMAIL

For general product information, contact COMSOL at [email protected].

COMSOL ACCESS AND TECHNICAL SUPPORT

To receive technical support from COMSOL for the COMSOL products, please
contact your local COMSOL representative or send your questions to
[email protected]. An automatic notification and a case number will be sent to you
by email. You can also access technical support, software updates, license information,
and other resources by registering for a COMSOL Access account.

14 | CHAPTER 1: INTRODUCTION
COMSOL ONLINE RESOURCES

COMSOL website www.comsol.com

Contact COMSOL www.comsol.com/contact
COMSOL Access www.comsol.com/access
Support Center www.comsol.com/support
Product Download www.comsol.com/product-download
Product Updates www.comsol.com/support/updates
COMSOL Blog www.comsol.com/blogs
Discussion Forum www.comsol.com/forum
Events www.comsol.com/events
COMSOL Application Gallery www.comsol.com/models
COMSOL Video Gallery www.comsol.com/video
Support Knowledge Base www.comsol.com/support/knowledgebase

ABOUT THE RAY OPTICS MODULE | 15

 Overview of the User’s Guide
The Ray Optics Module User’s Guide gets you started with modeling using
COMSOL Multiphysics. The information in this guide is specific to this module.
Instructions how to use COMSOL in general are included with the COMSOL
Multiphysics Reference Manual.

As detailed in the section Where Do I Access the Documentation and

Application Libraries? this information can also be searched from the
COMSOL Multiphysics software Help menu.

TABLE OF CONTENTS AND INDEX

To help you navigate through this guide, see the Contents and Index.

RAY OPTICS MODELING

The Ray Optics Modeling chapter provides an overview of ray tracing simulation. It
begins with an overview of the Essentials of Ray Tracing. It then describes the major
functionality groups that are included in The Geometrical Optics Interface, including
topics such as Geometry and Meshing, Ray Release Features, Study Types, and Results
Analysis and Visualization.

RAY OPTICS INTERFACES

The Ray Optics Interfaces chapter describes The Geometrical Optics Interface and
includes the Theory for the Geometrical Optics Interface.

MULTIPHYSICS INTERFACES AND COUPLINGS

The Multiphysics Interfaces and Couplings chapter describes The Ray Heating
Interface, a dedicated multiphysics interface for computing heat sources generated by
attenuation of rays in absorbing media, and includes the Theory for the Ray Heating
Interface.

16 | CHAPTER 1: INTRODUCTION
 2

Ray Optics Modeling

This chapter gives an overview of the most important considerations when

creating a ray optics model. The most significant modeling decisions include the
means of geometry setup, the choice of optical material properties, ray releases, and
handling of ray interactions with surfaces in the geometry.

In this chapter:

• Essentials of Ray Tracing

• Geometry and Meshing
• Domain Properties
• Boundary Conditions
• Ray Release Features
• Additional Variables Solved For
• Study Types
• Results Analysis and Visualization
• Variables and Nonlocal Couplings
• Material Libraries
 Essentials of Ray Tracing
The Geometrical Optics interface can be used to simulate cameras, telescopes, laser and
illumination systems, spectrometers, solar collectors, and much more. However,
certain essential elements of the ray optics approach are common to all of these
application areas. These universal concepts in ray optics simulation are outlined below.

In this section:

• The Ray Tracing Algorithm

• Basic Requirements of a Geometrical Optics Model

The Ray Tracing Algorithm

The Geometrical Optics interface solves for the position and wave vector of individual
rays. It also allows them to interact with boundaries that intersect their paths.

Ray tracing is usually a reasonable approach to modeling electromagnetic wave

propagation if the wavelength of the radiation is small compared to the smallest
geometric detail in the surroundings, since diffraction is ignored. As long as this
criterion is met, ray tracing can be used for nearly any part of the electromagnetic
spectrum, including radio waves, microwaves, visible light, and UV radiation. Note
that X-ray modeling may require the consideration of diffraction effects because X-rays
can interact with matter on an atomic level.

If the geometry is on the wavelength scale, you might also consider a multiscale
approach by first solving for the electric field in the frequency domain using the finite
element method (FEM) and then using the FEM solution to define a ray release.

While propagating through a homogeneous medium (one in which the refractive

index is spatially uniform), a ray simply goes in a straight line at speed c/n, where
c = 299,792,458 m/s is the speed of light in a vacuum and n (dimensionless) is the
absolute refractive index of the medium. In a graded-index medium, the ray can follow
a curved path, which is determined by integrating coupled first-order ordinary
differential equations over time.

To learn more about the equations of ray propagation and their

derivation, see Introduction to Geometrical Optics in the Theory for the
Geometrical Optics Interface chapter.

18 | CHAPTER 2: RAY OPTICS MODELING

The Geometrical Optics interface is compatible with the Ray Tracing, Time Dependent,
and Bidirectionally Coupled Ray Tracing study steps. These study types are very similar,
except that the Ray Tracing study step makes it possible to specify either a list of optical
path length intervals or a list of time steps. (Internally, the optical path lengths are
converted to the corresponding times, so this is just a matter of convenience.)

For simple ray tracing models, only the first and last path lengths or time steps might
be needed. The behavior of rays at any intermediate time can then be accurately
interpolated. If rays interact with boundaries in between the stored time steps, then the
exact time and position of each ray-boundary interaction is also stored and readily
available.

Finer stepping in time or path length may be needed when modeling ray propagation
in graded-index media, when rays pass through attenuating media and generate heat
in their surroundings, or when using specialized postprocessing features.

Every time the intersection of a ray with a surface is detected, a wide variety of
ray-boundary interactions may apply. These include specular reflection, diffuse
reflection, refraction, and several different types of absorption.

Along each ray, it is possible to evaluate expressions that involve variables defined on
the ray itself (such as optical path length, intensity, and wavelength) and variables
defined at the ray’s position in the modeling domain (such as temperature and
refractive index). For example, to determine the refractive index in an optically
dispersive medium, an expression is evaluated on each ray that combines the ray’s
wavelength or frequency with a function queried from the domain the ray occupies.
Similarly, when a ray hits a boundary, the new direction of the ray can depend on a
combination of ray variables (like wavelength) and surface variables (like surface
normal direction and Gaussian curvature).

Basic Requirements of a Geometrical Optics Model

Although geometrical optics models vary tremendously in application area and scope,
every geometrical optics model requires at least the following basic features:

• A geometry must be present, with at least one surface that can interact with the rays.
• The refractive index must be specified. If any domains are present, use the Medium
Properties node to either specify the refractive index directly or specify which
material controls it.
• The model must have at least one boundary condition, such as the Material
Discontinuity or Wall node. By default, every boundary adjacent to at least one

ESSENTIALS OF RAY TRACING | 19

 domain in the model is treated as a Material Discontinuity that can reflect and refract
rays. See Boundary Conditions for more information.
• Every model includes the Ray Properties node, a default node that cannot be
removed. This node defines the equations of ray propagation. If the radiation is
monochromatic, this is also where you specify the frequency or wavelength.
• Some rays must be released into the model. This requires at least one ray release
feature, such as the Release, Release from Boundary, or Release from Grid node.
See Ray Release Features for more details.
• The ray tracing algorithm detects ray-boundary interactions using the underlying
finite element mesh. At the very least, a surface mesh is needed. For multiphysics
modeling in which field variables like temperature are solved for, a domain mesh is
also necessary. Some related topics are discussed in the Geometry and Meshing.
section.
• A study is necessary to compute the ray paths. The Geometrical Optics interface is
compatible with the Time Dependent, Ray Tracing, and Bidirectionally Coupled Ray
Tracing study steps. Ray Tracing is recommended for most models. See Study Types
for more information.

Ray release feature: collimated light source, hexapolar grid

Reflected rays Focus

Boundary condition: specular reflection at clear aperture

Boundary condition: absorption on edges and sides

Figure 2-1: A simple ray tracing model, featuring specular reflection of a collimated beam
by an off-axis conic mirror. The color expression indicates the ray intensity.

20 | CHAPTER 2: RAY OPTICS MODELING

Geometry and Meshing
Electromagnetic radiation can interact with a wide variety of geometric entities. In ray
optics simulation, the most common elements include lenses, mirrors, prisms, beam
splitters, light pipes, and various obstructions. Rays may also interact with buildings,
vehicles, people, planets, and more.

In this section, various considerations for setting up the model geometry and the
associated finite element mesh are considered in more detail:

• Creating a Geometry
• Domain Selection
• Meshing and Discretization Error
• Meshing Guidelines

Creating a Geometry
The Geometrical Optics interface supports interaction with any geometric entity. The
geometry can be constructed from built-in primitives like circles and line segments,
loaded from various CAD formats, or constructed with parts from the Ray Optics Part
Library. Rays can interact with both deformed and undeformed geometries, for
example, when an object undergoes thermal expansion.

PART LIBRARIES
In ray optics models, it is often necessary to set up geometry sequences consisting of
entities that are more complex than simple geometry primitives such as spheres, cones,
and blocks. Instead it might be necessary to insert lenses with different thicknesses and
radii of curvature, off-axis conic mirrors, and parabolic concentrators. The entity
shapes might need to be resolved to extremely high precision. This can be conveniently
accomplished using the Part Libraries for the Ray Optics Module.

The Ray Optics Module Part Libraries contain fully parameterized sequences of
geometry instructions that produce more complex shapes frequently required for
geometrical optics simulation, including the following:

• Spherical lenses and mirrors

• Conic lenses and mirrors
• Aspheric lenses and mirrors (even, odd, and Q-type)

GEOMETRY AND MESHING | 21

 • Doublet and triplet lenses
• Cylindrical lenses
• Beam splitters
• Polygonal mirrors
• Prisms
• Circular and rectangular annuli
• Retroreflectors

For example, you can load the Spherical Lens 3D part into a model and then specify
the radii of curvature of each lens surface, along with the lens thickness and diameter.

Parts can include predefined selections that make it easy to apply boundary conditions
or material properties to multiple geometric entities at the same time. Some parts also
include multiple variants or different sets of inputs that you can choose between. For
example, you could either specify the off axis distance or off axis angle for a conic
mirror.

In addition, many parts automatically define work planes so that the parts and other
features can more easily be positioned and oriented with respect to each other.

Many Ray Optics tutorials use the Part Library to create their geometry
sequences. To learn more, see the following models:

• Newtonian Telescope: Application Library path Ray_Optics_Module/

Lenses_Cameras_and_Telescopes/newtonian_telescope
• Petzval Lens: Application Library path Ray_Optics_Module/
Lenses_Cameras_and_Telescopes/petzval_lens

Part Libraries in the COMSOL Multiphysics Reference Manual.

RAY TRACING IN AN IMPORTED MESH

It is possible to compute ray trajectories in an imported mesh. The mesh can be
imported from a COMSOL Multiphysics file (.mphbin for a binary file format
or .mphtxt for a text file format) or from a NASTRAN file (.nas, .bdf, .nastran,
or .dat).

22 | CHAPTER 2: RAY OPTICS MODELING

If the mesh is imported from a COMSOL Multiphysics file, the imported mesh always
uses a linear geometry shape order for the purpose of modeling ray-boundary
interactions, even if the model used to generate the mesh had a higher geometry shape
order.

If the mesh is imported from a NASTRAN file, the ray-boundary interactions can be
modeled using either a linear or higher geometry shape order. If Export as linear
elements is selected when generating the NASTRAN file, or if Import as linear elements
is selected when importing the file, then linear geometry shape order will be used.

Domain Selection
It is possible for rays to pass through domains in the geometry and to propagate in the
void region outside these domains. Boundary conditions can be specified at any
boundary, including boundaries that are not adjacent to any domain in the geometry.
This means that rays can be reflected or absorbed by a surface in 3D, or a line segment
in 2D, even if it is not attached to any other object. The ray tracing algorithm can also
detect boundary interactions in any order, without this order being specified.

Usually, the domain selection for the Geometrical Optics interface should include all
objects that the rays might pass through. In a lens system, this would mean that all
lenses are included but not necessarily the mount for these lenses.

Domains that are not included in the selection for the physics interface do not need to
be meshed. However, these domains would still require a mesh if some other variables,
like displacement and temperature, are solved for there. See Meshing Guidelines for
more details.

Some physics features require a domain mesh and will not function on
domains outside the physics interface selection. This includes all types of
Accumulator (Domain) feature, such as the dedicated Ray Heat Source
multiphysics feature.

Meshing and Discretization Error

In the Geometrical Optics interface, rays interact with a surface when they hit a mesh
element in that surface’s boundary mesh. When rays need information about the
surface they hit, such as the surface normal direction, which controls the direction of

GEOMETRY AND MESHING | 23

 reflected and refracted rays, this information is also evaluated on the boundary mesh.
Thus, having a high-quality mesh is an integral part of ray optics simulation.

The use of a boundary mesh to detect and apply ray-boundary interactions makes the
Geometrical Optics interface readily extensible to high-fidelity multiphysics
simulation, including such effects as translational motion, rotation, and structural
deformation of the geometry (including thermal stress). In addition, this
implementation makes it possible to trace rays through geometric entities of arbitrary
shape, not just simple shapes for which a parametric representation is readily available.

Because rays interact with a mesh representation of the geometry, the mesh must be
adequately refined so that the coordinates of points along the surface are sufficiently
accurate. The level of mesh refinement also affects the accuracy of the tangential and
normal unit vectors that are defined on the boundary elements, as well as the Gaussian
curvatures that may be used to calculate the intensity along rays. Very high accuracy
can be achieved with a coarse mesh on planar surfaces because even a small number of
linear boundary elements can represent a planar surface to machine precision.
Accurately discretizing the geometry becomes more important when the surfaces are
curved, as in spherical lenses and conic mirrors, or when the surfaces can be deformed.

In the following image, the radial position over one boundary element on a coarsely
meshed sphere (using quadratic elements) is compared to the exact value for an ideal
sphere. The error is on the order of 10-5.

Figure 2-2: Error in the radial displacement over one boundary element of a unit sphere.

A relative error like the one shown above (10-5) might be sufficiently small for some
simulation results, but in geometrical optics such an error might translate to tens of
additional wavelengths in spot size — large enough to invalidate the results of the
simulation entirely, unless adequate precautions are taken. Similarly, in models with

24 | CHAPTER 2: RAY OPTICS MODELING

mesh deformation, the degrees of freedom for the displacement field must be solved
for extremely accurately for the results of a coupled multiphysics model to be trusted.
A good practice, as with other types of simulation, is to perform a mesh refinement
study, ensuring that the results do not change appreciably when the mesh element size
is reduced further.

Note that ray tracing in COMSOL uses double precision floating point arithmetic, in
for which machine precision is of order 10−16.

Meshing Guidelines
• Flat edges (2D) and planar surfaces (3D) can be coarsely meshed, unless they meet
one of the criteria described below.
• Curved surfaces that can interact with rays should always be finely meshed. The
tighter the curvature of such surfaces, the finer the mesh should be.
• In 3D, the row of curved boundary elements adjacent to a curved edge are more
susceptible to mesh discretization error than boundary elements that are not
adjacent to an edge. For this reason, avoid unnecessary interior edges in 3D.
• Domains usually do not need to be as finely meshed as boundaries. A convenient
way to refine the mesh in the vicinity of the boundaries is to reduce the Curvature
factor in the Size settings window in the mesh sequence. This results in a finer mesh
only where the radius of curvature of the surface is small. You might also have to
reduce the Minimum element size to avoid warnings in the mesh sequence.
• Features that compute the density of some quantity on a domain or boundary
usually require a finer mesh, because the density term is piecewise discontinuous
across elements. This includes the Accumulator (Boundary), Accumulator
(Domain), and Deposited Ray Power (Boundary) features. If the mesh is too fine,
rays might entirely miss some elements, and then it is necessary to increase the
number of rays to avoid “holes” in the deposited power or other density field.
• Whenever possible, use parts from the Ray Optics Module Part Libraries, rather
than constructing complicated shapes from geometry primitives like the sphere and
cylinder. Many of these parts use high-accuracy geometry representations of the
surfaces which reduce the numerical error when rays interact with them.
• When using the built-in geometry Parts for aspheric lenses or mirrors, consider
enabling the built-in extra points (see the Input Parameters section), especially if
higher-order polynomial terms are included.

GEOMETRY AND MESHING | 25

 • Consider changing the Geometry shape order in the model component settings.
Using Cubic or Quartic causes the boundaries to be discretized using higher-order
polynomials, which can reduce error by several orders of magnitude.
Similarly, with physics interfaces that solve for a displacement field, such as Solid
Mechanics, locate the physics interface Discretization section. A higher shape order
such as Cubic Lagrange should be selected if rays are traced in the deformed geometry
• If the geometry uses parts or primitives, you can reduce discretization error by
selecting Use geometry normals for ray-boundary interactions in the physics interface
Ray Release and Propagation section. However, this only improves the accuracy if the
geometry is undeformed; it has no effect if the geometry is subjected to thermal
stress or some other type of deformation.

26 | CHAPTER 2: RAY OPTICS MODELING

Domain Properties
A fundamental step in setting up any ray optics model is identifying the material that
the rays are propagating through. You can control the refractive indices of the domains
in the selection for the Geometrical Optics interface. You can also control the refractive
index in the empty space between the selected domains, albeit with some restrictions
(see below).

Refractive Index
The refractive index is a complex-valued material property. Its real part indicates the
speed of light within the material; its imaginary part describes how quickly a beam will
attenuate (that is, decrease in intensity) as it passes through the material.

The refractive index and other properties of domains included in the Geometrical
Optics interface can be set using the Medium Properties node. It is also possible to
specify the refractive index of the void surrounding a Geometrical Optics model.

ABSOLUTE VS RELATIVE INDEX

When looking at refractive index data for a given material, it is essential to know
whether the refractive index is absolute or relative:

• An absolute refractive index is defined with respect to vacuum. That is, if the
absolute index of a material is 1.5, the speed of light in that material is c/1.5, where
c = 299,792,458 m/s is the speed of light in a vacuum.
• A relative refractive index is defined with respect to some other material, most often
air. If the relative index of a material is 1.5, then the speed of light in that material
is c/1.5, where c is the speed of light in air.

Furthermore, the speed of light in air is a function of temperature and pressure.

Typically, when defining the relative refractive index, the air is at a reference
temperature of 1 atm and a reference temperature of 20°C, although some glass
catalogs use a different reference temperature such as 15°C.

The settings for the Medium Properties always indicate whether the specified refractive
index is absolute or relative; and if the given refractive index is relative, a reference
temperature and reference pressure are always shown.

During results processing, the variable gop.nref is always the absolute refractive
index, regardless of whether the absolute or relative index was specified.

DOMAIN PROPERTIES | 27
 EXTERIOR AND UNMESHED DOMAINS
In the physics interface settings Material Properties of Exterior and Unmeshed Domains
section, the Optical dispersion model is specified. This determines the Refractive index
of exterior domains, real part. This refractive index is used in any domains outside of
the selection for the Geometrical Optics interface, as well as the void domain outside
the geometry. It is a spatially uniform, scalar-valued quantity. The refractive index
outside the domain selection cannot depend on field variables such as temperature or
pressure and cannot be a graded-index medium. For domains with
temperature-dependent or spatially nonuniform refractive indices, use the Medium
Properties node instead.

Scattering and Absorption

If you specify a complex-valued refractive index and are solving for ray intensity or
power (see the section Intensity, Polarization, and Power), you can make the intensity
and power vary along the ray path. If the medium properties are uniform, the intensity
and power usually follow exponential decay (for an absorbing medium) or growth (for
a gain medium). You can switch between absorbing and gain media by changing the
sign of the imaginary part of the refractive index.

An alternative way to specify the imaginary part of the refractive index is to enter a
value or expression for the internal transmittance for a given thickness of the material.
For example, many of the optical glasses from the Optical Material Library (see the
Material Libraries section) provide tabulated values of the internal transmittance as a
function of wavelength.

Usually the real and imaginary parts of the refractive index in a homogeneous medium
are material properties. Sometimes, however, the absorption is due to the presence of
small imperfections or particles in a medium. Some examples include dust, water
droplets, or smoke in air; bubbles or sediment in water; or colloidal suspensions. In
such cases, you can include the losses due to absorption and scattering by these
particles using the dedicated Scattering Domain node.

28 | CHAPTER 2: RAY OPTICS MODELING

Bo unda r y C o nd i t i on s
This section describes the boundary conditions of the Geometrical Optics interface in
greater detail:

• Reflection and Refraction

• Primary and Secondary Ray Releases
• Wall Conditions
• Special Boundary Conditions

Reflection and Refraction

The default boundary condition is the Material Discontinuity condition on all interior
and exterior boundaries. The Material Discontinuity causes rays to be reflected and
refracted if the two adjacent domains have different refractive indices.

The direction of the refracted ray is based on Snell’s law. If the ray intensity or power
is solved for, then they are reinitialized according to the Fresnel equations. You can
modify the application of Fresnel equations by adding one or more thin dielectric
layers to the surface.

Material Discontinuity Theory and Intensity, Wavefront Curvature, and

Polarization in Theory for the Geometrical Optics Interface.

SUPPRESSING THE RELEASE OF REFLECTED RAYS

The total number of released secondary rays in a model can sometimes grow rapidly
and exhaust all of the preallocated secondary degrees of freedom. For example, a single
ray reflecting back and forth between two Material Discontinuity boundaries can create
an inordinately large number of rays, each with extremely low intensity. Remember
that the total number of secondary rays that can be produced in the model is limited
by the number specified in the Maximum number of secondary rays field.

It can be useful to constrain the release of secondary rays at boundaries so that only
the most important rays are produced. If reflected rays are not of any interest, then in
the Rays to Release section, select Never from the Release reflected rays list. If reflected
rays are only relevant to the model under a certain condition, such as hitting a specific

BOUNDARY CONDITIONS | 29
 part of the surface or having a certain direction, instead select Based on logical
expression and then enter a user-defined Evaluation expression that must be satisfied.

If ray intensity is solved for, you can also specify a Threshold intensity. If a reflected ray
would have intensity below the threshold, it is not released at all. Similarly, if ray power
is solved for, you can specify a Threshold power.

Primary and Secondary Ray Releases

In the Geometrical Optics interface, a ray is released when it begins to propagate. The
physics features that control where and how rays are released are called ray release
features or simply release features. These features are separated into two categories,
called primary release features and secondary release features. Similarly, the rays
created by each feature are either primary rays or secondary rays.

PRIMARY RAYS
A primary release feature allows the initial position and direction of rays to be specified
directly. For the release positions, either specify the grid points directly (as in the
Release from Grid feature) or choose the geometric entities that produce the rays (as
in the Release, Release from Boundary, Release from Edge, and Release from Point
features). The initial direction can be specified directly or sampled from a distribution.

Primary rays are released directly by a release feature. They are called primary rays
because their release is not contingent on the prior existence of any other ray.

SECONDARY RAYS
Secondary rays are released when an existing ray is subjected to certain boundary
conditions. This existing ray might be a primary ray, or it could be a different
secondary ray that was released earlier in the simulation.

For example, the following diagram shows an incident ray being split into reflected and
refracted rays at a Material Discontinuity where the refractive indices on either side
differ. The Geometrical Optics interface always applies deterministic ray splitting at
such boundaries, so when one ray reaches the surface, two rays emerge from it. The
refracted ray is a continuation of the incident ray because it has the same index and uses
the same degrees of freedom. The reflected ray is a secondary ray.

Note that total internal reflection is automatically detected. In this case the incident
ray is simply reflected and no refracted ray appears. Therefore, total internal reflection
does not require the release of a secondary ray.

30 | CHAPTER 2: RAY OPTICS MODELING

 Released (primary) ray Reflected (secondary) ray

Material discontinuity

Refracted (primary) ray

ALLOCATION OF SECONDARY DOFS

The number of degrees of freedom allocated to secondary rays is determined when the
study begins, and if this maximum number is ever reached, then no additional
secondary rays will be released. In other words, if a large number of reflected rays have
been produced at material discontinuities in a model, at some point no more reflected
rays will be produced. In this case, a Warning will appear in the solver sequence. The
maximum number of secondary rays that can be produced in a model is controlled by
the Maximum number of secondary rays field in the physics interface settings Ray Release
and Propagation section.

Similarly, the Grating and Cross Grating boundary conditions can release both
reflected and transmitted rays of many diffraction orders. One of these reflected or
transmitted rays uses the degrees of freedom of the incident ray, and the rest must be
taken from the preallocated secondary degrees of freedom.

In the Sagnac Interferometer example, a single ray is split up into four

distinct rays due to reflection and refraction at the Material Discontinuity.
The phase difference between two of these rays is then computed.
Application Library path Ray_Optics_Module/
Spectrometers_and_Monochromators/sagnac_interferometer

BOUNDARY CONDITIONS | 31
 Wall Conditions
For most boundary conditions other than refraction, the Wall feature can be used. It
includes a wide variety of boundary conditions including the following:

• Specular reflection
• Diffuse reflection
• Isotropic reflection
• Combination of diffuse and specular reflection
• User-defined reflection
• Pass through
• Several varieties of absorption condition (see below)

In this context, diffuse reflection means Lambertian scattering, following the cosine
law; as opposed to isotropic scattering, in which reflected light has an equal probability
of being reflected in any direction to one side of the boundary.

If you choose Mixed diffuse and specular reflection, you can assign some probability that
the ray is reflected specularly; otherwise, it is reflected diffusely.

The Diffuse scattering and Mixed diffuse and specular reflection rely on pseudorandom
number generation, so they are not guaranteed to give exactly the same numerical
results in different software versions or on different architectures.

TYPES OF RAY ABSORPTION

There are several different types of absorption condition, classified by the type of
information that they retain about the ray.

• The Disappear condition annihilates the ray completely. After the ray disappears, its
position and other degrees of freedom evaluate to not-a-number (NaN).
• The Freeze condition retains the ray position and direction after the ray hits the
boundary, although these quantities no longer change over time.
• The Stick condition retains the ray position, but all other degrees of freedom are set
to zero.

If the model includes several different types of boundary condition and you want to
know what type of condition was applied to each ray, select the Store ray status data
check box. This will store a variable that indicates the final status of each ray: whether
it has not been released yet, is still propagating, or has been absorbed.

32 | CHAPTER 2: RAY OPTICS MODELING

CONDITIONAL RAY-WALL INTERACTIONS
The Wall feature also support conditional ray-wall interactions. For example, you could
cause rays to Freeze if a logical expression is satisfied — for example, having optical path
length greater than a specified threshold — and subject them to Specular reflection
otherwise. Together with the Mixed diffuse and specular reflection condition, it is
possible to prescribe up to 3 different types of wall interaction at a single boundary.

Special Boundary Conditions

More specialized boundary conditions are available for some application areas.

The Grating and Cross Grating boundary conditions, together with the Diffraction
Order (Grating) and Diffraction Order (Cross Grating) subnodes, respectively, are
used to separate rays by diffraction at a periodic structure. The Grating is periodic in
one direction and homogeneous in the orthogonal direction; the Cross Grating is
periodic in two directions. Either feature can release both reflected and transmitted
rays. You can specify the direction of periodicity and the size of the unit cell.

Czerny-Turner Monochromator: Application Library path

Ray_Optics_Module/Spectrometers_and_Monochromators/
czerny_turner_monochromator

White Pupil Échelle Spectrograph: Application Library path

Ray_Optics_Module/Spectrometers_and_Monochromators/
white_pupil_echelle_spectrograph

Cross Grating Échelle Spectrograph: Application Library path

Ray_Optics_Module/Spectrometers_and_Monochromators/
cross_grating_echelle_spectrograph

The Mirror boundary condition is a simplified Wall that only causes specular reflection.

The Scattering Boundary feature can be used to scatter rays in both transmission and
reflection.

The Axial Symmetry boundary condition is only available in 2D axisymmetric models.

It is automatically applied at the axis of symmetry and cannot be applied anywhere else.

Optical devices like the Linear Polarizer and Linear Wave Retarder are available when
ray intensity or power is computed. They do not have any effect on the ray direction
but they can affect the ray polarization.

BOUNDARY CONDITIONS | 33
 Ray Release Features
To trace rays it is first necessary to prescribe their initial position and direction. This
process is called releasing rays. The physics features used to enter this information are
called release features or ray release features. If other quantities are being solved for
along the rays, such as the intensity and polarization, then these quantities are
initialized by the ray release features.

In this section:

• Grid-Based Release
• Release from Domains, Boundaries, Edges, or Points
• Specialized Release Features
• Release Features for Multiscale Modeling

Grid-Based Release
Use the Release from Grid feature to specify the initial positions of rays using a grid of
points. It is useful to release rays from a grid when the initial ray position is known
exactly. A grid-based release may be used, for example, when rays are released from the
focus of a lens or when a system is excited by a laser. This is the easiest way to release
rays from known locations in the void region outside the geometry.

Alternatively, you can load the ray release positions and directions from a text file, using
the Release from Data File node.

Release from Domains, Boundaries, Edges, or Points

There are ray release features for every geometric entity level:

• Use the Release feature to release rays from domains in 2D or 3D.

• Use the Release from Boundary feature to release edges (2D) or surfaces (3D).
• Use the Release from Edge feature to release rays from edges in 3D.
• Use the Release from Point feature to release rays from points in 2D or 3D.

Releasing rays from a domain, surface, or edge initializes the ray position based on the
underlying finite element mesh, which discretizes the geometry. A side effect is that
the ray positions can change slightly when switching between geometry kernels. Ray

34 | CHAPTER 2: RAY OPTICS MODELING

release can be uniform or proportional to a user-defined expression, which is specified
in the Density proportional to text field.

Specialized Release Features

The Illuminated Surface is a specialized ray release feature that produces reflected or
refracted light directly at a surface. This can be used, for example, when the direction
of incident radiation is known, but its propagation is not interesting until it has already
been reflected once by an object.

The Solar Radiation computes the initial direction of rays as if they were solar rays
being released at a specified latitude, longitude, date, and time. Instead of the latitude
an longitude, you can also select from a list of built-in cities.

The following examples involve buildings or other objects being

illuminated by solar radiation in a specific direction:

• Solar Dish Receiver: Application Library path Ray_Optics_Module/

Solar_Radiation/solar_dish_receiver
• Vdara® Caustic Surface: Application Library path
Ray_Optics_Module/Solar_Radiation/vdara_caustic_surface

Release Features for Multiscale Modeling

The Ray Optics Module also supports multiscale electromagnetics simulation, in which
the electric field is first computed in the frequency domain over a length scale of several
wavelengths, then propagated over a much longer distance using a ray tracing
approach. For more details, see Release from Electric Field and Release from Far-Field
Radiation Pattern.

RAY RELEASE FEATURES | 35

 Additional Variables Solved For
The Geometrical Optics interface always solves for the ray position q (SI unit: m) and
wave vector k (SI unit: rad/m). This section lists the optional quantities that can be
solved for in addition to these required variables.

In this section:

• Modeling Polychromatic Radiation

• Intensity, Polarization, and Power
• Other Dependent Variables
• Order of Initialization of Auxiliary Dependent Variables

Modeling Polychromatic Radiation

By default, the rays released in a model are monochromatic with a ray frequency or
free-space wavelength defined in the Ray Properties settings window. However, you
can trace polychromatic rays by specifying different frequencies or vacuum
wavelengths in each ray release feature.

To allow the rays to be polychromatic, in the settings window for the Geometrical
Optics interface locate the Ray Release and Propagation section. By default,
Monochromatic is selected from the Wavelength distribution of released rays list. Select
Polychromatic, specify vacuum wavelength to release polychromatic rays by entering an
expression for the vacuum wavelength or sampling it from a distribution. Alternatively,
select Polychromatic, specify frequency to define an expression or distribution for the ray
frequency. These expressions are defined in the Vacuum Wavelength and Initial Ray
Frequency sections in each ray release feature.

When modeling polychromatic light, the number of degrees of freedom in the model
increases by one per ray because the wavelength or frequency is stored as an auxiliary
dependent variable on each ray.

Czerny-Turner Monochromator: Application Library path

Ray_Optics_Module/Spectrometers_and_Monochromators/
czerny_turner_monochromator

36 | CHAPTER 2: RAY OPTICS MODELING

Intensity, Polarization, and Power
Ray intensity is computed using a variant of the Stokes–Mueller calculus in which both
the amplitude and polarization are tracked along individual rays.

LIST OF AVAILABLE SETTINGS

To decide whether intensity is computed, select an option from the Intensity
Computation list. The following options are available.

• None: Does not compute any intensity information.

• Compute intensity: Solves for intensity, which typically increases as rays are focused
and decreases as they diverge. Also affected by reflection, refraction, and attenuating
media. Only valid when the media are homogeneous.
• Compute power: Solves for power, which is unaffected by the convergence or
divergence of rays but is still affected by reflection, refraction, and attenuating
media. Can be used to compute heat source terms in attenuating domains or heat
flux terms on absorbing boundaries that the rays hit.
• Compute intensity and power: Combines the capabilities of Compute intensity and
Compute power, at the cost of a few extra degrees of freedom per ray.
• Compute intensity in graded media: Similar to Compute intensity, but is also applicable
to graded-index media. The tradeoff is that this method is slower and less accurate
for homogeneous media.
• Compute intensity and power in graded media: Similar to Compute intensity in graded
media but can also be used to generate heat sources in attenuating domains and heat
flux terms at boundaries.

HANDLING POLARIZATION
Whenever intensity or power is solved for, the polarization of every ray is known. Rays
can have any degree of polarization, ranging from 0 (unpolarized) to 1 (fully polarized)
and anything in between. When rays have some degree of polarization, they can be
linearly, elliptically, or circularly polarized.

When rays are reflected and refracted at boundaries, the intensity, polarization, and
power are updated based on the Fresnel equations, which automatically take the
polarization direction into account.

The polarization is determined based on the Stokes parameters, which are allocated as
extra degrees of freedom along each ray. For more information, see The Stokes
Parameters in the Theory for the Geometrical Optics Interface chapter.

ADDITIONAL VARIABLES SOLVED FOR | 37

 In the following examples, ray polarization is manipulated in an
instructive way:

• Total Internal Reflection Thin-Film Achromatic Phase Shifter

(TIRTF APS): Application Library path Ray_Optics_Module/
Prisms_and_Coatings/achromatic_phase_shifter
• Linear Wave Retarder: Application Library path Ray_Optics_Module/
Tutorials/linear_wave_retarder

WAVEFRONT CURVATURE
When the ray intensity is solved for, it increases where rays are focused together and
decreases where rays diverge. This is accomplished by treating each ray as a wavefront
and storing its principal radii of curvature as extra degrees of freedom. In this way, all
released rays are treated as points on planar, spherical, or ellipsoid-shaped wavefronts.

For more information on wavefront radii of curvature and their effect on intensity, see
Principal Radii of Curvature in the Theory for the Geometrical Optics Interface
chapter.

COMPUTING DEPOSITED RAY POWER

The options Compute power, Compute intensity and power, and Compute intensity and
power in graded media all allow heat sources to be defined on domains or boundaries.
As rays propagate through an attenuating medium — that is, a medium where the
refractive index is complex-valued — some energy is lost from the ray. The
corresponding heat source on the surrounding domain can be computed using either
the Deposited Ray Power (Boundary) subnode or the Ray Heat Source multiphysics
node. A Ray Heat Source node is automatically created when you have selected The Ray
Heating Interface in the Model Wizard. The heat generated as rays propagate in an
attenuating medium can be used to define a heat source in the Heat Transfer in Solids
interface or another physics interface that computes a temperature field.

Thermally Induced Focal Shift in High-Power Laser Focusing Systems:

Application Library path Ray_Optics_Module/
Structural_Thermal_Optical_Performance_Analysis/
thermally_induced_focal_shift

38 | CHAPTER 2: RAY OPTICS MODELING

TOTAL POWER TRANSMITTED AND REFLECTED AT GRATINGS
The Grating feature is used to model the transmission and reflection of rays at
diffraction gratings. It includes a Diffraction Order (Grating) subnode to specify which
diffraction orders to release. When the ray power is solved for, the Store total
transmitted power and Store total reflected power check boxes are shown in the Grating
settings window. Selecting either of these check boxes causes an auxiliary dependent
variable to be declared, storing the total power of the transmitted and reflected rays of
all diffraction orders.

Diffraction Grating: Application Library path Ray_Optics_Module/

Verification_Examples/diffraction_grating

Other Dependent Variables

It is possible to define an auxiliary dependent variable for the optical path length by
selecting the Compute optical path length check box in the Additional Variables section
of the physics interface node’s Settings window. Initially, the optical path length is set
to 0 for all released rays. It is possible to reset the optical path length to 0 when the
rays interact with boundaries.

The phase of a ray is necessary for some applications that require information about
the instantaneous electric fields of multiple rays, such as interferometers. To define an
auxiliary dependent variable for phase, select the Compute phase check box in the
Intensity Computation section of the physics interface node’s Settings window. This
check box is only available if the ray intensity is computed.

The instantaneous phase can be used to visualize interference patterns where the rays
intersect a surface. See the Results Analysis and Visualization section.

Other dependent variables can also be assigned, including:

• The number of times each ray has been reflected.

• Help variables for more accurately tracing rays in strongly absorbing media, where
the real and imaginary parts of the refractive index are comparable in magnitude.
• Help variables used to apply perturbations in ray direction due to rough surfaces.
• User-defined Auxiliary Dependent Variable nodes.

ADDITIONAL VARIABLES SOLVED FOR | 39

 Order of Initialization of Auxiliary Dependent Variables
When rays are released, the variables defined for each ray are initialized in a specific
order. The initial values of ray variables can only depend on the values of variables that
have already been defined. The order of dependent variable initialization is governed
by the following rules:

• The initial ray position is always determined first.

• By default, user-defined auxiliary dependent variables (that is, those that are defined
by adding an Auxiliary Dependent Variable node to the physics interface) are
initialized after all other variables. They can instead be initialized immediately after
the ray position vector components by selecting the Initialize before wave vector
check box shown in the release feature Initial Value of Auxiliary Dependent Variables
section.
• If more than one user-defined auxiliary dependent variable is present, these variables
are initialized in the order in which the corresponding Auxiliary Dependent Variable
nodes appear in the Model Builder.
• The remaining degrees of freedom are defined in the following order (each listed
group is initialized simultaneously, and the variables within a group cannot reliably
be initialized in terms of each other):
- Help variable for the perturbation of initial ray direction at illuminated surfaces
- Ray frequency or vacuum wavelength
- The wave vector components, optical path length, and total power transmitted
and reflected by gratings
- The integral of the attenuation coefficient along ray paths and the components
of the unit vector that indicates the direction corresponding to one of the
principal radii of curvature
- The principal radii of curvature, initial principal radii of curvature, intensity,
Stokes parameters, and help variables for computing the curvature tensor and
intensity
- The total power transmitted by the ray

Items in each bullet point may not be initialized as functions of items in a later bullet
point. For example, the initial ray direction vector can depend on the ray frequency,
but the initial principal radii of curvature cannot depend on the total power
transmitted by the ray.

40 | CHAPTER 2: RAY OPTICS MODELING

Study Types
The Geometrical Optics interface is compatible with three study types: Ray Tracing,
Time Dependent, and Bidirectionally Coupled Ray Tracing. In this section some of the
more relevant study settings are explored in greater detail.

In this section:

• Ray Tracing Study Step

• Bidirectionally Coupled Ray Tracing Study Step

Ray Tracing Study Step

The Ray Tracing and Time Dependent study steps are very similar, and either one could
be used for the vast majority of geometrical optics models. The Ray Tracing study step
has some additional features that make it more convenient to use, such as more
reasonable default values and built-in stop conditions.

A NOTE ON GEOMETRIC NONLINEARITY

If a physics interface solves for the displacement field, such as the Solid Mechanics
interface, then the Include geometric nonlinearity check box appears in the Study
Settings section. It is very important to select this check box when tracing rays in a
system that is deformed due to external forces or thermal stress. If the check box is
cleared, then rays are instead traced in the undeformed geometry.

TIME STEPS AND OPTICAL PATH LENGTH INTERVALS

By default, the Ray Tracing study step computes ray trajectories from t = 0 to t = 1 ns
with a time step size of 0.01 ns. However, it is often useful to think of ray tracing in
terms of the maximum distance of ray propagation instead of the maximum time. To
express the duration of the study in terms of a maximum optical path length, change
the Time step specification setting from the default, Specify time steps, to Specify
maximum path length. Then select a Length unit (default: m), enter a set of Lengths
(default: range(0,0.01,1)), and enter a Characteristic group velocity (default:
c_const, a built-in constant for the speed of light in a vacuum). With the default solver
settings, the time-dependent solver must take at least one time step whenever the
optical path length of a ray moving at the Characteristic group velocity would have
reached one of the values in the list of Lengths.

STUDY TYPES | 41
 BUILT-IN STOP CONDITIONS
The Ray Tracing study step includes options to create a Stop Condition node in the
default solver sequence. The Stop Condition node can terminate the study before the
full range of specified times or optical path lengths has been simulated, if a condition
is met before then.

To use one of the built-in stop conditions, select one of the following options from the
Stop condition list in the Study Settings:

• None: The study ends at the specified maximum time or maximum optical path
length.
• No active rays remaining: The study terminates if all rays have been stopped or
absorbed. A ray can be absorbed by a boundary, or it can be removed by the Ray
Termination domain feature.
• Active rays have intensity below threshold: This option is only available when ray
intensity is computed. The study terminates if every ray in the model is either
stopped, absorbed, or has sufficiently low intensity.
• Active rays exceed maximum number of reflections: This option is only usable when
the Count reflections check box is selected in the physics interface settings Additional
Variables section. The study terminates if every ray in the model is either stopped,
absorbed, or has been reflected the specified number of times.

Stop Condition in the COMSOL Multiphysics Reference Manual.

COUPLED PHYSICS INTERFACES

If other physics interfaces are also solved for in the Ray Tracing study step, it is assumed
that these other fields vary over the same time scale as the ray propagation. This is
seldom true. If, instead, the coupled physics do not change over the time scale for ray
propagation, consider a study with two steps: Stationary for all other fields and Ray
Tracing just for the rays.

In the COMSOL Multiphysics Reference Manual:

• Ray Tracing
• Studies and Solvers

42 | CHAPTER 2: RAY OPTICS MODELING

Bidirectionally Coupled Ray Tracing Study Step
The Bidirectionally Coupled Ray Tracing study step is a dedicated study step for ray
heating and similar applications.

It should only be used if all of the following criteria are met:

1 Rays are being traced.

2 Some other field, such as temperature or structural displacement, is solved for in a
domain where the rays are being traced.
3 All fields, apart from the ray paths themselves, are stationary.
4 The ray paths are affected by the field being solved for. This could include rays
interacting with a deformed geometry, or a refractive index that depends on the
values of field variables like strain or temperature.
5 The rays generate enough heat to significantly affect one of the fields being solved
for in the domain, usually temperature.

If condition 4 is not satisfied, instead use a Stationary study step for the fields, followed
by a Ray Tracing study step for the rays.

In addition to the settings that are available for the Ray Tracing study step, it is possible
to specify a Number of iterations. The default value is 5. If the Bidirectionally Coupled
Ray Tracing study step is used with The Ray Heating Interface, the following iterative
solver loop is automatically set up to compute the ray trajectories and temperature:

1 Solve for the temperature field, assuming that the rays do not generate any heat
source.
2 Using the temperature computed during the previous step, compute the ray paths
and any heat sources that occur due to ray attenuation in an absorbing medium.
3 Using the heat source computed in the previous step, compute the temperature
field.
4 Alternate between Steps 2 and 3 for the specified Number of iterations, or specify a
Global variable whose convergence will be used to terminate the loop.

The result of the iterative solver loop is that the heat source generated by the
attenuation of rays is taken into account when computing the temperature. By
including a thermo-optic dispersion model, such as the Temperature-dependent
Sellmeier model, the temperature in turn affects the ray paths. Thus, a bidirectional
coupling is established between the two physics interfaces.

STUDY TYPES | 43
 Like any COMSOL Multiphysics simulation, it is possible to extend this bidirectional
coupling to include other physical effects. For example, to include structural
deformation due to thermal stress, add the Solid Mechanics physics interface and the
Thermal Expansion Multiphysics coupling.

Thermally Induced Focal Shift in High-Power Laser Focusing Systems:

Application Library path Ray_Optics_Module/
Structural_Thermal_Optical_Performance_Analysis/
thermally_induced_focal_shift

In the COMSOL Multiphysics Reference Manual:

• Bidirectionally Coupled Ray Tracing

• Studies and Solvers

44 | CHAPTER 2: RAY OPTICS MODELING

Results Analysis and Visualization
The results of a ray tracing simulation can be interpreted and visualized using a variety
of built-in features. In the following, some dedicated features are discussed in detail.

In this section:

• Ray Trajectories Plot

• Ray Plot
• Interference Pattern Plot
• Spot Diagram Plot
• Optical Aberration Plot and Aberration Evaluation
• Plotting Refractive Index and Abbe Number

Ray Trajectories Plot

The default Ray Trajectories plot shows each ray as a line, ribbon, or tube. The final ray
position can be rendered as a point or vector. It is also possible to modify the Ray
Trajectories plot by adding Color Expression, Deformation, Material Appearance, Filter,
Translation, or Transparency attributes (subnodes).

As an example, when the ray intensity is computed, you can also plot polarization
ellipses along the ray paths, using the built-in variables (gop.pax, gop.pay,gop.paz)
for the semi-major axis and (gop.pbx, gop.pby,gop.pbz) for the semi-minor axis.

Figure 2-3: Polarization ellipses as a ray crosses a linear polarizer, linear wave retarder,
and a second linear polarizer.

Linear Wave Retarder: Application Library path Ray_Optics_Module/

Tutorials/linear_wave_retarder

RESULTS ANALYSIS AND VISUALIZATION | 45

 Ray Plot
The Ray plot can be added to 1D plot groups. There are two primary ways to use this
plot:

1 Plot an expression versus time for all rays, or as a data series operation such as a sum
or average over all rays.
2 Plot two expressions against each other, creating a separate curve for each ray, at a
specified time.

The following plot is an example of usage case 2. Polychromatic light is transmitted

through a multilayer film, and then the intensity at the final time step is plotted against
vacuum wavelength for all rays.

Figure 2-4: Reflectance of a distributed Bragg reflector with 21 dielectric layers is plotted
as a function of free-space wavelength.

Distributed Bragg Reflector: Application Library path

Ray_Optics_Module/Prisms_and_Coatings/distributed_bragg_reflector

46 | CHAPTER 2: RAY OPTICS MODELING

Interference Pattern Plot
The pattern of fringes resulting from the interference of two or more rays can be
plotted using the dedicated Interference Pattern plot. The Interference Pattern plot is
available in 2D plot groups and requires a Cut Plane dataset pointing to a Ray dataset.
The interference pattern is then plotted using the locations and properties of rays as
they intersect the cut plane.

Figure 2-5: Interference pattern resulting from two point sources at slightly different
distances from a surface.

The Interference Pattern can only be plotted if the ray intensity and phase are solved
for; in the physics interface settings, select an appropriate intensity computation
method from the Intensity computation list, then select the Compute phase check box.

For the resulting interference pattern to be physically meaningful, it must be plotted

over a region with a length scale that is much smaller than the principal radii of
curvature of the incident wavefronts. This is equivalent to the assumption that the
wavefront associated with each ray subtends a very small solid angle, and is necessary
due to the approximation used to compute the incident intensity.

Interference Pattern Theory

RESULTS ANALYSIS AND VISUALIZATION | 47

 Spot Diagram Plot
After tracing rays in 3D, you can plot the intersection points of rays with a surface using
the dedicated Spot Diagram plot. This plot can only be added to a 2D Plot Group. The
Spot Diagram includes many options for filtering and sorting the intersection points. It
can also automatically locate a plane of best focus, and add text annotations.

The Spot Diagram can either use a Ray dataset, in which case it plots all applicable rays
at the same time, or an Intersection Point 3D dataset, where the intersection points with
a surface are plotted.

Figure 2-6: Spot diagrams for a Schmidt–Cassegrain telescope at three different field
angles. The color expression indicates the wavelength.

FILTERING RAYS
Use the Filters section to show or hide certain rays. This section contains check boxes
for different filter criteria. You can select several of these check boxes at once, and rays
will only be shown if they satisfy all of the selected criteria.

Figure 2-7: Filter options for the Spot Diagram plot.

48 | CHAPTER 2: RAY OPTICS MODELING

For example, you can select Filter by wavelength to hide rays of all wavelengths except
a selected value; then select Filter by release feature index to only show rays that are
released by a specific feature (for example, Release from Grid 1). In many examples of
cameras and telescopes in the Application Libraries, each release feature corresponds
to a distinct field angle for light entering the optical system.

The Filter by number of reflections check box should only be used if you previously
selected Count reflections in the physics interface Additional Variables section.

CREATING AND ORGANIZING AN ARRAY OF SPOTS

Use the Layout section to organize rays into different spots based on release feature or
wavelength. If the model has multiple wavelengths or release features, this could result
in an array of different spots being plotted at the same time. This section also has
settings for the positioning of these different spots relative to each other in the
Graphics window. For example, you can adjust the Horizontal padding factor to increase
the horizontal spacing between the spots.

Figure 2-8: Layout options for the Spot Diagram plot.

Figure 2-9: Spot diagrams of a Petzval lens system, with three different wavelengths in the
same spot (left) or sorted into separate spots for each wavelength (right).

RESULTS ANALYSIS AND VISUALIZATION | 49

 AUTOMATICALLY LOCATING THE FOCAL PLANE
The Focal Plane Orientation section can be used to automatically locate a plane of best
focus for the optical system. When you click Create Focal Plane Dataset, an Intersection
Point 3D dataset will be created at a local minimum of the root mean square (rms) spot
size. If the rays stop at a surface in the model and the plane of best focus is behind this
surface, the automatic focal plane calculation might fail. Consider deleting the
absorbing boundary, or replacing it with a Wall with the Pass through condition.

Figure 2-10: Focal Plane Orientation options for the Spot Diagram plot.

ADDING TEXT ANNOTATIONS TO A SPOT DIAGRAM

The Annotations section can add text above or below the spots in the Graphics window.
These annotations can show the position, rms spot size, and wavelength (or
wavelength range) of each spot. There is also an option to draw a circle in the plot,
which could be used to draw the Airy disk for scale.

Figure 2-11: Annotations settings for the Spot Diagram plot.

50 | CHAPTER 2: RAY OPTICS MODELING

Optical Aberration Plot and Aberration Evaluation
The Optical Aberration plot can be added in a 2D Plot Group whereas the Aberration
Evaluation node is available under Derived Values. Both of these postprocessing features
serve a similar purpose: to quantify the monochromatic aberrations of an imaging
system and express them in terms of Zernike polynomial coefficients. In order to use
the Optical Aberration plot, the following prerequisites must be met:

• The model component must be 3D.

• An instance of the Geometrical Optics interface must be present and solved for.
• The Compute optical path length check box must be selected in the Geometrical
Optics Settings window before solving.
• An Intersection Point 3D dataset must be created. This dataset must point to a Ray
dataset. Optionally, this dataset can be automatically generated (see below). In the
Settings window for the Intersection Point 3D dataset, Hemisphere must be selected
from the Surface type list. The Center is the location of the focus and the Axis
direction points from the focus toward the center of the exit pupil.
Using the hemisphere defined in the Intersection Point 3D dataset, a Gaussian reference
sphere is defined. The optical path difference among intersection points with this
sphere is then expressed as a sum of Zernike polynomials.

AUTOMATIC DATASET GENERATION

Alternatively, using the Optical Aberration plot, you can use the Create Reference
Hemisphere Dataset button in the Focal Plane Orientation section to automatically
generate a reference hemisphere at the location of the rms focus. However, note that
the defocus term Z(2,0) might be larger for the rms focus than it would for some other
definitions of the image plane, such as the marginal focus.

The monochromatic aberrations in a double Gauss lens system are shown in

Figure 2-12. The circle on the left shows the sum of all Zernike coefficients. The circle
on the right ignores the defocus term Z(2,0) and the piston term Z(0,0). Usually,
lower-order Zernike terms can be reduced by correcting misalignments in the optical
system or adjusting the nominal focus. Higher-order terms are consequences of the
type of lens or mirror used in an optical system; for example, a single spherical lens will
always demonstrate some spherical aberration, for any position of the nominal focus.

The Optical Aberration plot always shows the optical path differences on a unit circle.
The color expression is the wavefront error in microns. In the circle on the right in
Figure 2-12, the largest contribution is from the spherical aberration term Z(4,0).

RESULTS ANALYSIS AND VISUALIZATION | 51

 Figure 2-12: Monochromatic aberrations in a double Gauss lens system.

For a list of Zernike polynomials and their derivation, see Monochromatic

Aberrations and the Zernike Polynomials in Theory for the Geometrical
Optics Interface.

Double Gauss Lens: Application Library path Ray_Optics_Module/

Lenses_Cameras_and_Telescopes/double_gauss_lens

Plotting Refractive Index and Abbe Number

When trying to plot the refractive index and dispersion of materials in a Geometrical
Optics model, users frequently encounter two challenges:

• There is a specific syntax that controls which variables are defined in domains, and
which variables are defined only on the rays that pass through the domains.
• Real-world media are dispersive (their refractive indices depend on wavelength), so
there is not just a single numeric value of the refractive index at each location.

52 | CHAPTER 2: RAY OPTICS MODELING

RAY VARIABLES AND DOMAIN VARIABLES
For an instance of the Geometrical Optics interface with the default name gop, the
variable for refractive index defined on rays is gop.nref, while the variable for
refractive index in the domain is called gop.nref_local. As the rays propagate
through domains in the geometry, the value of gop.nref is determined by evaluating
gop.nref_local at a geometry point corresponding to each ray’s current location.

Similarly, when a ray is refracted at the boundary between two different media, the
implementation of Snell’s law requires the variable gop.nref_local to be evaluated
in the media on the side of the incident ray and the refracted ray, at the point where
the incident ray intersects the boundary.

During postprocessing, a ray variable such as gop.nref can only be evaluated on rays,
not at arbitrary points in the geometry where a ray might not be present. For example,
trying to use gop.nref in a Slice plot will usually result in an error message such as:

Cannot evaluate expression.

Undefined variable.
- Variable: comp1.gop.nref
- Geometry: geom1
- Domains: 1
Failed to evaluate expression.
- Expression: comp1.gop.nref
- Plot: Slice 1

A better choice for the Slice plot would be the variable gop.nref_local, which is a
domain variable and therefore exists at every point in the domain, regardless of
whether a ray is present there or not.

However, a Slice plot of gop.nref_local will not always work. The plot will appear
correctly if the refractive index is entered as a constant value in each domain; that is, if
dispersion is neglected. If the model uses optical dispersion relations that define the
refractive index as a function of the wavelength of the rays, you may get an error
message such as:

Illegal operator context.

Failed to evaluate expression.

- Expression: noenv(comp1.gop.lambda0)
Failed to evaluate variable.
- Variable: comp1.gop.mp1.lambda0
- Defined as: noenv(comp1.gop.lambda0)

A more reliable means of plotting the refractive index of dispersive media is presented
in the next section.

RESULTS ANALYSIS AND VISUALIZATION | 53

 EVALUATION OF REFRACTIVE INDEX OF DISPERSIVE MEDIA
All real materials exhibit some degree of optical dispersion, or wavelength dependence
of the refractive index. You can enter a value or expression for the refractive index
directly, define a function (such as an Analytic, Interpolation, or Piecewise function), or
use one of the built-in dispersion relations such as the Sellmeier dispersion formula.

In addition, the Geometrical Optics interface may be configured to trace either

monochromatic or polychromatic light, by selecting the desired option from the
Wavelength distribution of released rays list in the physics interface Ray Release and
Propagation list.

Most of the optical glass material data in the Optical Material Libraries use a dispersion
formula to define the refractive index. (See Material Libraries for a more
comprehensive list of the available materials.) For example, if the Sellmeier dispersion
formula is used, then the variable gop.nref_local is defined as

gop.nref_local = sqrt(1
+gop.ODsma1*gop.lambda0norm^2/(gop.lambda0norm^2-gop.ODsma4)
+gop.ODsma2*gop.lambda0norm^2/(gop.lambda0norm^2-gop.ODsma5)
+gop.ODsma3*gop.lambda0norm^2/(gop.lambda0norm^2-gop.ODsma6)
)

where the variables gop.ODsma<1,2,3,4,5,6> are the Sellmeier dispersion

coefficients.

For many commercial optical glasses, the dispersion formula defines the
refractive index relative to air at a specified temperature and pressure, and
then this so-called relative refractive index is automatically converted to
an absolute refractive index with respect to vacuum. Here the conversion
between relative and absolute index has been omitted for brevity.

The variable gop.lambda0norm is the vacuum wavelength of the ray expressed in

micrometers,

gop.lambda0norm = gop.mp1.lambda0/1[um]

and this scoped variable gop.mp1.lambda0 is defined as

gop.mp1.lambda0 = noenv(gop.lambda0)

The special operator noenv() allows a two-way exchange of information between the
ray and the domain it is in, for the purpose of evaluating variables. It allows a ray
variable (such as the wavelength) to be used in the definition of a domain variable,
which in turn can be evaluated at the ray’s location. However, this construction only

54 | CHAPTER 2: RAY OPTICS MODELING

works if the domain variable is evaluated at the location of a ray, not at any arbitrary
location in the geometry, which is why a Slice plot of gop.nref_local fails for
dispersive media.

The Medium Properties node automatically defines domain variables for the refractive
index at several spectral lines:

• gop.nrefd is the refractive index at the helium d-line, a yellow spectral line of
helium with a wavelength of approximately 587.56 μm.
• gop.nrefF is the refractive index at the hydrogen F-line, a blue spectral line of
hydrogen with a wavelength of approximately 486.13 μm.
• gop.nrefC is the refractive index at the hydrogen C-line, a red spectral line of
hydrogen with a wavelength of approximately 656.28 μm.

The Abbe number gop.Vd is also available as a built-in postprocessing variable. Its
definition is

gop.Vd=(gop.nrefd-1)/(gop.nrefF-gop.nrefC)

For displaying the refractive index in Slice or Volume plots, these variables for refractive
index at specific wavelength values can be used more reliably than gop.nref_local.

Alternatively, you can evaluate the refractive index of a dispersive medium at any other
wavelength value using the subst() operator, which can replace the ray wavelength in
the dispersion formula with a fixed value. For example, another way to plot the d-line
refractive index (rather than the built-in variable gop.nrefd) is using the expression

subst(gop.nref_local,gop.mp1.lambda0,587.56[nm])

where the third argument to the subst() operator could be replaced with any other
numeric value.

SUMMARY OF AVAILABLE VARIABLES

The built-in postprocessing variables gop.nrefd, gop.nrefF, and gop.nrefC can be
used to plot the refractive indices of all materials at fixed wavelengths of 587.56 μm,
486.13 μm, and 656.28 μm, respectively. The built-in variable gop.Vd is the Abbe
number in terms of the refractive indices at these three wavelengths. You can also plot
the refractive indices of domains at any other fixed value of the wavelength using an
expression of the form

subst(gop.nref_local,gop.mp1.lambda0,<wavelength>)

where <wavelength> should be replaced with a numeric value.

RESULTS ANALYSIS AND VISUALIZATION | 55

 In the COMSOL Multiphysics Reference Manual:

• Spot Diagram
• Interference Pattern
• Optical Aberration
• Aberration Evaluation
• Ray (Plot)
• Ray Trajectories
• Filter for Ray and Ray Trajectories
• Phase Portrait
• Poincaré Map
• Ray (Dataset) and Datasets
• Ray Bin
• Intersection Point 2D and Intersection Point 3D
• Ray Evaluation and Derived Values, Evaluation Groups, and Tables
• Plot Groups and Plots

56 | CHAPTER 2: RAY OPTICS MODELING

Va r i a ble s a nd N on l oc al Cou p l i n gs
This section lists some built-in nonlocal couplings and variables for the Geometrical
Optics interface.

In this section:

• Ray Statistics
• Global Variables
• Variables for Average Ray Position
• Using Ray Detectors
• Nonlocal Couplings

Ray Statistics
The Geometrical Optics interface defines a number of special variables, some of which
can only be used during results processing. These variables can be found in the Ray
statistics section in the Add/Replace Expression menus.

In the variable names below, gop is the physics interface tag. If multiple instances of
the Geometrical Optics interface are different, this tag may differ.

The following variables are defined for each ray:

• Ray index gop.pidx. Each ray is assigned a unique index starting from 1 up to the
total number of rays. This expression can be used as an input argument to a
function. For example, randomnormal(gop.pidx) would sample pseudorandom
numbers from a normal distribution with unit variance. The random numbers
would be distinct for each ray.
• Ray release feature gop.prf. If there are multiple release features in a model, it is
useful to visualize which rays correspond to each release feature. Rays from each
release feature are assigned a unique integer, starting at 1. This variable can also be
used to filter ray trajectories in postprocessing so that only the rays released by a
specific feature are shown.
• Ray release time gop.prt. Usually, all primary rays are released at t = 0. To allow
other release times to be specified, select the Allow multiple release times check box
in the physics interface Advanced Settings section.

VARIABLES AND NONLOCAL COUPLINGS | 57

 If the Store ray status data check box is selected in the physics interface Additional
Variables section, then the following additional variables are created:

• The release time of a given ray (variable name gop.rti). Unlike gop.prt, this
works for secondary rays and thus allows for extraction of the time at which a
secondary ray was released. This includes reflected rays at material discontinuities
and higher diffraction orders at gratings.
• Stop time gop.st, when a ray gets absorbed at a boundary or annihilated by the Ray
Termination feature.
• Final status gop.fs. This indicates the status of a ray at a given point in time. When
used during postprocessing, the value always indicates the status of the ray at the last
time step. The value is an integer that has one of the following values:
- 0 for unreleased rays.
- 1 for rays that are still propagating.
- 2 for frozen rays.
- 3 for stuck rays.
- 4 for rays that have disappeared.

Global Variables
The following quantities can be evaluated globally, using the Global plot or Global
Evaluation feature. They could be used in a Ray Evaluation, but they would just give the
same value for every ray.

• Total number of rays, gop.Nt. This total includes both primary and secondary rays,
and includes rays that have disappeared or have not been released.
• Total number of rays in selection, gop.Nsel. If a selection has been applied to the
Ray dataset, the number of rays in that selection can be evaluated.
• Transmission probability, gop.alpha is the ratio of the number of rays in a selection
to the total number of rays.

For each ray release feature, the total number of rays released by feature is also defined.
This variable has physics feature scope, so, for example, a Release from Grid node with
tag relg1 would define a variable called gop.relg1.Ntf.

This global variable is uniquely defined for each release feature and gives the total
number of rays that are released by that feature. This includes rays that have
disappeared or have otherwise stopped propagating due to interaction with the
surrounding boundaries. It does not include any secondary rays.

58 | CHAPTER 2: RAY OPTICS MODELING

Variables for Average Ray Position
Some other variables for average ray position, both absolute and relative, are also
defined. These variables exist with both physics scope and physics feature scope; for
example, the variable gop.rrms is the RMS spot size for all rays in the model, but
gop.relg1.rrms is the RMS spot size only for rays produced by Release from Grid 1.

• Components of the average ray position gop.qavex, gop.qavey, and gop.qavez,

which are simple arithmetic means over the respective ray coordinates.
• Relative ray position gop.deltaqx, gop.deltaqy, and gop.deltaqz. These are the
ray positions relative to the average ray position. for example, the relative
x-component gop.deltaqx is qx-gop.qavex.
• RMS spot size gop.rrms.
• Maximum relative radial position gop.rmaxall.

Using Ray Detectors

A Ray Detector is a domain or boundary feature that provides information about rays
arriving on a set of selected domains or surfaces from a release feature. The feature
provides convenient expressions that can be used in the Filter attribute of the Ray
Trajectories plot, which allows only the rays which reach the ray detector selection to
be shown. The following variables are defined by the Ray Detector feature, assuming a
physics interface with tag gop and a detector with tag rd1:

• gop.rd1.Ntf is number of transmitted rays from the release feature to the ray
detector at the end of the simulation.
• gop.rd1.alpha is the transmission probability from the release feature to the ray
detector.
• gop.rd1.rL is a logical expression for ray inclusion. This can be set in the Filter
node of the Ray Trajectories plot in order to visualize the rays which connect the
release feature to the detector.

The Ray Detector feature defines new variables but does not otherwise affect the
solution. Therefore, you can add a Ray Detector to a completed model and evaluate the
variables it defines without recomputing the solution. To do so, right-click the Study
node and select Update Solution. The new variables described above will be immediately
available for results processing.

VARIABLES AND NONLOCAL COUPLINGS | 59

 Nonlocal Couplings
The purpose of a model is often to compute the sum, average, maximum value, or
minimum value of a quantity over a group of rays, such as the average intensity or the
maximum path length. An instance of the Geometrical Optics interface with physics
interface tag gop creates the following nonlocal couplings:

The built-in nonlocal coupling names have recently changed, but the
couplings with the old names are still available. In their descriptions, the
couplings with the old names are marked as “deprecated”. Both the old
and new names are listed in Table 2-1.

• gop.sum(expr) evaluates the sum of the expression expr over the rays. The sum
includes all rays that are active, frozen, or stuck to boundaries. It excludes rays that
have not yet been released and those that have disappeared.
• gop.sum_all(expr) evaluates the sum of the expression expr over all rays,
including rays those that are not yet released or have disappeared. Since the
coordinates of unreleased and disappeared rays are not-a-number (NaN), the sum
may return NaN if the model includes unreleased or disappeared rays. An expression
such as gop.sum_all(isnan(qx)) can be used to compute the total number of
unreleased and disappeared rays.
• gop.ave(expr) evaluates the average of the expression expr over the active,
frozen, and stuck rays. Unreleased and disappeared rays contribute to neither the
numerator nor the denominator of the arithmetic mean.
• gop.ave_all(expr) evaluates the average of the expression expr over all rays. It is
likely to return NaN if the model includes unreleased or disappeared rays.
• gop.max(expr) evaluates the maximum value of the expression expr over all active,
frozen, and stuck rays.
• gop.max_all(expr) evaluates the maximum value of the expression expr over all
rays.
The treatment of NaN values in nonlocal maximum couplings can be
platform-dependent, so use caution when evaluating the maximum over all rays
including disappeared and unreleased rays.
• gop.min(expr) evaluates the minimum value of the expression expr over the
active, frozen, and stuck rays.

60 | CHAPTER 2: RAY OPTICS MODELING

• gop.min_all(expr) evaluates the minimum value of the expression expr over all
rays.
The treatment of NaN values in nonlocal minimum couplings can be
platform-dependent, so use caution when evaluating the minimum over all rays
including disappeared and unreleased rays.
• gop.max(expr, evalExpr) evaluates the expression evalExpr for the ray that has
the maximum value of the expression expr out of all active, frozen, and stuck rays.
For example, the expression gop.max(gop.I, qx) would evaluate the x-coordinate
qx of the ray with the greatest intensity gop.I.

• gop.max_all(expr, evalExpr) evaluates the expression evalExpr for the ray

that has the maximum value of the expression expr for all rays, including
disappeared and unreleased rays.
• gop.min(expr, evalExpr) evaluates the expression evalExpr for the ray that has
the minimum value of the expression expr out of all active, frozen, and stuck rays.
For example, the expression gop.min(gop.I, qx) would evaluate the x-coordinate
qx of the ray with the minimum intensity gop.I.

• gop.min_all(expr, evalExpr) evaluates the expression evalExpr for the ray

that has the minimum value of the expression expr for all rays, including
disappeared and unreleased rays.

The built-in nonlocal couplings are summarized in Table 2-1.

TABLE 2-1: BUILT-IN NONLOCAL COUPLINGS FOR THE GEOMETRICAL OPTICS INTERFACE

NAME DEPRECATED NAME DESCRIPTION

gop.sum(a) gop.gopop1(a) Sum over rays

gop.sum_all(a) gop.gopop_all1(a) Sum over all rays
gop.ave(a) gop.gopaveop1(a) Average over rays
gop.ave_all(a) gop.gopaveop_all1(a) Average over all rays
gop.max(a) gop.gopmaxop1(a) Maximum over rays
gop.max_all(a) gop.gopmaxop_all1(a) Maximum over all rays
gop.min(a) gop.gopminop1(a) Minimum over rays
gop.min_all(a) gop.gopminop_all1(a) Minimum over all rays
gop.max(a,b) gop.gopmaxop1(a,b) At maximum over rays
gop.max_all(a,b) gop.gopmaxop_all1(a,b) At maximum over all rays
gop.min(a,b) gop.gopminop1(a,b) At minimum over rays
gop.min_all(a,b) gop.gopminop_all1(a,b) At minimum over all rays

VARIABLES AND NONLOCAL COUPLINGS | 61

 Material Libraries
The Ray Optics Module features a material library with material properties for optical
glasses, metals, polymers, gases, and other compounds to assist in ray optics simulation.

The Optical material library contains material property data from the following
companies’ products:

CDGM GLASS CO., LTD

H-FK61, H-FK61B, H-FK71, H-FK95N, H-QK1, H-QK3L, H-K1, H-K2, H-K3,
H-K5, H-K6, H-K7, H-K8, H-K9L, H-K9LGT, H-K9L*, H-K9LA, H-K10, H-K11,
H-K12, H-K50, H-K51, H-ZPK1A, H-ZPK2A, H-ZPK3, H-ZPK5, H-ZPK7,
H-BaK1, H-BaK2, H-BaK3, H-BaK4, H-BaK5, H-BaK6, H-BaK7, H-BaK7GT,
H-BaK7A, H-BaK8, H-ZK1, H-ZK2, H-ZK3, H-ZK3A, H-ZK4, H-ZK5, H-ZK6,
H-ZK7, H-ZK7A, H-ZK8, H-ZK9B, H-ZK9A, H-ZK10, H-ZK10L, H-ZK11,
H-ZK14, H-ZK20, H-ZK21, H-ZK50, H-ZK50GT, H-LaK1, H-LaK2A, H-LaK3,
H-LaK4L, H-LaK5A, H-LaK6A, H-LaK7A, H-LaK8B, H-LaK8A, H-LaK10,
H-LaK11, H-LaK12, H-LaK50A, H-LaK51A, H-LaK52, H-LaK53B, H-LaK53A,
H-LaK54, H-LaK59A, H-LaK61, H-LaK67, H-KF6, H-QF1, H-QF3, H-QF6A,
H-QF8, H-QF14, H-QF50, H-QF50A, H-QF56, H-F1, H-F2, H-F3, H-F4, H-F5,
H-F13, H-F51, H-BaF2, H-BaF3, H-BaF4, H-BaF5, H-BaF6, H-BaF7, H-BaF8,
H-ZBaF1, H-ZBaF3, H-ZBaF4, H-ZBaF5, H-ZBaF16, H-ZBaF20, H-ZBaF21,
H-ZBaF50, H-ZBaF52, H-ZF1A, H-ZF1, H-ZF2, H-ZF3, H-ZF4A, H-ZF4AGT,
H-ZF5, H-ZF6, H-ZF7LA, H-ZF7LAGT, H-ZF10, H-ZF11, H-ZF12, H-ZF13,
H-ZF13GT, H-ZF39, H-ZF50, H-ZF52, H-ZF52GT, H-ZF52TT, H-ZF62,
H-ZF62GT, H-ZF71, H-ZF71GT, H-ZF72A, H-ZF72AGT, H-ZF73, H-ZF73GT,
H-ZF88, H-ZF88GT, H-LaF1, H-LaF2, H-LaF3B, H-LaF4, H-LaF4GT, H-LaFL5,
H-LaF6LA, H-LaF7, H-LaF10LA, H-LaF50B, H-LaF51, H-LaF52, H-LaF53,
H-LaF54, H-LaF55, H-LaF62, H-ZLaF1, H-ZLaF2A, H-ZLaF3, H-ZLaF4LA,
H-ZLaF4LB, H-ZLaF50E, H-ZLaF50D, H-ZLaF51, H-ZLaF52A, H-ZLaF52,
H-ZLaF53B, H-ZLaF53BGT, H-ZLaF55D, H-ZLaF56B, H-ZLaF66,
H-ZLaF66GT, H-ZLaF68C, H-ZLaF68N, H-ZLaF69, H-ZLaF69A, H-ZLaF71,
H-ZLaF71AGT, H-ZLaF73, H-ZLaF75A, H-ZLaF75B, H-ZLaF76, H-ZLaF76A,
H-ZLaF78B, H-ZLaF89L, H-ZLaF90, H-ZLaF92, H-TF3L, H-TF5, D-FK61,
D-FK61A, D-FK95, D-QK3L, D-K9, D-K9GT, D-K59, D-PK3, D-ZPK1A,
D-ZPK3, D-ZPK5, D-ZPK7, D-ZK2, D-ZK3, D-ZK79, D-LaK5, D-LaK6,
D-LaK52, D-LaK70, D-ZF10, D-ZF93, D-LaF50, D-LaF050, D-LaF53, D-LaF79,
D-ZLaF50, D-ZLaF52LA, D-ZLaF61, D-ZLaF67, D-ZLaF81, D-ZLaF85A,

62 | CHAPTER 2: RAY OPTICS MODELING

D-ZLaF85L, H-QK3LGTi, H-K90GTi, QF50GTi, HWS1, HWS2, HWS3, HWS4,
HWS5, HWS6, HWS7, HWS9, and HWS27.

The foregoing materials are the property of CDGM GLASS CO., LTD. For product
information concerning these materials, see http://www.cdgmgd.com/. For other
questions concerning the materials, please contact [email protected].

Product catalog: www.cdgmglass.com/Glass-Finder

CORNING INCORPORATED
Corning® Gorilla® Glass Victus®, Corning® Gorilla® Glass 3, Corning® Gorilla®
Glass 5, Corning® Gorilla® Glass 6, Antimicrobial Corning® Gorilla® Glass,
Corning® Astra™ Glass, Corning® Lotus™ NXT Glass, Corning® Eagle XG® Glass,
ULE® Corning Code 7972 Ultra Low Expansion Glass, Corning® Unicrown Glass,
Corning® UV Clear Glass, Corning® 1.6/41 White High Index Glass, Corning®
1.7/35 White High Index Glass, Corning® 1.7/42 White High Index Glass,
Corning® 1.8/35 White High Index Glass, Corning® 1.9/31 White High Index
Glass, Corning® Calcium Fluoride (CaF2) - Code 9575, Corning® HPFS® 8655
Fused Silica, and Corning® HPFS® 7980 Fused Silica.

Material Data supplied by Corning Incorporated, corning.com. ASTRA, CORNING,

EAGLE XG, HPFS, GORILLA, LOTUS, ULE, and VICTUS are trademarks or
registered trademarks of Corning Incorporated.

OHARA INC.
S-FPL51, S-FPL53, S-FPL55, S-FPM 2, S-FPM 3, S-FPM 4, S-FPM 5, S-FSL 5,
S-BSL 7, S-BSM 2, S-BSM10, S-BSM14, S-BSM15, S-BSM16, S-BSM18, S-BSM25,
S-BSM28, S-BSM71, S-BSM81, S-NSL 3, S-NSL36, S-BAL 2, S-BAL 3, S-BAL12,
S-BAL14, S-BAL35, S-BAL41, S-BAL42, S-BAM 4, S-BAM12,S-BAH10, S-BAH11,
S-BAH27, S-BAH28, S-PHM52, S-PHM52Q, S-PHM53, S-TIL 1, S-TIL 2,
S-TIL 6, S-TIL25, S-TIL26, S-TIL27, S-TIM 2, S-TIM 5, S-TIM 8, S-TIM22,
S-TIM25, S-TIM27, S-TIM28, S-TIM35, S-TIH 1, S-TIH 3, S-TIH 4, S-TIH 6,
S-TIH10, S-TIH11, S-TIH13, S-TIH14,S-TIH18, S-TIH23, S-TIH53, S-TIH53W,
S-TIH57, S-LAL 7, S-LAL 7Q, S-LAL 8, S-LAL 9, S-LAL10, S-LAL12, S-LAL12Q,
S-LAL13, S-LAL14, S-LAL18, S-LAL19, S-LAL20, S-LAL21, S-LAL54, S-LAL54Q,
S-LAL58, S-LAL59, S-LAL61, S-LAM 2, S-LAM 3, S-LAM 7, S-LAM54, S-LAM55,
S-LAM60, S-LAM61, S-LAM66, S-LAM73, S-LAH51, S-LAH52, S-LAH52Q,
S-LAH53, S-LAH53V, S-LAH55V, S-LAH55VS, S-LAH58, S-LAH59, S-LAH60,
S-LAH60MQ, S-LAH60V, S-LAH63, S-LAH63Q, S-LAH64, S-LAH65V,
S-LAH65VS, S-LAH66, S-LAH71, S-LAH79, S-LAH88, S-LAH89, S-LAH92,

MATERIAL LIBRARIES | 63
 S-LAH93,S-LAH95, S-LAH96, S-LAH97, S-LAH98, S-LAH99, S-FTM16,
S-NBM51, S-NBH 5, S-NBH 8, S-NBH51, S-NBH52V, S-NBH53V, S-NBH55,
S-NBH56, S-NBH57, S-NBH58, S-NPH 1, S-NPH 1W, S-NPH 2, S-NPH 3,
S-NPH 4, S-NPH 5, S-NPH 7, L-BSL 7, L-BAL35, L-BAL35P, L-BAL42,
L-BAL42P, L-BAL43, L-TIM28, L-TIM28P, L-LAL13, L-LAL15, L-LAM60,
L-LAM69, L-LAH53, L-LAH84, L-LAH85V, L-LAH90, L-LAH91, L-LAH94,
S-FPL51Y, S-FSL5Y, BSL7Y, BAL15Y, BAL35Y, BSM51Y, PBL1Y, PBL6Y, PBL25Y,
PBL26Y, PBL35Y, PBM2Y, PBM8Y, PBM18Y, SK-1300, and SK-1310.

The foregoing Material are the property of Ohara Inc. For product information
concerning these Material, see www.ohara-inc.co.jp/en/product/catalog/. For other
questions concerning the Material, please contact Ohara Inc. through
www.ohara-inc.co.jp/en/contact/.

Product catalog: www.oharacorp.com/catalog.html

SCHOTT AG
F2, F2HT, F5, FK5HTi, K10, K7, LAFN7, LASF35, LF5, LF5HTi, LLF1, LLF1HTi,
N-BAF10, N-BAF4, N-BAF51, N-BAF52, N-BAK1, N-BAK2, N-BAK4,
N-BAK4HT, N-BALF4, N-BALF5, N-BASF2, N-BASF64, N-BK10, N-BK7,
N-BK7HT, N-BK7HTi, N-F2, N-FK5, N-FK51A, N-FK58, N-K5,
N-KF9,N-KZFS11, N-KZFS2, N-KZFS4, N-KZFS4HT, N-KZFS5, N-KZFS8,
N-LAF2, N-LAF21, N-LAF33, N-LAF34, N-LAF35, N-LAF7, N-LAK10,
N-LAK12, N-LAK14, N-LAK21, N-LAK22, N-LAK33B, N-LAK34, N-LAK7,
N-LAK8, N-LAK9, N-LASF31A, N-LASF40, N-LASF41, N-LASF43, N-LASF44,
N-LASF45, N-LASF45HT, N-LASF46A, N-LASF46B, N-LASF9, N-LASF9HT,
N-PK51, N-PK52A, N-PSK3, N-PSK53A, N-SF1, N-SF10, N-SF11, N-SF14,
N-SF15, N-SF2, N-SF4, N-SF5, N-SF57, N-SF57HT, N-SF57HTultra, N-SF6,
N-SF66, N-SF6HT, N-SF6HTultra, N-SF8, N-SK11, N-SK14, N-SK16, N-SK2,
N-SK2HT, N-SK4, N-SK5, N-SSK2, N-SSK5, N-SSK8, N-ZK7, N-ZK7A, P-BK7,
P-LAF37, P-LAK35, P-LASF47, P-LASF50, P-LASF51, P-SF68, P-SF69, P-SF8,
P-SK57, P-SK57Q1, P-SK58A, P-SK60, SF1, SF10, SF11, SF2, SF4, SF5, SF56A,
SF57, SF57HTultra, SF6, and SF6HT.

The foregoing materials are the property of Schott AG. For product information
concerning these materials, see https://www.schott.com/advanced_optics/. For
other questions concerning the materials, please contact [email protected].

Product catalog: www.schott.com/products/optical-glass

64 | CHAPTER 2: RAY OPTICS MODELING

ORGANIC, INORGANIC, AND MISCELLANEOUS
The remaining categories of Organic, Inorganic, and Miscellaneous are derived from the
refractive index database. Authors: Mikhail Polyanskiy. Availability: https://
refractiveindex.info/. License: Public Domain through CCO 1.0 Universal.

MATERIAL LIBRARIES | 65
66 | CHAPTER 2: RAY OPTICS MODELING
 3

Ray Optics Interfaces

This chapter describes the physics interfaces found under the Optics>Ray Optics
branch ( ).

In this chapter:

• The Geometrical Optics Interface

• Theory for the Geometrical Optics Interface

67
 T he G e o m e t r i c al O p t i cs In t erface
The Geometrical Optics (gop) interface ( ), found under the Optics>Ray Optics
branch ( ) when adding a physics interface, computes the paths of electromagnetic
waves in the high frequency limit. The wavefronts are treated as rays that can propagate
through optically large geometries and reflect or refract at boundaries. The
Geometrical Optics interface includes built-in options to compute the ray intensity,
polarization, phase, and optical path length. It is also fully compatible with deformed
geometries such as systems with thermal stress.

When this physics interface is added, these default nodes are also added to the Model
Builder: Medium Properties, Material Discontinuity, and Ray Properties. Then, from the
Physics toolbar, add other nodes that implement, for example, light sources or
boundary conditions. You can also right-click Geometrical Optics to select physics
features from the context menu.

Geometrical Optics Physics Interface Settings

The Label is the physics interface name. The default is Geometrical Optics.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is gop.

RAY RELEASE AND PROPAGATION

The settings in this section affect the way in which primary and secondary rays are
released.

Wavelength Distribution of Released Rays

Select an option from the Wavelength distribution of released rays list: Monochromatic
(the default), Polychromatic, specify vacuum wavelength, or Polychromatic, specify
frequency.

• For Monochromatic, all rays in the model have the same vacuum wavelength and
frequency, which is entered as a value or expression in the Ray Properties node.

68 | CHAPTER 3: RAY OPTICS INTERFACES

• For Polychromatic, specify vacuum wavelength, a degree of freedom is allocated for
the vacuum wavelength of each ray in the model. These degrees of freedom are
initialized when the rays are released and are controlled by the Vacuum Wavelength
section in the settings for the ray release features, such as Release from Grid.
• For Polychromatic, specify frequency, a degree of freedom is allocated for the
frequency of each ray in the model. These degrees of freedom are initialized when
the rays are released, and are controlled by the Initial Ray Frequency section in the
settings for the ray release features, such as Release from Grid.

Maximum Number of Secondary Rays

The Maximum number of secondary rays limits the total number of secondary rays that
can be created by capping them at the specified number. Secondary rays are released
when an existing ray is subjected to certain boundary conditions. For example, when
a ray arrives at a Material Discontinuity or Scattering Boundary between different
media, the incident ray is refracted and a reflected ray is created; the degrees of freedom
for this reflected ray are taken from one of the available secondary rays, which are
preallocated when the study begins. Secondary rays are also used to model the
interaction of rays with diffraction gratings, using the Grating boundary condition.

If an insufficient number of secondary rays are preallocated, a reflected ray may not be
released when an existing ray undergoes refraction, even if some radiation should be
reflected at the material discontinuity, and a warning message will be generated by the
study. However, if a very large number of secondary rays are preallocated, then the
number of degrees of freedom may become unnecessarily large. Thus, the Maximum
number of secondary rays should only be large enough that all reflected rays which
significantly affect the solution can be released. Rays that undergo total internal
reflection at material discontinuities are not considered secondary rays and do not
require extra preallocated degrees of freedom.

Use Geometry Normals for Ray-Boundary Interactions

By default the Use geometry normals for ray-boundary interactions check box is cleared.
Then, whenever rays are reflected or refracted at boundaries, the surface normal is
computed using the underlying mesh discretization of the boundary rather than the
exact shape of the geometry itself. If this check box is selected, then the surface normal
will instead be computed from an analytic representation of the geometry surfaces, if
such an analytic representation can be obtained.

For the simple case of ray reflection by a parabolic edge in 2D, three example plots are
shown in Figure 3-1 below. In the leftmost plot, linear geometry shape order has been
specified; in other words, Linear has been selected from the Geometry shape order list

THE GEOMETRICAL OPTICS INTERFACE | 69

 in the settings for the model component. The left figure also uses mesh normals. The
reflected ray directions are visibly inaccurate because the boundary mesh is very coarse.
The center figure uses Linear geometry shape order and geometry normals. Although
the mesh is equally coarse, the reflected ray directions are much more accurate. The
rightmost figure uses Quadratic shape order; because the edge is parabolic, this shape
order results in reflected ray directions that are exact (to within machine precision) no
matter whether mesh normals or geometry normals are used, because quadratic
elements can perfectly represent a parabola.

Figure 3-1: Comparison of ray reflections at a parabolic boundary, using various

definitions of the mesh normal.

The Use geometry normals for ray-boundary interactions check box has no effect on the
solution if the mesh can deform. This is true, for example, when the geometry is
subjected to structural loads or thermal stresses. Then the mesh normal is always used.

Only Store Accumulated Variables in Solution

By default the Only store accumulated variables in solution check box is cleared. If this
check box is selected, then the degrees of freedom associated with individual rays (such
as the ray position, wave vector, intensity, power, and normalized Stokes parameters)
will be solved for but not stored in the solution data. However, certain physics features
define degrees of freedom in domains or on boundaries rather than on the individual
rays, and these degrees of freedom will be retained in the solution. These quantities,
called accumulated variables, are defined by the following features:

• Fluence Rate Calculation (domain)

• Deposited Ray Power (Domain)
• Deposited Ray Power (Boundary)
• Accumulator (Domain)
• Accumulator (Boundary)
• Nonlocal Accumulator (subnode to Release from Boundary)
• Ray Heat Source (multiphysics coupling)

70 | CHAPTER 3: RAY OPTICS INTERFACES

It might be useful to discard ray degrees of freedom and only retain the accumulated
variables when a very large number of rays are required to accurately solve for the
accumulated variable. By only storing the accumulated variables, the file size can be
significantly reduced.

While the Only store accumulated variables in solution check box is selected, if you run
a study for the first time, the default Ray dataset and Ray Trajectories plot will not be
created. If you later clear the check box and recompute the study, you may need to add
these Results features manually.

Annular Ultraviolet Reactor: Application Library path

Ray_Optics_Module/Ultraviolet_Sterilization/annular_ultraviolet_reactor

Annular Ultraviolet Reactor with Particle Tracing: Application

Library path Ray_Optics_Module/Ultraviolet_Sterilization/
annular_ultraviolet_reactor_particle

MATERIAL PROPERTIES OF EXTERIOR AND UNMESHED DOMAINS

Use this section to control the properties of the void exterior to the model geometry
as well as the properties of domains outside of the selection of the Geometrical Optics
interface. For further details about what constitutes a “void” region, see the Geometry
and Meshing section of the Ray Optics Modeling chapter.

Optical dispersion model

Select an option from the Optical dispersion model list: Absolute vacuum, Specify
refractive index (the default), or Air, Edlén (1953). This setting controls the refractive
index in the empty space outside the geometry, called the void region. It also controls
the refractive index for domains that are not included in the selection of the
Geometrical Optics interface.

• For Absolute vacuum, the Refractive index of exterior domains, real part
(dimensionless) is next = 1. If ray power and/or intensity is being computed, the
void region is assumed to be nonabsorbing.
• For Specify refractive index, enter a value or expression for the Refractive index of
exterior domains, real part. The default is next = 1. Enter a user defined value for
Refractive index of exterior domains, imaginary part if power and/or intensity is being
computed. The default is kext = 0.
• For Air, Edlén (1953), the temperature and pressure dependent Refractive index of
exterior domains, real part is computed using the model described in Ref. 2. The

THE GEOMETRICAL OPTICS INTERFACE | 71

 model accepts user defined values of the Exterior domain temperature
Text (SI unit: K) and Exterior domain pressure Pext (SI unit: Pa). The defaults are
Text = 293.15K and Pext = 1 atm. See Optical Dispersion Models, The Refractive
Index of Air for details.

For any type of optical dispersion model in the exterior and unmeshed domains, the
algorithm to trace rays assumes that the medium is homogeneous. If a graded-index
material is present, then its domain should be included in the selection of the
Geometrical Optics interface and the spatial dependence of the refractive index should
be controlled from the Medium Properties node.

INTENSITY COMPUTATION
The settings in this section control the treatment of ray intensity and polarization.
These settings are also important in multiphysics applications such as ray heating.

Intensity Computation
Select an option from the Intensity computation list: None (the default), Compute
intensity, Compute power, Compute intensity and power, Compute intensity in graded
media, or Compute intensity and power in graded media. For None no additional variables
are computed along the rays. For other options, the ray intensity or ray power is
computed as described below.

• For Compute intensity, auxiliary dependent variables are used to compute the
intensity and polarization of each ray. For a complete list of the auxiliary dependent
variables that are defined, see Intensity, Wavefront Curvature, and Polarization in
Theory for the Geometrical Optics Interface. This option is only valid for
computing intensity in homogeneous media; for graded-index media, use Compute
intensity in graded media instead. The refractive index can still change
discontinuously at boundaries, where the Fresnel equations are automatically used
to compute the intensity of the reflected and refracted rays.
The default name for the intensity variable is gop.I.
• For Compute power, the total power transmitted by each ray is defined as an auxiliary
dependent variable. Information about the ray polarization is also available. The
Deposited Ray Power (Boundary) subnode is available for the Wall feature. In
addition, if a heat transfer interface such as the Heat Transfer in Solids interface is
included in the model, the Ray Heat Source multiphysics node can be used to
compute the heat source due to attenuation of rays within domains.
The default name for the power variable is gop.Q.

72 | CHAPTER 3: RAY OPTICS INTERFACES

• The option Compute intensity and power functions as a combination of the Compute
intensity and Compute power options.
• The Compute intensity in graded media option functions like Compute intensity but
is valid for both homogeneous and graded-index media. If all media in the model
are homogeneous then it is recommended to select Compute intensity instead, since
it is the more accurate method.
• The Compute intensity and power in graded media option functions like Compute
intensity and power but is valid for both homogeneous and graded-index media. If
all media in the model are homogeneous then it is recommended to select Compute
intensity and power instead, since it is the more accurate method.

Compute Phase
The Compute phase check box is only shown if the ray intensity is computed. Select the
check box to allocate an auxiliary dependent variable for the phase of each ray. When
the phase of each ray is computed, it is possible to plot interference patterns and
visualize the instantaneous electric field components of polarized rays in
postprocessing. When this check box is selected, the total number of degrees of
freedom increases by 1 per ray. This option is based on the assumption that the
coherence length of the radiation is arbitrarily large.

Use Corrections for Strongly Absorbing Media

The Use corrections for strongly absorbing media check box is shown if the ray intensity
is computed. Select the check box to accurately model reflection and refraction of rays
at boundaries between strongly absorbing media, in which the imaginary part of the
refractive index is very large. This option allocates two or three auxiliary dependent
variables per ray based on space dimension. For more information about the way this
option affects the intensity calculation, see Refraction in Strongly Absorbing Media in
Theory for the Geometrical Optics Interface.

Tolerance for Curvature Tensor Computation

When the Intensity computation is set to Compute intensity in graded media or Compute
intensity and power in graded media enter a Tolerance for curvature tensor computation
(dimensionless). This tolerance is used internally when computing the principal radii
of curvature of propagating wavefronts in a graded medium. A larger tolerance makes
the solution less accurate but more stable.

THE GEOMETRICAL OPTICS INTERFACE | 73

 ADDITIONAL VARIABLES
The options in this section can be used to solve for additional variables other than those
that are used to compute intensity or define the ray frequency. By default, all of the
check boxes in this section are cleared so that these variables are not solved for.

Compute Optical Path Length

Select the Compute optical path length check box to allocate an auxiliary dependent
variable for the optical path length of each ray. The default variable name is gop.L. It
is possible to reset the optical path length to 0 when rays interact with boundaries.

Count Reflections
Select the Count reflections check box to allocate an auxiliary dependent variable for the
number of reflections undergone by each ray, including reflections by the Wall and
Material Discontinuity features. The default variable name is gop.Nrefl. The auxiliary
variable begins at 0 when rays are released and is incremented by 1 every time a ray is
reflected at a boundary.

Store Ray Status Data

Select the Store ray status data check box to add new variables for quantities that
cannot necessarily be recovered from the ray trajectory data alone. This is especially
true if automatic remeshing is used in a model. The variables created include the
following, all of which would be preceded by the physics interface tag (for example,
gop.):

• The release time of a given ray (variable name rti).

• The time at which a ray stopped at a boundary (variable name st).
• The final status of the ray (variable name fs). This indicates the status of a ray at the
final time step. The value is an integer which has one of the following values:
- 0 for unreleased rays
- 1 for rays that are still in the modeling domain
- 2 for frozen rays
- 3 for stuck rays
- 4 for rays that have disappeared

74 | CHAPTER 3: RAY OPTICS INTERFACES

To summarize the total number of rays having each final status, the following global
variables are also defined.
TABLE 3-1: GLOBAL VARIABLES BASED ON RAY STATUS

NAME DESCRIPTION

fac Fraction of active rays at final time step

fds Fraction of disappeared rays at final time step
ffr Fraction of frozen rays at final time step
fse Fraction of secondary rays released
fst Fraction of stuck rays at final time step
nsr Number of released secondary rays
nsrf Number of released secondary rays at final time step
nsu Number of unreleased secondary rays
nsuf Number of unreleased secondary rays at final time step

The global variable names in Table 3-1 all take the unreleased secondary rays into
account. For example, suppose an instance of the Geometrical Optics interface
includes 100 primary rays and 100 allocated secondary rays. At the last time step,
suppose that 80 of the primary rays have disappeared at boundaries and that 40
secondary rays have been emitted, all of which are still active. Then the variable
gop.fac, the fraction of active rays at the final time step, would have the value
(20 + 40)/(100 + 100) or 0.3.

ADVANCED SETTINGS
This section is only shown when Advanced Physics Options are enabled (click the Show
More Options button in the Model Builder toolbar, and select Advanced Physics
Options in the Show More Options dialog box). These options are hidden by default
because it is only necessary to modify them under very specific circumstances.

Wall Accuracy Order

The Wall accuracy order sets the accuracy order of the time stepping used for time steps
during which a ray-wall interaction happens. Select an order of 1 to use a forward Euler
step and compute the direction or ray propagation both before and after the wall
interaction. Select an order of 2 (the default) to use a second-order Taylor method to
compute the trajectory before the wall interaction. After the ray-wall interaction a
second-order Runge-Kutta method is used. Usually this setting has no effect on the
ray tracing calculation because the rays move along straight lines, but the higher wall
accuracy order is more accurate if graded-index media are present.

THE GEOMETRICAL OPTICS INTERFACE | 75

 Arguments for Random Number Generation
Select an option from the Arguments for random number generation list: Generate unique
arguments (the default), Generate random arguments, or User defined. This setting
determines how the additional argument to the random function is defined in features
such as the Wall boundary condition with the Diffuse scattering wall condition.
Typically the random numbers are functions of the ray index, time, a unique input
argument for different variable definitions, and another argument i, defined as follows:

• For Generate unique arguments the additional argument is based on the position of
each node in the Model Builder. As a result, random numbers generated in different
nodes are created independently of each other.
• For Generate random arguments the additional argument is randomly created each
time the study is run.
• For User defined the additional argument is defined by a user input in the Settings
window for each feature. Uncorrelated sets of random numbers can be obtained by
running a Parametric Sweep for different values of i.

Note that the random function is a pseudorandom number generator (PRNG), not a
true random number generator (RNG) in the sense that an observer with perfect
knowledge of the algorithm and a history of previously generated values could predict
the ensuing values. However, PRNG is sufficient for most purposes.

For any model using PRNG, it is recommended to test for statistical convergence by
varying the number of rays and observing the effect on some ray statistics, such as
average final ray position or transmission probability. This is analogous to the mesh
refinement studies that are typically recommended to validate finite element models.

Allow Multiple Release Times

The Allow multiple release times check box, which is cleared by default, allows an array
of release times for the rays to be specified in any of the ray release features. If the check
box is cleared, all rays are released at time t = 0.

Exclude Jacobian Contributions from Ray Equations

By default the Exclude Jacobian contributions from ray equations check box is cleared. If
this check box is selected, then expressions for the time derivatives of the degrees of
freedom on rays are not included in the Jacobian matrix. Excluding contributions to
the Jacobian can significantly improve solution time if the number of rays in the model
is very large. The drawback is that the Jacobian is not exact, so smaller time steps or
path length intervals may be needed to obtain an accurate solution. This drawback is
most noticeable when computing the ray intensity or power in an absorbing medium.

76 | CHAPTER 3: RAY OPTICS INTERFACES

Treat Unmeshed Releasing Entities as Errors
By default the Treat unmeshed releasing entities as errors check box is selected. Certain
ray release features, such as the Release and Release from Boundary features, compute
the ray release positions based on a geometric entity, which must be meshed. While this
check box is selected, failure to mesh the selections of such features will result in an
error message when running any study in the model. If this check box is cleared, failure
to mesh the selections of such features will result in them releasing zero rays, but will
otherwise not interrupt the computation. Other physics features will still be able to
release rays normally.

Maximum Number of Wall Interactions per Time Step

Enter a value for the Maximum number of wall interactions per time step. The default
value is 1000. If a ray undergoes more than the specified number of boundary
interactions in a single time step taken by the solver, the ray will disappear. This is
included as a safeguard to prevent rays from getting stuck in infinite loops if the time
between successive ray-wall interactions becomes infinitesimally small.

DEPENDENT VARIABLES
The dependent variables (field variables) are the components of the Ray position and
Wave vector. The name can be changed but the names of fields and dependent variables
must be unique within a model. In 3D, the default ray position vector components are
qx, qy, and qz; and the default wave vector components are kx, ky, and kz.

Theory for the Geometrical Optics Interface

List of Geometrical Optics Interface Physics Features

The Geometrical Optics Interface has the following nodes and subnodes (listed in
alphabetical order) available from the Physics ribbon toolbar (Windows users), Physics
context menu (Mac or Linux users), or right-click to access the context menu (all
users).

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

THE GEOMETRICAL OPTICS INTERFACE | 77

 • Accumulator (Boundary) • Nonlocal Accumulator
• Accumulator (Domain) • Photometric Data Import
• Auxiliary Dependent Variable • Ray Continuity
• Axial Symmetry • Ray Detector
• Blackbody Radiation • Ray Properties
• Circular Wave Retarder • Ray Termination
• Cross Grating • Release
• Deposited Ray Power (Boundary) • Release from Data File
• Deposited Ray Power (Domain) • Release from Boundary
• Diffraction Order (Grating) • Release from Edge
• Diffraction Order (Cross Grating) • Release from Electric Field
• Fluence Rate Calculation • Release from Far-Field Radiation
• Gaussian Beam Pattern

• Grating • Release from Grid

• Ideal Depolarizer • Release from Grid on Axis

• Ideal Lens • Release from Point

• Illuminated Surface • Release from Point on Axis

• Linear Polarizer • Release from Symmetry Axis

• Linear Wave Retarder • Scattering Domain

• Material Discontinuity • Scattering Boundary

• Medium Properties • Solar Radiation

• Mirror • Thin Dielectric Film

• Mueller Matrix • Wall

Ray Properties
The Ray Properties node is added by default. Only one instance of this node can exist
for each physics interface. The Ray Properties node defines the equations for ray
propagation and may also include inputs for some ray properties.

RAY PROPERTIES
This section is shown when Monochromatic is selected from the Wavelength distribution
of released rays list in the physics interface Ray Release and Propagation section.

78 | CHAPTER 3: RAY OPTICS INTERFACES

Otherwise the frequency or vacuum wavelength of released rays is controlled separately
by each ray release feature, such as Release or Release from Grid.

Select an option from the Ray property specification list: Specify vacuum wavelength (the
default) or Specify frequency.

• For Specify vacuum wavelength enter a Vacuum wavelength λ0 (SI unit: m). The
default is 660 nm.

• For Specify frequency enter a Ray frequency v (SI unit: Hz). The default is
4.54 × 1014 Hz.

Medium Properties
Use the Medium Properties node to specify the refractive index of the medium. An
instance of this feature is created by default, including all of the selected domains for
the physics interface. For the region outside the geometry and for any domains not
included in the physics interface selection, the refractive index is instead controlled by
the Optical dispersion model in the physics interface Material Properties of Exterior and
Unmeshed Domains section.

MEDIUM PROPERTIES
Use the settings in this section to specify how the real part of the refractive index is
defined. If the ray intensity or power is computed, then you can also specify the
imaginary part of the refractive index, which is used to describe absorption within the
medium.

Refractive Index of Domains

Choose an option from the Refractive index of domains list:

• If Specify absolute refractive index (the default) is selected, the Refractive index, real
part can be taken From material, or it can be entered directly. The default value is 1.
This index is considered absolute, that is, relative to vacuum.
• If Specify relative refractive index is selected, the Refractive index, real part can be
taken From material, or it can be entered directly. The default value is 1. Then enter
the Reference temperature Tref,rel (SI unit: K, default 293.15 K) and the Reference
pressure Pref,rel (SI unit: Pa, default 0). The given refractive index is understood to
be relative to air; that is, the absolute refractive index is the product of the specified
value with the refractive index of air at the reference temperature and pressure.

THE GEOMETRICAL OPTICS INTERFACE | 79

 • If Get dispersion model from list is selected, choose an option from the Optical
dispersion model list. The following options are available:
- Cauchy,
- Conrady,
- Herzberger,
- Schott,
- Schott extended,
- Sellmeier,
- Sellmeier modified, type 1,
- Sellmeier modified, type 2, and
- Temperature-dependent Sellmeier.

• If Get dispersion model from material is selected, the dispersion model in each
domain is automatically deduced from the Material nodes and their selections. This
allows different optical dispersion models to be used in different domains in the
geometry, using only a single Medium Properties node.

For example, suppose that a model contains two lenses consisting of two different
glasses, and that the first glass is defined using Sellmeier coefficients whereas the
second glass is defined using Schott coefficients. Then selecting Get dispersion model
from material will automatically express the refractive index using the Sellmeier
equation in the first lens and the Schott (polynomial) equation in the second. An
alternative way to use different optical dispersion models in different domains is to
use more than one instance of the Medium Properties node and select the dispersion
models manually.
For more information on each optical dispersion model, see Table 3-3 in Theory for
the Geometrical Optics Interface, Optical Dispersion Models section.
The coefficients for each of these dispersion models are taken From material by
default. Alternatively, User defined coefficients may be entered.

For the built-in optical dispersion models, the wavelength is always

assumed to be in units of microns (μm). For example, in the Schott
(polynomial) model, the coefficients A0, A1, A2, A3, and so on have units
of 1, μm, μm2, μm3, and so on. If another source were to provide these
coefficients using nanometers instead of microns, then some manual
conversion would be required.

80 | CHAPTER 3: RAY OPTICS INTERFACES

Similar to the option Specify relative refractive index, most of the built-in optical
dispersion models define the refractive index relative to air, and therefore they all
require a reference temperature and pressure to be specified. The only exception is the
Temperature-dependent Sellmeier, in which the refractive index is assumed to be
absolute, such that n = 1 is the refractive index of an ideal vacuum. To use any other
optical dispersion model to define an absolute index, set the reference pressure to zero.

Using the Temperature-dependent Sellmeier model also disables the settings for
selecting a Thermo-optic dispersion model (see the following subsection).

The conversion from relative to absolute refractive index is made using the Edlén
model (Ref. 2) for the refractive index of air. See Optical Dispersion Models for further
details.

Optical Attenuation Model

By default the Optical attenuation model list is not shown. However, it will be shown if
you select any option except None from the Intensity computation list in the physics
interface settings. This is because the speed of light in a medium only depends on the
real part of the refractive index, whereas the imaginary part affects the attenuation of
ray intensity or power if they are solved for.

The standard used by the Geometrical Optics interface is to treat absorbing media as
having complex-valued refractive indices of the form n - iκ, where n and κ are real
dimensionless numbers and κ > 0. The case κ < 0 would constitute a gain medium.

The Optical attenuation model list has the following options:

• For Extinction coefficient, specify the value of κ directly. This can be taken From
material or User defined; if User defined the default is 0. The Geometrical Optics
interface only recognizes isotropic κ where the absorption rate does not depend on
ray polarization.
• For Attenuation coefficient, specify the value of the Attenuation coefficient α
(SI unit: 1/m) which is related to κ by α = 4πκ/λ0. The default is 1 1/m.
• For Internal transmittance, 2 mm sample thickness, enter a value or expression for the
internal transmittance of a 2 mm thick sample of the material, neglecting Fresnel
losses. Internal transmittances are dimensionless and are denoted τi,d, where d is
replaced with the sample thickness, for example, τi,2. The default is 0.9. The
relationship between τi,d and κ is κ = -(λ0 log τi,d)/(4πd).
• The remaining options are internal transmittances at other values of the sample
thickness. Options for 5 mm, 10 mm, and 25 mm samples are available. Otherwise
the settings are the same as the Internal transmittance, 2 mm sample thickness option.

THE GEOMETRICAL OPTICS INTERFACE | 81

 Conversions Between Optical Attenuation Models
The material properties for different optical attenuation models are used to define the
corresponding properties for other attenuation models automatically. For example, if
you select Extinction coefficient and try to take the value of κ From material, but the
selected material only has internal transmittance data, then the value of the internal
transmittance data is automatically converted to a value of κ based on the wavelength.

If the selected material has internal transmittance data for two or more sample
thicknesses, the data for the greatest thickness is automatically used unless otherwise
specified. This default behavior is intended to give the best possible numerical
precision when the internal transmittance values are close to 1, implying weak
absorption. However, it may be preferable to use internal transmittance data for a
smaller sample thickness in IR or UV applications when most of the energy is absorbed.

How to Automatically Detect Optical Dispersion Models

The option Get dispersion model from material is unique because it allows a single
Medium Properties node to detect different optical dispersion models from a large
number of different materials having different domain selections. For example, the
materials used in the Petzval Lens tutorial are shown in Figure 3-2 below. To
automatically detect the optical dispersion models of the loaded glasses, take the
following steps:

1 Add materials to the model using the Optical Material Library. Most of the glasses
in this material library use an optical dispersion model to define the refractive index.
As the materials are added, the coefficients used by the optical dispersion model are
automatically loaded.
2 Locate the default Medium Properties node.
3 From the Refractive index of domains list, select Get dispersion model from material.
If the glasses loaded in step 1 also provide thermo-optic coefficients, it is important
to specify an accurate value of the Temperature as well.

82 | CHAPTER 3: RAY OPTICS INTERFACES

 1

Figure 3-2: Workflow for automatically detecting optical dispersion models from the glasses
in an optical prescription.

Thermo-Optic Dispersion Models

A temperature-dependent offset in the Refractive index may be specified using a
thermo-optic dispersion model. Two options are possible:

• None (the default): no offset will be applied.

• Schott thermo-optic: this is the only built-in thermo-optic dispersion model
available. The coefficients used to compute a temperature dependent offset in the
refractive index can either be taken From material (the default), or User defined. A
reference temperature is also required. This may also be taken From material (the
default), or User defined.

For further details, see Thermo-Optic Dispersion Models.

THE GEOMETRICAL OPTICS INTERFACE | 83

 Material Discontinuity
The Material Discontinuity node is the default feature on all boundaries.

The Geometrical Optics interface always applies reflection and refraction at boundaries
between different media using a deterministic ray splitting approach. The direction of
the refracted ray is computed using Snell’s law, based on the refractive index on either
side. If extra degrees of freedom have been allocated for secondary rays, a reflected ray
is also released. If the incident ray undergoes total internal reflection, no refracted ray
is produced and no secondary rays are needed to release the reflected ray.

You can use the Rays to Release section (see below) to decide whether to split the
incident ray into reflected and refracted rays, or to ignore the reflected ray.

If the ray intensity or power is solved for in the model, the Material Discontinuity feature
computes the new values of these variables for the reflected and refracted rays. This
reinitialization uses the Fresnel Equations while accounting for the incident ray
polarization and the influence of any dielectric films on the surface reflectance.

The maximum number of reflected rays can be controlled via the

Maximum number of secondary rays text field, which is found under the Ray
Release and Propagation section for the physics interface.

If a reflected ray should be produced, but all of the preallocated secondary

rays have already been released, the study will produce a Warning node.

The Accumulator (Boundary) subnode is available from the context menu (right-click
the parent node) or from the Physics toolbar, Attributes menu. The Thin Dielectric
Film subnode is also available if the following conditions are met:

• The ray intensity or power is being computed, and

• One of the following options is selected from the Thin dielectric films on boundary
list in the Coatings section: Add layers to surface or Add layers to surface, repeating.

COATINGS
This section is available if ray intensity or power is being computed in the model. Use
the options in this section to add thin dielectric layers to the boundary between the
media. These thin dielectric layers are used, together with the refractive indices on
either side, to modify the Fresnel equations for the reflection and transmission
coefficients at the boundary.

84 | CHAPTER 3: RAY OPTICS INTERFACES

The thicknesses of the thin dielectric layers must be small relative to the coherence
length of the radiation. They are usually comparable in length scale to the free-space
wavelength.

Select an option from the Thin dielectric films on boundary list. The default is None. The
following options are available.

• None: no dielectric films are on the boundary. The intensity of the reflected and
refracted light is computed solely using the Fresnel equations with the refractive
indices of the two adjacent domains.
• Add layers to surface: you can add any number of thin dielectric films to the
boundary by adding one or more Thin Dielectric Film subnodes; right click the
Material Discontinuity node or select Thin Dielectric Film from the Physics toolbar,
Attributes menu. If multiple thin films are added to a single surface, they are
arranged in the same order as the corresponding Thin Dielectric Film subnodes in the
Model Builder, from the upside of the boundary toward the downside.

Anti-Reflective Coating with Multiple Layers: Application Library path

Ray_Optics_Module/Prisms_and_Coatings/antireflective_coating_multilayer

• Add layers to surface, repeating: this option functions like the Add layers to surface
option described above. Add individual dielectric coatings to the boundary using
the Thin Dielectric Film subnode. In addition, enter a value or expression for the
Number of repeating unit cells N (dimensionless). The default value is 3. In the
settings windows for the Thin Dielectric Film subnodes, you can decide which layers
constitute a unit cell that will be repeated the specified number of times. Use this
option to specify periodic arrangements with a large number of repeatind dielectric
layers, without having to add each layer manually.

Distributed Bragg Reflector: Application Library path

Ray_Optics_Module/Prisms_and_Coatings/distributed_bragg_reflector

• Anti-reflective coating: the incident ray is refracted with unit transmittance and zero
reflectance.
• Single-layer coating, specified thickness: this is a simplified version of the Add layers to
surface option that only allows a single dielectric layer. Enter a value or expression
for the Film refractive index n (dimensionless). The default value is 1. Then enter a
value or expression for the Film thickness t (SI unit: m). The default value is 1 μm.

THE GEOMETRICAL OPTICS INTERFACE | 85

 • Specify reflectance: enter a value or expression for the reflectance at the boundary
directly. Enter a value or expression for the Reflectance R (dimensionless). The
default is 0.1. If the Specify different values for s- and p-polarization check box is
selected, you can instead enter separate values for the reflectance of p- and
s-polarized light, called Rp and Rs, respectively. The layers are considered
nonabsorbing so that the sum of the reflectance and transmittance is unity.
• Specify transmittance: enter a value or expression for the transmittance at the
boundary directly. Enter a value or expression for the Transmittance T
(dimensionless). The default is 0.9. If the Specify different values for s- and
p-polarization check box is selected, you can instead enter separate values for the
transmittance of p- and s-polarized light, called Tp and Ts, respectively. The layers
are considered nonabsorbing so that the sum of the reflectance and transmittance is
unity.

Real and Ideal Optical Coatings

The options Anti-reflective coating, Specify reflectance, and Specify transmittance all
include a check box called Treat as single layer dielectric film, which is cleared by
default. While this check box is cleared, the specified reflectance or transmittance
applies to all rays, regardless of wavelength or angle of incidence, and the remaining
inputs in this section are not shown. This makes complicated multilayer films easier to
set up because you can enter the reflectance or transmittance explicitly without having
to know the properties of each layer in the coating. If the coating properties depend
on the wavelength or angle of incidence, you can usually build this dependence into
the model by defining the film reflectance or transmittance as an Interpolation function
of the ray properties. The tradeoff is that accurate information about the
discontinuities in ray phase is lost, even if the Compute phase check box is selected in
the physics interface Intensity Computation section, because this information depends
on the complex-valued Fresnel coefficients.

If the Treat as single layer dielectric film check box is selected, then instead the Material
Discontinuity automatically computes the thickness of a single layer needed to produce
the given reflectance or transmittance. The drawback is that the specified value will
only be attained for rays at a specific wavelength and angle of incidence.

The following inputs are only shown when Treat as single layer dielectric film is selected.

For Specify reflectance and Specify transmittance enter the Film refractive index n
(dimensionless). The default is 1. If it is impossible to construct a single layer of this
refractive index and obtain the specified reflectance or transmittance, then the
boundary is treated as having no dielectric films. For Anti-reflective coating, the

86 | CHAPTER 3: RAY OPTICS INTERFACES

refractive index of the coating is not an input because it is always the geometric mean
of the refractive indices in the two adjacent domains.

For Anti-reflective coating, Specify reflectance, and Specify transmittance enter the
following:

• Vacuum wavelength for specified film properties λ0 (SI unit: m). The default is
660 nm.
• Angle of incidence for specified film properties θi (SI unit: m). The default is 0.
• Select an option from the Angle of incidence specification list: With respect to upside
(the default) or With respect to downside. You can see which side is the upside by
selecting the Show boundary normal check box (see Advanced Settings below); the
normal vector rendered in the Graphics window points from the upside to the
downside.
• Select an option from the Specified film behavior applies to list: S-polarized radiation
(the default) or P-polarized radiation.

RAYS TO RELEASE
Use the inputs in this section to control whether the incident ray is split into a reflected
and refracted ray at the material discontinuity. Select an option from the Release
reflected rays list: Always (the default), Never, or Based on logical expression.

• If Always is selected, whenever a ray hits the material discontinuity, the incident ray
is refracted across the boundary, and a reflected ray is produces using some of the
preallocated degrees of freedom for secondary rays.
• If Never is selected, the incident ray is refracted, but no reflected ray is produced.
This does not prevent a ray from undergoing total internal reflection at the
boundary.
• If Based on logical expression is selected, enter a value or expression for the Evaluation
expression e (dimensionless). The default is 1. A reflected ray is produced only if the
value of this expression is nonzero for the incident ray.

If the ray intensity is solved for in the model, enter a Threshold intensity Ith
(SI unit: W/m2). The default is 1 mW/m2. If the interaction of a ray with a material
discontinuity would create a reflected ray of intensity less than the threshold intensity,
the release of this reflected ray is suppressed.

If the ray power is solved for in the model, enter a Threshold power Qth (SI unit: W).
The default is 10-3 mW. If the interaction of a ray with a material discontinuity would
create a reflected ray of power less than the threshold power, the release of this

THE GEOMETRICAL OPTICS INTERFACE | 87

 reflected ray is suppressed. Suppressing the release of reflected rays of extremely low
intensity or power is a convenient way to avoid wasting computational resources.

If both the Evaluation expression and the Threshold intensity are specified, a reflected
ray is only released if the value of the expression is nonzero and the intensity of the
reflected ray exceeds the threshold. In other words, a ray must satisfy all of the criteria
in the Rays to Release section in order to be released, not just one of them.

AUXILIARY DEPENDENT VARIABLES, REFRACTED RAY

This section includes a check box (cleared by default) and a text field (default 0) for
every Auxiliary Dependent Variable in the model. Select the check box to activate the
text field, which controls the new value of this auxiliary dependent variable for the
refracted ray. If the check box is cleared, the new value equals that of the incident ray.

The value can be a function of any combinations of ray variables and variables defined
on the boundary. For example, to increment the value of psi by 1 when a ray touches
or crosses a boundary, enter psi+1 in the text field for psinew.

AUXILIARY DEPENDENT VARIABLES, REFLECTED RAY

This section is the same as Auxiliary Dependent Variables, Refracted Ray, except that
the expressions are used to auxiliary dependent variables for the reflected ray.

ADVANCED SETTINGS
Select the Show boundary normal check box to view the boundary normal in the
Graphics window.

Visualization of the boundary normal is important when adding multiple

Thin Dielectric Film subnodes, since the thin films are oriented from the
upside of the boundary to the downside in the same order as their
corresponding nodes in the Model Builder.

If the Compute optical path length check box is selected in the physics interface
Additional Variables section, select the Reset optical path length check box to set the
optical path length of reflected and refracted rays to 0. Otherwise both the reflected
and refracted ray inherit the optical path length of the incident ray.

Material Discontinuity Theory

88 | CHAPTER 3: RAY OPTICS INTERFACES

Thin Dielectric Film
Use the Thin Dielectric Film subnode to model the interaction of rays with walls or
material discontinuities that are coated with thin, transparent dielectric layers. The
presence of these layers modifies the Fresnel coefficients used to compute the intensity
or power of the reflected and refracted rays. The films must be very thin relative to the
coherence length of the radiation and are usually comparable to or smaller than the
wavelength. The subnode is available from the context menu (right-click the Material
Discontinuity or Wall parent node) or from the Physics toolbar, Attributes menu.

The Thin Dielectric Film subnode is applied to all boundaries in the selection of the Wall
or Material Discontinuity parent node. The effects of the Thin Dielectric Film subnodes
are applied when computing the coefficients of reflection and transmission, which are
necessary for the reinitialization of Stokes parameters during reflection and refraction
if the intensity is solved for in the model.

FILM PROPERTIES
Enter a Film refractive index n (dimensionless). The default value is 1. Then enter a Film
thickness t (SI unit: m). The default is 1 μm.

REPEATING MULTILAYER FILMS

If Add layers to surface, repeating is selected from the Thin dielectric films on boundary
list in the Coatings section in the settings window for the parent node, this section is
used to determine which layers are part of the repeating unit cell of a periodic
multilayer film. If the Repeat layer in multilayer films check box is selected, the layer is
repeated a number of times equal to the Number of repeating unit cells specified in the
parent node settings window. If the check box is cleared, the layer will appear only once
in multilayer films.

Thin Dielectric Film Theory

Wall
Use the Wall node to determine what happens to the rays when contact with a
boundary is made. Rays automatically detect walls that intersect their path, without the
order of the interactions being specified. At boundaries, rays can be reflected or
absorbed. You can also select two different conditions and assign a criterion or
probability to determine which one is applied.

THE GEOMETRICAL OPTICS INTERFACE | 89

 The Accumulator (Boundary) subnode is available from the context menu (right-click
the parent node) or from the Physics toolbar, Attributes menu. If the ray intensity or
power is computed, the Thin Dielectric Film subnode is also available. If the ray power
is computed, the Deposited Ray Power (Boundary) subnode is also available.

WALL CONDITION
Select a Wall condition: Freeze (the default), Specular Reflection, Stick, Disappear, Pass
through, Diffuse scattering, Isotropic scattering, Mixed diffuse and specular reflection, or
General reflection.

TABLE 3-2: WALL CONDITION OPTIONS

OPTIONS DESCRIPTION

Freeze Select to fix the ray position and wave vector at the instant a wall is
struck. So, the ray position no longer changes after contact with the
wall and the wave vector of the ray remains at the same value as when
the ray struck the wall. This boundary condition is typically used to
recover the ray intensity or phase at the instant contact was made with
the wall.
Specular Select to specularly reflect from the wall, so that the angle of incidence
Reflection and the angle of reflection are equal.
Stick Select to fix the ray position at the instant the wall is struck. The wave
vector components are set to zero.
Disappear This option means that the ray is not displayed once it has made
contact with the wall. Use it if the ray location after contact with the
wall is not of interest.
Pass through This option allows rays to cross the boundary unimpeded. It does not
change the ray direction at a boundary between different media; to
model refraction, instead use the Material Discontinuity feature.
Diffuse Select to reflect rays at a wall according to Lambert’s cosine law. That
scattering is, the probability of a reflected ray propagating in a given direction
within a solid angle dω is given by cos(θ)dω where θ is the angle
between the direction of the ray and the wall normal.
Isotropic Select to reflect rays at a wall isotropically in random directions.
scattering
Mixed Select to reflect rays at a wall either specularly or according to
diffuse and Lambert’s cosine law, based on a user-defined probability.
specular
reflection
General Select to allow an arbitrary direction vector to be specified after a ray
reflection makes contact with the wall. The ray direction vector components can
be functions of the incident ray direction, phase, or any other quantity.

90 | CHAPTER 3: RAY OPTICS INTERFACES

GENERAL REFLECTION SETTINGS
This section is available when General reflection (see Table 3-2) is selected as the Wall
condition.

Enter values for the Reflected ray direction vector Lp (dimensionless) either in Cartesian
coordinates (x, y, z) (the default) or select the Specify tangential and normal direction
vector components check box to enter coordinates in the tangent-normal coordinate
system (t1, t2, n). In this case the normal direction is selected so that an incident ray is
reflected back into the domain it previously occupied if the specified normal direction
vector component is positive. The tangential directions are oriented so that they form
a right-handed coordinate system, together with the normal direction.

PRIMARY RAY CONDITION

Use this section to apply two different wall conditions on the same set of boundaries
based on a probability or logical expression.

Select a Primary ray condition: None (the default), Probability, or Expression. When the
default, None, is kept, it means that the Wall condition is always respected by the
incident rays.

Probability
If Probability is selected, the Wall condition is applied with a certain probability. Enter
a value for the Probability, γ (dimensionless). If the Wall condition is not used, the ray
instead behaves according to the Otherwise setting.

The value of γ should always be between 0 and 1.

For example, if the Wall condition is set to:

• Freeze and γ is set to 0.1, then for every 10 rays that strike the wall, on average one
freezes and the remaining 9 rays behave according to the Otherwise setting.
• Stick and γ is set to 0.5 then on average half of the rays stick to the wall and the other
half behave according to the Otherwise setting.

Expression
If Expression is selected, the Evaluation expression e (dimensionless) is evaluated
whenever a ray strikes the wall. The default expression is 1. If the Evaluation expression
is nonzero, the ray behaves according to the Wall condition, otherwise the ray behaves
according to the Otherwise setting.

THE GEOMETRICAL OPTICS INTERFACE | 91

 Otherwise
The options available for the Otherwise setting are the same as for the Wall Condition,
except that General reflection and Mixed diffuse and specular reflection are not available.
The Otherwise setting can be used to make rays interact with a wall differently with a
certain probability or when a certain condition is satisfied. For example, to model
reflection at a partially specular surface in which 50% of the ray intensity is absorbed
and 30% is reflected specularly:

• Select Mixed diffuse and specular reflection as the Wall condition,

• set the Probability of specular reflection to 0.3/(1-0.5),
• set the Primary ray condition to Probability,
• set the Probability to 1-0.5, and
• select Freeze as the Otherwise option.

FREQUENCY REINITIALIZATION
This section is available when Polychromatic, specify frequency is selected from the
Wavelength distribution of released rays list in the physics interface Ray Release and
Propagation section, and when a Wall Condition (or Otherwise condition) other than
Freeze, Stick, or Disappear is selected.

This section allows the ray frequency to be changed when rays are reflected by the
boundary or pass through it. To do so, select the Reinitialize frequency check box; it is
cleared by default. Then select one of the following from the Distribution function list:

• None (the default): enter a value or expression for the Ray frequency ν (SI unit: Hz)
directly. The default is 3 × 1014 Hz.
• Normal: the new value of the ray frequency is sampled from a normal or Gaussian
distribution. Enter a value or expression for the Mean ray frequency μ (SI unit: Hz).
The default is 3 × 1014 Hz. Then enter a value or expression for the Ray frequency
standard deviation σ (SI unit: Hz). The default is 3 × 1013 Hz.
• Lognormal: the new value of the ray frequency is sampled from a lognormal
distribution; that is, its logarithm follows a normal or Gaussian distribution. A big
advantage of the Lognormal distribution (compared to the Normal distribution) is
that it statistical outliers will never cause the frequency to become negative. Enter a
value or expression for the Mean ray frequency μ (SI unit: Hz). The default is

92 | CHAPTER 3: RAY OPTICS INTERFACES

 3 × 1014 Hz. Then enter a value or expression for the Ray frequency standard
deviation σ (SI unit: Hz). The default is 3 × 1013 Hz.
• Uniform: the new value of the ray frequency will be sampled uniformly from an
interval. Enter a value or expression for the Minimum ray frequency νmin
(SI unit: Hz). The default is 3 × 1014 Hz. Then, enter a value or expression for the
Maximum ray frequency νmax (SI unit: Hz). The default is 3.5 × 1014 Hz.

WAVELENGTH REINITIALIZATION
This section is available when Polychromatic, specify vacuum wavelength is selected from
the Wavelength distribution of released rays list in the physics interface Ray Release and
Propagation section, and when a Wall Condition (or Otherwise condition) other than
Freeze, Stick, or Disappear is selected.

This section allows the wavelength to be changed when rays are reflected by the
boundary or pass through it. To do so, select the Reinitialize wavelength check box; it
is cleared by default. Then select one of the following from the Distribution function list:

• None (the default): enter a value or expression for the Vacuum wavelength λ0
(SI unit: Hz) directly. The default is 500 nm.
• Normal: the new value of the wavelength is sampled from a normal or Gaussian
distribution. Enter a value or expression for the Mean vacuum wavelength μ
(SI unit: m). The default is 500 nm. Then enter a value or expression for the Vacuum
wavelength standard deviation σ (SI unit: m). The default is 50 nm.
• Lognormal: the new value of the wavelength is sampled from a lognormal
distribution; that is, its logarithm follows a normal or Gaussian distribution. A big
advantage of the Lognormal distribution (compared to the Normal distribution) is
that it statistical outliers will never cause the wavelength to become negative. Enter
a value or expression for the Mean vacuum wavelength μ (SI unit: m). The default is
500 nm. Then enter a value or expression for the Vacuum wavelength standard
deviation σ (SI unit: m). The default is 50 nm.
• Uniform: the new value of the wavelength will be sampled uniformly from an
interval. Enter a value or expression for the Minimum vacuum wavelength λ0,min
(SI unit: m). The default is 500 nm. Then, enter a value or expression for the
Maximum vacuum wavelength λ0,max (SI unit: m). The default is 550 nm.

THE GEOMETRICAL OPTICS INTERFACE | 93

 ABSORPTION COEFFICIENT
This section is available when

• the ray intensity or power is solved for in the model, and

• the Wall condition is set to any option that can reflect rays, including the following:

- Specular Reflection,
- Diffuse scattering,
- Isotropic scattering,
- General reflection, or
- Mixed diffuse and specular reflection.

Select one of the following from the Compute reflected intensity using list: Absorption
coefficients or Reflection coefficients.

For Absorption coefficients enter the Absorption coefficient α (dimensionless). The

default is 0. The intensity or power of the reflected ray will be proportional to 1 − α.

For Reflection coefficients enter the Reflection coefficient r (dimensionless). The default
value is 1. The intensity or power of the reflected ray will be proportional to r2.

If the Wall condition is set to Mixed diffuse and specular reflection, this section is instead
called Absorption Coefficients and the absorption coefficients for the diffusely and
specularly reflected rays are specified separately. All of the text fields are given
subscripts s and d for specularly and diffusely reflected rays, respectively.

NEW VALUE OF AUXILIARY DEPENDENT VARIABLES

This section is available if an Auxiliary Dependent Variable has been added to the
model.

When a ray crosses or touches a boundary, the values of the user-defined auxiliary
dependent variables can be changed. The value can be a function of any combinations
of ray variables and variables defined on the boundary.

Select the Assign new value to auxiliary variable check box or boxes based on the
number of auxiliary variables in the model. Then enter the new value or expression in
the field. For example, if there is an auxiliary variable, psi, then enter a value for psinew
in the field. So, to increment the value of psi by 1 when a ray touches or crosses a
boundary, enter psi+1 in the text field for psinew.

94 | CHAPTER 3: RAY OPTICS INTERFACES

ADVANCED SETTINGS
If the Primary ray condition is set to Probability, or if the Diffuse scattering, Isotropic
scattering, or Mixed diffuse and specular reflection wall condition is used, then the Wall
feature generates random numbers. Random numbers are also used if Reinitialization
wavelength or Reinitialization frequency is activated.

If, in addition, the Arguments for random number generation setting is set to User defined
in the physics interface Advanced Settings section, the Advanced Settings section is
available. Enter the Additional input argument to random number generator i
(dimensionless). The default value is 1.

The Advanced Settings section is also shown if the Compute optical path length check
box is selected under the physics interface Additional Variables section. Select the Reset
optical path length check box to set the optical path length to 0 when a ray touches the
wall.

Mirror
Use the Mirror node to specularly reflect rays at a boundary. This feature functions as
a special case of the Wall node, except that the wall condition is always specular
reflection.

See Wall for information on the following sections: Absorption Coefficient, Frequency
Reinitialization, Wavelength Reinitialization, New Value of Auxiliary Dependent Variables,
and Advanced Settings.

Ideal Lens
Use the Ideal Lens boundary condition to treat a surface as if it were a lens, without
having to create a detailed lens geometry.

The selected boundary or boundaries for the Ideal Lens should generally be a planar
surface (in 3D) or a straight edge (in 2D).

THE GEOMETRICAL OPTICS INTERFACE | 95

 Device Properties
Select an option from the Ideal lens type list: Paraxial surface (the default), Thin lens, or
Thick lens.

• If Paraxial surface is selected, choose one of the following from the Paraxial surface
type list:
- Specify effective focal length (the default): enter a value or expression for the
Effective focal length feff (SI unit: m). The default is 100 mm.
- Specify optical power: enter a value or expression for the Optical power K
(SI unit: 1/m). The default is 0.1 m-1.

• If Thin lens is selected, enter the following:

- Radius of curvature, surface 1 R1 (SI unit: m). The default is 100 mm. This value
should be positive for a convex surface or negative for a concave surface.
- Radius of curvature, surface 2 R2 (SI unit: m). The default is 100 mm. This value
should be negative for a convex surface or positive for a concave surface.
- Refractive index n (dimensionless). The default is 1.5.

• If Thick lens is selected, enter the same parameters as for Thin lens, plus the additional
Center thickness Tc (SI unit: m). The default is 0.

Ideal Lens Reference

The ideal lens reference is the pole of the ideal lens. A rays passing through this point
will be undeviated. The Ideal lens reference can be selected from a number of options.

• Centroid of selection will compute the geometric center of all selected boundaries.
• From coordinates will give a point on the ideal lens that is a projection of a user
defined point onto a plane containing the ideal lens surface.
• Selected point will use the point indicated in the Reference Point Selection section.

See Wall for information on the section New Value of Auxiliary Dependent Variables.

Ideal Lens Theory

Axial Symmetry
The Axial Symmetry node is automatically added to 2D axisymmetric models. The
options available for the feature are the same as for Wall, except as follows.

96 | CHAPTER 3: RAY OPTICS INTERFACES

The Boundary Selection is locked and is only applicable to edges on the symmetry axis.
The only wall conditions available are Specular reflection (the default), Freeze, Stick, and
Disappear. In addition, the Primary Ray Condition section is not included. The Axial
Symmetry boundary condition does not affect ray intensity or polarization, except to
introduce a phase delay of π in reflected rays to compensate for the change in the
principal curvature direction.

The Axial Symmetry condition only applies to straight boundaries that are
both on the axis of symmetry and adjacent to a domain in the selection of
the physics interface. If rays are propagating in the void region outside of
the selected domains, it is possible for the rays to enter the region where
r < 0, even though no geometry is defined there.

To ensure that rays cannot propagate into the region where r < 0, draw
line segments on the parts of the symmetry axis where no geometric
entities exist, then add the Wall node to these edges. Keep in mind that
rays can only be subjected to boundary conditions if the corresponding
boundaries are meshed.

Release
Use the Release node to release rays within domains based on arbitrary expressions or
based on the positions of the mesh elements.

RELEASE TIMES
This section is only available when the Allow multiple release times check box has been
selected in the physics interface Advanced Settings section. Enter Release times
(SI unit: s) or click the Range button ( ) to select and define a range of specific times.
At each release time, rays are released with initial position and ray direction vector as
defined next.

INITIAL POSITION
Select an Initial position: Density (the default) or Mesh based.

Density
For Density enter a value for the Number of rays per release N (dimensionless). The
default is 100. Then enter a value or expression for the Density proportional to ρ
(dimensionless). The default is 1.

THE GEOMETRICAL OPTICS INTERFACE | 97

 The Density proportional to ρ can be an expression rather than a number; the resulting
ray distribution approximately has a number density that is proportional to this
expression. The resulting distribution looks a bit random, and it depends on the order
in which the mesh elements are numbered. The distribution is probably not exactly the
same in different COMSOL Multiphysics versions, but the total number of rays
released is always N.

The Density proportional to expression must be strictly positive.

Advanced Options for Density-Based Release

The following options can be adjusted to make the number density of rays more closely
conform to the Density proportional to expression.

Select a Release distribution accuracy order between 1 and 5 (the default is 5), which
determines the integration order that is used when computing the number of rays to
release within each mesh element. The higher the accuracy order, the more accurately
rays will be distributed among the mesh elements.

The Position refinement factor (default 0) must be a nonnegative integer. When the
refinement factor is 0, each ray is always assigned a unique position, but the density is
taken as a uniform value over each mesh element. If the refinement factor is a positive
integer, the distribution of rays within each mesh element is weighted according to the
density, but it is possible for some rays to occupy the same initial position. Further
increasing the Position refinement factor increases the number of evaluation points
within each mesh element to reduce the probability of multiple rays occupying the
same initial position.

Mesh Based
For Mesh based the rays are released from a set of positions determined by a selection
of geometric entities (of arbitrary dimension) in the mesh. Given a Refinement factor
between 1 and 5, the centers of the refined mesh elements are used. Thus, the number
of positions per mesh element is refine^dim, except for pyramids, where it is
(4*refine2-1)*refine/3.

98 | CHAPTER 3: RAY OPTICS INTERFACES

RAY DIRECTION VECTOR
Select an option from the Ray direction vector list: Expression (the default), Spherical,
Hemispherical, Conical, or Lambertian (3D only).

• For Expression a single ray is released in the specified direction. Enter coordinates
for the Ray direction vector L0 (dimensionless) based on space dimension.
• For Spherical a number of rays are released at each point, sampled from a spherical
distribution in wave vector space. Enter the Number of rays in wave vector space Nw
(dimensionless). The default is 50.
• For Hemispherical a number of rays are released at each point, sampled from a
hemispherical distribution in wave vector space. Enter the Number of rays in wave
vector space Nw (dimensionless). The default is 50. Then enter coordinates for the
Hemisphere axis r based on space dimension.
• For Conical a number of rays are released at each point, sampled from a conical
distribution in wave vector space. Enter the Number of rays in wave vector space Nw
(dimensionless). The default is 50. Then enter coordinates for the Cone axis r based
on space dimension. Then enter the Cone angle a (SI unit: rad). The default value is
π/3 radians.
• The Lambertian option is only available in 3D. A number of rays are released at each
point, sampled from a hemisphere in wave vector space with probability density
based on the cosine law. Enter the Number of rays in wave vector space Nw
(dimensionless). The default is 50. Then enter coordinates for the Hemisphere axis r
based on space dimension.

If Conical is selected in a 3D model, select an option from the Conical distribution list:

• Uniform density (the default): rays are released with polar angles from 0 to the
specified cone angle. The rays are distributed in wave vector space so that each ray
subtends approximately the same solid angle.
• Specify polar and azimuthal distributions: specify the Number of polar angles Nθ
(dimensionless) and the Number of azimuthal angles Nϕ (dimensionless). Rays are
released at uniformly distributed polar angles from 0 to the specified cone angle. A
single axial ray (θ = 0) is also released. For each value of the polar angle, rays are
released at uniformly distributed azimuthal angles from 0 to 2π. Unlike other
options for specifying the conical distribution, it is not necessary to directly specify
the Number of rays in wave vector space Nw (dimensionless), which is instead derived
from the relation Nw = Nθ × Nϕ + 1.
• Hexapolar: specify the Number of polar angles Nθ (dimensionless). In this
distribution, for each release point, one ray will be released along the cone axis. Six

THE GEOMETRICAL OPTICS INTERFACE | 99

 rays are released at an angle α/Nθ from the cone axis, then 12 additional rays at an
angle of 2α/Nθ, and so on. The total number of ray directions in the distribution is
Nw = 3Nθ(Nθ + 1) + 1.
• Flat: rays are released in a flat fan shape within the specified angle.
• Marginal rays only: the rays are all released at an angle α with respect to the cone axis.
The rays are released at uniformly distributed azimuthal angles from 0 to 2π.
• Marginal and axial rays only: the rays are all released at an angle α with respect to the
cone axis, except for one ray which is released along the cone axis. The marginal rays
are released at uniformly distributed azimuthal angles from 0 to 2π.
Hexapolar Uniform density Polar and azimuthal

Marginal and axial Marginal only Flat

Figure 3-3: Comparison of the available cone-based release distributions.

In 3D for the Conical distribution you can also let the Transverse direction be Automatic
(the default) or User defined. For User defined enter the components of et. This
controls, for example, the orientation of the ray fan when Flat is selected.

For Spherical, Hemispherical, Conical, and Lambertian, select an option from the
Sampling from Distribution list: Deterministic (the default) or Random. If Deterministic is
selected, the initial ray direction vectors are computed using an algorithm that seeks to
distribute the rays as evenly as possible in wave vector space. This algorithm will give
the same initial ray directions whenever the study is run. If Random is selected, the
initial direction of each ray is sampled from a probability distribution in wave vector
space using pseudorandom numbers. The result may be the same when rerunning the
study multiple times on the same computer, but the solution is likely to be different
on different architectures.

100 | CHAPTER 3: RAY OPTICS INTERFACES

INITIAL RAY FREQUENCY
This section is available when Polychromatic, specify frequency is selected from the
Wavelength distribution of released rays list in the physics interface Ray Release and
Propagation section.

Select a Distribution function: None (the default), Normal, Lognormal, Uniform,

Blackbody, or List of values.

When None is selected, enter an initial value ν0 (SI unit: Hz). The default value is
4.54 × 1014 Hz.

Select Normal to create a normal distribution function, Lognormal to create a

log-normal distribution function, Uniform to create a uniform distribution function, or
Blackbody to sample values according to Planck’s law for blackbody radiation. For any
of these distributions, select an option from the Sampling from distribution check box:
Deterministic (the default) or Random. For Random sampling the mean and standard
deviation may not be exactly equal to the expected values but will statistically converge
as the number of rays is increased. The Number of values sets the number of values that
are sampled from the distribution function at each release point.

For the Normal or Lognormal distribution enter a user-defined Mean ray frequency
(default 4.54 × 1014 Hz) and Ray frequency standard deviation (default 1014 Hz). For
the Uniform distribution enter the Minimum ray frequency νmin (default 4.3 × 1014 Hz)
and Maximum ray frequency νmax (default 7.5 × 1014 Hz). Select List of values to enter
a list of distinct frequency values.

For Blackbody, enter the Blackbody temperature Tbb (SI unit: K). The default value of
5780 K is an approximate value for treating the sun as a blackbody radiation source.
Then select an option from the Distribution type list: Specify fraction of total emissive
power (the default), or Specify bandwidth.

• If Specify fraction of total emissive power is selected, enter the Fraction of total
emissive power fbb (dimensionless). The default is 0.95.
• If Specify bandwidth is selected, enter the Minimum ray frequency νmin and Maximum
ray frequency νmax (SI unit: Hz). The defaults are 4.3 × 1014 Hz and
7.5 × 1014 Hz, respectively.

VACUUM WAVELENGTH
This section is available when Polychromatic, specify vacuum wavelength is selected from
the Wavelength distribution of released rays list in the physics interface Ray Release and
Propagation section.

THE GEOMETRICAL OPTICS INTERFACE | 101

 Select a Distribution function: None (the default), Normal, Lognormal, Blackbody,
Uniform, or List of values.

When None is selected, enter a value or expression for the Vacuum wavelength λ0
(SI unit: m). The default is 660 nm. All rays released by this feature will have the same
wavelength.

Select Normal to create a normal distribution function, Lognormal to create a

log-normal distribution function, Uniform to create a uniform distribution function, or
Blackbody to sample values according to Planck’s law for blackbody radiation. For any
of these distributions, select an option from the Sampling from distribution check box:
Deterministic (the default) or Random. For Random sampling the mean and standard
deviation may not be exactly equal to the expected values but will statistically converge
as the number of rays is increased. The Number of values sets the number of values that
are sampled from the distribution function at each release point.

For the Normal or Lognormal distribution enter a user-defined Mean vacuum wavelength
(default 660 nm) and Vacuum wavelength standard deviation (default 100 nm). For the
Uniform distribution enter the Minimum vacuum wavelength λ0,min (default 400 nm)
and Maximum vacuum wavelength λ0,max (default 700 nm). Select List of values to enter
a list of distinct wavelength values directly.

For Blackbody, enter the Blackbody temperature Tbb (SI unit: K). The default value of
5780 K is an approximate value for treating the sun as a blackbody radiation source.
Then select an option from the Distribution type list: Specify fraction of total emissive
power (the default), or Specify bandwidth.

• If Specify fraction of total emissive power is selected, enter the Fraction of total
emissive power fbb (dimensionless). The default is 0.95.
• If Specify bandwidth is selected, enter the Minimum vacuum wavelength λ0,min and
Maximum vacuum wavelength λ0,max (SI unit: m). The defaults are 400 nm and
700 nm, respectively.

INITIAL PHASE
This section is available when the Compute phase check box is selected under the
physics interface Intensity Computation section. Enter an Initial phase Ψ0 (SI unit: rad).
The default value is 0.

102 | CHAPTER 3: RAY OPTICS INTERFACES

INITIAL INTENSITY
This section is available when the ray intensity is solved for in the model and Expression
is selected as the Ray direction vector. Enter a value for the Initial intensity I0
(SI unit: W/m2). The default is 1000 W/m2.

INITIAL RADII OF CURVATURE

This section is available when the ray intensity is solved for in the model and Expression
is selected as the Ray direction vector. Select a Wavefront shape. In 3D the available
options are Plane wave (the default), Spherical wave, and Ellipsoid. In 2D the available
options are Plane wave (the default) and Cylindrical wave.

• For an idealized plane wave the radii of curvature would be infinite. However,
because the algorithm used to compute intensity requires finite values, when Plane
wave is selected the initial radii of curvature are instead given an initial value that is
108 times greater than the characteristic size of the geometry.
• For a Spherical wave or Cylindrical wave, enter the Initial radius of curvature r0
(SI unit: m).
• For an Ellipsoid, enter the Initial radius of curvature, 1 r1,0 (SI unit: m) and the Initial
radius of curvature, 2 r2,0 (SI unit: m). Also enter the Initial principal curvature
direction, 1 e1,0 (dimensionless).

For spherical and cylindrical waves the Initial radius of curvature must be
nonzero. To release a ray such that the initial wavefront radius of
curvature is zero, instead select a different option such as Conical from the
Ray direction vector list.

Principal Radii of Curvature

INITIAL POLARIZATION
This section is available when the ray intensity or power is solved for.

Select an Initial polarization type: Unpolarized (the default), Fully polarized, or Partially
polarized.

THE GEOMETRICAL OPTICS INTERFACE | 103

 Select an Initial polarization: Along principal curvature direction (the default) or User
defined.

• For Fully polarized and Partially polarized rays in 3D enter an Initial polarization
parallel to reference direction a1,0 (dimensionless), Initial polarization perpendicular
to reference direction a2,0 (dimensionless), and Initial phase difference δ0
(SI unit: rad).
• For Fully polarized and Partially polarized rays in 2D enter an Initial polarization, in
plane axy,1 (dimensionless), Initial polarization, out of plane az,0 (dimensionless), and
Initial phase difference δ0 (SI unit: rad).
• For User defined also enter an Initial polarization reference direction u
(dimensionless).

For Partially polarized, also enter an Initial degree of polarization P0 (dimensionless).

TOTAL SOURCE POWER

This section is available:

• when the ray intensity is solved for, and

• when Spherical, Hemispherical, or Conical is selected as the Ray direction vector.
• It is also available when the ray power is solved for, and then any choice of Ray
direction vector displays this section.

Select an option from the Intensity initialization list: Uniform distribution (the default)
or Weighted distribution. If any Photometric Data Import nodes have been added to
the model then they can also be selected from the list.

If Uniform distribution or Weighted distribution is selected, enter a Total source power

Psrc (SI unit: W). The default is 1 W. In 2D, instead enter the Total source power per
unit thickness Psrc (SI unit: W/m). The default is 1 W/m. For Weighted distribution,
also enter an expression for the Power weighting factor Pwt. The weighting factor may
have any unit; the default expression is 1. The released rays will have initial intensity

104 | CHAPTER 3: RAY OPTICS INTERFACES

and power proportional to the weighting factor, while still adding up to the specified
source power.

For example, if you release 1000 rays in a Spherical distribution with an

initial power of 10 W, and enter a weighting factor of gop.niz+1, then
the sum of the power over all released rays, gop.sum(gop.Q), will equal
10 W. Since gop.niz is the z-component of the ray direction vector, rays
close to the z-axis will each have power of about 0.02 W, those around
the negative z direction will have almost no power, and those in the
xy-plane will each have power of about 0.01 W.

If any Photometric Data Import feature is selected from the list, the source power is
instead obtained directly from the imported photometric data (IES) file. Enter values
or expressions for the components of the Photometric horizontal ph (dimensionless)
and Photometric zero pz (dimensionless). By default, these vectors point in the
directions of the positive x- and z-axes, respectively.

The relationship between the photometric horizontal, photometric

vertical, and ray direction vector, and its effect on the initial ray intensity,
is explained in ANSI/IESNA LM-63-02 (R2008), IESNA Standard file
format for the electronic transfer of photometric data and related
information, Illuminating Engineering Society (2002).

The Photometric Data Import feature does not support the options
TILT=INCLUDE or TILT=<FILENAME> that are included in some IES files.
Only TILT=NONE is allowed.

INITIAL VALUE OF AUXILIARY DEPENDENT VARIABLES

For each of the Auxiliary Dependent Variable nodes added to the model, select a
Distribution function for the initial value of the auxiliary dependent variables and
whether the initial value of the auxiliary dependent variables should be a scalar value
or sampled from a distribution function.

The number of rays simulated can increase substantially and the following options are
available for each Auxiliary Dependent Variable added to the model.

THE GEOMETRICAL OPTICS INTERFACE | 105

 When None is selected, enter an initial value. The symbol for the initial value is the
auxiliary variable name followed by a subscript 0, so for the default name rp the initial
value has symbol rp0.

For the initial value of the auxiliary dependent variables, select Normal to create a
normal distribution function, Lognormal to create a log-normal distribution function,
or Uniform to create a uniform distribution function. For any of these distributions,
select an option from the Sampling from distribution check box: Deterministic (the
default) or Random. For Random sampling the mean and standard deviation may not be
exactly equal to the specified values but will statistically converge as the number of rays
is increased. The Number of values sets the number of values that are sampled from the
distribution function at each release point.

For the Normal or Lognormal distribution enter the Mean (default 0) and Standard
deviation (default 1). For the Uniform distribution enter the Minimum (default 0) and
Maximum (default 1). Select List of values to enter a set of numerical values directly.

By default auxiliary dependent variables are initialized after all other degrees of
freedom. Select the Initialize before wave vector check box to compute the initial value
of the auxiliary dependent variable immediately after computing the initial wave
vectors of the rays. By selecting this check box it is possible to define the initial ray
direction as a function of the auxiliary dependent variables.

Release from Boundary

Use the Release from Boundary node to determine how to release rays on an interior or
exterior boundary in a specific direction.

The Nonlocal Accumulator subnode is available from the context menu (right-click
the parent node) or from the Physics toolbar, Attributes menu.

See Release for information on the following sections: Release Times, Initial Ray
Frequency, Vacuum Wavelength, Initial Phase, Initial Intensity, Total Source Power, Initial
Polarization, and Initial Value of Auxiliary Dependent Variables.

COORDINATE SYSTEM SELECTION

It is possible to specify the initial ray direction in terms of the global coordinates or in
another coordinate system defined for the model Component. Select an option from
the Coordinate system list. By default Global coordinate system is selected. If other
coordinate systems are defined, they can also be selected from the list. When specifying
the initial ray direction (see the Ray Direction Vector section), direction components can

106 | CHAPTER 3: RAY OPTICS INTERFACES

be specified using the basis vectors of whichever coordinate system has been selected
from the list.

When a coordinate system other than Global coordinate system is selected from the
Coordinate system list, arrows will appear in the Graphics window to indicate the
orientation of the basis vectors of the coordinate system on the selected boundaries.

INITIAL POSITION
Select an Initial position: Uniform distribution (the default for 2D components)
Projected plane grid (the default for 3D components), Mesh based, or Density. Mesh
based and Density have the same settings as described for the Release node.

For 2D components, if Uniform distribution is selected, enter the Number

of rays per release N (dimensionless). The union of the selected
boundaries is divided into N segments of approximately equal length, and
a ray is placed in the middle of each segment.

For 3D components, if Projected plane grid is selected, enter the Number

of rays per release N (dimensionless). The rays are distributed on a plane
grid in planes that are approximately tangential to the selected boundaries
(for a plane boundary, you get a uniform distribution).

RAY DIRECTION VECTOR

Select an option from the Ray direction vector list: Expression (the default),
Hemispherical, Conical, or Lambertian (3D only).

• For Expression a single ray is released in the specified direction. Enter coordinates
for the Ray direction vector L0 (dimensionless) based on space dimension. Select the
Specify tangential and normal vector components check box to specify the initial
direction using a coordinate system based on the directions tangential and normal
to the surface (t1, t2, n).
• For Hemispherical a number of rays are released at each point, sampled from a
hemispherical distribution in wave vector space. Enter the Number of rays in wave
vector space Nw (dimensionless). The default is 50. Then enter coordinates for the
Hemisphere axis r based on space dimension. Select the Specify tangential and normal
vector components check box to specify the hemisphere axis using a coordinate
system based on the directions tangential and normal to the surface (t1, t2, n).

THE GEOMETRICAL OPTICS INTERFACE | 107

 • For Conical a number of rays are released at each point, sampled from a conical
distribution in wave vector space. Enter the Number of rays in wave vector space Nw
(dimensionless). The default is 50. Then enter coordinates for the Cone axis r based
on space dimension. Then enter the Cone angle α (SI unit: rad). The default value is
π/3 radians. Select the Specify tangential and normal vector components check box to
specify the cone axis using a coordinate system based on the directions tangential
and normal to the surface (t1, t2, n).
• The Lambertian option is only available in 3D. A number of rays are released at each
point, sampled from a hemisphere in wave vector space with probability density
based on the cosine law. Enter the Number of rays in wave vector space Nw
(dimensionless). The default is 50. Then enter coordinates for the Hemisphere axis r
based on space dimension. Select the Specify tangential and normal vector
components check box to specify the hemisphere axis using a coordinate system
based on the directions tangential and normal to the surface (t1, t2, n).

When the Specify tangential and normal vector components check box is
selected, arrows indicating the normal direction on the selected
boundaries will appear in the Graphics window.

Note that the normal direction may be opposite the built-in variable for
the boundary normal (for example, nx, ny, and nz) to ensure that a
positive value causes rays to be released into the simulation domain. This
often occurs when the release is applied to exterior boundaries.

When the normal direction used by the Release from Boundary feature is
opposite the normal vector defined by the geometry, the tangential
directions are similarly inverted to ensure that the boundary coordinate
system is right-handed.

If Conical is selected in a 3D model, select an option from the Conical distribution list:

• Uniform density (the default): rays are released with polar angles from 0 to the
specified cone angle. The rays are distributed in wave vector space so that each ray
subtends approximately the same solid angle.
• Specify polar and azimuthal distributions: specify the Number of polar angles Nθ
(dimensionless) and the Number of azimuthal angles Nϕ (dimensionless). Rays are
released at uniformly distributed polar angles from 0 to the specified cone angle. A
single axial ray (θ = 0) is also released. For each value of the polar angle, rays are
released at uniformly distributed azimuthal angles from 0 to 2π. Unlike other

108 | CHAPTER 3: RAY OPTICS INTERFACES

 options for specifying the conical distribution, it is not necessary to directly specify
the Number of rays in wave vector space Nw (dimensionless), which is instead derived
from the relation Nw = Nθ × Nϕ + 1.
• Hexapolar: specify the Number of polar angles Nθ (dimensionless). In this
distribution, for each release point, one ray will be released along the cone axis. Six
rays are released at an angle α/Nθ from the cone axis, then 12 additional rays at an
angle of 2α/Nθ, and so on. The total number of ray directions in the distribution is
Nw = 3Nθ(Nθ + 1) + 1.
• Flat: rays are released in a flat fan shape within the specified angle.
• Marginal rays only: the rays are all released at an angle α with respect to the cone axis.
The rays are released at uniformly distributed azimuthal angles from 0 to 2π.
• Marginal and axial rays only: the rays are all released at an angle α with respect to the
cone axis, except for one ray which is released along the cone axis. The marginal rays
are released at uniformly distributed azimuthal angles from 0 to 2π.

The available Conical distribution options are illustrated in Figure 3-3.

In 3D for the Conical distribution you can also let the Transverse direction be Automatic
(the default) or User defined. For User defined enter the components of et. This
controls, for example, the orientation of the ray fan when Flat is selected.

For Hemispherical, Conical, and Lambertian, select an option from the Sampling from
Distribution list: Deterministic (the default) or Random. If Deterministic is selected, the
initial ray direction vectors are computed using the same algorithm, which seeks to
distribute the rays as evenly as possible in wave vector space, whenever the study is run.
If Random is selected, the initial direction of each ray is sampled from a probability
distribution in wave vector space using pseudorandom numbers.

INITIAL RADII OF CURVATURE

This section is available when the ray intensity is solved for in the model and Expression
is selected as the Ray direction vector. Select a Wavefront shape. In 3D the available

THE GEOMETRICAL OPTICS INTERFACE | 109

 options are From surface curvature (the default), Plane wave, Spherical wave, and
Ellipsoid. In 2D the available options are Plane wave (the default) and Cylindrical wave.

• For a Spherical wave or Cylindrical wave, enter the Initial radius of curvature r0
(SI unit: m).
• For an Ellipsoid (3D only), enter the Initial radius of curvature, 1 r1,0 (SI unit: m)
and the Initial radius of curvature, 2 r2,0 (SI unit: m). Also enter the Initial principal
curvature direction, 1 e1,0 (dimensionless).

For spherical and cylindrical waves the Initial radius of curvature must be
nonzero. To release a ray such that the initial wavefront radius of
curvature is zero, instead select a different option such as Conical from the
Ray direction vector list.

Principal Radii of Curvature

Release from Symmetry Axis

Use the Release from Symmetry Axis node to release rays from specified edges along the
axis of symmetry in a 2D axisymmetric geometry. It functions like the Release from
Boundary node, except that it can only be applied to the axis of symmetry and that the
options in the Ray Direction Vector section are slightly different.

See Release for information on the following sections: Release Times, Initial Ray
Frequency, Vacuum Wavelength, Initial Phase, Initial Intensity, Initial Radii of Curvature,
Total Source Power, Initial Polarization, and Initial Value of Auxiliary Dependent Variables.

INITIAL POSITION
Select an Initial position: Uniform distribution (the default), Mesh based, or Density. Mesh
based and Density have the same settings as described for the Release node.

If Uniform distribution is selected, enter the Number of rays per release N

(dimensionless). The default is 100. The set of the selected boundaries is divided into
N segments of approximately equal length, with one ray at the middle of each segment.

110 | CHAPTER 3: RAY OPTICS INTERFACES

RAY DIRECTION VECTOR
Select an option from the Ray direction vector list: Expression (the default), Spherical,
or Conical.

• For Expression a single ray is released in the specified direction. Enter the radial and
axial components of the Ray direction vector L0 (dimensionless).
• For Spherical a number of rays are released at each point, from the negative
z direction to the positive z direction in the half plane with r > 0. These rays are
sampled from an anisotropic hemispherical distribution in wave vector space, so that
rays are more heavily weighted toward the r direction rather than the ±z directions.
In 3D this would correspond to a spherical distribution of rays such that each ray
subtends approximately the same solid angle in wave vector space.
Enter the Number of rays in wave vector space Nw (dimensionless). The default is 50.
Then enter coordinates for the Hemisphere axis r based on space dimension. Select
the Specify tangential and normal vector components check box to specify the
hemisphere axis using a coordinate system based on the directions tangential and
normal to the surface (t1, t2, n).
• For Conical a number of rays are released at each point, sampled from a conical
distribution in wave vector space. Enter the Number of rays in wave vector space Nw
(dimensionless). The default is 50. Then enter coordinates for the Cone axis r based
on space dimension. Then enter the Cone angle α (SI unit: rad). The default value is
π/3 radians. Unlike the Spherical distribution, the rays in the Conical distribution
each subtend the same angle in 2D, not the same solid angle in 3D.

For Hemispherical and Conical, select an option from the Sampling from Distribution list:
Deterministic (the default) or Random. If Deterministic is selected, the initial ray
direction vectors are computed using the same algorithm, which seeks to distribute the
rays as evenly as possible in wave vector space, whenever the study is run. If Random is
selected, the initial direction of each ray is sampled from a probability distribution in
wave vector space using pseudorandom numbers.

Release from Edge

Use the Release from Edge node to release rays from positions along a set of edges in a
3D geometry.

The Initial Position section is described below. For all other sections, see Release.

THE GEOMETRICAL OPTICS INTERFACE | 111

 INITIAL POSITION
Select an Initial position: Mesh based (the default), Uniform distribution, or Density. Mesh
based and Density have the same settings as described for the Release node. If Uniform
distribution is selected, enter the Number of rays per release N (dimensionless). The
union of the selected edges is divided into N segments of approximately equal length,
and a ray is released from the middle of each segment.

Release from Point

Use the Release from Point node to release rays from a set of points in a geometry. By
default one ray is released at every selected point, although it is possible to release
multiple rays at each point if the initial direction or an auxiliary dependent variable is
sampled from a distribution.

For information on the settings, see Release.

Release from Point on Axis

The Release from Point on Axis feature is similar to the Release from Point feature,
except that the points must be on the axis of symmetry in a 2D axisymmetric geometry.

See Release from Boundary for information on the Ray Direction Vector section.

For all other sections, see Release from Point.

Release from Grid

Use the Release from Grid node to release rays from a user-defined grid of points.

The Initial Coordinates section is described below. For all other sections, see Release.

INITIAL COORDINATES
Select an option from the Grid type list: All combinations (the default), Specified
combinations, Cylindrical (3D only), or Hexapolar (3D only).

Linear, Rectangular, and Arbitrary Grids

For All combinations and Specified combinations enter Initial coordinates based on space
dimension (qx,0, qy,0, and qz,0 for 3D components) for the ray positions or click the
Range button ( ) to select and define a range of specific coordinates.

If Specified combinations is selected, the number of initial coordinates entered for each
space dimension must be equal, and the total number of rays released is equal to the

112 | CHAPTER 3: RAY OPTICS INTERFACES

length of one of the lists of initial coordinates. If All combinations is selected, the total
number of rays released is equal to the product of the lengths of each list of initial
coordinates.

For example, suppose a 2D component includes a Release from Grid node with the
following initial coordinates:

• qx,0 = range(0,1,3)
• qy,0 = range(2,2,8)

If All combinations is selected, a total of 16 rays will be released, including every possible
combination of the initial x- and y-coordinates. If Specified combinations is selected, 4
rays will be released with initial positions (0,2), (1,4), (2,6), and (3,8).

Cylindrical Grids
For Cylindrical, enter coordinates for the Center location qc (SI unit: m). By default, the
distribution is centered at the origin. Then enter the components of the Cylinder axis
direction rc (dimensionless). The rays will be released at specified radial distances and
angles in the plane containing the point qc and orthogonal to the direction rc.

Select an option from the Radial distribution list: Uniform radius intervals (the default),
Uniform number density, or User defined.

For Uniform radius intervals or Uniform number density enter a value or expression for
the Radius Rc (SI unit: m). The default is 1 m. Then enter a positive integer for the
Number of radial positions Nc (dimensionless). The default is 5.

For User defined enter a list of Radial coordinates qr (SI unit: m) directly. An arbitrary
number of radial coordinates can be entered in the list. The default is 1 m.

The effect of the Radial distribution setting on the resulting ray placement is illustrated
in Figure 3-4. For Uniform radius intervals, the distances between the concentric rings
of rays are all equal, but the number density of rays is greater at the center of the
distribution than at the outer edge. The option Uniform number density corrects this
imbalance by defining nonuniform increments in the radial position between the
concentric rings.

Enter a positive integer for the Number of angles Nϕ (dimensionless). The default is 10.

THE GEOMETRICAL OPTICS INTERFACE | 113

 The total number of rays released (before accounting for distributions of ray direction
or auxiliary dependent variables at each release point) is N c × N φ + 1 because a single
ray is also released at the center of the distribution.

Figure 3-4: Comparison of the cylindrical release in which rays are positioned at uniform
radius intervals (left), with uniform number density (center), or at user-defined radii
(right).

Hexapolar Grids
For Hexapolar, enter coordinates for the Center location qc (SI unit: m). By default, the
distribution is centered at the origin. Then enter the components of the Cylinder axis
direction rc (dimensionless). The rays will be released at specified radial distances and
angles in the plane containing the point qc and orthogonal to the direction rc.

Enter a value or expression for the Radius Rc (SI unit: m). The default is 1 m. Then
enter a positive integer for the Number of radial positions Nc (dimensionless). The
default is 5.

Unlike the Cylindrical grid types described in the previous section, each concentric ring
of the Hexapolar grid releases a different number of rays, as shown in Figure 3-5. One
ray is always released at the center. The first ring of rays surrounding the center has 6
rays arranged in a regular hexagon. Each ring of rays beyond the first has 6 more rays
than the ring preceding it, with the grid points arranged in a regular polygon. The
radius increments between consecutive rings are uniform.

The total number of grid points generated is

N grid = 1 + 3N c ( N c + 1 )

114 | CHAPTER 3: RAY OPTICS INTERFACES

Thus the relationship between the number of radius intervals and the total number of
grid points is quadratic.

Figure 3-5: Comparison of hexapolar grids with 2 rings (left), 5 (middle), and 10 (right).

Previewing Grid Points

In the Initial Coordinates section, you can click the Preview Initial Coordinates and
Preview Initial Extents buttons to visualize the ray release positions. Clicking
Preview Initial Coordinates will cause a point to appear in the Graphics window for every
release position. Clicking Preview Initial Extents will cause a bounding box to appear,
indicating the spatial extents of the released rays. Examples are shown in Figure 3-6
and Figure 3-7.

Figure 3-6: Graphics window after clicking the Preview Initial Coordinates button.

THE GEOMETRICAL OPTICS INTERFACE | 115

 Figure 3-7: Graphics window after clicking the Preview Initial Extents button.

Release from Grid on Axis

The Release from Grid on Axis feature is similar to the Release from Grid feature, except
that it can only release rays from a grid of points along the axis of symmetry in a 2D
axisymmetric geometry; that is, qr,0 = 0 for all released rays.

See Release from Boundary for information on the Ray Direction Vector section.

For all other sections, see Release from Grid.

Release from Data File

Use a Release from Data File node to specify the initial ray position, direction, and values
of auxiliary dependent variables using data from a text file.

See Release for information on the following sections: Release Times, Initial Phase,
Initial Intensity, Initial Radii of Curvature, Total Source Power, and Initial Polarization.

Formatting Guidelines for Files Containing Ray Data

The imported data file should be a text file (*.txt) arranged in a spreadsheet format;
that is, each row corresponds to a distinct ray and should have the same number of
columns as all other rows. Columns can be separated by spaces, tabs, or a combination
of the two. Begin a line with the percent (“%”) character to include comments or
empty lines in the data file.

116 | CHAPTER 3: RAY OPTICS INTERFACES

For example, a data file containing the following text would release rays at the positions
(0.1, 0.2, 0.6) and (0.2, 0.4, 0.8) in a three-dimensional geometry:

% Ray release positions

% qx0 qy0 qz0
0.1 0.2 0.6
0.2 0.4 0.8

INITIAL POSITION
Browse your computer to select a text file, then click Import to import the data. To
remove the imported data, click Discard. Enter the Index of first column containing
position data i to indicate which column represents the first coordinate of the ray
position vectors. The default value, 0, indicates the first column.

TRANSFORMATIONS
The distribution of loaded ray release positions may be scaled, rotated, and translated
before the rays are released.

To scale the distribution of release positions, enter a value or expression for the Scale
factor R (dimensionless). The default is 1. This scale factor can be used to correct unit
discrepancies between the data file and the model geometry. For example, if the
geometry length unit is in meters but the data file lists coordinates in millimeters, enter
a scale factor of 0.001.

To rotate the distribution, enter the Euler angles (Z-X-Z) α, β, and γ (in 3D) or the
Rotation angle α (in 2D). The default values are all 0.

In 3D, α is the rotation angle about the space-fixed z-axis, then β is the rotation angle
about the transformed x-axis (or x’-axis), and finally γ is the rotation angle about the
transformed z-axis (or z’’-axis). Positive values indicate counterclockwise rotations.

Enter values or expressions for the components of the Displacement vector Δq

(SI unit: m). The default is not to apply any translation.

If a translation and another type of transformation (scaling or rotation) are applied,

then the translation is always applied after the other transformations.

RAY DIRECTION VECTOR

Select an option from the Ray direction vector list: Expression (the default), From file,
Spherical, Hemispherical, Conical, or Lambertian (3D only).

• For From file, enter the Index of first column containing direction data i. The default
is 3. The columns are zero-indexed; that is, an index of 0 corresponds to the first

THE GEOMETRICAL OPTICS INTERFACE | 117

 column. If you chose to rotate the loaded ray release positions in the Initial Position
section, select the Rotate initial directions check box to apply the same rotation to
the initial direction of the released rays. The check box is cleared by default.
• For all other settings, the settings are the same as for the Release node.

INITIAL RAY FREQUENCY

This section is available when Polychromatic, specify frequency is selected from the
Wavelength distribution of released rays list in the physics interface Ray Release and
Propagation section.

Select a Distribution function: None (the default), Normal, Lognormal, Uniform, List of
values, or From file.

• When From file is selected, enter the Index of column containing frequency data. The
default value is 3. The columns are zero-indexed; that is, an index of 0 corresponds
to the first column.
• For all other options, the settings are the same as for the Release node.

VACUUM WAVELENGTH
This section is available when Polychromatic, specify vacuum wavelength is selected from
the Wavelength distribution of released rays list in the physics interface Ray Release and
Propagation section.

Select a Distribution function: None (the default), Normal, Lognormal, Uniform, List of
values, or From file.

• When From file is selected, enter the Index of column containing wavelength data. The
default value is 3. The columns are zero-indexed; that is, an index of 0 corresponds
to the first column.
• For all other options, the settings are the same as for the Release node.

INITIAL VALUE OF AUXILIARY DEPENDENT VARIABLES

This section is available if an Auxiliary Dependent Variable has been added to the
model.

118 | CHAPTER 3: RAY OPTICS INTERFACES

For each of the active Auxiliary Dependent Variable nodes in the model, choose an
option from the Distribution function list: From file, None (the default), Normal,
Lognormal, Uniform, or List of Values.

• When From file is selected, enter the Index of column containing data. The default
value is 3. The columns are zero-indexed; that is, an index of 0 corresponds to the
first column.
For all other options, the settings are the same as for the Release node.

Photometric Data Import

The Photometric Data Import node is only available in 3D, and only if the
ray intensity is solved for in the model.

Use the Photometric Data Import node to import photometric data files that can be
used to initialize the ray intensity and power as a function of the initial ray direction.
The Photometric Data Import node supports the *.ies file extension, the standard
photometric data file format of the Illuminating Engineering Society of North America
(IESNA).

PHOTOMETRIC DATA IMPORT

Enter a File name, or click the Browse button to locate a photometric data file. When a
valid file path has been entered, click the Import button to load the photometric data
into the model. A shortened version of the name of the photometric data file will then
be displayed, along with a Discard button. Click the Discard button to discard the
loaded photometric data and clear the File name field.

Photometric Data Formatting

The format of the imported photometric data file must be consistent with ANSI/
IESNA LM-63-02 (R2008), IESNA Standard file format for the electronic transfer

THE GEOMETRICAL OPTICS INTERFACE | 119

 of photometric data and related information, Illuminating Engineering Society
(2002).

After creating an instance of the Photometric Data Import feature and

loading an appropriately formatted IES file into the model, initialize the
ray intensity as a function of initial ray direction by changing the following
settings in a ray release feature such as the Release or Release from Grid
feature:

• Select Spherical, Hemispherical, Conical, or Lambertian from the Ray

direction vector list in the physics feature Ray Direction Vector section.
• Locate the Total Source Power section. From the Intensity initialization
list, select an instance of the Photometric Data Import feature.
• Enter values or expressions for the Photometric horizontal and
Photometric zero directions.

Gaussian Beam
Use the Gaussian Beam feature to release rays with a Gaussian distribution of intensity
or power. Since the Geometrical Optics interface does not consider diffraction effects,
the Gaussian Beam node can be used in two different asymptotic limits: to release a
collimated beam over short distances, or to treat the beam waist as a point source and
trace rays over very long distances.

This feature is only available when either the ray intensity or power is solved for; it will
not be available if None is selected from the physics interface Intensity computation list.

The Gaussian Beam node does not require a selection because it is not necessary to
release the rays at a boundary or even within a domain; the waist may be located at an
arbitrary location in the geometry (2D or 3D) or an arbitrary location along the z-axis
(2D axisymmetric).

See Release for information on the following sections: Release Times, Initial Phase, and
Initial Value of Auxiliary Dependent Variables.

BEAM POSITION AND DIRECTION

Enter the Beam waist position qw (SI unit: m) based on space dimension. By default
the beam waist is located at the origin. Then enter the Beam axis r (dimensionless)

120 | CHAPTER 3: RAY OPTICS INTERFACES

based on space dimension. By default this is the positive z direction in 3D or the
positive x direction in 2D.

In a 2D axisymmetric geometry, the position and orientation of the

Gaussian beam is more restricted compared to 2D and 3D geometries.
The beam must be symmetric about the z-axis or else it would not actually
be a Gaussian beam in the 3D volume of revolution.

Enter the Beam waist z-coordinate zw (SI unit: m). The default is 0. Then,
from the Beam direction list, select Positive z-axis (the default) or Negative
z-axis.

BEAM DIMENSIONS
Select an option from the Beam type list:

• Point source (small Rayleigh range), the default, is appropriate when the rays in the
model propagate very large distances compared to the beam waist radius or Rayleigh
range. The beam waist is treated as a point source, from which rays emanate outward
in a cone. The symmetry axis of the cone is the Beam axis from the previous section.
• Collimated source (large Rayleigh range) is appropriate when the rays propagate for
only a short distance compared to the Rayleigh range. The rays are released from a
grid of points and initially propagate in the same direction.

Releasing Rays from a Point Source

For a Point source, enter the Number of polar angles Nθ (dimensionless). The default
is 10. In a 2D or 2D axisymmetric geometry, the actual number of rays released is equal
to Nθ, while in 3D the total number of rays released is 3Nθ(Nθ + 1) + 1. because rays
released at a larger polar angle require a finer sampling of azimuthal angles to ensure
uniform sampling within the solid angle subtended by the beam.

To control the width of the beam, select one of the following from the Beam
dimensions list:

• Specify beam divergence half-angle (the default), then enter a value or expression for
the Beam divergence half-angle α (SI unit: rad). The default is 1 mrad.
• Specify beam waist radius, then enter a value or expression for the Beam waist radius
w0 (SI unit: m). The default is 1 mm. In 2D, this option is instead called Specify
beam waist half-width and the input is called the Beam waist half-width because the

THE GEOMETRICAL OPTICS INTERFACE | 121

 2D beam is uniform in the out-of-plane direction, not rotationally symmetric about
the beam axis.
• Specify Rayleigh range, then enter a value or expression for the Rayleigh range zR
(SI unit: m). The default is 1 mm.

Releasing Rays from a Collimated Source

If instead Collimated source is selected, enter the Number of radial positions Nc
(dimensionless) in a 3D or 2D axisymmetric geometry, or the Number of rays per
release N (dimensionless). For any dimension, the default is 10. In 3D the total
number of rays released is 3Nc(Nc + 1) + 1 because rays released at greater radial
distances from the center of the beam waist require a finer sampling of azimuthal
angles, to ensure uniform sampling within the beam cross section. Then enter the Beam
waist radius w0 (SI unit: m). The default is 1 mm. In 2D this is called the Beam waist
half-width.

Beam Cutoff Ratio

For all of the above settings, enter a value or expression for the Beam cutoff ratio c
(dimensionless). The default is 1. The cutoff ratio does not affect the other beam
parameters but it affects how rays are sampled from the beam.

Theoretically the irradiance and power of light within a Gaussian beam remains
nonzero out to an arbitrarily large radial distance. But for practical purposes the beam
must be truncated at some point, so that rays will only be released within a finite
distance from the center of the beam waist (for collimated sources) or within a finite
angle (for point sources).

Changing the value of c changes the maximum distance or angle from the beam axis
at which rays are released. For the default value of 1, rays will be released at radial
positions out to the waist radius w0 (for a collimated source) or in a cone with angles
up to the beam divergence half-angle α (for a point source). In a 3D or 2D
axisymmetric beam, the released rays then comprise a fraction of 1 - e2 or about 87%
of the total beam power. In a 2D sheet beam, this fraction is about 95%.

If the value of c is increased, rays will be released at greater distances from the beam
axis (for a collimated beam) or at greater angles (for a point source) and the total power
over all of the released rays will be greater than the fractions given above. Setting c = 2
results in 99.97% of the total beam power being sampled by the rays. , but most of the
released rays will individually have rather low power. At about c = 4.2 the released
power among all the rays matches the specified total source power to double precision.

122 | CHAPTER 3: RAY OPTICS INTERFACES

INTENSITY AND POWER
Select one of the following options from the Intensity specification list:

• Specify total beam power (the default), then enter a value or expression for the Total
source power Psrc (SI unit: W). The default is 1 W.
• Specify encircled power, then enter a value or expression for the Encircled power Pcir
(SI unit: W). The default is 1 W.
• Specify Intensity at beam waist, then enter a value or expression for the Peak intensity
at beam waist I0 (SI unit: W/m2). The default is 1 W/mm2.

INITIAL POLARIZATION
This section is a simplified version of the Initial Polarization section for the Release
node. The main differences are that the Gaussian Beam must always release polarized
rays, and that the initial polarization direction must be entered directly rather than
based on the directions of the principal wavefront curvatures.

• In 3D enter the following:

- Components of the Initial polarization reference direction u (dimensionless). By
default the vector points in the positive y direction. This must not be parallel to
the beam axis.
- Initial polarization parallel to reference direction a1,0 (dimensionless, default 1).
- Initial polarization perpendicular to reference direction a2,0 (dimensionless,
default 0).
• In 2D axisymmetric models enter the following:
- Initial polarization, in plane arz,0 (dimensionless, default 1)
- Initial polarization, out of plane a φ, 0 (dimensionless, default 0).
• In 2D enter the following:

- Initial polarization, in plane axy,0 (dimensionless, default 1)

- Initial polarization, out of plane az,0 (dimensionless, default 0)
For all dimensions, enter the Initial phase difference δ0 (SI unit: rad, default 0).

VACUUM WAVELENGTH
This section is shown if Polychromatic, specify vacuum wavelength is selected from the
Wavelength distribution of released rays list in the physics interface Ray Release and
Propagation section. Enter a value or expression for the Vacuum wavelength λ0
(SI unit: m). The default is 660 nm.

THE GEOMETRICAL OPTICS INTERFACE | 123

 INITIAL RAY FREQUENCY
This section is shown if Polychromatic, specify frequency is selected from the Wavelength
distribution of released rays list in the physics interface Ray Release and Propagation
section. Enter a value or expression for the Initial ray frequency ν0 (SI unit: Hz). The
default is 4.54 × 1014 Hz.

Gaussian Beam Theory

Illuminated Surface
Use the Illuminated Surface node to release rays under the assumption that the emitted
rays are specularly reflected from an external radiation source. It is assumed that the
entire selected surface has direct line of sight to the external radiation source; that is,
shadowing due to other geometric entities is not taken into account.

See Release for information on the following sections: Release Times, Initial Ray
Frequency, Vacuum Wavelength, Initial Phase, Initial Polarization, and Initial Value of
Auxiliary Dependent Variables.

See Release from Boundary for information on the Initial Position section.

RAY DIRECTION VECTOR

Select the Rays to release: Reflected (the default) or Refracted.

• For Reflected select the Reflection coefficient specification: Using absorption coefficient
(the default) or Using refractive indices.
• For Refracted enter a value for Refractive index of exterior domains next
(dimensionless). The default is 1. This index is absolute (that is, relative to vacuum).

Select an Incident ray direction vector: User-defined direction (the default), User-defined
point source, or Solar radiation (3D components only).

Then based on the selection enter the applicable information:

For User-defined direction enter the components of the Incident ray direction vector Li
(dimensionless) based on space dimension.

• For User-defined point source enter the Point source location rsrc (SI unit: m).
• For Solar radiation see Solar Radiation for information about the options.

124 | CHAPTER 3: RAY OPTICS INTERFACES

ANGULAR PERTURBATIONS
If User-defined direction or Solar radiation is selected from the Incident ray direction
vector list, you can include small perturbation factors designed to take the finite size of
the sun (or another light source) into account. Select an option from the Corrections
for finite source diameter list: None (the default), Create light cones at release points, or
Sample from conical distribution.

• For None no perturbation to the initial ray direction is applied.

• For Create light cones at release points a cone-shaped distribution of rays is released
at each point. Enter a value for the Number of rays in wave vector space
(dimensionless) Nw (dimensionless). The default is 50. A cone-shaped distribution
containing Nw rays is released at each point, which may cause a very large number
of rays to be released.
• For Sample from conical distribution a single ray is released at each release point. This
ray is given a random perturbation with uniform probability density within a
cone-shaped distribution in wave vector space.

For Create light cones at release points and Sample from conical distribution, specify the
Maximum disc angle Ψm (SI unit: rad). The default is 4.65 mrad, approximately the size
of the solar disc when viewed from Earth.

For Create light cones at release points and Sample from conical distribution, select a Limb
darkening model: None (the default), Empirical power law, Linear, or User defined. The
limb darkening model reduces the intensity of solar radiation that is released from the
periphery of the solar disk, relative to radiation that is released near the center. For
Linear enter a Limb darkening coefficient β (dimensionless). The default value is 0.8. For
User defined enter a Limb darkening coefficient fL (dimensionless). The default value is 1.

Select the Include surface roughness check box to include an additional perturbation
term based on uncertainty in the orientation of the surface normal, which is common
in imperfect reflecting surfaces. Enter a value or expression for the Surface slope error
σ (SI unit: rad). The default is 1 mrad. The surface normal is then perturbed by a
random angle, which is sampled from a Rayleigh distribution.

INCIDENT RAY INTENSITY

This section is available when the ray intensity is solved for. Enter a value for the
Incident ray intensity Ii (SI unit: W/m2). The default is 1000 W/m2.

THE GEOMETRICAL OPTICS INTERFACE | 125

 TOTAL SOURCE POWER
This section is available:

• when the ray intensity is solved for, and

• when Spherical, Hemispherical, or Conical is selected as the Ray direction vector.
• It is also available when the ray power is solved for, and then any choice of Ray
direction vector displays this section.

Enter a Total source power Psrc (SI unit: W). The default is 1 W. In 2D, instead enter
the Total source power per unit thickness Psrc (SI unit: W/m). The default is 1 W/m.

Illuminated Surface Theory

The following examples involve buildings or other objects being

illuminated by solar radiation in a specific direction:

• Solar Dish Receiver: Application Library path Ray_Optics_Module/

Solar_Radiation/solar_dish_receiver
• Vdara® Caustic Surface: Application Library path
Ray_Optics_Module/Solar_Radiation/vdara_caustic_surface

Solar Radiation
The Solar Radiation node uses the solar position to specify the initial direction vector
of rays. The node is selected from the Physics toolbar, Global menu.

See Release for information on the following sections: Release Times, Initial Ray
Frequency, Vacuum Wavelength, Initial Phase, and Initial Value of Auxiliary Dependent
Variables.

INITIAL COORDINATES
These settings are the same as for the Release from Grid node.

North, West, and the up directions correspond to the x, y, and z

directions, respectively.

126 | CHAPTER 3: RAY OPTICS INTERFACES

SOURCE
Select an option from the Location defined by list: Coordinates (the default) or City.

For City select a predefined city and country combination from the list. Click to select
the Include daylight saving time (Time zone + 1) check box to add one hour to the
default setting for the city selected. For example, if New York City, USA is selected and
the default standard time zone is UTC −5 hours. If the check box is selected, the
daylight savings time is used instead (UTC −4 hours). This feature does not
automatically deduce whether daylight savings should be applied based on the selected
city and date; you must specify it explicitly.

If Coordinates is selected, or your city is not listed, in the Location table define the:

• Latitude, a decimal value, positive in the northern hemisphere (the default is Las
Vegas, USA latitude, 36.1). Enter a value without a unit to avoid double conversion.
This is because the latitude value is expected to represent degrees but the model’s
unit for angles may be different (for example, the SI unit for the angle is radians).
• Longitude, a decimal value, positive at the East of the Prime Meridian (the default is
Las Vegas, USA longitude, −115.2). Enter a value without a unit to avoid double
conversion. This is because the latitude value is expected to represent degrees but
the model’s unit for angles may be different (for example, the SI unit for the angle
is radians).
• Time zone, the number of hours to add to UTC to get local time (the default is Las
Vegas, USA time zone, −8). For example in New York City, USA the time zone is UTC
−5 hours (standard time zone) or UTC −4 hours (with daylight savings time).

For either selection, in the Date table enter the:

• Day, the default is 01. Enter a value without a unit to avoid double conversion. This
is because the value is expected to represent days but the model’s unit for time may
be different (for example, the SI unit for time is second).
• Month, the default is 10 (October). Enter a value without a unit to avoid double
conversion. This is because the value is expected to represent months but the
model’s unit for time may be different (for example, the SI unit for time is second).
• Year, the default is 2014. Enter a value without a unit to avoid double conversion.
This is because the value is expected to represent years but the model’s unit for time
may be different (for example, the SI unit for time is second). The solar position is
accurate for a date between 2000 and 2199.

THE GEOMETRICAL OPTICS INTERFACE | 127

 For either selection, in the Local time table enter the:

• Hour, the default is 12. Enter a value without a unit to avoid double conversion. This
is because the value is expected to represent hours but the model’s unit for time may
be different (for example, the SI unit for time is second).
• Minute, the default is 0. Enter a value without a unit to avoid double conversion.
This is because the value is expected to represent minutes but the model’s unit for
time may be different (for example, the SI unit for time is second).
• Second, the default is 0.

The sun position is updated if the location, date, or local time changes
during a simulation. Because the time scale of ray propagation is typically
many orders of magnitude smaller than the time scale for significant
changes in solar position, modeling ray propagation at different times of
day typically requires a Parametric Sweep over the larger time scale. Note
that no validity range is prescribed on the time inputs. It is possible to
enter values that exceed the expected boundary. For example, entering 5h
2min 81s is equivalent to 5h 3min 21s. This makes it easier, for example,
to run a Parametric Sweep over the time of day using only a single
parameter.

For either selection, in the Solar irradiance field Is (SI unit: W/m2) define the incident
radiative intensity coming from the sun.

ANGULAR PERTURBATIONS
Select an option from the Corrections for finite source diameter list: None (the default),
Create light cones at release points, or Sample from conical distribution.

• For None no perturbation to the initial ray direction is applied.

• For Create light cones at release points a cone-shaped distribution of rays is released
at each point. Enter a value for the Number of rays in wave vector space
(dimensionless) Nw (dimensionless). The default is 50. A cone-shaped distribution
containing Nw rays is released at each point, which may cause a very large number
of rays to be released.
• For Sample from conical distribution a single ray is released at each release point. This
ray is given a random perturbation with uniform probability density within a
cone-shaped distribution in wave vector space.

128 | CHAPTER 3: RAY OPTICS INTERFACES

For Create light cones at release points and Sample from conical distribution, specify the
Maximum disc angle Ψm (SI unit: rad). The default is 4.65 mrad.

For Create light cones at release points and Sample from conical distribution, select a Limb
darkening model: None (the default), Empirical power law, Linear, or User defined. The
limb darkening model is typically used to reduce the intensity of solar radiation that is
released from the periphery of the solar disk, relative to radiation that is released near
the center. For Linear enter a Limb darkening coefficient β (dimensionless). The default
value is 0.8. For User defined enter a Limb darkening coefficient fL (dimensionless). The
default value is 1.

The options for accounting for finite source diameter and solar limb
darkening are discussed in the Illuminated Surface Theory section.

Blackbody Radiation

This feature is only available in 3D.

The Blackbody Radiation node releases rays from selected boundaries with the following
properties:

1 The rays are diffusely emitted from one side of the surface, meaning that the initial
ray directions are random and their distribution follows Lambert’s cosine law,
2 If the Geometrical Optics interface has been set up to trace polychromatic light rays,
the vacuum wavelength or frequency of the released rays is random and their
distribution follows Planck’s Law, and
3 If ray intensity or power is solved for, the total source power is defined using the
Stefan-Boltzmann law.

See Release for information on the Release Times and Initial Value of Auxiliary Dependent
Variables sections.

MODEL INPUT
Enter a value or expression for the Temperature T (SI unit: K), which is used in the
Stefan-Boltzmann law for ray power and the Planck equation to sample wavelength or
frequency values. The default is 293.15 K. If the temperature has already been solved
for in another physics interface, then it may be selected from the list.

THE GEOMETRICAL OPTICS INTERFACE | 129

 INITIAL POSITION AND DIRECTION
Enter an integer value for the Number of rays per release N (dimensionless). The default
is 1000. The released rays are always randomly distributed over the selected surfaces
with uniform probability.

The Blackbody Radiation node always releases rays in random directions following
Lambert’s cosine law, which states that the flux across a differential solid angle element
is proportional to the cosine of the angle between that solid angle element and the wall
normal direction. While the settings window for the Blackbody Radiation node is
shown, the orientation of the surface normal is indicated by arrows shown in the
Graphics window. You can reverse the direction of the surface normal by selecting or
toggling the Reverse direction check box.

Blackbody Radiation Theory

Ray Termination
Use the Ray Termination feature to remove rays from the model as they exit the
geometry, or to annihilate rays of sufficiently low intensity or power. Only one Ray
Termination node can be added to the model, but it is possible to combine
position-based and intensity-based criteria for removing rays if the ray intensity or
power is computed by the model.

TERMINATION CRITERIA
Select an option from the Spatial extents of ray propagation list: Bounding box, from
geometry (the default), Bounding box, user defined, or None. This list controls whether
rays are terminated based on their positions.

• For Bounding box, from geometry, a ray is immediately terminated if it reaches the
surface of the bounding box of the geometry, defined as the smallest rectangle or
rectangular prism with edges parallel to the axes of the global coordinate system that
encloses all of the geometric entities. This bounding box is extended by 5% in each
direction so that rays can still interact with boundaries that coincide with the
perimeter of the bounding box. Therefore rays can propagate a short distance away
from the geometry before disappearing.

130 | CHAPTER 3: RAY OPTICS INTERFACES

• For Bounding box, user defined, a ray is immediately terminated if it reaches the
surface of a box of user-specified dimensions. To define the box, enter maximum
and minimum values of each spatial coordinate.
• For None, the rays that propagate outward from the geometry will continue to
propagate until the study is complete. If the maximum path length for the
Ray Tracing study step is too large, this may result in trajectories that are extremely
long relative to the characteristic geometry size.

If rays are released from outside the bounding box, they will not immediately
disappear; instead, they will only disappear if they enter the bounding box and then
exit it again.

Intensity- and Power-Based Termination Criteria

If the ray intensity or power is being computed in the model, you can then terminate
rays based on either or both of these quantities.

Select an option from the Additional termination criteria list. None (the default) is always
available. If ray intensity is solved for, then Intensity is available. If ray power is solved
for, then Power is available. If both ray intensity and power are solved for, the Intensity
and Power option is also shown.

• For Intensity or Intensity and power, enter the Threshold intensity Ith (SI unit: W/
m2). The default is 1 mW/m2. Rays will be removed from the model if their
intensity decreases below the specified value.
• For Power or Intensity and power, enter the Threshold power Qth (SI unit: W). The
default is 10-3 mW.

The Ray Termination feature can terminate rays at the exact time at which
their intensity or power reaches the specified threshold (to within
numerical precision), if there exists a closed-form analytic expression for
this time. Such a closed-form analytic solution exists if:

• The rays represent planar wavefronts in an absorbing medium, or

• The rays represent spherical or ellipsoidal wavefronts in a nonabsorbing
medium.

Otherwise, the rays are stopped at the first discrete time step taken by the
solver for which the intensity or power is less than the threshold value.

THE GEOMETRICAL OPTICS INTERFACE | 131

 Ray Termination Theory

Scattering Domain
Use the Scattering Domain feature to model extinction of light by a medium containing
a number of small spherical particles. As rays propagate through the selected domains,
some energy is either converted into heat within the particles (absorption) or sent in a
different direction (scattering). The overall removal of light from the initial beam
propagation direction is called extinction.

SCATTERING MODEL
The Scattering model list is only shown if the ray intensity or power is solved for. If so,
choose either Deterministic or Random (the default). If ray intensity or power is not
solved for, this feature always behaves as if Random is selected.

• If Deterministic is selected, the ray intensity and power will continuously decrease as
rays propagate through the selected domains, but the total number or rays will
remain the same. If the number density and size of the scattering particles is spatially
uniform, the intensity and power will decay exponentially.
• If Random is selected, rays will have some probability of disappearing as they
propagate through the selected domains. If the ray intensity and power are solved
for, they will not continuously decrease along the ray trajectories.

Use the Deterministic option for a reproducible result where nothing is random and the
ray variables vary continuously; the Random option is more of a Monte Carlo approach.

Choose an option from the Cross section calculation list. The correct choice depends
on the size of the scattering particles in comparison with the wavelength. The
individual scattering particles are said to be optically large if the particle diameter is
much larger than the wavelength of the radiation. The following options are available:

• Small spheres (Rayleigh theory) (the default) is an approximate solution for optically
small particles. The particle must be much smaller than the wavelength both inside
and outside the particle, an important consideration for tiny metallic particles.
• General spheres (Mie theory) is a full solution to the electromagnetic wave equation
as a plane wave interacts with a sphere. It is the most accurate solution across a wide
range of optical sizes, and the only approach that gives excellent accuracy at

132 | CHAPTER 3: RAY OPTICS INTERFACES

 intermediate optical sizes, when the wavelength and particle diameter are similar in
size. The calculation becomes more time-consuming for extremely large particles.
• Large conducting spheres (isotropic reflection) represents the asymptotic limit of
scattering by spheres much larger than the wavelength. The total extinction cross
section is double the geometrical cross section of the sphere because the sphere
blocks a certain amount of light and also causes diffraction along its perimeter.
• If Specify efficiency factors is selected, enter values or expressions for the
dimensionless extinction and scattering efficiency factors.
• If Specify cross section is selected, enter values of expressions for the scattering and
absorption cross sections directly.

For all choices of Cross section calculation except Specify cross section, enter the Radius
of scattering particle R (SI unit: m). The default is 1 μm.

For all choices of Cross section calculation, enter the Number density of scattering
particles N (SI unit: 1/m3). The default is 1020 m-3.

If Small spheres (Rayleigh theory) or General spheres (Mie theory) is selected, enter the
Particle refractive index, real part np,r (dimensionless). The default is 1.4. Also enter
the Particle refractive index, imaginary part np,i (dimensionless). The default is 10-6.
Internally the complex-valued refractive index is defined as n = np,r - inp,i, so positive
values here indicate that the scattering particle is absorbing.

If Specify efficiency factors is selected, enter the Extinction efficiency Qext

(dimensionless). The default is 2. Also enter the Scattering efficiency Qsca
(dimensionless). The default is 2. The absorption efficiency Qabs is then the difference
between the extinction efficiency and the scattering efficiency.

If Specify cross section is selected, enter the Extinction cross section σext (SI unit: m2).
The default is 10-15 m2. Also enter the Scattering cross section σsca (SI unit: m2). The
default is 10-15 m2. The absorption cross section σabs is then the difference between
the extinction cross section and the scattering cross section. The scattering efficiency
(cross section) should not be greater than the extinction efficiency (cross section).

Volumetric Scattering Theory

THE GEOMETRICAL OPTICS INTERFACE | 133

 Scattering Boundary
The scattering boundary can be used to scatter rays in reflection and transmission from
any boundary using one of several scattering models:

• Diffuse scattering will scatter rays according to Lambert’s cosine law. That is, the
probability of a ray propagating in a given direction within a solid angle dω is given
by cos θ dω where θ is the angle between the direction of the ray and the wall
normal.
• Isotropic scattering will cause rays to be randomly scattered in all directions.
• Surface slope error will cause the surface normal to be perturbed by a random angle
sampled from a Rayleigh distribution based on the value or expression for the
Surface slope error σ (SI unit: rad). The default is 1 mrad.
Rays can be released from a Scattering Boundary in both reflection and transmission.
That is, the options for which Rays to release include Reflected and transmitted (the
default), Reflected, or Transmitted.

If both reflected and transmitted rays are chosen, then, unless total internal reflections
occurs, the primary rays are expected to be transmitted.

WAVELENGTH OR FREQUENCY REINITIALIZATION

This section is available when the distribution of released rays is either Polychromatic,
specify vacuum wavelength, or Polychromatic, specify frequency. If activated the
wavelength or frequency of a ray can be reset as it interacts with the selected
boundaries. The new wavelength or frequency distribution function can be:

• None – the wavelength or frequency is set to a user defined expression.

• Normal – the wavelength or frequency distribution will follow a normal (Gaussian)
distribution with a user defined Mean vacuum wavelength or Mean frequency μ, and a
Vacuum wavelength standard deviation or Frequency standard deviation σ.
• Lognormal – the wavelength or frequency distribution will follow a log normal
(Gaussian) distribution with a user defined Mean vacuum wavelength or Mean
frequency μ, and a Vacuum wavelength standard deviation or Frequency standard
deviation σ
• Uniform – the wavelength or frequency will be drawn from interval specified by the
Minimum vacuum wavelength λ0,min or Minimum frequency νmin and the Maximum
vacuum wavelength λ0,max or Maximum frequency νmax

134 | CHAPTER 3: RAY OPTICS INTERFACES

NEW VALUES OF AUXILIARY DEPENDENT VARIABLES
See Wall for information on the section New Value of Auxiliary Dependent Variables.

ADVANCED SETTINGS
All of the scattering models (and wavelength or frequency reinitialization) generate
random numbers. The option to enter an Additional input argument to random number
generator i (dimensionless, default 1) is possible if in the Advance Settings section for
the Geometrical Optics interface, the Arguments for random number generation setting
is set to User defined. The Advanced Settings section is also shown if the Compute optical
path length check box is selected under the physics interface Additional Variables
section. Select the Reset optical path length check box to set the optical path length to
0 when a ray interacts with a scattering boundary.

Surface Scattering Theory

Release from Electric Field

This feature is only available if the ray intensity or power is solved for.
Select an appropriate option from the Intensity computation list in the
physics interface Intensity Computation section.

Use the Release from Electric Field node to release rays from selected boundaries, with
initial intensity and polarization based on the incident electric field at that boundary.
The electric field may be user defined or solved for in a previous study.

See Release for information on the Initial Values of Auxiliary Dependent Variables section
and the optional Release Times section. See Release from Boundary for information on
the Initial Position section.

RAY DIRECTION VECTOR

Select an option from the Ray direction vector list: Parallel to Poynting vector, Outward
normal (the default), or Inward normal. For Outward normal and Inward normal the
direction of the released rays is indicated by an arrow in the Graphics window.

The Parallel to Poynting vector option can be useful when the outgoing wave makes an
oblique angle with an absorbing boundary, such as a Port, Scattering Boundary
Condition, or Matched Boundary Condition; or when the simulation domain is

THE GEOMETRICAL OPTICS INTERFACE | 135

 surrounded by a Perfectly Matched Layer. In those scenarios, releasing the rays normal
to the surface might not be accurate.

For the Parallel to Poynting vector option, it is not sufficient to only know
the electric field at the surface; it must be differentiable in both the
tangential and normal directions. Usually this means that selecting an
Electric field defined by one of the Electromagnetic Waves interfaces will
work, but selecting a Tangential electric field will fail.

INCIDENT ELECTRIC FIELD

Enter values or expressions in the table for the Electric field E (SI unit: V/m) based on
space dimension. If the electric field is computed by another physics interface then it
can be selected from the list.

The Use frequency from the coupled physics interface as the ray frequency check box is
cleared by default. If this check box is cleared, the wavelength or frequency of the
released rays may be specified depending on the option selected from the Wavelength
distribution of released rays list in the physics interface Ray Release and Propagation
section:

• If Polychromatic, specify vacuum wavelength is selected, enter the Vacuum wavelength

λ0 (SI unit: m). The default is 660 nm.
• If Polychromatic, specify frequency is selected, enter the Ray frequency ν (SI unit: 1/
s). The default is 4.54 × 1014 Hz.

If this check box is selected, then the frequency or wavelength of the released rays will
be determined by the Frequency Domain or Wavelength Domain study in the Values of
Variables not Solved For section of the Time Dependent solver settings. Typically this
Frequency Domain or Wavelength Domain study would first be used to solve for the
electric field, using either the Electromagnetic Waves, Frequency Domain interface or
the Electromagnetic Waves, Beam Envelopes interface.

Selecting this check box may also cause the frequency or wavelength specified in the
Ray Properties node to be ignored while the Release from Electric Field node is active.

In the Geometrical Optics interface, ray intensity is stored in the Stokes

parameters for each ray. For more information, see Intensity, Wavefront
Curvature, and Polarization in Theory for the Geometrical Optics
Interface.

136 | CHAPTER 3: RAY OPTICS INTERFACES

 Ray Release Based on a Plane Electromagnetic Wave: Application
Library path Ray_Optics_Module/Tutorials/plane_em_wave_to_ray_release

Release from Far-Field Radiation Pattern

This feature requires at least one of the following licenses: RF Module,

Wave Optics Module.

In addition, this feature is only available if the ray intensity or power is

solved for. Select an appropriate option from the Intensity computation list
in the physics interface Intensity Computation section.

Use the Release from Far-Field Radiation Pattern node to release rays with an intensity
and power distribution based on a far-field radiation pattern. The far-field radiation
pattern must first be solved for in a previous study.

The Release from Far-Field Radiation Pattern node can be used for multiscale
electromagnetics modeling. First, the Electromagnetic Waves, Frequency Domain
interface should be used to compute the electric field in the immediate vicinity of a
radiation source, such as an antenna or waveguide. The Far-Field Domain node and its
default subnode, Far-Field Calculation, should be added to this instance of the
Electromagnetic Waves, Frequency Domain interface. In this wavelength-scale model,
the mesh must be fine enough to resolve individual oscillations of the electric field,
fulfilling the Nyquist criterion. Typically this requires 10 linear elements or 5
second-order elements per wavelength.

Then, in the same model component or a different component, the Geometrical

Optics interface can be used to extend the far field over an arbitrarily large number of
wavelengths, because the ray tracing approach does not require a finite element mesh
fine enough to resolve individual oscillations of the electric field.

The Release from Far-Field Radiation Pattern node is only available in 3D. However, the
far-field function used to initialize the ray intensity and polarization can be defined
either in a 3D model component or a 2D axisymmetric model component. If the
previous model component was a 2D axisymmetric component, the intensity
distribution of released rays will also be assumed to be axisymmetric.

See Release for information on the Initial Values of Auxiliary Dependent Variables section
and the optional Release Times section.

THE GEOMETRICAL OPTICS INTERFACE | 137

 INITIAL COORDINATES
Select an option from the Grid type list: All combinations (the default) or Specified
combinations.

Linear, Rectangular, and Arbitrary Grids

Enter Initial coordinates based on space dimension (qx,0, qy,0, and qz,0) for the ray
positions or click the Range button ( ) to select and define a range of specific
coordinates.

If Specified combinations is selected, the number of initial coordinates entered for each
space dimension must be equal, and the total number of rays released is equal to the
length of one of the lists of initial coordinates. If All combinations is selected, the total
number of rays released is equal to the product of the lengths of each list of initial
coordinates.

Previewing Grid Points

In the Initial Coordinates section, you can click the Preview Initial Coordinates and
Preview Initial Extents buttons to visualize the ray release positions. Clicking
Preview Initial Coordinates will cause a point to appear in the Graphics window for every
release position. Clicking Preview Initial Extents will cause a bounding box to appear,
indicating the spatial extents of the released rays.

RAY DIRECTION VECTOR

Select an option from the Ray direction vector list: Spherical (the default), Hemispherical,
or Conical.

• For Spherical a number of rays are released at each point, sampled from a spherical
distribution in wave vector space. Enter the Number of rays in wave vector space Nw
(dimensionless). The default is 50.
• For Hemispherical a number of rays are released at each point, sampled from a
hemispherical distribution in wave vector space. Enter the Number of rays in wave
vector space Nw (dimensionless). The default is 50. Then select an option from the
Hemisphere axis list: From Euler angles or User Defined. For User Defined enter
coordinates for the Hemisphere axis r based on space dimension.
• For Conical a number of rays are released at each point, sampled from a conical
distribution in wave vector space. Enter the Number of rays in wave vector space Nw
(dimensionless). The default is 50. Then select an option from the Cone axis list:
From Euler angles or User Defined. For User Defined enter coordinates for the Cone
axis r based on space dimension. Then enter the Cone angle α (SI unit: rad). The
default is π/3 radians.

138 | CHAPTER 3: RAY OPTICS INTERFACES

Note that unlike the Release from Grid feature, the options Expression and Lambertian
are not available here. In addition, if Conical is selected, the cone must be isotropic.
This is because each ray is assumed to subtend approximately the same solid angle, so
direction distributions that give unequal weight to some directions (like Lambertian)
or options with no direction distribution at all (like Expression) cannot initialize the ray
intensity and power in a consistent and physically meaningful way.

FAR-FIELD VARIABLE NAME

Enter an expression for the Far-field variable name. The default expression is Efar. This
expression must match the variable name in a corresponding Far-Field Calculation
feature.

SOURCE ORIENTATION
Use the settings in this section to orient the ray source. This allows, for example, the
release of rays from an antenna to be modeled with the antenna oriented in many
different directions, while only solving for the far-field radiation pattern once.

Enter the Euler angles (Z-X-Z). The labels Z-X-Z indicate the order in which rotation
about the different axes is performed. First, the local coordinate system of the radiation
pattern is rotated about its Z-axis by the angle α. Then it is rotated about its X-axis
(which is now at an angle α to the global x-axis) by the angle β. Finally, this local
coordinate system is rotated about its new Z-axis by the angle γ. All three inputs are
plane angles (SI unit: rad) with default values of 0.

ADVANCED SETTINGS
This section is only shown if Monochromatic is selected from the Wavelength distribution
of released rays list in the physics interface Ray Release and Propagation section.

The Use frequency from the far-field calculation as the ray frequency check box is selected
by default. While this check box is selected and the Release from Far-Field Radiation
Pattern feature is active, the frequency or vacuum wavelength of released rays will be
determined by the Frequency Domain or Wavelength Domain study in the Values of
Variables not Solved For section of the Time Dependent solver settings. Usually, this
previous study was used to solve for the far-field radiation pattern using the
Electromagnetic Waves, Frequency Domain interface.

THE GEOMETRICAL OPTICS INTERFACE | 139

 While this check box is selected and the Release from Far-Field Radiation Pattern node
is active, the frequency or wavelength specified in the Ray Properties node is ignored.

In the Geometrical Optics interface, ray intensity is stored in the Stokes

parameters for each ray. For more information, see Intensity, Wavefront
Curvature, and Polarization in the theory section.

Ray Release from a Dipole Antenna Source (3D): Application Library

path Ray_Optics_Module/Tutorials/
ray_release_from_dipole_antenna_source_3d

Ray Release from a Dipole Antenna Source (2D Axisymmetric):

Application Library path Ray_Optics_Module/Tutorials/
ray_release_from_dipole_antenna_source_2daxi

Grating
Use the Grating node to treat a boundary as a diffraction grating that can release
reflected and transmitted rays into multiple diffraction orders. A Diffraction Order
(Grating) subnode for reflected and transmitted rays of order m = 0 is added by
default. Change the settings for this default subnode to release rays of a different
diffraction order. You can also release rays of multiple diffraction orders from the same
boundary by adding more Diffraction Order subnodes from the from the context menu
(right-click the parent node) or from the Physics toolbar, Attributes menu.

The Accumulator (Boundary) subnode is also available from the context menu
(right-click the parent node) or from the Physics toolbar, Attributes menu.

Releasing Secondary Rays

When two or more rays are released from a Grating boundary, one of the released rays
uses the same degrees of freedom as the incident ray. The remaining degrees of
freedom must be preallocated in memory as secondary rays. The total number of
secondary rays that can be released in the model is controlled by the Maximum number
of secondary rays field in the physics interface Ray Release and Propagation section.

For example, if both reflected and transmitted rays of diffraction orders m = −1, m = 0,
and m = 1 are released from a boundary, then for every incident ray, a total of five
secondary rays are released. If the diffraction order m = −1 is the first subnode to
appear in the Model Builder, then the transmitted ray of order m = −1 uses the same

140 | CHAPTER 3: RAY OPTICS INTERFACES

degree of freedom as the incident ray, and the other rays are secondary rays. If the
Maximum number of secondary rays is 500 and more than 100 rays interact with the
grating, then no more secondary rays are emitted and a warning is generated by the
solver, indicating that the Maximum number of secondary rays should be increased.

DEVICE PROPERTIES
Select an option from the Rays to release list: Reflected and transmitted (the default),
Reflected, or Transmitted. This setting affects the release of rays of all diffraction orders.

Enter the a Grating constant d (SI unit: m), which is the distance between consecutive
lines in the grating, that is, the length of one unit cell. The default is 5 µm. Then select
an option from the Interpretation of grating constant list: Unit cell width on surface (the
default) or Projected unit cell width.

The choice of Interpretation of grating constant is particularly important

for curved diffraction gratings.

• If Unit cell width on surface is selected, the value of d is just the distance from one
unit cell to the next in a direction tangent to the selected surface.
• If Projected unit cell width is selected, enter the components of the Projection
direction np (dimensionless). The default is the positive y direction in 2D or the
positive z direction in 3D. Then the specified Grating constant is interpreted as the
distance between the projections of grating grooves onto a tangent plane with a
surface normal in the direction of np. If d represents the projected unit cell spacing,
then the actual cell spacing in the direction tangent to the surface is d/cos φ, where
φ is the angle between the projection direction and the surface normal.

Select Use relative order numbers if the grating blaze angle is known, and diffraction
into the highest-efficiency orders is desired. Then enter the Grating blaze angle θB
(SI unit: rad). The default is 0. The absolute order number will be computed based on
the blaze angle and the refractive indices of the adjacent domains.

If Use relative order numbers is not selected, then the Automatic Diffraction Order
Calculation section will be present. This will be discussed below.

If the ray power is solved for, the Store total transmitted power and Store total reflected
power check boxes are shown. Select them to declare auxiliary dependent variables for
the total power of all transmitted and reflected diffraction orders, respectively.

THE GEOMETRICAL OPTICS INTERFACE | 141

 GRATING ORIENTATION
In 2D, select an option from the Direction of periodicity list: Forward (the default) or
Reverse. For either setting, an arrow appears in the Graphics window, lying tangent to
the selected edges. Reflected and refracted rays from positive diffraction orders have a
larger ray direction component in the direction of this arrow, compared to rays from
negative diffraction orders.

In 3D, select an option from the Grating orientation list: Specify direction of periodicity
(the default) or Specify direction of grating lines.

• If Specify direction of periodicity is selected, select an option from the Direction of

periodicity list: User defined (the default) or Parallel to reference edge. For User
defined enter the components of the direction of periodicity Tp (dimensionless)
directly. The default is the positive x-axis. For Parallel to reference edge, the Reference
Edge Selection section is shown in the Settings window. Select a single edge, which
must be adjacent to at least one boundary in the selection for the Grating feature.
• If Specify direction of grating lines is selected, select an option from the Direction of
grating lines list: User defined (the default) or Parallel to reference edge. For User
defined enter the components of the direction of grating lines Tl (dimensionless)
directly. The default is the positive y-axis. For Parallel to reference edge, the Reference
Edge Selection section is shown in the Settings window. Select a single edge, which
must be adjacent to at least one boundary in the selection for the Grating feature.

142 | CHAPTER 3: RAY OPTICS INTERFACES

Regardless of how Tp or Tl has been specified, the direction of periodicity projected
onto the grating surface Tg will be computed. This is shown in Figure 3-8 for 2D and
Figure 3-9 for 3D.

Figure 3-8: Diagram of diffraction grating orientation in 2D.

Figure 3-9: Diagram illustrating the options to specify grating orientation in 3D.

AUTOMATIC DIFFRACTION ORDER CALCULATION

The Automatic Diffraction Order Calculation section is not shown if the Use
relative order numbers check box is selected; see the Device Properties
section.

Use the Add Diffraction Orders button to automatically add diffraction orders to the
Grating feature. The other settings in this section allow you to automatically generate

THE GEOMETRICAL OPTICS INTERFACE | 143

 the diffraction orders that would be produced by a ray of specified wavelength at a
specified angle of incidence. The recommended workflow is to first configure all of the
inputs in this section, and then click Add Diffraction Orders.

For each diffraction order a corresponding Diffraction Order (Grating) subnode will
be added to this node. If some Diffraction Order subnodes are already present, then
clicking Add Diffraction Orders will delete them and replace them with the automatically
generated nodes. A warning will be issued if this operation might result in the creation
of a very large number of subnodes.

From the Automatic diffraction order type list select either Specify minimum wavelength
(the default) or Specify maximum frequency. The Minimum wavelength or Maximum
frequency can be either taken From Physics (the default) or User defined. If From Physics
is selected, then the algorithm that generates the diffraction orders will first search
through all ray release features to determine the minimum wavelength.

In addition, specify the Angle of incidence θi (SI unit: rad). In 3D, instead enter both
the Angle of incidence, in plane θi,in (SI unit: rad) and the Angle of incidence, out of plane
θi,out (SI unit: rad). Also enter the Refractive index, incoming n1 and Refractive index,
outgoing n2. The defaults are 0. If the Rays to release are Reflected then n2 = n1. See
Figure 3-10 and Figure 3-11 for details.

Figure 3-10: The definition of the angle of incidence in 2D.

144 | CHAPTER 3: RAY OPTICS INTERFACES

Figure 3-11: The definition of the in-plane and out-of-plane angles of incidence in 3D.

Diffraction Grating Theory

Diffraction Order (Grating)

The Diffraction Order subnode is available from the context menu (right-click the
Grating parent node) or from the Physics toolbar, Attributes menu. Each Diffraction
Order subnode can either release transmitted rays, reflected rays, or both transmitted
and reflected rays of a specified diffraction order, depending on the option selected
from the Rays to release list in the parent Grating node.

DEVICE PROPERTIES
If the Use relative order numbers is not selected in the settings for the parent Grating
node, then enter a Diffraction order m (dimensionless). The default is m = 0 for the first
Diffraction Order added. When more diffraction orders are added to the model, the new
default is the highest diffraction order currently present, plus 1.

If Use relative order numbers is selected in the settings for the parent Grating node, then
enter a Relative diffraction order Δm (dimensionless). As above, the default is Δm = 0
if no other Diffraction Order subnode exists; otherwise the default is 1 greater than the
highest relative order that is currently present.

When the ray intensity or power is solved for, enter the Reflectance R and the
Transmittance T (both dimensionless). If the reflectance or transmittance of a certain
diffraction order is exactly zero, that diffraction order will not be released.

THE GEOMETRICAL OPTICS INTERFACE | 145

 AUXILIARY DEPENDENT VARIABLES, REFRACTED RAY
This section is available if an Auxiliary Dependent Variable has been added to the
model. The settings are the same as for the Material Discontinuity feature.

AUXILIARY DEPENDENT VARIABLES, REFLECTED RAY

This section is available if an Auxiliary Dependent Variable has been added to the
model. The settings are the same as for the Material Discontinuity feature.

If relative order numbers are being used then the absolute order number
can be tracked using an auxiliary dependent variable.

Cross Grating

This feature is only available in 3D.

Use the Cross Grating node to treat a boundary as a diffraction grating that can release
reflected and transmitted rays into multiple diffraction orders. Unlike a standard
Grating, a Cross Grating can have two directions of periodicity; the boundary is treated
as a general two-dimensional periodic substructure. A Diffraction Order (Cross
Grating) subnode for reflected and transmitted rays in order m = n = 0 is added by
default. Change the settings for this default subnode to release rays into different
diffraction orders. You can also release rays of multiple diffraction order pairs from the
same boundary by adding more Diffraction Order subnodes from the from the context
menu (right-click the parent node) or from the Physics toolbar, Attributes menu.

The Accumulator (Boundary) subnode is also available from the context menu
(right-click the parent node) or from the Physics toolbar, Attributes menu.

Releasing Secondary Rays

When two or more rays are released from a Cross Grating boundary, one of the released
rays uses the same degrees of freedom as the incident ray. The remaining degrees of
freedom must be preallocated in memory as secondary rays. The total number of
secondary rays that can be released in the model is controlled by the Maximum number
of secondary rays field in the physics interface Ray Release and Propagation section. See
the Grating feature for further details.

146 | CHAPTER 3: RAY OPTICS INTERFACES

DEVICE PROPERTIES
Select an option from the Rays to release list: Reflected and transmitted (the default),
Reflected, or Transmitted. Then enter the Grating constant, direction 1 d1 (SI unit: m)
and the Grating constant, direction 2 d2 (SI unit: m). The defaults are d1 = 1 μm and
d2 = 2 μm.

If the ray power is solved for, the Store total transmitted power and Store total reflected
power check boxes are shown. Select them to declare auxiliary dependent variables for
the total power of all transmitted and reflected diffraction orders, respectively.

GRATING ORIENTATION
The Cross Grating requires two grating orientations to be specified. For each of Grating
orientation 1 and Grating orientation 2 select either Specify direction of periodicity (the
default) or Specify direction of grating lines. The Grating orientation 2 may also be
Orthogonal to grating orientation 1.

• If Specify direction of periodicity is selected, select an option from the Direction of

periodicity 1 (or 2) list: User defined (the default) or Parallel to reference edge. For
User defined enter the components of the direction of periodicity Tp,1 or Tp,2
(dimensionless) directly. The defaults for Tp,1 and Tp,2 are the positive x-axis and
positive y-axis respectively. For Parallel to reference edge, the Reference Edge
Selection, Direction 1 and/or Reference Edge Selection, Direction 2 sections are shown.
Select a single edge (for each direction), which must be adjacent to at least one
boundary in the selection for the Cross Grating feature and not parallel to each other.

• If Specify direction of grating lines is selected, select an option from the Direction of
grating lines 1 (or 2) list: User defined (the default) or Parallel to reference edge. For
User defined enter the components of the direction of grating lines Tl,1 or Tl,2
(dimensionless) directly. The defaults for Tl,1 and Tl,2 are the positive y-axis and
positive x-axis respectively. For Parallel to reference edge, the Reference Edge
Selection, Direction 1 and/or Reference Edge Selection, Direction 2 sections are shown.
Select a single edge (for each direction), which must be adjacent to at least one
boundary in the selection for the Cross Grating feature and not parallel to each other.
Similar to the Grating, regardless of the grating orientations are specified, the
directions of periodicity projected onto the grating surface Tg,1 and Tg,2 will be
computed. This is shown in Figure 3-12, along with typical paths for incident
reflected, and transmitted rays.

THE GEOMETRICAL OPTICS INTERFACE | 147

 Figure 3-12: Diagram illustrating the cross grating orientation. The directions of
periodicity and/or lines (which need not be orthogonal) are projected onto the cross grating
surface.

Diffraction Grating Theory

Diffraction Order (Cross Grating)

The Diffraction Order subnode is available from the context menu (right-click the Cross
Grating parent node) or from the Physics toolbar, Attributes menu. Each Diffraction
Order subnode can either release transmitted rays, reflected rays, or both transmitted
and reflected rays of the specified order, depending on the option selected from the
Rays to release list in the parent Cross Grating node.

DEVICE PROPERTIES
Enter the Diffraction orders m and n (dimensionless). The default is m = n = 0 for the
first Diffraction Order subnode added. When more diffraction orders are added to the
model, the new default created by adding 1 to either m or n, whichever is lowest. If m
and n are equal, then m is incremented by 1.

When the ray intensity or power is solved for, enter the Reflectance R and the
Transmittance T (both dimensionless). If the reflectance or transmittance of a certain
diffraction order is exactly zero, that diffraction order will not be released.

AUXILIARY DEPENDENT VARIABLES, REFRACTED RAY

This section is available if an Auxiliary Dependent Variable has been added to the
model. The settings are the same as for the Material Discontinuity feature.

AUXILIARY DEPENDENT VARIABLES, REFLECTED RAY

This section is available if an Auxiliary Dependent Variable has been added to the
model. The settings are the same as for the Material Discontinuity feature

148 | CHAPTER 3: RAY OPTICS INTERFACES

Linear Polarizer
Use the Linear Polarizer node to define a linear polarizer on a boundary selection. Add
this node from the Optical Devices submenu when ray intensity or power is solved for
in the model. The Accumulator (Boundary) subnode is available from the context
menu (right-click the parent node) or from the Physics toolbar, Attributes menu.

DEVICE PROPERTIES
Select a Transmission axis specification:

• User defined (the default). Then enter coordinates for the Transmission axis T
(dimensionless).
• Parallel to reference edge (for 3D components). The Reference Edge Selection section
is then shown. Add an edge to this selection to specify the transmission axis
orientation. Only a single edge may be selected. The edge must be adjacent to a
boundary in the physics feature selection.
• In-plane or Out-of-plane (for 2D and 2D axisymmetric components)

NEW VALUE OF AUXILIARY DEPENDENT VARIABLES

This section is available if an Auxiliary Dependent Variable has been added to the
model. The settings are the same as for the Wall feature.

Linear Polarizer Theory

Ideal Depolarizer
Use the Ideal Depolarizer node to reinitialize the Stokes parameters so that the
outgoing ray is completely unpolarized. Add this node from the Optical Devices
submenu when ray intensity or power is solved for in the model. The Accumulator
(Boundary) subnode is available from the context menu (right-click the parent node)
or from the Physics toolbar, Attributes menu.

THE GEOMETRICAL OPTICS INTERFACE | 149

 NEW VALUE OF AUXILIARY DEPENDENT VARIABLES
This section is available if an Auxiliary Dependent Variable has been added to the
model.

Ideal Depolarizer Theory

Linear Wave Retarder

Use the Linear Wave Retarder node to define a linear wave retarder on a boundary
selection. Add this node from the Optical Devices submenu when ray intensity or power
is solved for in the model. The Accumulator (Boundary) subnode is available from the
context menu (right-click the parent node) or from the Physics toolbar, Attributes
menu.

DEVICE PROPERTIES
Select a Fast axis specification:

• User defined (the default). Then enter coordinates for the Fast axis F
(dimensionless).
• Parallel to reference edge (for 3D components). The Reference Edge Selection section
is then shown. Add an edge to this selection to specify the fast axis orientation. Only
a single edge may be selected. The edge must be adjacent to a boundary in the
physics feature selection.
• In-plane or Out-of-plane (for 2D and 2D axisymmetric components)

For any selection, enter a value for the Retardance δ (SI unit: rad).

NEW VALUE OF AUXILIARY DEPENDENT VARIABLES

This section is available if an Auxiliary Dependent Variable has been added to the
model. The settings are the same as for the Wall feature.

Linear Wave Retarder Theory

150 | CHAPTER 3: RAY OPTICS INTERFACES

Circular Wave Retarder
Use the Circular Wave Retarder node to define a circular wave retarder on a boundary
selection. Add this node from the Optical Devices submenu when ray intensity or power
is solved for in the model. The Accumulator (Boundary) subnode is available from the
context menu (right-click the parent node) or from the Physics toolbar, Attributes
menu.

DEVICE PROPERTIES
Enter a value for the Retardance δ (SI unit: rad).

By default the Sense of rotation depends on ray trajectory check box is not selected. This
means a positive value of the retardance always corresponds to clockwise rotation of
the polarization ellipse. Select the check box then a positive value of the retardance
corresponds to clockwise rotation of the polarization ellipse only when the ray crosses
the boundary in the direction indicated by the normal vector in the Graphics window.

The Reverse sense of rotation check box is available when the Sense of rotation depends
on ray trajectory check box is selected. Click to select this check box to reverse the
direction of the normal vector shown in the Graphics window.

NEW VALUE OF AUXILIARY DEPENDENT VARIABLES

This section is available if an Auxiliary Dependent Variable has been added to the
model. The settings are the same as for the Wall feature.

Circular Wave Retarder Theory

Mueller Matrix
Use the Mueller Matrix node to define Mueller matrices for customized optical
components. Add this node from the Optical Devices submenu when ray intensity or
power is solved for in the model. The Accumulator (Boundary) subnode is available
from the context menu (right-click the parent node) or from the Physics toolbar,
Attributes menu.

DEVICE PROPERTIES
Select a Reference axis specification:

• User defined (default). Enter coordinates for the Reference axis R (dimensionless).

THE GEOMETRICAL OPTICS INTERFACE | 151

 • Parallel to reference edge (for 3D components). The Reference Edge Selection section
is then shown. Add an edge to this selection to specify the reference direction
orientation. Only a single edge may be selected. The edge must be adjacent to a
boundary in the physics feature selection.
• In-plane or Out-of-plane (for 2D and 2D axisymmetric components).

For any choice, select Isotropic (the default), Diagonal, Symmetric, or Full from the list.
Then enter a value or expression for the Mueller matrix M in the matrix or field.

NEW VALUE OF AUXILIARY DEPENDENT VARIABLES

This section is available if an Auxiliary Dependent Variable has been added to the
model. The settings are the same as for the Wall feature.

Theory of Mueller Matrices and Optical Components

Fluence Rate Calculation

The Fluence Rate Calculation node is used to compute the fluence rate within domains.
Fluence rate is defined as the total radiant power that would hit a small sphere located
within the domain, divided by the cross sectional area of the sphere. The sphere is
assumed to be much smaller than any geometric details but much larger than the
wavelength of the radiation, so that Mie scattering can be neglected.

This feature is available when the ray power is solved for. The variable for fluence rate
is <phys>.<name>.E0, where <phys> is the name of the Geometrical Optics physics
interface and <name> is the name of the physics feature, for example gop.frc1.E0.

Fluence Rate Calculation Theory

Deposited Ray Power (Domain)

The Deposited Ray Power node is used to compute a heat source based on the
attenuation of rays as they propagate through a domain with a complex refractive
index. This feature is available when the ray power is solved for. The variable for
deposited power is <name>.Qsrc, where <name> is the name of the Geometrical
Optics physics interface, for example gop.Qsrc.

152 | CHAPTER 3: RAY OPTICS INTERFACES

 Attenuation Within Domains

Deposited Ray Power (Boundary)

When the ray power is solved for, the Deposited Ray Power subnode is available from
the context menu (right-click the Wall, parent node) or from the Physics toolbar,
Attributes menu. The Deposited Ray Power subnode computes the total incident energy
flux on a surface based on the power of incident rays.

SMOOTHING
Select the Compute smoothed accumulated variable check box to enter a Smoothing
radius r (SI unit: m) The smoothing radius functions like the corresponding setting for
the Accumulator (Boundary) node.

Accumulator (Domain)
Use the Accumulator node to define additional degrees of freedom on a domain. Each
Accumulator defines a variable, called the accumulated variable, on each domain
element in the selection list. The values of the accumulated variables are determined
by the properties of rays in each domain element.

ACCUMULATOR SETTINGS
Select an Accumulator type: Density (default) or Count.

• For Density the accumulated variable is divided by the volume of the mesh element
where it is defined.
• For Count the accumulated variable is unaffected by the element size.

Select an option from the Accumulate over list: Elements (the default) or Elements and
time.

• For Elements the value of the accumulated variable in an element is the sum of the
source terms of all rays in that element. If the Accumulator type is set to Density, this
sum is divided by the mesh element volume.
• For Elements and time the time derivative of the accumulated variable in an element
is the sum of the source terms of all rays in that element. If the Accumulator type is

THE GEOMETRICAL OPTICS INTERFACE | 153

 set to Density, this sum is divided by the mesh element volume. As each ray
propagates through a series of mesh elements, it leaves behind a contribution to the
accumulated variable that remains even after the ray has moved on.

Enter the Accumulated variable name. The default is rpd. The accumulated variable is
defined as <name>.<varname>, where <name> is the physics interface name and
<varname> is the accumulated variable name. For example, in an instance of the
Geometrical Optics interface with default name gop and default accumulated variable
name rpd, the variable would be named gop.rpd.

Enter a Source R. The unit of the source depends on the settings in the Units section.
The source term is used to calculate the accumulated variable in a manner specified by
the Accumulate over and Accumulator type settings.

If Elements and time is selected from the Accumulate over list, select an option from the
Source interpolation list: Constant, Linear (the default), Quadratic, or Exponential. The
Source interpolation determines what functional form the Source is assumed to follow
during each time step taken by the solver. This information is used to compute the
accumulated variable in mesh elements that the rays pass through during each time
step.

UNITS
Select a Dependent variable quantity from the list; the default is Dimensionless [1]. To
enter a unit, select None from the list and in the Unit field enter a value, for example, K,
m/s, or mol/m^3.

Accumulator (Boundary)
The Accumulator subnode is available from the context menu (right-click the Wall or
Material Discontinuity parent node) or from the Physics toolbar, Attributes menu. Each
Accumulator subnode defines a variable, called the accumulated variable, on each
boundary element in the selection of the parent node. The accumulated variables are
incremented when rays interact with the boundaries where they are defined.

ACCUMULATOR SETTINGS
Select an option from the Accumulator type list: Density (default) or Count.

• For Density the accumulated variable is divided by the surface area of the boundary
element where it is defined.
• For Count the accumulated variable is the sum of the source terms of all rays that hit
the boundary element, and is unaffected by the boundary element size.

154 | CHAPTER 3: RAY OPTICS INTERFACES

Select an option from the Accumulate over list: Ray-wall interactions (the default) or
Rays in boundary elements.

• For Ray-wall interactions the accumulated variable is affected by all rays that hit the
boundary element.
• For Rays in boundary elements the accumulated variable is only affected by rays that
freeze or stick to the boundary element.

Enter the Accumulated variable name. The default is rpb. The accumulated variable is
defined as <scope>.<name>, where <scope> includes the name of the physics
interface node, parent boundary condition, and the Accumulator node, and <name> is
the accumulated variable name.

For example, if the Accumulator subnode is added to a Wall node in an instance of the
Geometrical Optics interface using the default variable name rpb, the accumulated
variable name might be gop.wall1.bacc1.rpb.

Enter a Source R. The unit of the source term depends on the settings in the Units
section. Whenever a ray collides with a boundary element in the selection of the parent
node, the accumulated variable in that element is incremented by the source term. If
the Accumulator type is set to Density, the source term is divided by the area of the
boundary element (in 3D) or the length of the boundary element (in 2D).

For example, if the source is 1 and Count is selected from the Accumulator type list, a
variable is created to store the number of times each boundary element is hit by a ray.

UNITS
Select a Dependent variable quantity from the list; the default is Dimensionless [1]. To
enter a unit, select None from the list and in the Unit field enter a value, for example, K,
m/s, or mol/m^3.

SMOOTHING
The accumulated variables are computed using discontinuous shape functions. Select
the Compute smoothed accumulated variable check box to compute a smoothed
accumulated variable by computing the average value of the variable within a sphere of
a user-defined radius. Then enter a Smoothing radius r (SI unit: m). The default is 1 m.

Accumulator Theory: Boundaries

THE GEOMETRICAL OPTICS INTERFACE | 155

 Nonlocal Accumulator
Use the Nonlocal Accumulator subnode to communicate information from a ray’s
current position to the position from which it was released.

The subnode is available from the context menu (right-click the Release from
Boundary parent node) or from the Physics toolbar, Attributes menu.

Each Nonlocal accumulator subnode defines a variable, called the accumulated variable,
that is computed using variables defined on rays released by the parent Release from
Boundary node or on domains and boundaries encountered by such rays.

ACCUMULATOR SETTINGS
Select an Accumulator type: Density (default) or Count.

• For Density the accumulated variable is divided by the volume of the mesh element
where it is defined.
• For Count the accumulated variable is unaffected by the element size.

Select an option from the Accumulate over list: Elements (default) or Elements and
time.

• For Elements the accumulated variable is proportional to the instantaneous value of

the Source term R for all applicable rays.
• For Elements and time the time derivative of the accumulated variable is proportional
to the instantaneous value of the Source term R for all applicable rays, and thus the
accumulated variable considers the time history of rays in the modeling domain
instead of just their current values.

Enter the Accumulated variable name. The default is rpi.

Enter a Source R. The unit of the source depends on the settings in the Units section.
The source term is used to calculate the accumulated variable in a manner specified by
the Accumulate over and Accumulator type settings.

Select a Source geometric entity level: Domains, Boundaries, or Domains and boundaries.

• If Domains is selected, rays only contribute to the accumulated variable on their

releasing surface while they are active (still propagating through a domain).
• If Boundaries is selected, the rays only contribute to the accumulated variable if they
have become stuck or frozen to a boundary somewhere in the model.
• If Domains and boundaries is selected, all of the active, stuck, and frozen rays released
by a feature can contribute to the accumulated variable.

156 | CHAPTER 3: RAY OPTICS INTERFACES

UNITS
Select a Dependent variable quantity from the list; the default is Dimensionless [1]. To
enter a unit, select None from the list and in the Unit field enter a value.

SMOOTHING
Select the Compute smoothed accumulated variable check box to enter a Smoothing
radius r (SI unit: m). The default is 0.1 m.

Auxiliary Dependent Variable

Use the Auxiliary Dependent Variable node to solve additional first-order ordinary
differential equations for each ray released.

AUXILIARY DEPENDENT VARIABLE

Enter a Field variable name. The default is rp and can be changed to anything provided
it does not conflict with the name of the variables for the position or momentum
degrees of freedom, or with other auxiliary dependent variables.

Enter a Source R. The unit of the source depends on the unit of the auxiliary dependent
variable and the option selected from the Integrate list described below.

Select a way to Integrate the equation you have defined: With respect to time or Along
ray trajectory. For example, to reproduce the variable for optical path length, you could
set the Source to 1 and set Integrate to Along ray trajectory. To compute the length of
the ray trajectory you can set the Source to 1 and set Integrate to Along ray trajectories.

UNITS
These settings are the same as for Accumulator (Domain).

Ray Continuity
Use the Ray Continuity node to specify that rays should cross a pair boundary as if it
were invisible. Pair boundaries appear when the geometry sequence ends in Form
Assembly instead of Form Union. Such boundaries require special handling because the
mesh elements on either side of the pair boundary are not required to match up exactly.

PAIR SELECTION
Select one or more identity pairs to allow rays to cross between the source and
destination boundaries of these pairs. Such identity pairs are typically created

THE GEOMETRICAL OPTICS INTERFACE | 157

 automatically on interior boundaries when the geometry sequence ends in a Form
Assembly node instead of a Form Union node.

The Ray Continuity node does not cause the ray position components to
change discontinuously; the source and destination boundaries for the
identity pairs must be overlapping.

If the geometry sequence ends in a Form Union node, the Ray Continuity
node usually is not needed because rays can freely cross interior
boundaries where no boundary condition has been applied.

Ray Detector
Use the Ray Detector feature to compute information about rays that are located in a
set of selected domains or on a set of selected boundaries. The detector may detect all
rays or only the rays released by a specified release feature. Computed variables are the
number of rays transmitted, the transmission probability, and a logical expression
which can be used to filter the rendered rays during results processing.

The Ray Detector feature is also capable of computing the total transmitted ray power.
If the ray power is solved for, then a global variable for the total power transmitted to
the detector is available.

RAY DETECTOR
Select an option from the Release feature list: All (the default), or any of the features
which are capable of releasing rays. When All is selected, the variables computed are
summed over all release features in the model.

158 | CHAPTER 3: RAY OPTICS INTERFACES

Theory for the Geometrical Optics
Interface
The Geometrical Optics Interface theory is described in this section. The following
sections are listed in the order in which they appear:

• Note on Sign Conventions • Illuminated Surface Theory

• The Electromagnetic Wave • Theory of Mueller Matrices and
Equation Optical Components
• Introduction to Geometrical Optics • Thin Dielectric Film Theory
• Optical Dispersion Models • Diffraction Grating Theory
• Thermo-Optic Dispersion Models • Ideal Lens Theory
• Optical Attenuation Models • Surface Scattering Theory
• Initial Conditions: Position • Volumetric Scattering Theory
• Initial Conditions: Direction • Interference Pattern Theory
• Initial Conditions: Wavelength or • Monochromatic Aberrations and the
Frequency Zernike Polynomials
• Blackbody Radiation Theory • Accumulator Theory: Domains
• Gaussian Beam Theory • Accumulator Theory: Boundaries
• Material Discontinuity Theory • References for the Geometrical
• The Fresnel Equations Optics Interface

• Intensity, Wavefront Curvature, and

Polarization
• Wavefront Curvature Calculation in
Graded Media
• Refraction in Strongly Absorbing
Media
• Attenuation Within Domains
• Ray Termination Theory

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 159

 Note on Sign Conventions
When describing the propagation of a plane electromagnetic wave, different references
will employ different sign conventions. The choice of sign convention can have
far-reaching implications for the interpretation of phase delays and material properties,
among other things.

A plane electromagnetic wave is sometimes defined as

E = E 0 exp [ i ( k ⋅ r – ωt ) ] (3-1)

and at other times as

E = E 0 exp [ i ( ωt – k ⋅ r ) ] (3-2)

where

• E (SI unit: V/m) is the electric field along the ray,

• E0 (SI unit: V/m) is the electric field amplitude at r = 0 and t = 0,
• k (SI unit: rad/m) is the wave vector,
• r (SI unit: m) is the position vector,
• ω (SI unit: rad/s) is the angular frequency, and
• t (SI unit: s) is time.

For example, Refs. 1 and 9 use the sign convention in Equation 3-1, whereas Refs. 10,
12, 15, 17, and 20 use the sign convention in Equation 3-2.

Suppose that the wave propagates in the positive x direction. Then the wave vector
may be defined as

ωn
k =  --------, 0, 0
c

where n (dimensionless) is the refractive index of the medium, assumed to be isotropic.

For an absorbing or gain medium, n is complex-valued. Using Equation 3-1, the
electric field along a plane wave in an absorbing or gain medium is

ω ω
E = E 0 exp [ – iωt ] exp i ---- xRe ( n ) exp – ---- xIm ( n )
c c

which has an exponentially decaying magnitude (absorbing medium) if the imaginary

part of n is positive, and an exponentially growing magnitude (gain medium) if the

160 | CHAPTER 3: RAY OPTICS INTERFACES

imaginary part of n is negative. In contrast, when starting from Equation 3-2, the
electric field is

ω ω
E = E 0 exp [ iωt ] exp – i ---- xRe ( n ) exp ---- xIm ( n )
c c

So the medium is absorbing if the imaginary part of n is negative, or a gain medium if

the imaginary part of n is positive.

The Geometrical Optics interface uses the convention of Equation 3-2. Therefore the
complex-valued refractive index may be written as n − iκ, where both n and κ are
real-valued. The variable κ, sometimes called the extinction coefficient, should then be
positive when defining an absorbing medium, for example when specifying the
Refractive index, imaginary part in the Medium Properties node.

The Electromagnetic Wave Equation

The relationships between the electric and magnetic fields are together known as
Maxwell’s Equations (Ref. 10),

∂B
∇ × E = – -------
∂t
∂D
∇ × H = -------- + J
∂t
∇⋅D = ρ

∇⋅B = 0

where

• E (SI unit: V/m) is the electric field,

• D (SI unit: C/m2) is the electric displacement,
• H (SI unit: A/m) is the magnetic field,
• B (SI unit: Wb/m2) is the magnetic flux density or magnetic induction,
• ρ (SI unit: C/m3) is the charge density, and
• J (SI unit: A/m2) is the current density.

The above equations are formulated in SI units. In Gaussian units (Ref. 9), the
equations may be presented with additional factors of 4π and some additional divisions
by the speed of light in a vacuum c.

The electric and magnetic quantities are further governed by the constitutive relations

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 161

 D = εE
B = μH

where ε (SI unit: F/m) is the electric permittivity and μ (SI unit: H/m) is the
magnetic permeability. Generally ε and μ are tensors of rank 2 but they can be treated
as scalar quantities if the medium is isotropic.

In a medium that is charge-free, current-free, homogeneous, and isotropic, the first

two of Maxwell’s Equations may be simplified to

∂H
∇ × E = – μ --------
∂t
∂E
∇ × H = ε -------
∂t

One of the two remaining field variables can be eliminated by substitution, giving the
so-called “curl-curl” equation for the remaining field,

1 ∂2E
∇ × --- ∇ × E + ε ---------
- = 0
μ ∂t 2
1 ∂ 2 H-
∇ × --- ∇ × H + μ ---------- =0
ε ∂t 2

If the medium is homogeneous, meaning it has spatially independent dielectric

properties, then the factors of ε and μ can be taken outside the curl operator, and then
using the identity

∇ × ∇ × E = ∇(∇ ⋅ E ) – ∇2E

and recalling that the fields are divergence-free, the electromagnetic wave equation
may be written as

∂ 2 E-
∇ 2 E – με --------- = 0
∂t 2
∂ H-
2
∇ 2 H – με ---------- =0
∂t 2

Thus, either the electric field or magnetic field can be obtained by solving a vector
Helmholtz equation. Here the electric field will be used.

The vector Helmholtz equation permits a variety of different solutions, one of which
is a plane electromagnetic wave,

E = E 0 cos ( ωt – k ⋅ r )

162 | CHAPTER 3: RAY OPTICS INTERFACES

where

• k (SI unit: rad/m) is the wave vector,

• r (SI unit: m) is the position vector,
• ω (SI unit: rad/s) is the angular frequency,
• t (SI unit: s) is time, and
• E0 (SI unit: V/m) is the complex-valued electric field amplitude at r = 0 and t = 0.

It is more common to express E as the exponential of a complex phase,

E = E 0 exp [ i ( ωt – k ⋅ r ) ]

Although the real operator is not explicitly applied here, it should be understood that
only the real part of this expression for E has any physical significance, and that the
imaginary part is just a mathematical construction.

The form of the electric field is a set of plane surfaces of constant phase that propagate
forward at the phase velocity

ω
v p = ---- (3-3)
k

Where k is the magnitude of the wave vector k. This is the speed of light in the
medium.

Differentiating E gives

∂E
------- = iωE
∂t

∇ ⋅ E = –i ( k ⋅ E )

However, since E is divergence-free, k and E must be orthogonal. The same can be

deduced about k and H. In addition, from Maxwell’s Equations,

∇ × H = iωεE

So that k, E, and H are all orthogonal.

Substituting the electromagnetic plane wave definition back into the electromagnetic
wave equation yields

k 2 E – μεω 2 E = 0

So the relationship between the angular frequency and wave vector norm is

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 163

 k = ω με (3-4)

Comparing Equation 3-3 and Equation 3-4 shows that

1
v p = ----------
με

Now express the permittivity and permeability as

ε = εr ε0 μ = μr μ0

where

• ε0 = 8.854187817 × 10-12 F/m is the vacuum permittivity,

• μ0 = 1.256637062 × 10-6 H/m is the vacuum permeability,
• εr (dimensionless) is the relative permittivity of the medium, and
• μr (dimensionless) is the relative permeability of the medium.

Noting that

1
c = ----------------
μ0 ε0

where c = 2.99792458 × 108 m/s is the speed of light in a vacuum, the phase velocity
is

c
v p = ---------------
μr εr

At optical frequencies, many materials are nonmagnetic, μr = 1. In that case, the phase
velocity is just c/n, where

n = εr

is the refractive index of the material. Thus, for linear nonmagnetic materials,

ω
k = ---- n
c

If the medium absorbs energy from the propagating wave or adds energy to it, then
the refractive index becomes a complex-valued quantity, denoted n - iκ, where κ is
called the extinction coefficient. For absorbing media, κ > 0, whereas for gain media,
κ < 0.

164 | CHAPTER 3: RAY OPTICS INTERFACES

Introduction to Geometrical Optics
The previous sections show that Maxwell’s Equations allow a plane wave solution,

E = E 0 e iΨ

where E0 is the slowly varying amplitude and the phase Ψ is a function of the position
vector q and time t. If the field is observed only at locations that are very far from any
source, the phase can be expressed as

Ψ = ωt – k ⋅ q + α

where k is the wave vector, ω is the angular frequency, and α is an arbitrary phase shift.
In an isotropic medium, the wave vector and the angular frequency are further related
by the expression

ck
ω = ------------
n(q)

where n(q) is the refractive index of the medium. It follows that the wave vector and
angular frequency can be expressed in terms of the phase:

∂Ψ ∂Ψ
k = -------- ω = – --------
∂q ∂t

Following Landau and Lifshitz in Ref. 1, the wave vector and frequency are analogous
to the generalized momentum p and Hamiltonian H of a solid particle,

∂S ∂S
p = ------- H = – -------
∂q ∂t

where S is the integral of the Lagrangian along the particle’s trajectory. From this
analogy, it follows that the ray trajectory can be computed by solving six coupled
first-order ordinary differential equations for the components of k and q:

∂ω
------ = ∂ω
dk dq
------ = – ------- -------
dt ∂q dt ∂k

In regions of constant refractive index, the simplified equations of motion are:

dk dq ck-
------ = 0 ------ = ---------
dt dt n k

These first-order equations for the ray trajectory are only valid in regions that are many
wavelengths away from any point source; that is, Ψ » 1 .

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 165

 Optical Dispersion Models
The Refractive index, real part, n (dimensionless) may be expressed as a function of
wavelength using an optical dispersion model. The Optical dispersion model for
domains outside the selection can be specified in the Geometrical Optics interface.
Domains that are included in the Geometrical Optics interface selection have
dispersion models that are specified in the Medium Properties feature.

THE REFRACTIVE INDEX OF AIR

A dispersion model for the refractive index of air can be used in exterior and unmeshed
domains. It can also be used to convert from relative to absolute refractive indices (see
below). The Edlén (1953) air model, described in Ref. 2, which gives the refractive
index of air nair as a function of wavelength, temperature, and pressure is

n air, STP ( λ 0 ) – 1 P
n air ( λ 0, T, P ) = 1 + --------------------------------------------------------------------
–3
---------- (3-5)
1 + 3.4785 ×10 ( T – T std ) std P

where T is the temperature (SI unit: K) and P is the pressure (SI unit: Pa). The
refractive index of air at standard temperature and pressure, nair,STP, is given by

–8  2,949,810λ′ - 25,540λ
2
′2 
n air, STP ( λ 0 ) = 1 + 10 ×  6432.8 + ----------------------------------
2
+ -------------------------- 
2
 146λ′ – 1 41λ′ – 1 

where, by definition, the standard temperature is Tstd = 15°C and the standard
pressure is Pstd = 101,325 Pa (and with a CO2 concentration of 0.03%) with the
wavelength given in microns; that is λ´ = λ/1[μm].

RELATIVE VS. ABSOLUTE REFRACTIVE INDICES

It is important to appreciate that the refractive indices given by manufacturers are often
relative to air at a certain temperature and pressure. Therefore, each of the optical
dispersion models (except Temperature-dependent Sellmeier) require a reference
temperature Tref (SI unit: K) and pressure Pref (SI unit: Pa) to be specified. This is
interpreted to be the temperature and pressure of the air in which the refractive indices
used to generate the model were measured. The reference pressure is absolute. If the
reference pressure Pref ≠ 0, then it is assumed that the dispersion model will give
relative refractive indices nrel (dimensionless). The conversion of refractive index from
relative (nrel) to absolute (n) is made using a model for the refractive index of air
(nair). That is,

166 | CHAPTER 3: RAY OPTICS INTERFACES

 n = n rel × n air

with nair defined in Equation 3-5 with T = Tref and P = Pref.

AVAILABLE OPTICAL DISPERSION MODELS

Each model requires a set of coefficients to be specified. These can be determined, for
example, by making a least squares fit to a set of refractive index measurements at
discrete wavelengths. At least one optical dispersion model (Temperature-dependent
Sellmeier) allows the refractive index to be expressed as a function of both wavelength
and temperature. All other optical dispersion models may be used together with a
thermo-optic dispersion model to calculate a temperature-dependent shift in the
refractive index. This is discussed in Thermo-Optic Dispersion Models.

The built-in optical dispersion models are listed in Table 3-3.

TABLE 3-3: LIST OF OPTICAL DISPERSION MODELS

NAME EXPRESSION

Cauchy B C
n rel = A + -----2- + -----4-
λ λ
Conrady B C
n rel = A + ---- + ---------
-
λ λ 3.5
Herzberger A1 A2 2 4
n rel = A 0 + --------------------------- + ----------------------------------
2 2
+ A3 λ + A4 λ
2
λ – 0.028 ( λ – 0.028 )
Schott
2 2 A2 A3 A4 A5
n rel = A 0 + A 1 λ + ------
2
- + ------- + ------- + -------
4 6 8
λ λ λ λ
Schott
2 2 4 A3 A4 A5 A6 A7 A8
extended n rel = A 0 + A 1 λ + A 2 λ + ------2- + ------4- + ------6- + ------8- + -------
10
- + --------
12
λ λ λ λ λ λ
Sellmeier 2 2 2
2 A1 λ A2 λ A3 λ
n rel = 1 + ------------------
2
- + ------------------
2
- + ------------------
2
-
λ – B1 λ – B2 λ – B3

Sellmeier 2 2 2 2
n rel – 1 A1 λ A2 λ A3 λ
modified, ------------------
- = ------------------
- + ------------------
- + ------------------
-
2 2 2 2
type 1 n rel + 2 λ – B1 λ – B2 λ – B3

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 167

 TABLE 3-3: LIST OF OPTICAL DISPERSION MODELS

NAME EXPRESSION

Sellmeier 2 2
2 B1 λ B2 λ
modified, n rel = A + ------------------
2
- + ------------------
2
-
type 2 λ – C1 λ – C2

Temperature 2
3 4 4
S ( T )λ
 --------------------------  Sij Tj , λi ( T )  λij Tj
- dependent i
n2 = 1 + - , Si ( T ) = =
Sellmeier 2
λ – λ (T)
2
i=1 i j=0 j=0

All optical dispersion models assume that coefficients have been determined with
expressions where the wavelengths have units of μm. For example, in the Schott
model, the coefficients have units 1, μm, μm2, μm3, and so on. The input wavelength
can still have any unit; the normalization to microns is done automatically.

The Temperature-dependent Sellmeier model (Ref. 3, Ref. 4) assumes coefficients based

on absolute temperature specified in Kelvin (K).

Thermo-Optic Dispersion Models

The Refractive index, real part, n (dimensionless) may be expressed as a function of
both wavelength and temperature by using an expression of the form

n (λ,T) = n ( λ ) + Δn (λ,T) .

The wavelength-dependent (non-temperature-dependent) component of this

equation, n(λ), might be any one of the optical dispersion models discussed above, or
it could be specified by other means. The change in refractive index as a function of
temperature, Δn(λ,Τ) can be computed using a thermo-optic dispersion model.

SCHOTT THERMO-OPTIC DISPERSION MODEL

An expression for the change in refractive index as a function of temperature, Δn(λ,T)
is given in Ref. 5. It may be derived by integrating the dispersion formula for the
thermo-optic coefficient (dn/dT). That is,

2
dn (λ,T)- n (λ,T 0) – 1  2 E 0 + 2E 1 ΔT
------------------- = -------------------------------  D 0 + 2D 1 ΔT + 3D 2 ΔT + --------------------------------,
dT 2n (λ,T 0)  2
λ –λ
2

TK

which gives

168 | CHAPTER 3: RAY OPTICS INTERFACES

 2
n (λ,T 0) – 1  2 3 E 0 ΔT + E 1 ΔT 
Δn (λ,T) = -----------------------------  D 0 ΔT + D 1 ΔT + D 2 ΔT + ---------------------------------------
- .
2n (λ,T 0)  2
λ –λ
2

TK

In the above formula T0 is the reference temperature against which the temperature
difference ΔT = T − T0 is computed and D0, D1, D2, E0, E1, and λTK, are glass-specific
coefficients.

Refractive Index at Common Spectral Lines

Predefined postprocessing variables are available for the refractive index at frequently
used spectral lines (Ref. 6):

• nd is the d-line refractive index (yellow helium spectral line, 587.56 μm)
• nF is the F-line refractive index (blue hydrogen spectral line, 486.13 μm)
• nC is the C-line refractive index (red hydrogen spectral line, 656.28 μm)

In addition, the dimensionless Abbe number Vd is automatically defined,

nd – 1
V d = --------------------
nF – nC

The Abbe number indicates the sensitivity of the glass refractive index to changes in
wavelength across the visible part of the electromagnetic spectrum. A greater Abbe
number means that a singlet lens of the material would exhibit less chromatic
aberration, while a singlet lens of a lower Abbe number would be more susceptible to
chromatic aberration.

For more instructions on postprocessing of the refractive index in

Geometrical Optics models, see Plotting Refractive Index and Abbe
Number in the Ray Optics Modeling chapter.

Optical Attenuation Models

Most real-world media absorb some energy from the electromagnetic waves that pass
through them, converting this energy into another form such as heat. A medium might
also redirect some energy without absorbing it, a phenomenon called scattering which
will be explored in a later section.

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 169

 In the Geometrical Optics interface, an absorbing medium has a complex-valued
refractive index of the form n − iκ., where n and κ are real numbers and κ > 0. As a ray
propagates through an absorbing medium, the built-in variables that track ray intensity
and power may decrease.

There is yet another phenomenon that influences the calculation of the

built-in variable for ray intensity: the convergence or divergence of the
wavefront. For simplicity this is not considered here (the rays are assumed
to represent plane waves) but will be revisited in a later section.

Recall that the electric field amplitude of a plane electromagnetic wave in a medium of
complex-valued refractive index n − iκ is

ω ω
E = E 0 exp [ iωt ] exp – i ---- xRe ( n ) exp ---- xIm ( n )
c c

Noting that the intensity is proportional to E ⋅ E* , the imaginary terms in the

exponents vanish, leaving only the attenuation term,

ω ω
I = I 0 exp 2 ---- xIm ( n ) = I 0 exp – 2 ---- κx
c c

Then, using the relationship

ω
---- = k 0 = 2π
------
c λ0

the intensity decays exponentially according to

4πκ
I = I 0 exp – ---------- x
λ0

In the physics interface Medium Properties node, you can control how the value of κ
is defined by selecting different options from the Optical attenuation model list.

EXTINCTION COEFFICIENT
For Extinction coefficient, enter the value of κ directly.

ATTENUATION COEFFICIENT
For Attenuation coefficient, enter the value or expression of the coefficient α, defined as

170 | CHAPTER 3: RAY OPTICS INTERFACES

 4πκ
α ≡ ----------
λ0

Thus

I = I 0 exp [ – αx ]

INTERNAL TRANSMITTANCE
The Optical attenuation model list also includes the following options:

• Internal transmittance, 2 mm sample thickness

• Internal transmittance, 5 mm sample thickness
• Internal transmittance, 10 mm sample thickness
• Internal transmittance, 25 mm sample thickness

The internal transmittance is the fraction of the ray intensity that is transmitted, rather
than absorbed, through a sample of the given thickness d, neglecting Fresnel losses at
the surfaces of the sample. Setting x = d in the previous equations,

I 4πκ
τ i, d ≡ ----- = exp – ---------- d
I0 λ0

Solving this equation for κ yields

λ0
κ = – ----------- log τ i, d (3-6)
4πd

This result is positive or zero because the internal transmittance cannot exceed unity.

Internal Transmittance Data and Numerical Precision

If internal transmittance data is tabulated for different glasses and different
wavelengths, and all internal transmittance values are reported to the same number of
digits, loss of precision may occur if the internal transmittance data for a very thin
sample of a weakly absorbing glass, or a thick sample of a very strongly absorbing glass,
is used.

To see this more concretely, suppose that internal transmittance data is available for
10 mm and 25 mm samples of a glass at λ0 = 600 nm, each reported to three digits.
Suppose the reported values are τi,10 = 0.998 and τi,25 = 0.995. Because of roundoff,
we may presume the actual value of τi,10 to be between 0.9975 and 0.9985, and

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 171

 similarly for τi,25 to be between 0.9945 and 0.9955. Substituting these lower and
upper bounds into Equation 3-6 then gives the following results:

• For the 10 mm sample, 7.2 × 10-9 < κ < 12.0 × 10-9

• For the 25 mm sample, 8.6 × 10-9 < κ < 10.5 × 10-9

Therefore it would be preferable to use the data for the thicker sample if it is available.

Conversely, if the reported transmittance values were very close to zero (perhaps in the
IR or UV range) it would be beneficial to use data for the thinnest sample available.

If a material uses internal transmittance data and one of the options Extinction
coefficient or Attenuation coefficient is selected, the transmittance data will
automatically be converted to this coefficient. If two or more sets of transmittance data
for different sample thicknesses are defined for the material, the default behavior is to
use the data for the thicker sample unless otherwise specified.

Initial Conditions: Position

The ray release positions may be mesh based, from a user-specified grid, uniformly
distributed, loaded from a file, or based on an analytic expression.

LOADING INITIAL COORDINATES FROM FILE

If the initial ray positions are loaded from a text file using the Release from Data File
node, the initial positions may be scaled, rotated, translated, or some combination of
these. Assuming q0 to be the loaded ray release position, q0 can be rotated using the
rotation matrix A

q 0' = Aq 0

where A in 3D is given by

cos α cos γ – sin α cos β sin γ – cos α sin γ – sin α cos β cos γ sin α sin β
A = sin α cos γ + cos α cos β sin γ – sin α sin γ + cos α cos β cos γ – cos α sin β
sin β sin γ sin β cos γ cos β

and in 2D by

A = cos α – sin α
sin α cos α

172 | CHAPTER 3: RAY OPTICS INTERFACES

In 3D, α, β, and γ are the Euler angles:

• First α rotates the position about the space-fixed z-axis.

• Then β rotates the position about the x'-axis, that is, the body-fixed X-axis after
rotation about the z-axis.
• Finally γ rotates the position about the z''-axis, that is, the body-fixed Z-axis after
rotation about the z-axis and x'-axis.

In 2D, α is the angle of counterclockwise rotation about the origin.

All components of the ray release position vector are multiplied by the Scale factor,

q 0'' = Rq 0'

Finally, the initial positions may be translated by a displacement vector Δq,

q 0''' = q 0'' + Δq

The primes indicate that translation is applied after any rotation or dilation (scaling).
However, because the dilation applies equally to all vector components, the dilation
and rotation commute with each other, so their relative order does not matter.

Initial Conditions: Direction

There are several options available for specifying the initial direction of rays. These
options work by assigning values to the degrees of freedom corresponding to the wave
vector k (SI unit: rad/m) of each ray.

EXPRESSION
The default is to enter an expression for the Ray direction vector L0 (dimensionless).
This vector need not have a magnitude of unity because it is always normalized
automatically. The initial wave vector is

ωnL 0
k = ---------------
c L0

where ω (SI unit: rad/s) is the angular frequency, c = 299,792,458 m/s is the speed of
light in a vacuum, and n (dimensionless) is the refractive index.

LOADING INITIAL DIRECTION FROM FILE

If the initial ray positions are loaded from a text file, then optionally the initial ray
direction can also be loaded from the same file. If the direction is loaded from the file

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 173

 and a rotation has been applied to the ray release positions, then optionally the same
rotation can be applied to the initial direction.

SPHERICAL
When Spherical is selected the initial wave vectors are sampled from a distribution in
wave vector space at each release point. The number of rays released from each point
is usually equal to the specified value Nw (dimensionless), although it may be larger if
the initial values of any auxiliary dependent variables are also sampled from a
distribution.

In 2D the initial wave vector components are

ωn
k x = -------- cos θ
c
ωn
k y = -------- sin θ
c

where θ goes from 0 to 2π in Nw steps. In 3D the initial wave vector components are
sampled according to the expressions

ωn
k x = -------- sin θ cos ϕ
c
ωn
k y = -------- sin θ sin ϕ
c
ωn
k z = -------- cos θ
c

The azimuthal angle ϕ is uniformly distributed from 0 to 2π. The polar angle θ is
sampled from the interval [0, π] with probability density proportional to sin θ. The
polar angle is arbitrarily chosen as the angle that the initial wave vector makes with the
positive z-axis, but any direction could be chosen because the sphere is isotropic.
Therefore each ray subtends approximately the same solid angle in wave vector space.

HEMISPHERICAL
The Hemispherical option is the same as the Spherical option, except that in 2D θ goes
from 0 to π and in 3D θ goes from 0 to π/2. The angle is measured from the direction
given by the Hemisphere axis setting.

CONICAL
The Conical option is the same as the Spherical option, except that θ goes from 0 to α.
The angle is measured from the direction given by the Cone axis setting.

174 | CHAPTER 3: RAY OPTICS INTERFACES

LAMBERTIAN
The Lambertian option releases rays within a hemisphere in 3D, but the probability
distribution function is different from that of the Hemispherical option. Recall that for
an isotropic hemispherical distribution the polar angle θ has a probability density
proportional to sin θ; for the Lambertian distribution the probability density is instead
proportional to sin θ cos θ. Because of this extra cosine term, distributions following
this probability density are said to follow Lambert’s cosine law.

Initial Conditions: Wavelength or Frequency

As indicated in the sections Optical Dispersion Models and Thermo-Optic Dispersion
Models, the refractive index of real materials is often a function of the vacuum
wavelength or frequency of the propagating rays.

In the Geometrical Optics interface, you can make the released rays monochromatic
or polychromatic. If the released rays are polychromatic, you can either specify their
frequency ν (SI unit: Hz) or their vacuum wavelength λ0 (SI unit: m). If the medium
is not a vacuum, the actual wavelength of the rays in the medium λ is automatically
computed by dividing the vacuum wavelength by the absolute refractive index.

To trace polychromatic light, in the settings for the Geometrical Optics interface,
locate the Ray Release and Propagation section. From the Wavelength distribution of
released rays list, select either Polychromatic, specify vacuum wavelength or
Polychromatic, specify frequency; the default is Monochromatic.

Then, when releasing rays into the model (for example, with a Release or Release from
Grid node), you can enter a value or expression for the wavelength or frequency
directly, or sample multiple values from a distribution, by selecting an option from the
Distribution Function list in the Vacuum Wavelength or Initial Ray Frequency section.

In ray release features, different options that control the number of

released rays or the number of values in a distribution are multiplicative,
not additive. For example, if you use the Release node and set the Number
of rays per release equal to 100, then release a Normal distribution of
frequency values and set the Number of values to 1000, the total number
of released rays will be 100,000.

NORMAL DISTRIBUTION
The wavelength or frequency is sampled from the distribution function

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 175

 1 ( λ0 – μ )2 1 ( ν – μ ) 2-
f ( λ 0 ) = --------------- exp – ----------------------
- f ( ν ) = --------------- exp – -------------------
σ 2π 2σ 2 σ 2π 2σ 2

where μ is the mean and σ is the standard deviation.

LOGNORMAL DISTRIBUTION
The wavelength or frequency is sampled from the distribution function

1 ( log λ 0 – log M ) 2
f ( λ 0 ) = --------------------------------------- exp – ------------------------------------------
λ 0 log GSD 2π 2log 2 GSD

1 ( log ν – log M ) 2
f ( ν ) = --------------------------------------- exp – ---------------------------------------
λ 0 log GSD 2π 2log 2 GSD

where M is the median and GSD is the geometric standard deviation, which are related
to the arithmetic mean μ and the standard deviation σ by the relations

μ = M exp  --- log 2 GSD

1
σ = μ exp ( log 2 GSD ) – 1
2 

UNIFORM
The vacuum wavelength or frequency has a uniform probability within an interval
bounded by the specified maximum and minimum values, and zero probability outside
this interval.

BLACKBODY

This subsection refers to the Blackbody option in the Initial Ray Frequency
or Vacuum Wavelength section of most general-purpose ray release features
such as Release and Release from Grid. These settings only affect the
distribution of the spectral variable and do not affect, for example, the
initial ray intensity or power. The dedicated Blackbody Radiation node
also controls the ray direction distribution and the total emitted power;
for more information, see Blackbody Radiation Theory.

The Blackbody option samples the vacuum wavelength and frequency of released rays
as if they were released from a blackbody radiation source at a specified temperature;
that is, the probability distribution of wavelength or frequency follows Planck’s law.

176 | CHAPTER 3: RAY OPTICS INTERFACES

To be able to speak of both wavelength and frequency at the same time, first define the
dimensionless variable x (not to be confused with the spatial x-coordinate)

hν hc
x = ----------- = -----------------
kB T kB λ0 T

• h = 6.62607015 × 10-34 J·s is Planck’s constant,

• c = 299,792,458 m/s is the speed of light in a vacuum,
• kB = 1.380649 × 10-23 J/K is the Boltzmann constant,
• T (SI unit: K) is the surface temperature,
• ν (SI unit: Hz) is the ray frequency, and
• λ0 (SI unit: m) is the vacuum wavelength.

Then the dimensionless form of Planck’s law states that the normalized probability
distribution of this variable is

15 x 3
f ( x ) = -----4- -------------
-
π ex – 1

It might be useful to exclude outliers corresponding to extremely large or small values

of the wavelength or frequency when sampling from this distribution function. You can
choose to specify maximum and minimum values of the wavelength or frequency,
which are first converted to upper and lower limits on x; or you may choose to Specify
fraction of total emissive power and then specify a fraction slightly less than 1, so that
the lower and upper tails of the distribution are automatically excluded.

Blackbody Radiation Theory

The Blackbody Radiation node releases rays diffusely from a surface (following
Lambert’s cosine law) with a wavelength or frequency distribution following Planck’s
law (if release of polychromatic rays has been enabled) and a total source power based
on the Stefan-Boltzmann law (if ray power is solved for).

IDEAL BLACKBODY RADIATION SOURCE

An ideal blackbody (Ref. 7) is an object that absorbs all radiation that is incident upon
it, regardless of the body’s temperature or the wavelength of the radiation.

Consider a blackbody and some other object (not a blackbody) of the same size and
shape, at the same temperature. Because, by definition, the blackbody absorbs more
radiation than the other object, it must also emit more radiation than the other object

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 177

 in order to maintain thermal equilibrium. Therefore an ideal blackbody is both the
most efficient absorber and emitter of radiation.

SPECTRAL EMISSIVITY AND GRAYBODY RADIATION

Real-world objects usually do not behave as ideal blackbody radiation sources. The
spectral emissivity ε(λ,T) of an object (dimensionless) is the radiant energy flux per unit
surface area the object releases, divided by the radiant energy flux per unit surface area
that an ideal blackbody at the same temperature would release. By definition, ε = 1 for
an ideal blackbody.

A graybody is a radiation-emitting body whose emissivity is a constant, independent of

temperature and wavelength, but not necessarily equal to 1. Blackbody radiation is
thus a special case of graybody radiation, which itself is a special case out of all possible
spectral emissivity functions.

RADIOMETRIC VS ACTINOMETRIC QUANTITIES

The amount of radiation released by an object may be quantified in different ways. In
the COMSOL implementation, radiometric quantities are used, meaning that the
radiation is described by the amount of energy it transmits. Typical units include watts,
watts per square meter, watts per square meter per steradian, watts per square meter
per unit wavelength, and so on.

An alternative way to describe the amount of radiation is by using actinometric

quantities, which describe the number of photons released rather than the amount of
energy. From quantum mechanics, the energy of one photon, the quantum of the
electromagnetic field, is defined as E = hν, where ν is the frequency (SI unit: Hz) and
h = 6.62607015 × 10-34 J·s is Planck’s constant. Currently the Geometrical Optics
interface does not automatically define any variables in actinometric units.

EXITANCE AND SPECTRAL EXITANCE

The total radiant exitance M eb (SI unit: W/m2) is the amount of electromagnetic
energy emitted from the surface of the blackbody radiation source per unit time, per
unit surface area.

The energy released from a blackbody source follows a wavelength distribution. For
some value of the wavelength λ, imagine a sensor that detects 100% of the radiation
having wavelengths between λ and λ + dλ, where dλ is an infinitesimally small change
in the wavelength, but detects 0% of radiation outside this range. Then the amount of
energy detected by this sensor is M eb, λ dλ , where M eb, λ is the spectral radiant exitance.

178 | CHAPTER 3: RAY OPTICS INTERFACES

The units of spectral radiant exitance are energy per unit time per unit surface area, per
unit wavelength.

The most concise way to write the unit is W/m3 but sometimes it is more conveniently
expressed with the area and wavelength units written separately, for example, W/
(cm2·μm). The spectral radiant exitance is usually a function of wavelength and
temperature, so it may be written as M eb, λ ( λ, T ) .

For any two wavelengths λ1 and λ2 (assume λ1 < λ2), the total energy of emitted
radiation between these wavelengths is

λ2
λ Meb, λ ( λ, T )dλ
1

The total energy released over all wavelengths is, by definition, the total radiant
exitance,

∞
M eb = 0 Meb, λ ( λ, T )dλ
The spectral radiant exitance may also be written in terms of a spectral quantity other
than wavelength, so long as that other quantity has a one-to-one relationship with
wavelength. Other quantities that have been used include the frequency, the logarithm
of frequency or wavelength, and other powers of frequency or wavelength. In fact
there are infinitely many choices for the arbitrarily high powers of λ and ν, of which
the first power of wavelength is historically the most common choice (Ref. 7).

The spectral radiance in terms of frequency is written M eb, ν ( ν, T ) (SI unit: W·s/m2),

∞
M eb = 0 Meb, ν ( ν, T )dν
The total energy between two wavelengths λ1 and λ2 should be the same when
converting these wavelengths to frequencies ν1 = c/λ1 and ν2 = c/λ2,

λ2 ν1
λ Meb, λ ( λ, T )dλ = ν Meb, ν ( ν, T )dν
1 2
(3-7)

The order of the limits of integration is reversed because if λ1 < λ2, then ν1 > ν2.

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 179

 SPECTRAL RADIANCE
The radiance of a blackbody is the amount of energy per unit time that is released from
the surface of the body and passes through a certain solid angle in the half-space
adjacent to the body, per unit time, per unit solid angle, per unit projected surface area
of the radiation source,

1 d ( M e, λ ( λ, T ) )
b
L eb, λ ( λ, T ) = ------------ ------------------------------------- (3-8)
cos θ dΩ

The term cos θ appears in the denominator because the radiance is measured per unit
projected surface area from the point of view of an observer at polar angle θ, rather
than the actual surface area from which the radiation is emitted.

The spectral radiance in terms of wavelength L eb, λ ( λ, T ) has SI units of W/(m3·sr),

or it may be written with the area units and wavelength units separated, for example
W/(cm2·nm·sr). Similarly the spectral radiance in terms of frequency L eb, ν ( ν, T ) has
SI units of (W·s)/(m2·sr) but can also be written using different area units.

In general, the spectral radiance may be a function of the polar angle θ. However, the
ideal blackbody radiation source is assumed to be a Lambertian source, that is, a diffuse
emitter of radiation, meaning the direction distribution of released radiation follows
the cosine law. This cosine term exactly cancels out the cosine term appearing in the
denominator of Equation 3-8. Therefore the spectral emittance of an ideal blackbody
radiation source is independent of the viewing angle.

In terms of the spectral emittance, the spectral exitance is

M eb, λ ( λ, T ) =  Leb, λ ( λ, T ) cos θ dΩ

Ω

where the integration is over the half-space (hemisphere) into which the surface area
element emits radiation. Since the differential solid angle element in spherical polar
coordinates is

dΩ = sin θd θdφ

the relationship between spectral emittance and spectral exitance becomes

2π π ⁄ 2
M eb, λ ( λ, T ) = 0 0 L eb, λ ( λ, T ) sin θ cos θ dθ dφ = πL eb, λ ( λ, T )

Or alternatively,

180 | CHAPTER 3: RAY OPTICS INTERFACES

 1
L eb, λ ( λ, T ) = --- M eb, λ ( λ, T )
π

Although any differential area on the surface of the blackbody releases radiation into
the hemispherical half-space above it (subtending a solid angle of 2π steradians), the
conversion between exitance and radiance only requires a multiplication by π rather
than 2π. This has been shown in the above derivation to follow from the treatment of
an ideal blackbody as a Lambertian source.

PLANCK’S LAW OF BLACKBODY RADIATION

The spectral exitance of an ideal blackbody radiation source follows Planck’s law. In
terms of the wavelength,

2πhc 2
exp  --------------- – 1
hc –1
M eb, λ ( λ, T ) = ---------------- (3-9)
λ5  k B λT

Or in terms of the frequency,

2πhν 3-
exp  ----------- – 1
hν –1
M eb, ν ( ν, T ) = ---------------- (3-10)
c2 kB T

Similarly, the spectral radiance may be expressed as a function of wavelength or

frequency,

2hc 2
- exp  --------------
hc - –1
L eb, λ ( λ, T ) = ------------ –1
λ5  k B λT

2hν 3
- exp  ----------
hν- –1
L eb, ν ( ν, T ) = ------------ –1
c 2  k BT


Dismissing an Apparent Paradox

One might be inclined to substitute λ = c/ν into Equation 3-9 or substitute ν = c/λ
into Equation 3-10. Doing so would give seemingly contradictory expressions for the
spectral exitance, but doing so would be a mistake because the exitance is not energy
flux itself but rather a distribution function of the energy flux (Ref. 7). A better
comparison is to integrate Equation 3-9 and Equation 3-10 over matching wavelength
and frequency ranges, respectively, noting from Equation 3-7 that these integrals
should agree.

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 181

 λ2 λ2
2πhc 2
exp  --------------- – 1
hc –1
λ Meb, λ ( λ, T ) dλ = λ ----------------
1
λ5 1
 k B λT
dλ

ν1 ν1
2πhν 3
exp  ----------- – 1
hν –1
ν 2
M eb, ν ( ν, T ) dν = ν ----------------
2
c2
-
kB T
dν

Then in the first integral, the following substitutions are made:

c c c c
λ = --- dλ = – -----2- dν λ 1 = ----- λ 2 = -----
ν ν ν1 ν2

The integral then becomes

λ2 ν2
2πhν 5
exp  ----------- – 1  – -----
hν –1 c- 
λ 1
M eb, λ ( λ, T ) dλ = ν ----------------
1
c3
-
kB T  ν 2 dν

Using the negative sign to reverse the limits of integration and then simplifying yields

λ2 ν1
2πhν 3
exp  ----------- – 1
hν –1
λ Meb, λ ( λ, T ) dλ = ν ----------------
1
c2
-
2
kB T
dν

The integrand on the right-hand side matches the original definition of M eb, ν ( ν, T ) ,
so there is no contradiction. The key takeaway is the following:

The spectral exitance and spectral radiance are distribution functions of a

spectral variable. Any substitution of a different spectral variable must be
accompanied by the appropriate Jacobian of transformation.

This fundamental truth about the nature of spectral exitance is the reason why
phenomena such as Wien peaks appear to change location when the exitance is
expressed in terms of a different independent variable.

SAMPLING FROM THE PLANCK DISTRIBUTION

To simulate the emission of radiation from a blackbody radiation source, the
wavelength or frequency of a representative population of rays must be sampled from
the Planck function. To describe how values might be sampled from this distribution
function, it is convenient to introduce the dimensionless variable x (not to be confused
with the spatial x-coordinate)

hν hc
x = ----------- = ---------------
kB T k B λT

182 | CHAPTER 3: RAY OPTICS INTERFACES

The Jacobian of transformation is

h hc
dx = ----------- dν = – ----------------- dλ
kB T kB λ2 T

In terms of this dimensionless variable, the spectral exitance is

2πk B 4 T4
x3 -
M eb, x ( x, T ) = ---------------------
3 2
-------------
x
h c e –1

And the total exitance is thus

2πk B 4 T4 ∞
x3
M eb ( T ) = ---------------------
h c23 0 -------------
ex – 1
- dx

It can be shown by taking a series expansion of the integrand and then integrating
every term by parts (Ref. 7) that the value of the integral is

∞
x3 π 4-
0 -------------
ex – 1
- dx = -----
15

although there is no closed-form expression for this integral if the upper limit of
integration is finite; in that case, numerical integration is required. The expression for
the total exitance is then

2π 5 k B 4 T4
M eb ( T ) = ------------------------ = σT 4
15h 3 c 2

where σ is the Stefan-Boltzmann constant,

2π 5 k B 4
W -
σ = ------------------
- ≈ 5.67 × 10 – 8 ---------------
15h c 2 3 m2 K4

Thus an integration of Planck’s law confirms the Stefan-Boltzmann law.

It is convenient to normalize the spectral exitance so that it represents a probability

distribution function of x,

15 x 3 -
f ( x ) = -----4- -------------
π ex – 1

In the COMSOL implementation, when polychromatic radiation is permitted in a

model (by selecting either Polychromatic, specify vacuum wavelength or Polychromatic,

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 183

 specify frequency from the physics interface Wavelength distribution of released rays list),
values of x are first sampled from the above probability distribution function and then
converted to the vacuum wavelength or frequency as needed.

Gaussian Beam Theory

The Gaussian Beam node releases rays with a Gaussian distribution of intensity or
power. It can only be used when solving for either intensity or power, or both.

The Geometrical Optics interface does not consider diffraction effects.

Therefore, using the Gaussian Beam node involves two key assumptions:

• The distances the rays propagate in the model should either be much
larger than or much smaller than the Rayleigh range of the beam, which
will be defined in a later section.
- For ray propagation over relatively short distances, choose Collimated
source (large Rayleigh range) from the Beam type list.
- For ray propagation over relatively large distances, choose Point
source (small Rayleigh range) from the Beam type list.
• Paraxial approximation: the beam divergence angle should be small.

EFFECT OF SPACE DIMENSION

The Gaussian Beam node is available in 3D, 2D axisymmetric, and 2D geometries. The
interpretation of the beam is different in 2D than in 3D or 2D axisymmetric models.

• In 3D, the beam irradiance distribution is assumed to be rotationally symmetric

about the beam axis, the average direction of the released rays. The beam axis may
be in any arbitrary direction.
• In 2D axisymmetric models, the irradiance distribution is still assumed to be
rotationally symmetric about the beam axis, but the beam axis is restricted to either
the positive or negative z direction. If any other direction were chosen as the beam
axis, the beam would not actually be Gaussian in the corresponding 3D solid of
revolution.
• In 2D models, the irradiance distribution exhibits plane symmetry across the beam
axis but is uniform in the out-of-plane direction. Thus in the vicinity of the beam
waist, the beam appears more as a sheet rather than a cylinder.

184 | CHAPTER 3: RAY OPTICS INTERFACES

GAUSSIAN BEAM EQUATION IN 3D
Without loss of generality we assume the beam is propagating in the positive
z direction with a waist centered at the origin, although in a 3D geometry any other
waist position and beam axis may be specified.

The electric field amplitude of a paraxial Gaussian beam is most conveniently expressed
in cylindrical polar coordinates,

w0 ρ2 ρ2
E ( ρ, z ) = E 0 ------------ exp – --------------2- – jkz – jk ---------------- + jη ( z ) (3-11)
w( z) w( z) 2R ( z )

where

• ρ (SI unit: m) is the radial coordinate (with the beam centered around ρ = 0),
• z (SI unit: m) is the axial coordinate (chosen so that z = 0 is the beam waist),
• E(ρ,z) (SI unit: V/m) is the electric field at an arbitrary location (ρ,z),
• E0 = E(0,0) is the electric field at the center of the beam waist, and
• k (SI unit: rad/m) is the wave number in the medium through which the beam
propagates, k = 2π/λ = 2πn/λ0 where n (dimensionless) is the absolute refractive
index of the medium and λ0 (SI unit: m) is the vacuum wavelength.

The beam parameter w0 is called the waist radius. Of course, there is no well-defined
cutoff distance at which the electric field goes to zero, so there is a certain degree of
arbitrariness involved in assigning a value to the beam radius. As will be shown in the
following sections, the beam waist radius is defined such that the fraction of the total
beam power that passes through a circle or radius w0 centered at the origin is 1 - e-2.

The remaining functions in Equation 3-11 are defined as follows:

The function w(z) is the beam radius at any axial coordinate z,

w = w 0 1 +  ------
z 2
 z R

The function R(z) describes how quickly the beam phase varies with radial position in
a plane located at any axial coordinate z,

zR 2
R ( z ) = z 1 +  ------
z

The term η(z) is an additional phase delay known as the Guoy phase,

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 185

 η ( z ) = atan  ------
z
zR

In these definitions, zR (SI unit: m) is the Rayleigh range,

πw 02
z R = -----------
λ

The Rayleigh range is the axial distance from the beam waist at which the radius has
increased by a factor of 2 , or equivalently the distance from the beam waist at which
the cross sectional area of the beam has doubled.

The exponent in Equation 3-11 can be separated into amplitude and phase terms,

w0 ρ2
E ( ρ, z ) = E 0 ------------ exp – --------------2- exp [ – jδ ]
w(z) w( z)

where δ (SI unit: rad) is the phase factor,

ρ2
δ = kz + k ---------------- – η ( z ) (3-12)
2R ( z )

GAUSSIAN BEAM EQUATION FOR IRRADIANCE

It is also possible to write Equation 3-11 in terms of the irradiance rather than electric
field amplitude. As the irradiance I (SI unit: W/m2) is proportional to E ⋅ E* , the
phase factor vanishes,

w0 2 ρ2
I ( ρ, z ) = I 0  ------------ exp  – 2 --------------2- (3-13)
w(z) w(z)

ASYMPTOTIC LIMITS OF THE PARAXIAL GAUSSIAN BEAM

In the asymptotic limit as the z-coordinate is much greater than the Rayleigh range,
the beam radius approaches a linear function of z,

w ( z ) = w 0 1 +  ------ ≈ w 0 ------
z 2 z
z » zR
zR zR

so a beam with radius w(z) approaches a biconical shape (with both cone vertices
meeting at the origin) in the limits z « – z R and z » z R . These asymptotes each make
an angle of α (SI unit: rad) with the beam axis, where

186 | CHAPTER 3: RAY OPTICS INTERFACES

 λ
α = -----------
πw 0

α is called the beam divergence half-angle, to avoid confusion with references such as
Ref. 8 that may define a full angle equal to 2α.

Making the substitution w ( z ) ≈ w 0 z ⁄ z R into Equation 3-13 then yields

zR 2 zR 2 ρ 2
I ( ρ, z ) = I 0  ------ exp – 2  -------  ---
 z  w 0  z 

At this point it is convenient to convert to spherical polar coordinates,

ρ = r sin θ z = r cos θ

where r (SI unit: m) is the distance from the origin and θ (SI unit: rad) is the polar
angle measured from the positive z-axis. Thus

zR 2 z R 2 r sin θ 2
I ( r, θ ) = I 0  --------------- exp – 2  -------  ---------------
 r cos θ  w 0  r cos θ

Note also from the definitions of the Rayleigh range and the divergence half-angle that

w0
α = -------
zR

Thus

zR 2 tan θ 2
I ( r, θ ) = I 0  --------------- exp – 2  -------------
 r cos θ  α 

Finally, recall that the Gaussian beam is paraxial (small θ). Taylor series expansions of
the cosine and tangent functions reveal that

tan θ = θ + O ( θ 3 )
cos θ = 1 + O ( θ 2 )

So that for small θ,

zR 2 θ 2
I ( r, θ ) = I 0  ------ exp – 2  --- (3-14)
 r  α

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 187

 Equation 3-14 shows that the far-field limit of a Gaussian beam ( z » z R ) is a cone of
half-angle α in which the ray intensity or power is proportional to a Gaussian function
of the angle between each ray and the beam axis.

ASYMPTOTIC LIMIT OF PHASE

The Geometrical Optics interface includes the option to treat rays as fully coherent
wavefronts in which the instantaneous phase is solved for, by selecting the Compute
phase check box. When the rays are treated as a collimated source, their phase is
initialized as if all rays are released at locations along the beam waist. Substituting z = 0
into Equation 3-12 for the phase factor simply yields δ = 0, so the initial phase of all
released rays is equal.

Now consider once again the asymptotic limit of z » z R .

w ( z ) = w 0 1 +  ------ ≈ w 0 ------
z 2 z
 z R zR
zR 2
R ( z ) = z 1 +  ------ ≈ z
z
π
lim atan  ------ = ---
z
z→∞ zR 2

So for large z, Equation 3-12 becomes

ρ2 π
δ = kz + k ------ – ---
2z 2

Again converting to spherical polar coordinates,

1 π
δ = kr cos θ + --- kr sin θ tan θ – ---
2 2

And in fact a Taylor series expansion shows that

1 1 1
cos θ + --- sin θ tan θ = 1 + --- θ 4 + ------ θ 6 + O ( θ 8 )
2 8 24

So the phase of the released rays can be considered uniform, so long as

kLθ 4 « 1

where L is the maximum distance the rays propagate in the model.

188 | CHAPTER 3: RAY OPTICS INTERFACES

TOTAL BEAM POWER AND ENCIRCLED POWER
The total power transmitted by the Gaussian beam Psrc (SI unit: m) is found by
integrating the irradiance from Equation 3-13 over a plane perpendicular to the z-axis,

2π ∞ w0 2
ρ2 π
P src = 0 0 I0  -----------
w ( z )
- exp  – 2 --------------2- ρ dρ dφ = --- w 02 I 0
w( z) 2

As expected, the total power crossing a plane perpendicular to the z-axis does not
depend on the value of z, otherwise energy would be accumulating at some location
in the beam.

Compare the total source power Psrc with the encircled power Pcir (SI unit: W/m2)
that passes through a circle of radius w(z) centered at the z-axis,

2π w ( z ) w0 2 ρ2 π
P cir = 0 0 I 0  ------------ exp  – 2 --------------2- ρ dρ dφ = --- w 02 I 0 ( 1 – e – 2 )
 w ( z )  w(z)  2

Thus the radio of the encircled power to the total source power is

P cir
----------
- = 1 – e –2 ≈ 0.865
P src

In addition, by substituting ρ = 0 and ρ = w(z) into Equation 3-13, it follows that

( w ( z ), z )-
I-------------------------
= e –2
I ( 0, z )

Thus the beam radius w(z) is sometimes called the 1 ⁄ e 2 radius. If the above integral
is repeated, replacing the upper limit integration with cw(z) where c is a positive
constant (called the Beam cutoff ratio in the settings window), the result is

2π cw ( z ) w0 2 ρ2 π
P cir = 0 0 I 0  ------------ exp  – 2 --------------2- ρ dρ dφ = --- w 02 I 0 [ 1 – exp ( – 2c 2 ) ]
 w ( z )  w( z)  2

SUMMARY OF BEAM PARAMETERS (3D)

When the wavelength of the beam λ and any one of the three beam parameters (waist
radius w0, divergence half-angle α, and Rayleigh range zR) are defined, the other two
beam parameters can be derived:

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 189

 λ zR λ
w 0 = ------- = ----------
πα π
λ λ-
α = ----------- = ---------
πw 0 πz R
w0 λ πw 02
z R = ------- = ---------2- = -----------
α πα λ

For a beam of given waist radius w0 and any of the three intensity-related parameters
(intensity at the beam waist I0, total source power Psrc or encircled power Pcir), the
remaining two parameters can be derived:

2P src 2P cir
I 0 = -------------- = -------------
- [ 1 – exp ( 2c 2 ) ] –1
πw 02 πw 02
π
P src = --- w 02 I 0 = P cir [ 1 – exp ( 2c 2 ) ] – 1
2
π
P cir = P src [ 1 – exp ( 2c 2 ) ] = --- w 02 I 0 [ 1 – exp ( 2c 2 ) ]
2

GAUSSIAN BEAM EQUATION IN 2D

The Gaussian beam in 2D has a different functional form because instead of having
rotational symmetry about the beam axis, it is uniform in the out-of-plane direction,

w0 y2 y2
E ( x, y ) = E 0 ------------ exp – ---------------2 – jkx – jk ---------------- + jη ( x )
w(x) w( x ) 2R ( x )
2
w ( x ) = w 0 1 +  ------
x 2
R ( x ) = x 1 +  z-----R- η ( x ) = --- atan  ------
1 x
 z R  x 2 zR

Comparing the 2D form with the 3D form from Equation 3-11, the key differences
are a square root on the term w0/w(x) and a factor of 1/2 in the Guoy phase. In this
expression x has been considered the propagation direction but the beam may actually
propagate in any direction in the xy-plane.

The corresponding expression for the beam irradiance is

w0 y2
I ( x, y ) = I 0 ------------ exp – 2 ---------------2
w( x) w( x)

The total beam power per unit distance in the out-of-plane direction is

190 | CHAPTER 3: RAY OPTICS INTERFACES

 P src ∞ w0 y2 π
----------- =
Δz –∞ I0 ------------
w(x)
exp – 2 --------------2- dy = I 0 w 0 ---
w(x) 2

The 2D analogue to the encircled power from the previous section, the beam power
per unit distance that passes between coordinates −cw(x) and cw(x) is

P cir cw ( x ) w0 y2
---------
Δz
- =
–cw ( x ) I0 ------------
w(x )
exp – 2 --------------2- dy
w( x)

Because the irradiance is an even function of y, this may be simplified to

P cir cw ( x ) w0 y2 π
---------
Δz
- = 2I 0
0 ------------
w(x)
exp – 2 --------------2- dy = I 0 w 0 --- erf ( 2c )
w( x) 2

Thus the ratio of the encircled power to the total source power is

P cir
----------
- = erf ( 2c )
P src

where the error function erf is a built-in function in COMSOL,

x
2
erf ( x ) ≡ -------
π 0 exp ( –x'2 ) dx'
For c = 1 the encircled power is about 95% of the total beam power.

Material Discontinuity Theory

At a Material Discontinuity between two different media, the wave vector is
reinitialized using Snell’s law. First, the angle of incidence θi is computed:

ni ⋅ ns
θ i = acos  ----------------
 ni ns 

where ni is a unit vector in the direction of the incident ray and ns is a unit vector
normal to the material discontinuity.

At a boundary between two isotropic, nonabsorbing media, the refracted ray

propagates in the direction nt given by the following relations:

n t = ηn i + γn s

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 191

 γ = – η cos θ i + cos θ t

n1
η = ------
n2

θ t = asin ( η sin θ i )

where the ray propagates from the medium with refractive index n1 into the medium
with refractive index n2. Because the two media are nonabsorbing, the quantities n1,
n2, θi, and θt are real-valued. If the media are weakly absorbing, the real parts of n1
and n2, are used. If the media are strongly absorbing, several corrections to Snell’s law
and the Fresnel equations must be made, as discussed in Refraction in Strongly
Absorbing Media.

If the maximum number of secondary rays has not yet been released, a reflected ray is
also released at the material discontinuity with the initial direction

n r = n i – 2n s cos θ i

If the ray intensity is computed, the release of secondary rays may be suppressed when
the intensity of the reflected ray is less than the user-defined Threshold intensity Ith
(SI unit: W/m2). Similarly, when ray power is computed, a Threshold power Qth
(SI unit: W) can be specified.

The Fresnel Equations

At a boundary between two media, the tangential components of the electric field E
are conserved. If there are no surface currents, the tangential components of the
magnetic field H are also conserved.

If the incident ray is s-polarized, meaning the electric field is perpendicular to the plane
of incidence, then from the conservation of the tangential components of E and H,

E st = E si + E sr (3-15)

n 2 E st cos θ t = n 1 ( E si – E sr ) cos θ i (3-16)

Substitution of Equation 3-16 into Equation 3-15 yields the ratios

E sr n 1 cos θ i – n 2 cos θ t E st 2n 1 cos θ i

--------- = -----------------------------------------------
- -------- = -----------------------------------------------
-
E si n 1 cos θ i + n 2 cos θ t E si n 1 cos θ i + n 2 cos θ t

192 | CHAPTER 3: RAY OPTICS INTERFACES

Similarly, if the incident ray is p-polarized, meaning that the electric field lies in the
plane of incidence and the magnetic field lies perpendicular to it, then from the
conservation of the tangential components of E and H,

E pt cos θ t = ( E pi – E pr ) cos θ i (3-17)

n 2 E pt = n 1 ( E pi – E pr ) (3-18)

Substitution of Equation 3-18 into Equation 3-17 yields the ratios

E pr n 2 cos θ i – n 1 cos θ t E pt 2n 1 cos θ i

--------- = -----------------------------------------------
- --------- = -----------------------------------------------
-
E pi n 2 cos θ i + n 1 cos θ t E pi n 2 cos θ i + n 1 cos θ t

Finally, we define the reflection coefficient as the ratio of the reflected electric field
amplitude to the incident electric field amplitude, and similarly define the transmission
coefficient as the ratio of the refracted electric field amplitude to the incident electric
field amplitude.

n 2 cos θ i – n 1 cos θ t 2n 1 cos θ i

r p = ------------------------------------------------ t p = ------------------------------------------------
n 2 cos θ i + n 1 cos θ t n 2 cos θ i + n 1 cos θ t
n 1 cos θ i – n 2 cos θ t 2n 1 cos θ i
r s = ------------------------------------------------ t s = ------------------------------------------------
n 1 cos θ i + n 2 cos θ t n 1 cos θ i + n 2 cos θ t

The above definition of rp is consistent with Ref. 9 but shows the opposite sign
compared to Ref. 10. This is because Ref. 9 defines the direction of the in-plane field
component such that the s-polarized component, p-polarized component, and ray
direction vector (in that order) always comprise a right-handed coordinate system;
whereas in Ref. 10 the vectors written in this order would comprise a left-handed
coordinate system for the reflected ray.

Intensity, Wavefront Curvature, and Polarization

The following describes the algorithm used to compute the ray intensity and
polarization when Intensity computation is set to Compute intensity or Compute intensity
and power in the physics interface Intensity Computation section. In this algorithm, each
ray is treated as a propagating wavefront subtending a small solid angle. The algorithm
is only valid when the rays propagate in homogeneous media. The following auxiliary
dependent variables are defined for each ray:

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 193

 In 3D:

• The initial ray intensity I0 and normalized Stokes parameters sn1, sn2, and sn3. By
allocating four degrees of freedom in this way it is possible to characterize any
intensity magnitude and polarization state.
• The integral A of the attenuation coefficient along the ray path.
• Two initial principal radii of curvature, r1,i and r2,i.
• Two principal radii of curvature, r1 and r2, which represent the maximum and
minimum radii of curvature of the intersection of the wavefront with an arbitrary
plane.
• Three components of a unit vector e1 in the direction corresponding to the first
principal radius of curvature. This information is used to reinitialize the principal
radii of curvature at curved boundaries.

In 2D:

• The initial ray intensity I0 and normalized Stokes parameters sn1, sn2, and sn3.
• The integral A of the attenuation coefficient along the ray path.
• One initial principal radius of curvature, r1,i.
• One radius of curvature r1 of the wavefront. All wavefronts are assumed to be
converging or diverging cylindrical waves, so it is not necessary to define a second
radius of curvature.

A different algorithm is used to compute the intensity and polarization of each ray if
the Intensity computation is instead set to Compute intensity in graded media or Compute
intensity and power in graded media. This algorithm accounts for the effect of graded
media on the ray intensity, but it is also slower and less accurate for homogeneous
media. The following auxiliary dependent variables are defined for each ray:

In 3D:

• The intensity help variable Γ and the normalized Stokes parameters sn01, sn02, and
sn03, which characterize the intensity and polarization of the ray. They are
reinitialized at material discontinuities and walls.
• The integral A of the attenuation coefficient along the ray path.
• Two principal curvature calculation help variables α1 and α2 and the rotation angle
φ which indicates the orientation of the principal curvature directions. An additional
help variable is used internally to detect poles in the local coordinate system
definition and to redefine the local coordinate system accordingly.

194 | CHAPTER 3: RAY OPTICS INTERFACES

In 2D:

• The intensity help variable Γ and the normalized Stokes parameters sn01, sn02, and
sn03, which characterize the intensity and polarization of the ray. They are
reinitialized at material discontinuities and walls.
• The integral A of the attenuation coefficient along the ray path.
• One principal curvature calculation help variable α1.

For the options Compute Power, Compute intensity and power, and Compute intensity
and power in graded media an additional auxiliary dependent variable is defined to
indicate the total power transmitted by each ray. Like intensity, power is affected by
absorbing media and by reflection or refraction at boundaries. Unlike intensity, power
does not change due to the focusing or divergence of a ray bundle.

THE STOKES PARAMETERS

The four Stokes parameters completely characterize the intensity and polarization of a
fully polarized, partially polarized, or unpolarized ray. The Stokes parameters can be
interpreted as indicators of the ray intensity that would be measured when sending a
ray through various arrangements of polarizers and wave retarders.

The Stokes parameters of a ray are defined as in Ref. 9 as follows. Consider a ray
propagating along the z-axis of a Cartesian coordinate system, with electric field
components polarized within the xy-plane. In 3D models the x-axis is parallel to the
first principal curvature direction e1. In 2D models the x-axis is always oriented in the
out-of-plane direction.

Let I(θ,ε) be the intensity of radiation polarized in the xy-plane at an angle θ

counterclockwise from the x-axis, when a phase angle of ε is introduced between the
x- and y-components. For example, I(0,0) is the intensity that would be measured after
sending a ray through a linear polarizer with a transmission axis parallel to the x-axis,
and I(π/4, π/2) is the intensity that would be measured after sending a ray through a
system of polarizers and wave retarders that only transmits radiation with right-handed
circular polarization. The four Stokes parameters are then defined as follows:

• s 0 = I ( 0, 0 ) + I ( π ⁄ 2, 0 ) is the sum of the intensity of linearly polarized light in the

x and y directions. It represents the total intensity of the ray.
• s 1 = I ( 0, 0 ) – I ( π ⁄ 2, 0 ) is the difference in intensity between light that is linearly
polarized in the x direction and light that is linearly polarized in the y direction.

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 195

 • s 2 = I ( π ⁄ 4, 0 ) – I ( 3π ⁄ 4, 0 ) is the difference in intensity between linearly
polarized light in the direction of the line y = x and linearly polarized light in the
direction of the line y = −x.
• s 3 = I ( π ⁄ 4, π ⁄ 2 ) – I ( 3π ⁄ 4, π ⁄ 2 ) is the difference in intensity between light with
right-handed circular polarization and light with left-handed circular polarization.

For example, a ray with linear polarization parallel to the x-axis has Stokes parameters
s0 = s1 = I and s2 = s3 = 0. A fully polarized ray is characterized by the relation

s 02 = s 12 + s 22 + s 32

A partially polarized or unpolarized ray is characterized by the relation

s 02 ≥ s 12 + s 22 + s 32

The degree of polarization P is defined as

s 12 + s 22 + s 32
P = ----------------------------------
s0

A degree of polarization P = 1 corresponds to fully polarized radiation, whereas a

degree of polarization P = 0 corresponds to unpolarized radiation.

In the COMSOL implementation of ray intensity calculation, the degrees of freedom

are not the Stokes parameters themselves, but rather normalized Stokes parameters
sn1, sn2, and sn3. Then the actual Stokes parameters are recovered through the
relations s1 = Isn1, s2 = Isn2, and s3 = Isn3. The reason for this decoupling is that the
normalized Stokes parameters remain bounded, enabling accurate reinitialization of
ray power, even if a ray is reflected or refracted very close to a caustic, where the
intensity (in the geometrical optics limit) can become arbitrarily large.

PRINCIPAL RADII OF CURVATURE

In 3D models, each ray is treated as a wavefront for which two principal radii of
curvature, r1 and r2, are defined. In addition, the values of the two principal radii of
curvature are stored as r1,i and r2,i whenever the ray reaches a boundary.

Within domains, the equations

dr 1 dr
--------- = – 1 --------2- = – 1
ds ds

196 | CHAPTER 3: RAY OPTICS INTERFACES

are solved. Negative radii of curvature indicate that the wavefront is expanding as the
ray propagates, while positive radii of curvature indicate that the wavefront is
converging. A continuous locus of points at which either of the principal radii of
curvature equals zero is called a caustic surface. The unit vector e1 is defined so that
r1 is the radius of curvature of the intersection of the wavefront with the plane tangent
to e1 and the wave vector k. Similarly, if a vector e2 is defined so that

k × e1
e 2 = ------------------
k × e1

Then r2 is the radius of curvature of the intersection of the wavefront with a plane
tangent to e2 and k. The components of e1 are stored as auxiliary dependent variables
for each ray. The components of e2 can then be derived from e1 and k at any time.

The principal radii of curvature are reinitialized at material discontinuities, and the
orientation of the unit vectors e1 and e2 may change. If the unit vector in the direction
of the incident ray ni is not parallel to the surface normal ns, then it is possible to
define a unique tangent plane, called the plane of incidence, that contains the incident,
reflected, and refracted rays. The unit vector normal to this plane, u0, is defined as

ni × ns
u 0 = -------------------
ni × ns

If the ray is normal to the surface, then the incident ray is parallel to the refracted ray
and antiparallel to the reflected ray, and u0 can be any arbitrary unit vector orthogonal
to ni. In addition to the unit vector normal to the plane of incidence, the following
unit vectors tangent to the incident wavefront, refracted wavefront, reflected
wavefront, and surface normal are defined:

ui = ni × u0
ut = nt × u0
ur = nr × u0
us = ns × u0

where the subscripts i, t, and r denote the incident, refracted, and reflected wavefronts,
respectively. For a wavefront propagating in a direction n, with principal curvatures
k1 = 1/r1 and k2 = 1/r2 defined for directions e1 and e2, the principal curvatures in two
other orthogonal directions e1' and e2' (both orthogonal to n) are

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 197

 k 1' = k 1 cos2 θ + k 2 sin2 θ (3-19)
k 2' = k 1 sin2 θ + k 2 cos2 θ
k 12' = ( k 1 – k 2 ) cos θ sin θ

where θ is the angle of rotation about n which transforms the vectors e1 and e2 to e1'
and e2', respectively. Because k1' and k2' are not principal curvatures, it is necessary to
include off-diagonal elements of the wavefront curvature tensor equal to k12'.

The following algorithm is used to reinitialize the principal radii of curvature of the
wavefront and their orientations. The reinitialization of the curvature variables follows
the method of Stavroudis in Ref. 11.

1 Given ni and ns, compute unit vectors in the directions of the reflected and
refracted rays, nr and nt.
2 Compute the vectors u0, ui, ut, ur, and us.
3 Compute the angle of rotation θ(i) needed to transform the local coordinate system
with axes parallel to e1, e2, and ni to a local coordinate system with axes parallel to
u0, ui, and ni. (Note: the superscript (i) is used to distinguish this coordinate system
rotation angle from the angle of incidence of the ray, often denoted θi.)
4 Transform the principal curvatures to the local coordinate system defined by u0, ui,
and ni by using Equation 3-19, substituting θ(i) for θ.
5 Given the two principal curvatures of the surface, k1,s and k2,s with directions e1,s
and e2,s, transform the curvature variables to a local coordinate system with axes
parallel to u0, us, and ns. Let the new curvature variables be denoted by k1,s', k2,s',
and k12,s'.
6 The curvatures of the wavefront and the surface have now been defined in
coordinate systems that share the axis u0 and only differ by a rotation by the angle
of incidence θi about u0. Defining the variables η and γ as in Material Discontinuity
Theory, compute the curvature variables of the refracted ray in a coordinate system
defined by u0, ut, and nt using the equations

k 1, t' = ηk 1' + γk 1, s'

η cos θ i γ
k 12, t' = ------------------ k 12' + -------------- k 12, s'
cos θ t cos θ t
η cos2 θ i γ -
k 2, t' = -------------------
- k ' + --------------- k '
cos2 θ t 2 cos2 θ t 2, s

198 | CHAPTER 3: RAY OPTICS INTERFACES

7 Obtain the principal curvatures of the refracted ray by rotating the coordinate
system defined by u0, ut, and nt by an angle θ(t) about nt. The angle θ(t) is defined
as
2k 12, t'
θ ( t ) = --- atan  ---------------------------
1
2 k 1, t ' – k 2, t '

8 Reinitialize the principal curvatures of the refracted ray:

(t) (t) (t) (t)
k 1, t = k 1, t' cos2 θ + 2k 12, t' cos θ sin θ + k 2, t' sin2 θ
(t) (t) (t)
k 2, t = k 1, t ' sin2 θ – 2k 12, t' cos θ sin θ + k 2, t' cos2 θ ( t )

9 Invert the principal curvatures to obtain the principal radii of curvature of the
refracted ray. Rotate u0 about nt by the angle θ(t) to obtain the reinitialized
principal curvature direction e1,t.
10 If a reflected ray is released, compute the curvature variables of the reflected ray in
a coordinate system defined by u0, ur, and nr using the equations

k 1, r' = k 1' – 2k 1, s' cos θ i

k 12, r' = – k 12' + 2k 12, s'

2
k 2, r' = k 2' – -------------- k 2, s'
cos θ i

11 Obtain the principal curvatures of the reflected ray by rotating the coordinate
system defined by u0, ur, and nr by an angle θ(r) about nr. The angle θ(r) is defined
as
2k 12, r'
θ ( r ) = --- atan  ----------------------------
1
2 k 2, r' – k 1, r'

12 Initialize the principal curvatures of the reflected ray:

(r) (r) (r) (r)
k 1, r = k 1, r' cos2 θ + 2k 12, r' cos θ sin θ + k 2, r' sin2 θ
(r) (r) (r)
k 2, r = k 1, r ' sin2 θ – 2k 12, r' cos θ sin θ + k 2, r' cos2 θ ( r )

13 Invert the principal curvatures to obtain the principal radii of curvature of the
reflected ray. Rotate u0 about nr by the angle θ(r) to obtain the initialized principal
curvature direction e1,r.

In 2D components, only one principal radius of curvature is computed; each ray is

treated as a cylindrical wave. In addition, no auxiliary degrees of freedom for the

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 199

 principal curvature direction are required because the out-of-plane direction can
always be treated as one of the axes of the local coordinate system that defines the
orientation of the wavefront.

In 2D axisymmetric model components, radii of curvature are computed for the

in-plane direction, that is the rz-plane, as well as the out-of-plane or azimuthal
directions. During ray-boundary interactions, these radii of curvature are reinitialized
as if the ray interacted with a 3D surface of revolution, which may have finite radii of
curvature in both the in-plane and out-of-plane directions.

STOKES VECTOR CALCULATION

The values of the intensity I and normalized Stokes parameters are stored as the
auxiliary dependent variables I0, sn1, sn2, and sn3 when a ray is released. These
auxiliary variables are updated when the ray hits a boundary. At any point along the
ray’s trajectory, each Stokes parameter is equal to

r 1, i r 2, i
s 0 = I = I 0 ------------------ s 1 = Is n1 s 2 = Is n2 s 3 = Is n3
r1 r2

In 2D, the second principal radius of curvature r2 is treated as an arbitrarily large

distance that remains constant for each ray. Because s1 and s2 are based on the
differences in intensity between orthogonal polarizations of radiation, a local
coordinate system must be defined for the ray. Because auxiliary degrees of freedom
have already been allocated for e1 as explained in Principal Radii of Curvature, a local
coordinate system is defined with axes parallel to e1, e2, and k. The axes parallel to e1
and e2 then function as the x- and y-axes in The Stokes Parameters, respectively.

STOKES VECTOR REINITIALIZATION

Reflection and Refraction of S- and P-polarized Rays

Recall from The Fresnel Equations that the ratios of the electric field amplitudes of the
incident, reflected, and refracted rays can be expressed as a set of Fresnel coefficients,

n 2 cos θ i – n 1 cos θ t 2n 1 cos θ i

r p = ------------------------------------------------ t p = ------------------------------------------------
n 2 cos θ i + n 1 cos θ t n 2 cos θ i + n 1 cos θ t
n 1 cos θ i – n 2 cos θ t 2n 1 cos θ i
r s = ------------------------------------------------ t s = ------------------------------------------------
n 1 cos θ i + n 2 cos θ t n 1 cos θ i + n 2 cos θ t

where the subscripts s and p denote s- and p-polarized rays, or rays with linear
polarizations perpendicular to and parallel to the plane of incidence, respectively. The
incident ray is assumed to move from a region of refractive index n1 toward a region

200 | CHAPTER 3: RAY OPTICS INTERFACES

of refractive index n2. The angles θi and θt are the angle of incidence and angle of
refraction, respectively.

If n1 > n2 and the angle of incidence exceeds the critical angle θ = asin(n2/n1), the
incident ray undergoes total internal reflection, and the reflected ray has intensity equal
to that of the incident ray.

The intensity of the transmitted and reflected waves are related to the intensity of the
incident wave Ii by the equations

n 2 cos θ t 2
Ir = Ii rp 2 I t = -------------------- I i t p
n 1 cos θ i

for p-polarized rays, and

n 2 cos θ t 2
Ir = Ii rs 2 I t = -------------------- I i t s
n 1 cos θ i

for s-polarized rays. The coefficients being multiplied by the incident ray intensity are
called the reflectance R and transmittance T:

n 2 cos θ t 2
Rp = rp 2 T p = -------------------- t p
n 1 cos θ i
n 2 cos θ t 2
Rs = rs 2 T s = -------------------- t s
n 1 cos θ i

Phase Shift Calculation for Reflected and Refracted Rays

Any fully polarized ray can be resolved into a pair of s- and p-polarized rays, with a
phase shift δ between them. The reflected s- and p-polarized rays at a material
discontinuity may then undergo different phase shifts; this may, for example, cause a
linearly polarized incident ray to be yield a reflected ray with elliptical polarization. In
addition, a phase shift may be applied to the transmitted ray if one or more Thin
Dielectric Film subnodes are added to a Material Discontinuity node. The phase shifts
of s- and p-polarized rays are computed from the complex-valued Fresnel coefficients:

Ψ p, r = arg ( r p ) Ψ p, t = arg ( t p )
Ψ s, r = arg ( r s ) Ψ s, t = arg ( t s )

Unpolarized and Partially Polarized Rays

An unpolarized ray has the following properties:

• The instantaneous intensity of the ray changes at random.

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 201

 • At a given instant in time, the ray has an equal probability of being polarized in any
direction orthogonal to the ray trajectory.
• The previous two properties are not affected by the transmission of the ray through
a wave retarder or a polarization rotator.

An unpolarized ray has Stokes parameters s1 = s2 = s3 = 0.

A partially polarized ray is neither completely deterministic nor completely random. It

is characterized by a degree of polarization between 0 and 1.

Often, the ray being reflected or refracted at a boundary is unpolarized or partially

polarized; or the polarization direction neither lies in the plane of incidence nor
perpendicular to it. In this case, a more general formulation for the new values of the
Stokes parameters is needed. The Stokes parameters of the reflected and refracted rays
can be expressed in terms of the Stokes parameters of the incident ray, the
transmittance or reflectance at the boundary, and the phase delay introduced by the
boundary (if any) (Ref. 15):

s 0t Ts + Tp Ts – Tp 0 0 s0
s 1t 1 s T – T T
p s + T p 0 0 s
= --- ' 1
s 2t 2 0 0 2 T s T p cos δ 2 T s T p sin δ s 2
s 3t 0 0 – 2 T s T p sin δ 2 T s T p cos δ s3

s 0r Rs + Rp Rs – Rp 0 0 s0
s 1r 1 Rs – Rp Rs + Rp 0 0 s1
= ---
s 2r 2 0 0 – 2 R s R p cos δ – 2 R s R p sin δ s 2
s 3r 0 0 2 R s R p sin δ – 2 R s R p cos δ s 3

where the subscripts r and t represent reflection and refraction, respectively; and the
subscripts s and p indicate the transmittance or reflectance for s- and p-polarized rays.

Wavefront Curvature Calculation in Graded Media

It is possible to compute the ray intensity by changing Intensity computation to
Compute intensity or Compute intensity in graded media in the physics interface Intensity
Computation section. The options Compute intensity and power and Compute intensity
and power in graded media can also be used to compute intensity, with the only
difference being that these options define an additional auxiliary dependent variable

202 | CHAPTER 3: RAY OPTICS INTERFACES

for the total power transferred by the ray. The setting Compute intensity is more robust
and accurate than Compute intensity in graded media, but is only applicable to
homogeneous media. The setting Compute intensity in graded media can be used for
both homogeneous and graded media, but it introduces more numerical error than
Compute intensity.

ASSUMPTIONS FOR COMPUTING INTENSITY IN HOMOGENEOUS MEDIA

When Compute intensity is selected, the ray intensity is computed along each ray path
using the following assumptions:

1 The wavefront represented by each ray subtends a constant plane angle (2D) or solid
angle (3D), which can change only when rays are reflected or refracted at
boundaries.
2 Given the radii of curvature at two positions within the same domain and the
intensity at one of these two positions, the intensity at the other position can be
computed using the relationship
r 1, 0 r 2, 0
I = I 0 ---------------------
r1 r2

3 Except when rays are reflected or refracted at boundaries, the principal radii of
curvature change at a constant rate:
dr 1 dr
--------- = --------2- = – 1
ds ds

4 The principal curvature directions do not change, except when rays are reflected or
refracted at boundaries.

These assumptions are not valid in graded media, in which the refractive index changes
continuously as a function of position. It is possible for the angle between two different
rays to change as they propagate through the medium, so the solid angle subtended by
the wavefront is no longer constant. As a result, the ray intensity cannot be expressed
using the ratio of initial and final principal radii of curvature.

CURVATURE TENSOR DEFINITION IN GRADED MEDIA

The calculation of ray intensity in graded media is based on the concept of a curvature
tensor K, defined in terms of the principal curvatures κ1 and κ2 and the corresponding
principal curvature directions e1 and e2:

K = κ 1 e 1 e 1T + κ 2 e 2 e 2T

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 203

 Because e1 and e2 are orthogonal, it follows that κ1 and κ2 are eigenvalues of K. It
also follows that K is singular because there is no contribution that is orthogonal to
both e1 and e2.

The signs of the principal curvatures are chosen so that positive curvature indicates that
the wavefront is converging, whereas negative curvature indicates that the wavefront
is diverging.

The selection of the coordinate system in which the curvature tensor is defined is
crucial. It is convenient to describe the coordinate system so that two of the coordinate
axes lie in the plane containing e1 and e2, since this reduces the number of nonzero
terms in K. Further reduction in the number of nonzero terms can be achieved if K
can be defined in a coordinate system in which e1 and e2 are basis vectors. This is
possible in 2D because one of the principal curvature directions is always parallel to the
out-of-plane direction, but in 3D it is not feasible because the principal curvature
directions can change as rays propagate through a graded medium.

In the most general 3D case, the curvature tensor is described using the following
symbols, each of which corresponds to a different orthonormal basis:

• KX: curvature tensor defined in the Cartesian coordinate system with basis vectors
x, y, and z.
• KW: curvature tensor defined in the coordinate system consisting of the two
principal curvature directions e1 and e2 and the direction of propagation t.
• KS: curvature tensor defined in a coordinate system in which one basis vector is the
direction of propagation t. The other two basis vectors s1 and s2 can be determined
arbitrarily, as long as they are orthogonal to each other and to t, such that
s1 × s2 = t. It is more convenient to operate in this coordinate system if the basis
vectors can be expressed strictly in terms of x, y, z, and t.

The basis vectors s1 and s2 are defined as follows:

t×z
s 1 = ------------- s2 = t × s1
t×z

To avoid poles in the definition of the basis vectors, the following alternative
definitions are used when the rays propagate nearly parallel to the z-axis:

t×x
s 1 = -------------- s2 = t × s1
t×x

The relationship between s1, s2, e1, and e2 is given by

204 | CHAPTER 3: RAY OPTICS INTERFACES

 e 1 = s 1 cos φ – s 2 sin φ

e 2 = s 1 sin φ + s 2 cos φ

where the rotation angle φ is an auxiliary dependent variable that is stored for each ray.

The relationship between KS and KW is given by

T
KS = Q0 KW Q0

where Q0 is the rotation matrix:

s1 ⋅ e1 s1 ⋅ e2 s1 ⋅ t cos φ sin φ 0
Q 0 = s 2 ⋅ e 1 s 2 ⋅ e 2 s 2 ⋅ t = – sin φ cos φ 0
t ⋅ e1 t ⋅ e2 t ⋅ t 0 0 0

The relationship between KX and KS is given by

T
K X = QK S Q

where Q is the rotation matrix:

x ⋅ s1 x ⋅ s2 x ⋅ t
Q = y ⋅ s1 y ⋅ s2 y ⋅ t
z ⋅ s1 z ⋅ s2 z ⋅ t

In 2D, one of the principal curvature directions is always known, so it is possible to

define s1 and s2 so that they coincide with e1 and e2. The rotation matrix Q0 is the
identity, and the auxiliary dependent variable for the angle φ can be omitted.

In 2D, one of the principal curvature directions is always known, so it is possible to

define s1 and s2 so that they coincide with e1 and e2. The rotation matrix Q0 is the
identity, and the auxiliary dependent variable for the angle φ can be omitted.

DERIVATIVES OF THE CURVATURE TENSOR

Following Ref. 13, the derivative of the curvature tensor along the ray trajectory in a
graded medium is given by the expression

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 205

 5
dK
----------S- =
ds  Mi
i=1

where the terms Mi correspond to the different ways in which the curvature can
depend on the medium properties. They are defined using the following expressions:

M1 = KS ⋅ KS
1
M 2 = – --- ( t ⋅ ∇ S n )K S
n
1
M 3 = – --- Π ( ∇ S ⊗ ∇ S n )
n
2
M 4 = -----2- Π ( ∇ S n ⊗ ∇ S n )Π
n
1
M 5 = – --- [ ( K∇ S n ) ⊗ t + t ⊗ ∇ S nK ]
n

where Π is the projection matrix:

Π = I – (t ⊗ t)

The gradient operator ∇S consists of derivatives that are taken with respect to the local
coordinates with basis vectors s1, s2, and t.

DERIVATIVES OF THE PRINCIPAL CURVATURES

By application of the chain rule, the derivatives of the nonzero elements of KS can be
expressed in terms of the principal curvatures that occupy the diagonal elements of KW
and the rotation angle φ:

∂κ ∂K S, 11 ∂K S, 12 ∂K S, 22
--------1- = -----------------
- cos 2 φ – 2 -----------------
- sin φ cos φ + -----------------
- sin 2 φ
∂s ∂s ∂s ∂s

∂κ ∂K S, 11 ∂K S, 12 ∂K S, 22
--------2- = -----------------
- sin 2 φ + 2 -----------------
- sin φ cos φ + -----------------
- cos 2 φ
∂s ∂s ∂s ∂s

∂φ- ∂K S, 12 1 ∂K S, 11 ∂K S, 22
= ------------------ ------------------ cos ( 2φ ) + ---  ------------------ – ------------------ sin ( 2φ )
1
-----
∂s κ2 – κ1 ∂s 2 ∂s ∂s

NUMERICAL STABILIZATION
The principal curvatures κ1 and κ2 are not ideal choices for the auxiliary dependent
variables stored by each ray because their values can become arbitrarily large as rays
approach caustics. Similarly, the principal radii of curvature r1 and r2 are not ideal

206 | CHAPTER 3: RAY OPTICS INTERFACES

choices because their values can become arbitrarily large when a diverging wavefront
begins to converge while propagating through a graded medium.

Instead, the auxiliary dependent variables stored by each ray are the help variables α1
and α2. In 2D, only one help variable is allocated because the out-of-plane principal
radius of curvature is assumed to be infinite. The principal curvatures are expressed in
terms of the help variables using the expression

κ i = κ 0 tan α

where κ0 = 1 1/m and i ∈ [1, 2]. The derivatives are then related by the expression

dα 1 dκ i 2
---------i = -----
- --------- cos α i
ds κ 0 ds

Similarly, the intensity may become infinitely large at caustics, and its reciprocal
becomes infinitely large as rays undergo attenuation, so a help variable Γ is used to
represent the intensity of the ray:

I = I 0 tan Γ

where I0 = 1 W/m2. The relationship between the ray intensity help variable and the
principal curvatures is

dΓ 1
------- = --- sin ( 2Γ ) [ κ 1 + κ 2 – 2κk 0 ]
ds 2

where k0 is the free space wave number and κ is the imaginary part of the refractive
index. Three additional auxiliary dependent variables are used to store information
about the remaining Stokes parameters.

Refraction in Strongly Absorbing Media

The direction of ray propagation may be described via the loci of points having
constant phase or amplitude. In a nonabsorbing medium, one which may be described
by a real-valued refractive index, changes in the ray intensity are caused only by the
convergence or divergence of the wavefront, so the surfaces in which the phase is
constant are parallel to the surfaces in which the amplitude is constant.

In a weakly absorbing medium, in which the attenuation of the power transmitted by

each ray decays over a length scale much larger than the wavelength, it may be assumed
that the surfaces of constant amplitude are still parallel to the surfaces of constant

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 207

 phase. The directions of the reflected and refracted rays at material discontinuities, as
well as the Fresnel coefficients that govern their intensity, may be computed using the
real parts of the refractive indices on either side of the boundary. However, as the
imaginary part of the refractive index increases in magnitude, this assumption may
begin to present a considerable source of error because the surfaces of constant
amplitude and surfaces of constant phase are no longer parallel.

In the Settings window for the Geometrical Optics interface, select the Use corrections
for strongly absorbing media check box in the physics interface Intensity Computation
section to modify Snell’s law and the Fresnel equations to accurately model refraction
between media with complex refractive indices. This check box is available when
Intensity computation is set to Compute intensity or Compute intensity and power.

Following Chang and others in Ref. 14, the wave vector in an absorbing medium is
treated as a bivector with complex components,

k = k 0 ( N e – iK f )

where k is the wave vector, k0 is the wave number in free space, and e and f are unit
vectors with real-valued components indicating the normal direction to the surfaces of
constant phase and surfaces of constant amplitude, respectively. The real-valued
quantities N and K, sometimes called the apparent refractive indices, are related to the
complex refractive index by the relations

2 2 2 2
N –K = n –κ

NK e ⋅ f = nκ

where n − iκ is the complex refractive index of the medium. From these relations the
apparent refractive indices can be computed as long as the dot product e ⋅ f is known.
To store information about the value of this dot product, auxiliary dependent variables
for the components of f are stored when the Use corrections for strongly absorbing
media check box is selected.

At material discontinuities, the normal vectors to the surfaces of constant amplitude

and phase follow a modified form of Snell’s law that uses the apparent refractive
indices,

N 1 sin θ 1 = N 2 sin θ 2

K 1 sin Ψ 1 = K 2 sin ψ 2

208 | CHAPTER 3: RAY OPTICS INTERFACES

where θ and ψ are the acute angles between the surface normal and the normal vectors
to the surfaces of constant phase and surfaces of constant amplitude, respectively. The
real part of the apparent refractive index in the second domain is a root of the quartic
equation

AN 24 + BN 22 + C = 0

A = 1
B = – [ N s2 + K s2 + ( n 22 – κ 22 ) ]
C = N s2 K s2 + ( n 22 – κ 22 )N 22 – ( nk – N s K s cos φ ) 2

where Ns = N1sin θ1, Ks = K1sin ψ1, and the angle φ is the azimuthal angle between
the normal to surfaces of constant amplitude and the plane of incidence. This equation
is obviously quadratic in N 22 and can thus be solved using the quadratic formula.

The reinitialized unit normal vectors to the surfaces of constant amplitude and phase
are

N1 N1
e 2 = ------- e 1 +  cos θ 2 – ------- cos θ 1 n
N2 N2

K1 K1
f 2 = ------- f 1 +  cos ψ 2 – ------- cos ψ 1 n
K2  K2 

whereas for the reflected ray the reinitialized unit normal vectors are

e 2 = e 1 – 2n cos θ

f 2 = f 1 – 2n cos ψ

Attenuation Within Domains

Rays can gradually lose energy as they propagate through a domain in which the
refractive index is complex. The convention of the Geometrical Optics interface is that
a lossy medium has refractive index given by n − iκ. Given the electric field E0 at one
location, the electric field of a ray with infinite wavefront radii of curvature as it
propagates through a homogeneous, lossy medium is

k 0 KL
E = E 0 exp  – --------------- cos α
 N 

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 209

 where L is the optical path length between the points at which E and E0 are measured
and k0 is the wave number in free space. In weakly absorbing media, N = n, K = κ, and
α = 0. If the Use corrections for strongly absorbing media check box is selected, N is the
real part of the apparent refractive index, K is the complex part of the apparent
refractive index, and α is the angle between the surfaces of constant amplitude and
surfaces of constant phase; both of these concepts are explained in Refraction in
Strongly Absorbing Media. The ray intensity and power are both proportional to the
square of the electric field norm, so these quantities change according to the relations

2k 0 KL
I = I 0 exp  – ------------------- cos α
 N 
2k KL
P = P 0 exp  – ------------------- cos α
0
N

When computing the ray intensity in absorbing media, the principal radii of curvature
of the wavefront must also be considered. The two contributing factors are accounted
for by allocating auxiliary dependent variables for the initial Stokes parameters
following the most recent reinitialization of the wave vector, which are expressed as
solutions to first-order differential equations of the form

ds i, 0
------------
- = – 2k 0 Kcs i, 0 i ∈ { 0, 1, 2, 3 } (3-20)
dt

where c is the speed of light in the medium and si,0 is value of one of the Stokes
parameters of the ray before the effects of changes in the principal radii of curvature
have been applied.

The Geometrical Optics interface defines an additional dependent variable A

(dimensionless) for the path integral of the attenuation coefficient,

dI 0
--------- = 0 (3-21)
dt
I = I 0 exp ( A )
dA
-------- = – 2k 0 Kc
dt

Despite requiring an extra degree of freedom per ray, the advantage of Equation 3-21
over Equation 3-20 is that the dependent variables in Equation 3-21 all vary linearly
within a homogeneous medium. This makes Equation 3-21 less prone to numerical
stiffness; it remains more accurate when the solver takes long time steps, compared to
Equation 3-20, potentially reducing solution time and improving accuracy.

210 | CHAPTER 3: RAY OPTICS INTERFACES

Fluence Rate Calculation Theory
The fluence rate E0 (SI unit: W/m2) is a volume quantity that indicates potential
exposure to radiation. The fluence rate at any point in a domain is the total incident
power that would hit a small spherical target at that location, from all directions,
divided by the cross sectional area of the sphere (Ref. 16).

For the purpose of defining fluence rate the “small sphere” is an opaque detector
whose radius is much smaller than any geometry details in the model, but substantially
larger than the wavelength of radiation. Thus, the absorption cross section of the
sphere is equal to its geometrical cross section and it is not necessary to consider Mie
scattering theory.

In the path of a collimated beam of uniform intensity I (SI unit: W/m2), the fluence
rate is simply I; however, the fluence rate distribution may be more complicated if there
are multiple beams intersecting at a point (since fluence rate takes contributions from
radiation arriving in all directions), or if the radiation is diffusely emitted or reflected.

The Fluence Rate Calculation node is a special type of Accumulator (Domain) that can
be added to a model when the ray power is solved for, meaning that one of the options
Compute power, Compute intensity and power, or Compute intensity and power in graded
media should be selected from the physics interface Intensity Computation list. This
feature defines an accumulated variable in each domain mesh element that gets the
following contribution from rays as they propagate through the element:

∂E 0 1
---------- = ----------------------
∂t meshvol  Qi ci
i

where the sum is taken over the rays within each mesh element, Qi (SI unit: W) is the
ray power, and ci (SI unit: m/s) is the speed of light of the ray. The division by the
mesh element volume meshvol compensates for the tendency of a larger domain
element to intersect a greater number of rays. Rather than a time integral, the above
can be thought of as a path integral of ray power from the entrance point to the exit
point as each ray traverses the mesh element.

The key distinction between fluence rate and deposited power is that fluence rate can
be nonzero even if the medium is not absorbing; it represents the potential for a small
object to be irradiated even if that object can only block a tiny fraction of the radiation
in the domain.

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 211

 Ray Termination Theory
The Ray Termination feature can annihilate rays when their intensity or power
decreases below a specified threshold.

If the ray intensity is computed, then the threshold ray intensity Ith (SI unit: W/m2)
can be specified. Then the termination criterion is

I < I th

If the ray power is computed, then the threshold ray power Qth (SI unit: W) can be
specified. Then the termination criterion is

Q < Q th

It is also possible to specify termination criteria based on both intensity and power; in
this case, the rays terminate if either criterion is satisfied.

Whenever possible, the Ray Termination feature attempts to stop rays at the exact
instant when the ray intensity or power reaches a specified threshold value. A
closed-form analytic solution for the exact stop time is available under the following
circumstances:

• The rays represent planar wavefronts in an absorbing medium, or

• The rays represent spherical or ellipsoidal wavefronts in a nonabsorbing medium.

If neither of these special conditions is met, the ray can only terminate if the
termination criterion is met at the beginning of a time step taken by the solver; the ray
intensity or power at the instant the ray is stopped may be less than the threshold.

PLANAR WAVEFRONTS IN AN ABSORBING MEDIUM

In this special case, the principal wavefront radii of curvature are so large that the ray
intensity can only change significantly in an attenuating medium. The exact time ts
(SI unit: s) at which the ray intensity reaches the threshold value is

t s = t 0 + τ log  ------
I
 I th

where

• t0 (SI unit: s) is the previous time step, reflection time, or refraction time,
• I (SI unit: W/m2) is the intensity at t0, and
• Ith (SI unit: W/m2) is the threshold intensity.

212 | CHAPTER 3: RAY OPTICS INTERFACES

If the termination criterion is based on the ray power Q (SI unit: W) instead of ray
intensity, the analogous expression for the stop time is

t s = t 0 + τ log  --------
Q
 Q th

The characteristic time for ray attenuation τ (SI unit: s) is

λ0 N
τ = --------------
4πcK

where

• λ0 (SI unit: m) is the free-space wavelength of the radiation,

• N − iK (dimensionless) is the apparent refractive index of the medium, where N and
K are real numbers, and
• c = 299,792,458 m/s is the speed of light in a vacuum.

DIVERGING WAVEFRONTS IN A NONABSORBING MEDIUM

In a nonabsorbing medium in which the wavefront is nonplanar, the ray can only
terminate due to the intensity decreasing below the threshold, since the power does
not change. The time ts (SI unit: s) at which the ray intensity reaches the threshold in
a nonattenuating, homogeneous medium is

t s = t 0 + ----------  r 1 + r 2 + ( r 1 – r 2 ) 2 + 4 ------ r 1 r 2 
1 I
2V g I th

where

• t0 (SI unit: s) is the previous time step, reflection time, or refraction time,
• Vg (SI unit: m/s) is the group velocity magnitude,
• r1 (SI unit: m) is the first principal radius of curvature of the wavefront at t0,
• r2 (SI unit: m) is the second principal radius of curvature of the wavefront at t0,
• I (SI unit: W/m2) is the intensity at t0, and
• Ith (SI unit: W/m2) is the threshold intensity.

In 2D models, or in 3D models where the wavefronts are cylindrical (one extremely

large radius of curvature and one finite radius of curvature), the stop time is instead

t s = t 0 + ------  r 1 + ------ r 1 
1 I
Vg  I th 

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 213

 where r1 (SI unit: m) is the finite principal radius of curvature of the cylindrical
wavefront.

Illuminated Surface Theory

The Illuminated Surface is used to simulate the emission of rays from a surface, under
the assumption that the emitted rays are specularly reflected from an external radiation
source. It is assumed that the entire selected surface has direct line of sight to the
external radiation source; that is, shadowing due to other geometric entities is not
taken into account.

If the Incident ray direction vector is a User defined direction or based on Solar position,
the incident rays are treated as plane waves arriving from a distant source, and the
intensity of the incident rays Ii is specified directly.

If the Incident ray direction vector is computed using a User defined point source, the
principal radii of curvature of the incident wavefront are set equal to the distance from
the source point to the initial ray position,

r 1 = r 2 = – r src – q

where q is the initial ray position, rsrc is the position vector of the point source, and
r1 and r2 are the principal radii of curvature of the incident wavefront. In 2D, only one
principal radius of curvature is defined. The incident ray intensity is derived from the
total source power Psrc,

P src
I i = ----------------------------
4π r src – q

The principal radii of curvature of the reflected rays are then computed using the
approach outlined in the Principal Radii of Curvature section. Because no thin films
are present on the surface, the intensity of the reflected ray is equal to the intensity of
the incident ray. The effect of specular reflection is to negate the third and fourth
Stokes parameters of the ray.

ANGULAR PERTURBATIONS
It is often necessary to take the finite size of a radiation source into account when
modeling the reflection or refraction of rays at an illuminated surface. This is
particularly true in solar energy applications, in which the finite size of the solar disk
significantly affects the focusing capability of solar concentrating devices.

214 | CHAPTER 3: RAY OPTICS INTERFACES

To release rays with a distribution of initial directions based on the finite size of the
radiation source, select Create light cones at release points or Sample from conical
distribution option from the Corrections for finite source diameter check box in the
settings window for the Illuminated Surface feature.

Selecting Create light cones at release points causes a cone of rays to be released at each
release point. Selecting Sample from conical distribution causes only a single ray to be
released at each point, but each ray’s initial direction is sampled form a cone-shaped
probability distribution function. The maximum angle between any incident ray and
the specified incident ray direction is the user-defined maximum disk angle
ψm (SI unit: rad).

The options described above are also available in the settings window for the Solar
Radiation feature, in which they have the same effects on the initial direction of the
solar radiation.

Limb Darkening
The default value of the maximum disk angle, or the maximum angle between the
direction of the incident rays and the specified Incident ray direction vector, is
ψm = 4.65 mrad based on the following expressions (see for example Ref. 22):

d sun
ψ m = asin  -----------
2L 0
9
d sun = 1.392 × 10 m

11
L 0 = 1.496 × 10 m

When the finite size of the sun is taken into account, radiation from the center of the
solar disc tends to be greater in intensity than radiation from the edges of the disc, a
phenomenon known as limb darkening. If Empirical power law is selected from the Limb
darkening model list in the settings window for the Illuminated Surface or Solar Radiation
feature, the solar radiation is assumed to follow a power law distribution. Given the
intensity I(0) of radiation along a line intersecting the center of the sun, the incident
solar radiation from any other point on the sun is given by the expression

I ( ψs ) α(λ )
-------------- = μ 0
I(0)

where

μ = cos ψ s

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 215

 and ψs is the angle between a line of sight to the center of the sun and a line from the
center of the sun to another visible point on its surface. The exponent α can vary
from 0, for which all solar radiation is of uniform intensity, to 1, at which the intensity
of solar radiation falls off linearly as a function of radial position on the visible solar
disc. Based on Ref. 22 the exponent α is given as a function of the free-space
wavelength λ0:

 –1 –1 –1
 – 0.023 + 0.292λ 0 λ 0 ≤ 2.4μm

a ( λ 0 ) =  0.3778 + 0.125λ –1 2.4μm
–1 –1
≤ λ 0 ≤ 2.8μm
–1
 0
 – 0.507 + 0.441λ – 1 2.8μm
–1 –1
≤ λ0
 0

If Linear is selected from the Limb darkening model list, the distribution of solar intensity
is instead

I ( ψs ) 1 + βμ
------------- = ----------------
I(0) 1+β

Surface Roughness
It is also possible to include the effect of surface roughness when computing the
incident ray direction vector at illuminated surfaces by selecting the Include surface
roughness check box. The value of the Surface slope error σ is used as the standard
deviation in a Rayleigh distribution of perturbation angles about the unperturbed ray
direction vector, from which a user-defined number of rays are sampled. For more
details see Surface Slope Error.

Theory of Mueller Matrices and Optical Components

Optical devices such as polarizers and wave retarders are often used to manipulate the
polarization of radiation. These devices can be modeled by multiplying the Stokes
vector of the incident ray si by a Mueller matrix M,

s = Ms i

As explained in The Stokes Parameters, the Stokes vector requires a set of orthogonal
coordinate axes to be defined at the location of a propagating ray. The x-axis of this
local coordinate system, e1, is the first principal curvature direction in 3D models and
the out-of-plane direction in 2D models.

216 | CHAPTER 3: RAY OPTICS INTERFACES

Similarly, the Mueller matrix of an optical device is defined with respect to a local
coordinate system. This system need not coincide with the coordinate system that
defines the Stokes vector of the incident ray. Therefore it is usually necessary to apply
a rotation to one of these local coordinate systems before applying the Mueller matrix:

–1
s = R MRs i

where R is the rotation matrix

1 0 0 0
R = 0 cos ( 2θ ) sin ( 2θ ) 0
0 – sin ( 2θ ) cos ( 2θ ) 0
0 0 0 1

where θ is the angle by which the local coordinate system of the ray must be rotated
so that the x-axes of the two local coordinate systems both lie in the plane of incidence
and the y-axes are parallel.

Mueller matrices of some common optical devices are tabulated in Ref. 15.

LINEAR POLARIZER THEORY

A Linear Polarizer is an optical device that produces linearly polarized light. An ideal
linear polarizer transmits radiation that is polarized in one direction, indicated by the
transmission axis T, while completely preventing the transmission of radiation
polarized in the orthogonal direction.

In a local coordinate system in which the x-axis is parallel to the transmission axis, the
Mueller matrix of the ideal linear polarizer is

1 1 0 0
1 1 1 0 0
M = ---
2 0 0 0 0
0 0 0 0

LINEAR WAVE RETARDER THEORY

A Linear Wave Retarder is an optical device that separates a ray into two rays with
orthogonal linear polarizations, retards the phase of one of these rays with respect to
the other, and then recombines the two rays. The direction of polarization that
undergoes this retardation is called the slow axis, and the orthogonal direction is called

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 217

 the fast axis F. The phase delay between the two orthogonal components is the
retardance δ of the device.

In a local coordinate system in which the x-axis is parallel to the fast axis, the Mueller
matrix of an ideal linear wave retarder is

1 0 0 0
M = 0 1 0 0
0 0 cos δ sin δ
0 0 – sin δ cos δ

For example, a quarter-wave retarder (δ = π/2) converts circularly polarized light to

linearly polarized light, whereas a half-wave retarder (δ = π) converts left-handed
circularly polarized light to right-handed circularly polarized light.

CIRCULAR WAVE RETARDER THEORY

A Circular Wave Retarder, or polarization rotator, is an optical device that separates a
ray into two rays with left-handed and right-handed circular polarizations, retards the
phase of one of these rays with respect to the other, and recombines the two rays. A
right circular polarization rotator is represented by the Mueller matrix

1 0 0 0
M = 0 cos δ sin δ 0
0 – sin δ cos δ 0
0 0 0 1

The circular retarder rotates the polarization ellipse of the incident ray by an angle δ/2
about the origin.

IDEAL DEPOLARIZER THEORY

The Ideal Depolarizer converts radiation of any polarization to completely unpolarized
radiation. The Mueller matrix of an ideal depolarizer is

1 0 0 0
M = 0 0 0 0
0 0 0 0
0 0 0 0

218 | CHAPTER 3: RAY OPTICS INTERFACES

Thin Dielectric Film Theory
When ray intensity or power is solved for, thin dielectric films can be used to change
the reflection and transmission coefficients at boundaries. Individual dielectric layers
can be applied to a boundary by using the Thin Dielectric Film subnode, which can be
added to the Material Discontinuity and Wall nodes. The Thin Dielectric Film subnode
can be added to the same boundary multiple times to create multilayer films. In
general, the thickness of these thin dielectric films is comparable in magnitude to the
wavelength of the rays; a minimum requirement is that the thickness of each layer is
much less than the coherence length of the incident rays.

COMPUTING THE FRESNEL COEFFICIENTS OF BOUNDARIES WITH THIN

DIELECTRIC FILMS
When the Thin Dielectric Film feature is used to model dielectric layers, the equivalent
Fresnel coefficients for ray propagation through the layer are computed using a
recursive algorithm. These equivalent Fresnel coefficients are, in general,
complex-valued reflection and transmission coefficients that account for the phase
delay and amplitude change in the reflected and transmitted rays resulting from the
multiple reflections that the ray undergoes within the dielectric layer, all of which are
considered to be coherent with each other. As these multiple reflected and refracted
rays propagate into the domains adjacent to the film, they can interfere constructively
or destructively with each other, which affects the total electric field amplitude that
propagates into each domain. This is illustrated in Figure 3-13 below. Due to the rays’
capability to interfere constructively or destructively with each other, the optical
thickness of the thin film has a significant effect on the amplitude and phase of the
reflected and refracted rays.

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 219

 Figure 3-13: One incident ray enters a thin dielectric film and produces a large number
of coherent reflected and refracted rays in the adjacent semi-infinite media. These rays can
interfere constructively or destructively with each other.

Consider, as shown above, a thin film of thickness d and refractive index n2 bounded
by semi-infinite domains with refractive indices n1 and n3. Also consider a linearly
polarized ray with electric field amplitude E0 that propagates through the domain of
refractive index n1 and interacts with the thin film. Let the coefficients of refraction
and transmission for a ray moving from the domain of refractive index na to the
domain of refractive index nb be denoted rab and tab, respectively. When the ray arrives
at the film, the refracted ray undergoes a series of reflections at the boundaries between
the film and either adjacent domain. Following Ref. 9, the amplitudes of the refracted
rays in either domain form a pair of converging geometric series with sums

∞
E1 r 12 + r 23 e 2iβ
r eq = ------- = r 12 + t 12 t 21 r 23 e 2iβ
E0  ( r 21 r 23 e 2iβ ) k = ------------------------------------
1 + r 12 r 23 e 2iβ
-
k=0

for rays returning to the domain of refractive index n1, and

∞ iβ
E3 t 12 t 23 e
t eq = ------- = t 12 t 23 e iβ
E0  ( r 21 r 23 e 2iβ ) k = ------------------------------------
1 + r 12 r 23 e 2iβ
-
k=0

for rays entering the domain of refractive index n3. The angle β is the phase delay that
is introduced when a ray propagates from one surface of the film to the other,

220 | CHAPTER 3: RAY OPTICS INTERFACES

 2πn 2 d cos θ 2
β = --------------------------------- (3-22)
λ0

where λ0 is the free-space wavelength and θ2 is the angle between the wave vector of
a ray within the thin film and the normal to the film. The dielectric film can be modeled
as a material discontinuity with transmission and reflection coefficients teq = E3 / E0
and req = E1 / E0.

Several derivations for the equivalent Fresnel coefficients are also given by
Heavens in Ref. 17, including recursive algorithms and matrix
formulations. However, the sign conventions and symbols used in this
section have been selected to be consistent with Born and Wolf (Ref. 9)
and therefore differ from those in Ref. 17.

AUTOMATIC SETUP OF SINGLE-LAYER FILMS

The Material Discontinuity node includes options for automatic setup of single-layer
films with specified reflectance R or transmittance T. In either case the layers are
assumed to be nonabsorbing, so that R + T = 1. To enable these options, select Specify
reflectance or Specify transmittance from the Thin dielectric films on boundary list, and
then select the Treat as single layer dielectric film check box.

The specified reflectance R of a thin film with refractive index n is valid only for certain
values of the angle of incidence θi and free-space wavelength λ0. After computing the
Fresnel coefficients for rays entering and leaving the layer, the reflectance can be
expressed as

2 2
r 12 + r 23 e 2iβ 2 r 12 + r 23 + 2r 12 r 23 cos ( 2β )
R = ------------------------------------
- = -------------------------------------------------------------------------
2 2
-
1 + r 12 r 23 e 2iβ r 12 r 23 + 2r 12 r 23 cos ( 2β ) + 1

Solving this expression for β and substituting the result into Equation 3-22 yields

2 2 2 2
λ0  r 12 r 23 R – r 12 – r 23 + R
d = -------------------------- acos  -----------------------------------------------------------
-
4πn cos θ 2  2 2
2r r ( 1 – R ) 
12 23

For nonzero angles of incidence, the layer thickness that yields a specified reflectance
or transmittance depends on the ray polarization. The coefficients of reflection for rays

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 221

 propagating into or out of the layer are either those for S- or P-polarized radiation,
depending on the selected option in the Specified film behavior applies to list.

In some cases, the specified transmittance or reflectance will be impossible

to achieve, given the refractive indices of the layer and the surrounding
media. In such cases, rays will pass through the boundary as if no dielectric
layers were present.

AUTOMATIC SETUP OF ANTI-REFLECTIVE COATINGS

To automatically compute the refractive index and thickness of a single-layer
anti-reflective coating between two dielectric media, in the settings for the Material
Discontinuity node select Anti-reflective coating from the Thin dielectric films on
boundary list. By default the anti-reflective coating is ideal, in the sense that is always
has a reflectance of zero for rays of any wavelength and any angle of incidence.
However, if you select the Treat as single layer dielectric film check box, then instead
the single-layer coating only has a reflectance of zero for rays at a specific angle of
incidence θi (SI unit: rad) and vacuum wavelength λ0 (SI unit: m).

For oblique incidence, the refractive index of the ideal single-layer coating differs for
s- and p-polarized rays. For s-polarized rays, the refractive index n2 of a film between
media with refractive indices n1 and n3 is

2 2
n2 = ( n 1 sin θ i ) 2 + n 1 cos θ i n 3 – ( n 1 sin θ i )

For p-polarized rays, the refractive index is

2 2
2 cos θ i n a n 1 ( n 3 + n 3 – 4 cos θ i sin θ i n a n 1 n 3 )
n 2 = ------- -----------------------------------------------------------------------------------------------------------------------
2 n a cos θ i

where

2 2
n 3 – ( n 1 sin θ i )
n a ≡ ----------------------------------------
2
n3

Either equation returns the value of n2 for which r12 = r23 for rays of the specified
polarization at the specified angle of incidence. An ideal anti-reflective coating is
obtained when the magnitude of the equivalent reflection coefficient is 0. This can be

222 | CHAPTER 3: RAY OPTICS INTERFACES

obtained when r12 = r23 and the phase delay δ is equal to π/2. Substituting this
requirement into Equation 3-22 yields an expression for d:

λ0
d = -------------------------
4n 2 cos θ 2

SETUP OF REPEATING MULTILAYER FILMS

It is possible to identify a unit cell consisting of a small number of thin dielectric films
and to repeat these layers a specified number of times. To set up repeating multilayer
films, select Add layers to surface, repeating from the Thin dielectric films on boundary
list.

Select the Repeat layer in multilayer films check box in the Repeating Multilayer Films
section of the Thin Dielectric Film settings window to include the layer in the unit cell
of a repeating multilayer film. It is possible to include some layers in the unit cell while
excluding others. This could be used, for example, to ensure that a repeating multilayer
film begins and ends with the same layer, since in this case the last layer is not included
in any unit cell. However, the Thin Dielectric Film nodes for all layers in the unit cell
must be adjacent to each other in the Model Builder.

Enter a value or expression for the Number of repeating unit cells N in the Material
Discontinuity settings window. If there are p Thin Dielectric Film nodes in which the
Repeat layer in multilayer films check box is selected and q Thin Dielectric Film nodes in
which the Repeat layer in multilayer films check box is cleared, the total number of
dielectric layers in the multilayer film Nt will be Np + q.

The recursive algorithm described in the section Computing the Fresnel Coefficients
of Boundaries with Thin Dielectric Films is not well-suited to the calculation of the
effective Fresnel coefficients for an extremely large number of thin dielectric films in a
periodic arrangement. Instead, the following approach is used, in which the equivalent
Fresnel coefficients are expressed in terms of the product of a large number of 2-by-2
matrices.

Let the electric field amplitude of the incident, reflected, and transmitted rays entering
i r i
layer m be denoted E m , E m , and E m + 1 , respectively, where m ∈ [ 1, N t ] . The
ratios of the incoming and outgoing electric field amplitudes in consecutive layers m
and m+1 are

i i
Em 1 E
= ------------------------- C m m + 1
r t( m – 1 ) ( m ) r
Em Em + 1

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 223

 Where Cm is a 2-by-2 matrix:

C m, 11 C m, 12 exp ( – iδ m – 1 ) r ( m – 1 ) ( m ) exp ( – iδ m – 1 )
Cm = =
C m, 21 C m, 22 r ( m – 1 ) ( m ) exp ( iδ m – 1 ) exp ( iδ m – 1 )

where δm−1 is the phase shift within the layer, with δ0 = 0.

For a multilayer film containing Nt layers including N unit cells, the relationship
between the incoming and outgoing electric field amplitudes can be expressed using
the product of Nt + 1 matrices and Nt + 1 transmission coefficients:

N +1 –1 Nt + 1
E0
i  t    Ei
=  ∏ t ( m – 1 ) ( m ) 
∏ C m Nt + 1
r     r
E0 m = 1  m = 1  ENt + 1

To simplify the matrix multiplication, the product of the matrices can be expressed in
the form

Nt + 1

∏ Cm =
N
C in ( C cell ) C out
m=1

where Cin is the product of the matrices for all layers prior to the unit cell, Cout is the
product of the matrices for all layers following the unit cell, and Ccell is the matrix for
all layers within one unit cell, including the interface between one unit cell and the
next. Next an eigenvalue decomposition is performed on Ccell:

C cell = XDX – 1

where D is a diagonal matrix with nonzero elements D11 and D22. The ratio of the
incoming and outgoing electric field amplitudes can then be expressed as

N +1 –1
E0
i  t  N
D 11 0
i
ENt + 1
=  ∏ t ( m – 1 ) ( m ) C in X X – 1 C out
r   N r
E0 m = 1  0 D 22 ENt + 1

r
Because there is no reflected ray beyond the last dielectric layer, E N t + 1 = 0 and the
equivalent Fresnel coefficients of the multilayer film can now be obtained:

224 | CHAPTER 3: RAY OPTICS INTERFACES

 r i
E0 ENt + 1
r eq = ------i- t eq = ----------------
i
-
E0 E0

Diffraction Grating Theory

The Grating and Cross Grating features are boundary conditions that release rays into
one or more diffraction orders whenever a ray hits a boundary. In reality, the grating
surface is a periodic substructure in which the unit cells can be smaller than the
electromagnetic wavelength, but the substructure geometry is not explicitly modeled
as part of the ray optics simulation. Based on the unit cell size and wavelength, rays
that hit the grating will propagate into specific directions (the diffraction orders) in
which waves from neighboring unit cells would constructively interfere.

Although the substructure of a grating or cross grating may contain many fine details,
typically on the micron scale, from a ray optics perspective it is not necessary to
explicitly model the substructure geometry. Instead the grating is approximated as a
featureless surface where diffraction orders are released in directions computed from
the specified unit cell size and the electromagnetic wavelength.

DEFINITIONS
Unless otherwise stated, in this section, the term “grating” will be used to refer to a
surface whose substructure is periodic in one direction and homogeneous in the
orthogonal direction. The term “cross grating” will be used for a surface whose
substructure exhibits periodicity in two different directions.

Consider an electromagnetic wave incident on the grating or cross grating surface with
vacuum wavelength λ0 (SI unit: m). It therefore has angular frequency ω = 2π/λ0
(SI unit: rad/m). Assuming the surrounding media are transparent or weakly
absorbing, the ray propagates in the direction of its wave vector k (SI unit: rad/m)
which has magnitude k = ωn/c. Here n (dimensionless) is the absolute refractive index
of the material and c = 299,792,458 m/s is the speed of light in a vacuum.

GRATING
The effect of a grating is to offset the in-plane components of the wave vector by an
integer multiple of the reciprocal lattice vector G (SI unit: rad/m), given by

Tg
G = 2π ------ ,
d

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 225

 Here d (SI unit: m) is the grating constant, that is, the width of each unit cell of the
periodic substructure; and Tg (dimensionless) is a unit vector that lies in the grating
surface and indicates the direction of periodicity. By definition, Tg is orthogonal to the
surface normal ns.

The wave vector k of the incident ray can be treated as the vector sum of two
constituent vectors, one parallel to the surface (kp) and one perpendicular to it (kn):

kp = ns × ( k × ns ) k n = ( k ⋅ n s )n s ,

The in-plane component of the outgoing wave vector is given by

k p, out = k p + mG (3-23)

where m (dimensionless) is an integer known as the diffraction order. For m = 0 the

ray is specularly reflected if it propagates into the same medium as the incident ray, or
refracted (as in Snell’s law) if it propagates into the medium on the opposite side of the
grating. The diffraction order m can be any integer but (as detailed in a later section)
rays of a given diffraction order can only propagate if the out-of-plane wave vector
component remains real-valued while the in-plane component satisfies Equation 3-23.

CROSS GRATING
In a Cross Grating, the in-plane component of the wave vector is offset by integer
multiples of two different reciprocal lattice vectors corresponding to two directions of
periodicity in the grating surface,

k p, out = k p + mG 1 + nG 2 (3-24)

where m and n are the diffraction order numbers. The two reciprocal lattice vectors
G1 and G2 (SI unit: rad/m) are defined as

d2 × ns ns × d1
G 1 = 2π --------------------------------- G 2 = 2π ---------------------------------
d1 ⋅ ( d2 × ns ) d2 ⋅ ( ns × d1 ) (3-25)
d 1 ≡ d 1 T g, 1 d 2 ≡ d 2 T g, 2

where d1 and d2 (SI unit: m) are the dimensions of one unit cell in the
two-dimensional periodic substructure, and Tg,1 and Tg,2 (dimensionless) are unit
vectors indicating the corresponding directions of periodicity.

Note that Tg,1 and Tg,2 are not required to be perpendicular to each other, but they
cannot be parallel or antiparallel. By definition, both Tg,1 and Tg,2 are orthogonal to

226 | CHAPTER 3: RAY OPTICS INTERFACES

the surface normal ns. For the simplified case where Tg,1 and Tg,2 are perpendicular,
Equation 3-25 simplifies to

T g, 1 T g, 2
G 1 = 2π ----------- G 2 = 2π -----------
d1 d2

Similar to a grating with a single direction of periodicity, the two diffraction orders m
and n can take on any integer values, but that diffraction order can only propagate if
the out-of-plane wave vector component can be real-valued while the in-plane
component satisfies Equation 3-24.

WAVE VECTOR REINITIALIZATION

The magnitude of the outgoing perpendicular component of the reflected and
transmitted wavefront can be found from the requirement that the magnitude of the
outgoing wave vector kout is equal to the wave number; that is, ||kout|| = k0n, where
k0 = ω/c is the wave number in free space. The projections of the reflected and
transmitted wave vectors onto the surface normal kn,r and kn,t therefore have
magnitudes

2 2
k n, r = ( k 0 n 1 ) – k p, out
(3-26)
2 2
k n, t = ( k 0 n 2 ) – k p, out

where n1 and n2 are the refractive indices on the incoming (reflected) and outgoing
(transmitted) side of the grating, respectively. The outgoing reflected and transmitted
wave vectors kr and kt are therefore given by

k r = k p, out + k n, r n s
k t = k p, out + k n, t n s

A given order will only exist if the magnitude of the parallel (in-plane) wave vector is
less than the magnitude of each out-going wave vector. That is

k
p, out
< k0 n1

for rays on the incoming side (reflected rays) and

k p, out < k0 n2

for rays on the outgoing side (transmitted rays). Each is equivalent to requiring one of
the square roots in Equation 3-26 to return a positive real value.

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 227

 RELATIVE ORDER NUMBERS
For the Grating (but not the Cross Grating) it is possible to specify relative diffraction
order numbers instead of absolute orders. This is particularly useful for blazed gratings.

In a blazed grating, the main detail of the grating microstructure is an array of facets
that make an angle θB (SI unit: rad) with respect to the grating normal. This angle is
known as the blaze angle. The absolute order number m in which maximum
diffraction efficiency occurs can be found from the general grating equation

mλ 0 = nd g cos γ ( sin α + sin β ) ,

where the angles α and β (SI unit: rad) are defined as

α = θB + θ β = θ B – θ′

where θ and γ are the angles of incidence with respect to the grating normal in-plane
and out-of-plane respectively. At the peak of the blaze function the angle of diffraction
with respect to the grating normal (θ') is equal to θ. Therefore, for any given
wavelength λ, the absolute order number is given by

2nd g
m = round ------------- cos γ sin θ B cos θ + Δm ,
λ

where Δm is the order number relative to the order in which this wavelength is closest
to the blaze peak. The in-plane and out-of-plane angles can be derived from the
incoming wave vector components. First, we have the component in the direction of
the grating normal

kn = k ⋅ ns .

The in-plane and out-of-plane components are

k p, in = k p ⋅ T g

and

k p, out = k p ⋅ ( n s × T g )

where kp is the incoming wave vector projected onto the grating surface. This gives
the angle θ and γ as

k p, in
θ = – atan  ------------ – θ B
 kn 

228 | CHAPTER 3: RAY OPTICS INTERFACES

and

k p, out cos θ B
γ = – atan  ------------------------------- .
 kn 

Ideal Lens Theory

The Ideal Lens feature can be used to treat a flat surface (in 3D) or an edge (in 2D) as
if it were a lens, allowing it to focus or defocus rays.

Optical Power
The optical power K (SI unit: 1/m) of a lens is

1
K = --------
f eff

where feff is the effective focal length (SI unit: m). For a thin lens, we have

K = K1 + K2

where

n – n1 n – n2
K 1 = ---------------- K 2 = ----------------
R1 R2

with n1 and n2 being the refractive index before and after the lens respectively, and
where n is the refractive index of the lens. The radii of curvature of the first and second
lens surfaces is R1 and R2 (SI unit: m) respectively.

For a lens with thickness Tc (SI unit: m), the optical power is instead

K = K 1 + K 2 + K 12

where

n – n1 n – n2 Tc
K 1 = ---------------- K 2 = ---------------- K 12 = ------ K 1 K 2
R1 R2 n

Refracted Ray Direction

For each point r (SI unit: m) on the lens surface, define an orthonormal coordinate
system with unit vectors er, et, and ns (the surface normal), where

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 229

 δr
e r = --------- et = er × ns
δr

where the subscripts r and t refer to the radial and tangential directions respectively.
The distance δr is defined with respect to a reference point on the lens surface as
δr = r - rref.

The incoming ray direction can be split into the radial and tangential components as
follows:

ni ⋅ er ni ⋅ eφ
tan θ i, r = ---------------- tan θ i, φ = ---------------- ,
ni ⋅ ns ni ⋅ ns

where ni is the incident ray direction. The ideal lens will only affect the radial
component of the ray directions; that is

n 2 tan θ r = n 1 tan θ i, r – K δr

and

n 2 tan θ t = n 1 tan θ i, t

and therefore the transmitted ray direction is

n t = n s + tan θ r e r + tan θ t e t

Surface Scattering Theory

Several boundary features can be used to scatter rays. These include diffuse and
isotropic reflection from a Wall, and perturbations due to surface slope error with an
Illuminated Surface. These scattering models can be used in both transmission and/or
reflection with the Scattering Boundary feature.

DIFFUSE AND ISOTROPIC SCATTERING

Diffuse (or Lambertian) scattering follows the cosine law. That is, the flux dn of rays
across a surface element A whose directions lie within a small solid angle dω is
proportional to the cosine of the polar angle θ:

dn ∝ A cos θ dω

or, substituting the expression for the differential solid angle,

dω = sin θ dθ dϕ

230 | CHAPTER 3: RAY OPTICS INTERFACES

gives

dn ∝ A cos θ sin θ dθ dϕ

where ϕ is the azimuthal angle.

For isotropic scattering the flux of scattered rays is the same for any differential solid
angle; that is

dn ∝ A dω

or

dn ∝ A sin θ dθ dϕ

Diffuse and Isotropic Probability Distributions

In 3D, the normalized probability distribution functions f(θ,ϕ) for diffuse and
isotropic scattering are given by

1 1
f diffuse ( θ, ϕ ) = --- cos θ sin θ f isotropic ( θ, ϕ ) = ------ sin θ
π 2π

where

π
θ ∈ 0, --- ϕ ∈ [ 0, 2π ]
2

In 2D these distributions become

1 1
f diffuse ( θ ) = --- cos θ f isotropic ( θ ) = ---
2 π

where

π π
θ ∈ – ---, ---
2 2

The direction components of a ray after it scatters diffusely or isotropically are given by

δt 1 = sin θ cos ϕ
δt 2 = sin θ sin ϕ
δn = cos θ

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 231

 where δn is the component in the direction of the surface normal (ns) and where δt1
and δt2 are the components in the two directions t1 and t2 orthogonal to the surface
normal. That is, after scattering the reflected and transmitted ray directions are

n t = δt 1 t 1 + δt 2 t 2 + δnn s n r = δt 1 t 1 + δt 2 t 2 – δnn s

where

ns = t1 × t2

SURFACE SLOPE ERROR

Following Ref. 20 rays will be deviated upon reflection or refraction from a surface
with a non-zero slope error. It will be assumed that the tangential deviation is
negligible and that the radial distribution is Gaussian. Therefore, given an angular
slope error σ, the probability distribution function for the polar angle is a Rayleigh
distribution of the form

θ θ2
f ( θ ) = -----2- exp  – ---------2- ,
σ  2σ 

which can be used to get the directions components δn, δt1 and δt2 as shown in the
previous section. The effect of applying a surface slope error is to perturb the surface
normal ns to a value given by

n s, p = δt 1 t 1 + δt 2 t 2 + δnn s

form which the equations of reflection and refraction proceed as normal.

Volumetric Scattering Theory

The Scattering Domain feature is used to model the interaction of rays with a dilute
phase of particles in the medium the rays pass through. Some examples of this include
the following:

• Water droplets, smoke, or dust in air,

• Bubbles in water, and
• Colloidal suspensions of metallic nanoparticles in water.

TYPES OF SCATTERING BEHAVIOR

When a beam of light passes through a cloud of particles, some energy is subtracted
from the forward-propagating beam. Two other things can happen to the subtracted

232 | CHAPTER 3: RAY OPTICS INTERFACES

energy: it is either sent out in a different direction (scattering) or converted into
another form of energy within the particles, most often heat (absorption). The overall
subtraction of energy from the forward-propagating direction is called extinction. By
the principle of conservation of energy, the fundamental relationship between the
different scattering phenomena is

extinction = scattering + absorption (3-27)

PROPERTIES OF SCATTERING PARTICLES

The scattering particles are assumed to be homogeneous isotropic spheres. The space
between them is assumed to be large compared to the electromagnetic wavelength and
particle size. In addition, the particles are also assumed to be oriented randomly, not
in any specific pattern (such as a rectangular array). Thus there is no coherent phase
relationship between different particles, so scattering by many particles can be
understood by first considering a single isolated sphere.

Simplifying assumptions can be made if the particles are very large or very small when
compared to the wavelength of the electromagnetic radiation. When the particles are
optically very small, the simplified Rayleigh scattering model can be used. For particles
much larger than the wavelength, the ratio of extinction cross section to geometrical
cross section approaches an asymptotic limit of 2.

When the particles are neither optically very large nor very small, the more general Mie
scattering theory must be used. Mie theory provides an exact solution for the
interaction of a plane wave with a sphere, but the calculations involved are much more
computationally intensive than either of the asymptotic limits for large or small
particles.

When considering whether the particles are small enough to apply Rayleigh theory,
remember to consider the wavelength both inside and outside the particle; this could
make a big difference if the particles are made of a highly conductive material.

CROSS SECTIONS AND EFFICIENCY FACTORS

Let np (dimensionless) be the absolute refractive index of a spherical particle, and na
(dimensionless) the absolute refractive index of the surrounding medium. For this
analysis, na is assumed to be real (non-absorbing) but np may be complex-valued,

n p = n p, r – in p, i

where np,i > 0 for absorbing particles.

The wavelength in the medium surrounding the particle λ (SI unit: m) is

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 233

 λ0
λ = ------
na

where λ0 (SI unit: m) is the vacuum wavelength.

For optical scattering calculations, it is convenient to work with dimensionless

variables, such as the dimensionless size parameter x,

2πR
x = -----------
λ

where R (SI unit: m) is the radius of the scattering particle.

The size parameter may also be expressed in terms of the wave number in the
surrounding medium k (SI unit: rad/m),

2π
k = ------
λ

Thus,

x = kR

For brevity, in the remainder of this section the relative refractive index of the
scattering particle, that is, the ratio of the particle refractive index to that of the
surroundings, will be denoted n,

np
n ≡ ------
na

although caution should be used because n denotes the absolute index in other
sections of this manual.

Consider an incident beam of light with cross sectional area A0 (SI unit: m2) and
intensity I0 (SI unit: W/m2). This beam illuminates a surface, conferring an amount
of energy equal to I0A0. Now suppose that a single spherical particle obstructs part of
the beam, so that the energy reaching the surface is now I0A1. Then the difference
A0 - A1 is the extinction cross section σext (SI unit: m2) of the particle.

In other words, the extinction cross section of a particle is the surface area on a
detector that, if subtracted, causes the same reduction in total energy at that detector
as the presence of the particle in the beam.

Looking closer at the particle, the extinction or energy removal can be separated into
two phenomena: scattering and absorption. The amount of energy absorbed by the

234 | CHAPTER 3: RAY OPTICS INTERFACES

particle and converted to other forms of energy (predominantly heat) is I0σabs, where
I0σabs (SI unit: m2) is the absorption cross section. The amount of energy that
radiates outward from the particle in directions other than the direction of the incident
beam is I0σsca, where σsca (SI unit: m2) is the scattering cross section.

By the principle of conservation of energy exemplified in Equation 3-27, the three

cross sections follow the relationship

σ ext = σ sca + σ abs (3-28)

For each cross section, it is convenient to define a dimensionless efficiency factor by

dividing by the geometrical cross-sectional area of the particle,

σ ext σ sca σ abs

Q ext = ----------2 Q sca = ----------2 Q abs = ----------2 (3-29)
πR πR πR

where

• Qext is the extinction efficiency,

• Qsca is the scattering efficiency, and
• Qabs is the absorption efficiency.

Like the scattering cross sections, the efficiency factors follow a relationship based on
the energy conservation principle,

Q ext = Q sca + Q abs (3-30)

It is convenient describe the scattering behavior in terms of the dimensionless

quantities x, Qext, Qsca, and Qabs, rather than the particle radius, wavelength, and
cross sections. Another dimensionless variable is sometimes needed: the product of the
size parameter with the relative refractive index is denoted y,

y ≡ nx

RAYLEIGH SCATTERING THEORY

Rayleigh scattering theory (not to be confused with Rayleigh-Gans theory) is a
simplified scattering model for optically small particles. For Rayleigh scattering theory
to be applicable, the particle must be small relative to the wavelength outside of it,

x«1

and also small relative to the wavelength inside it,

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 235

 nx«1

It is possible for the former to be true, but not the latter, if the scattering centers are
small metallic particles with refractive indices that have very large imaginary parts.

The scattering and absorption efficiency factors in the Rayleigh limit are

8 n2 – 1 2
Q sca = --- x 4 ---------------
- (3-31)
3 n2 + 2

2
Q abs = – 4xIm  ---------------
n – 1
- (3-32)
 n 2 + 2

As usual, the sum of these two efficiency factors yields the extinction efficiency factor.

The absorption efficiency factor is thus zero if the refractive index of the scattering
particles is real-valued. However, if the refractive index has any imaginary part, the
absorption efficiency factor tends to dominate because it is only proportional to the
first power of the size parameter x, whereas the scattering efficiency scales with the
fourth power of x; when applying Rayleigh scattering theory, x tends to be very small.

MIE SCATTERING THEORY

The Mie scattering theory represents an exact solution to the scattering of a plane wave
by a homogeneous, isotropic spherical particle. In principle, this exact solution can be
used for any value of the size parameter x. In practice, some expressions for the
efficiency factors can become ill-conditioned or numerically unstable at extremely large
or small x, or if the refractive index has a very large imaginary part.

A complete derivation for the scattering amplitudes will not be given here; instead, the
reader is directed to Ref. 20. The extinction and scattering efficiency factors are
defined as infinite series,

∞
2
Q ext = -----2
x  ( 2l + 1 )Re ( al + bl ) (3-33)
l=1

∞
2
Q sca = -----2
x  ( 2l + 1 ) ( al 2 + bl 2 ) (3-34)
l=1

The absorption efficiency factor if the difference between these two expressions. The
coefficients al and bl are expressions involving the Riccatti–Bessel functions

236 | CHAPTER 3: RAY OPTICS INTERFACES

 ψ l' ( y )ψ l ( x ) – nψ l ( y )ψ l' ( x )
a l = ---------------------------------------------------------------------
ψ l' ( y )ζ l ( x ) – nψ l ( y )ζ l' ( x )
(3-35)
nψ l' ( y )ψ l ( x ) – ψ l ( y )ψ l' ( x )
b l = ---------------------------------------------------------------------
nψ l' ( y )ζ l ( x ) – ψ l ( y )ζ l' ( x )

These functions can be expressed in terms of the spherical Bessel functions of the first
and third kind, or in terms of the corresponding Bessel functions,

πx
ψ l ( x ) = xj l ( x ) = ------J l + 1 ⁄ 2 ( x )
2

ζ l ( x ) = xh l( 2 ) ( x ) = πx
------H l( 2+)1 ⁄ 2 ( x )
2

In Ref. 21, spherical Hankel functions of the first kind h l( 1 ) ( x ) are used.
This is tied to the author’s choice of sign convention when setting up the
electromagnetic wave equation. The convention used in COMSOL is
generally to represent the transient part of an electromagnetic wave as
exp(iωt), whereas Ref. 21 uses exp(-iωt). Ref. 20 uses exp(iωt) and also
arrives at h l( 2 ) ( x ) .

The primes in Equation 3-35 indicate differentiation with respect to the argument of
the function,

dψ l ( y ) dψ l ( y )
ψ l' ( y ) = ----------------- ψ l' ( y ) = -----------------
dy dy
dψ l ( x ) dψ l ( y )
ζ l' ( x ) = ----------------- ζ l' ( y ) = -----------------
dx dy

The general workflow of Mie scattering calculation is to first compute the

Riccatti-Bessel functions ψl and ζl; then the scattering coefficients al and bl; and finally
the scattering efficiency factors Qext and Qsca.

The scattering efficiency factors are computed using the built-in

COMSOL functions mieextinction(n,x) and miescattering(n,x).
The first input argument is the relative refractive index of the scattering
particle and the second input argument is the size parameter. Although
primarily used in ray optics, these functions can be accessed in any model.

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 237

 Efficient evaluation of the efficiency factors from Mie theory means that the scattering
coefficients al and bl from Equation 3-35 must be computed as quickly and robustly
as possible. Although they are infinite series, in practice the terms start to become
negligibly small when the index l is somewhat larger than x, so the series can be taken
to a finite upper bound L; one recommendation for the value of L (Ref. 21) is

L = x + 4x 1 ⁄ 3 + 2

For example, if the size parameter is 1000, the first 1042 terms in the infinite series will
be computed.

Recurrence Relations
One approach to computing the scattering coefficients is via upward recurrence. The
first few Riccatti-Bessel functions are

ψ –1 ( x ) = cos x ψ 0 ( x ) = sin x
ζ –1 ( x ) = cos x – i sin x ζ 0 ( x ) = sin x + i cos x

and the remainder can be obtained via the recurrence relations

2l + 1
ψ l + 1 ( x ) = --------------- ψ l ( x ) – ψ l – 1 ( x )
x
2l + 1
ζ l + 1 ( x ) = --------------- ζ l ( x ) – ζ l – 1 ( x )
x

LARGE CONDUCTING SPHERES: THE EXTINCTION PARADOX

When scattering particles are very large ( x » 1000 ), the extinction efficiency factor
from Mie theory asymptotically approaches a value of 2. In other words, the extinction
cross section of a large particle is double its geometrical cross section.

This observation is sometimes called the extinction paradox because it is contrary to

most day-to-day observations, in which a large object only obstructs the light that
intersects its geometrical cross section.

The apparent contradiction can be resolved (see for example Ref. 20) by noting that
the interpretation of the extinction cross section requires observation of the scattered
light to be made at a very large distance from the scattering particle, where regions of
light and shadow are not so distinct. According to Babinet’s principle, when a certain
amount of light is blocked by the surface of an object, an equal amount of light is
diffracted around the edges of the object. For close-range interactions such as light
illuminating a piece of furniture in a room, it is extremely difficult to distinguish the
diffracted light from the light that continues in the forward scattering direction (θ = 0)

238 | CHAPTER 3: RAY OPTICS INTERFACES

because the diffracted light concentrated in a very narrow cone about the forward
scattering direction. The scattering amplitudes for diffracted light as a function of
scattering angle are

J 1 ( x sin θ )
S 1( d ) ( θ ) = S 2( d ) ( θ ) = x 2 ---------------------------
x sin θ

where J1 is the Bessel function of the first kind; so if x is very large, most of the energy
is contained within a very small angle.

Interference Pattern Theory

The Interference Pattern plot can be used to visualize the fringes resulting from the
interference of two or more rays at a surface. The interference pattern is only valid
when it is plotted over an area with a length scale that is much smaller than the
principal radius of curvature of any incident wavefront. This is a consequence of the
treatment of each ray as a wavefront that subtends a small solid angle. Furthermore,
this plot type requires calculation of the ray intensity and instantaneous phase.

Given the phase Ψ0 of a ray at the point (qx, qy) where the ray intersects a plane, the
phase Ψ at a nearby point (qx + Δx, qy + Δy) is computed as follows. Without loss of
generality assume that the x- and y-axes are oriented so that they are parallel to the
projections of the principal curvature directions onto the cut plane. For a spherical
wavefront, the orientations of these axes may be determined arbitrarily as long as they
are orthogonal. Let ni,x and ni,y denote the x- and y- components, respectively, of the
normalized ray direction vector. Let r1 and r2 denote the principal radii of curvature
in the directions parallel to the x- and y-axes, respectively. For an incident ray with
wave vector magnitude k, the phase at (qx + Δx, qy) is

Ψ ( q x + Δx, q y ) = Ψ ( q x, q y ) + kΔr 1

Where Δr1 is the change in the principal radius of curvature of the wavefront when
moving between the two points.

2 2 2 2
( r 1 + Δr 1 ) = [ r 1 – ( n i, x r 1 ) ] + ( n i, x r 1 + Δx )

Further simplification yields

2 2
Δr 1 = r 1 + 2n i, x r 1 Δx + Δx – r 1

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 239

 This equation yields an indeterminate form in the limit of infinite radii of curvature,
making it unsuitable for computing interference patterns involving plane waves. The
equation can be made more robust for plane waves by applying a Taylor series
approximation about Δx = 0,

2 2 1 2 3
r 1 + 2n i, x r 1 Δx + Δx = r 1 + n i, x Δx + --------- ( 1 – n i2, x )Δx + O ( Δx )
2r 1

Discarding terms of order Δx3 or higher yields

1 2
Δr 1 = n i, x Δx + --------- ( 1 – n i2, x )Δx
2r 1

Similarly, the change in principal radius of curvature due to a small perturbation in the
y direction is

1 2
Δr 2 = n i, y Δy + --------- ( 1 – n i2, y )Δy
2r 2

The rationale for discarding higher-order terms is that they all involve division by
higher powers of r1 and r2. Therefore this series approximation is only valid when the
perturbations in the x and y directions are much smaller than the principal radii of
curvature. To assume otherwise would mean that the incident wavefronts subtend very
large solid angles.

Combining the perturbations in the two principal radii of curvature yields an

expression for the phase at any point (qx + Δx, qy + Δy) on the cut plane,

1 2 1 2
Ψ = Ψ 0 + k n i, x Δx + --------- ( 1 – n i2, x )Δx + n i, y Δy + --------- ( 1 – n i2, y )Δy
2r 1 2r 2

The incident Stokes vector is then converted to an incident electric field using the
relation

n ε
I = --- -----0- ( E ⋅ E∗ )
2 μ0

and the electric field at the intersection of the ray with the cut plane can be used to
compute the electric field at a nearby point using the relation

E ( q x + Δx, q y + Δy ) = E ( q x, q y ) exp [ Ψ ( q x + Δx, q y + Δy ) – Ψ 0 ]

240 | CHAPTER 3: RAY OPTICS INTERFACES

where it is again assumed that the ray subtends a small solid angle so that the intensity
of the incident ray is approximately uniform over the region in which the interference
pattern is plotted. It is then straightforward to compute the sum of the electric fields
of all incident rays and derive the total incident intensity from the electric field norm.

Results Analysis and Visualization in the Ray Optics Modeling chapter.

Monochromatic Aberrations and the Zernike Polynomials

A standard way to quantify monochromatic aberrations is to express the optical path
difference of all incident rays as a linear combination of Zernike polynomials. With the
Ray Optics Module, the Zernike coefficients can be reported in a table using the
Aberration Evaluation feature. Alternatively, a superposition of different Zernike
polynomials can be shown on a unit circle using the Optical Aberration plot.

Several different standards exist for naming, normalizing, and organizing the Zernike
polynomials. The approach used in this section follows the standards published by the
International Organization for Standardization (ISO, Ref. 23) and the American
National Standards Institute (ANSI, Ref. 24).

Each Zernike polynomial Z nm can be expressed as

Z nm = N nm R nm ( ρ )M ( mθ )

where

• N nm is the normalization term,

• R nm is the radial term,
• M(mθ) is the meridional term or azimuthal term,
• ρ is the radial parameter, given by ρ = r/a where r is the distance from the aperture
center and a is the aperture radius, so that 0 ≤ ρ ≤ 1,
• θ is the meridional parameter or azimuthal angle, 0 ≤ θ ≤ 2π,
• the lower index n is a nonnegative integer, n = 0,1,2,…, and
• the upper index m is an integer, m = −n, −n + 2, …, n − 2, n so that n − |m| is
always even.

The normalization term N nm is

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 241

 N nm = ( 2 – δ 0, m ) ( n + 1 )

where δ0,m is the Kronecker delta,

1 i = j
δ i, j = 
0 i≠j

The radial term is given by the equation

0.5 ( n – m )
( – 1 ) s ( n – s )!
R nm ( ρ ) =  -------------------------------------------------------------------------------------------------
s! [ 0.5 ( n + m ) – s ]! [ 0.5 ( n – m ) – s ]!
- ρ n – 2s
s=0

where “!” denotes the factorial operator; for nonnegative integers,

 1 n = 0
n! = 
 1 × 2 × 3 ×…×n n>0

The meridional term is given by the equation

 cos ( mθ ) m≥0
M ( mθ ) = 
 sin ( m θ ) m<0

The Zernike polynomials thus defined are normalized Zernike polynomials. They are
orthogonal in the sense that any pair of Zernike polynomials satisfy the equation

1 2π

 ρ dρ  Znm Zn'm' dθ = πδ n, n' δ m, m'

0 0

The normalized Zernike polynomials up to fifth order, along with their common
names, are given in Table 3-4.

TABLE 3-4: ZERNIKE POLYNOMIALS

TERM EXPRESSION COMMON NAME

Z 00 1 Piston

Z 1–1 2ρ sin ( θ ) Vertical tilt

Z 11 2ρ cos ( θ ) Horizontal tilt

242 | CHAPTER 3: RAY OPTICS INTERFACES

TABLE 3-4: ZERNIKE POLYNOMIALS

TERM EXPRESSION COMMON NAME

Z 2– 2 6ρ 2 sin ( 2θ ) Oblique astigmatism

Z 20 2 Defocus
3 ( 2ρ – 1 )

Z 22 6ρ 2 cos ( 2θ ) Astigmatism

Z 3– 3 8ρ 3 sin ( 3θ ) Oblique trefoil

Z 3– 1 8 ( 3ρ 3 – 2ρ ) sin ( θ ) Vertical coma

Z 31 8 ( 3ρ 3 – 2ρ ) cos ( θ ) Horizontal coma

Z 33 8ρ 3 cos ( 3θ ) Horizontal trefoil

Z 4– 4 10ρ 4 sin ( 4θ ) Oblique quatrefoil

Z 4– 2 10 ( 4ρ 4 – 3ρ 2 ) sin ( 2θ ) Oblique secondary astigmatism

Z 40 5 ( 6ρ 4 – 6ρ 2 + 1 ) Spherical aberration

Z 42 10 ( 4ρ 4 – 3ρ 2 ) cos ( 2θ ) Secondary astigmatism

Z 44 10ρ 4 cos ( 4θ ) Horizontal quatrefoil

Z 5– 5 12ρ 5 sin ( 5θ )

Z 5– 3 12 ( 5ρ 5 – 4ρ 3 ) sin ( 3θ )

Z 5– 1 12 ( 10ρ 5 – 12ρ 3 + 3ρ ) sin ( θ )

Z 51 12 ( 10ρ 5 – 12ρ 3 + 3ρ ) cos ( θ )

Z 53 12 ( 5ρ 5 – 4ρ 3 ) cos ( 3θ )

Z 55 12ρ 5 cos ( 5θ )

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 243

 Figure 3-14: Zernike polynomials on the unit circle.

Accumulator Theory: Domains

The Accumulator (Domain) node is used to transfer information from rays to the
domains they occupy or pass through. Each Accumulator defines a variable, called the
accumulated variable, in the selected domains. The accumulated variable is discretized
using constant shape functions, so its value is uniform over every mesh element and
may be discontinuous between adjacent mesh elements.

The name of the accumulated variable is specified in the Accumulated variable name
field in the Accumulator Settings section of the settings window. The default variable
name, rpd, will be used in the remainder of this section when referring to the
accumulated variable.

244 | CHAPTER 3: RAY OPTICS INTERFACES

ACCUMULATOR TYPE
The options in the Accumulator type list are Density and Count. If Density is selected, the
source term is divided by the area or volume of the mesh element when calculating
each ray’s contribution to the accumulated variable. If Count is selected, no division by
the area or volume of the mesh element occurs.

The equations in the following section are valid for the Density type. The
corresponding value of the accumulated variable for the Count type is

rpd count = rpd density × V

where V is the mesh element volume (in 3D) or area (in 2D).

ACCUMULATION OVER ELEMENTS

When Elements is selected from the Accumulate over list, the value of the accumulated
variable in a mesh element is the sum of the source terms Ri evaluated for all rays in
that mesh element:

N
1
rpd = ----
V  Ri
i=1

where N is the total number of rays in the element and V is the area or volume of the
mesh element. In other words, the contribution of each ray to the accumulated
variable is distributed uniformly over the mesh element the ray is in, regardless of the
ray’s exact position within the element.

If Elements and time is selected from the Accumulate over list, then the sum of the
source terms for rays in the mesh element is used to define the time derivative of the
accumulated variable, rather than its instantaneous value:

N
( rpd )
d---------------- 1
dt
= ----
V  Ri
i=1

Thus the value of the accumulated variable depends on the time history of rays in the
mesh element, instead of the instantaneous positions of the rays. As each ray
propagates, it will leave behind a trail based on its contributions to the accumulated
variables in mesh elements it has traversed. The algorithm for accumulating over time
takes into account the fraction of a time step taken by the solver that the ray spends in
each mesh element, even if it crosses between elements during the time step.

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 245

 Accumulator Theory: Boundaries
The Accumulator (Boundary) feature transfers information from rays to the
boundaries they hit or pass through. Each Accumulator defines a variable, called the
accumulated variable, on the selected boundaries. The accumulated variable is
discretized using constant shape functions, so its value if uniform over every mesh
element and may be discontinuous between adjacent mesh elements.

The name of the accumulated variable is specified in the Accumulated variable name
field in the Accumulator Settings section of the settings window. The default variable
name, rpb, will be used in the remainder of this section when referring to the
accumulated variable.

The options in the Accumulator type list are Density and Count. If Density is selected, the
source term is divided by the surface area or length of the boundary mesh element
when calculating each ray’s contribution to the accumulated variable. If Count is
selected, no division by the surface area or length of the boundary element occurs.

The equations in the following section are valid for the Density type. The
corresponding value of the accumulated variable for the Count type is

rpb count = rpb density × V

where V is the boundary element surface area (in 3D) or length (in 2D).

When Rays in boundary elements is selected from the Accumulate over list, the
accumulated variable in a boundary element gets incremented by the source term R
whenever a ray freezes or sticks to the boundary:

R
rpb new = rpb + ----
V

where division by the mesh element area or length occurs because the accumulator is
assumed to be of type Density. Thus the source term evaluated for an incident ray is
uniformly distributed over the boundary element it freezes or sticks to.

If instead Ray-wall interactions is selected from the Accumulate over list, then the
accumulated variable gets incremented regardless of what type of ray-wall interaction
occurs. Thus, it is possible for the same ray to increment the accumulated variable in
many different boundary elements, or even in the same element multiple times.

246 | CHAPTER 3: RAY OPTICS INTERFACES

BUILT-IN GLOBAL VARIABLES
By default, the boundary Accumulator defines the following global variables:
TABLE 3-5: BUILT-IN GLOBAL VARIABLES FOR BOUNDARY ACCUMULATORS

NAME EXPRESSION

<scope>.<name>_ave Average of accumulated variable

<scope>.<name>_int Integral of accumulated variable
<scope>.<name>_max Maximum of accumulated variable
<scope>.<name>_min Minimum of accumulated variable
<scope>.<name>_sum Sum of accumulated variable over elements

Here, <scope> includes the physics interface name and the names the Accumulator
and parent feature. For example, the average of the accumulated variable over a
boundary may be called gop.wall1.bacc1.rpb_ave, where gop is the name of the
Geometrical Optics interface, wall1 is the name of the parent Wall node, bacc1 is the
name of the Accumulator node, and rpb is the accumulated variable name. These
variables are all available in the Add/Replace Expression menus during results evaluation.

These global variables are computed by defining a set of nonlocal couplings on the
selection of the parent physics feature, such as the Wall feature to which the
Accumulators are added. The following expressions for the global variables are used.
TABLE 3-6: BUILT-IN GLOBAL VARIABLE DEFINITIONS FOR BOUNDARY ACCUMULATORS

NAME EXPRESSION

<scope>.<name>_ave <wscope>.aveop(<scope>.<name>)
<scope>.<name>_int <wscope>.intop(<scope>.<name>)
<scope>.<name>_max <wscope>.maxop(<scope>.<name>)
<scope>.<name>_min <wscope>.minop(<scope>.<name>)
<scope>.<name>_sum <wscope>.intop(<scope>.<name>/<scope>.meshVol)

Here, <wscope> is the scope of the parent boundary feature, for example, gop.wall1.

References for the Geometrical Optics Interface

1. L.D. Landau and E.M. Lifshitz, The Classical Theory of Fields, 4th ed.,
Butterworth-Heinemann, Oxford, 1975.

2. B. Edlén, “The dispersion of standard air”, J. Opt. Soc. Am,, vol. 2, pp. 339–344,
1953.

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 247

 3. W.J. Tropf, “Temperature-dependent refractive index models for BaF2, CaF2,
MgF2, SrF2, LiF, NaF, KCL, ZnS, and ZnSe”, Opt. Eng., vol. 34, pp. 1369–1373,
1995.

4. B.J. Frey, D.B. Leviton, and T.J. Madison, “Temperature-dependent refractive

index of silicon and germanium”, Optomechanical Technologies for Astronomy, Proc.
SPIE, vol. 6273, 2006.

5. H.J. Hoffmann, W.W. Jochs, and G. Westenberger, “Dispersion formula for the
thermo-optic coefficient of optical glass”, Proc. SPIE, vol. 1327, 1990.

6. R. Ditteon, Modern Geometrical Optics, Wiley, 1998.

7. S.M. Stewart and R.B. Johnson, Blackbody Radiation: A History of Thermal

Radiation Computational Aids and Numerical Methods, CRC Press, 2017.

8. D.C. O’Shea, Elements of Modern Optical Design, Wiley, 1985.

9. M. Born and E. Wolf, Principles of Optics, 7th ed., Cambridge, 1999.

10. P. Yeh, Optical Waves in Layered Media, Wiley, 2005.

11. O.N. Stavroudis, The Optics of Rays, Wavefronts, and Caustics, Academic, 1972.

12. B.E.A. Saleh and M.C.Teich, Fundamentals of Photonics, Wiley, 1991.

13. R.M. More and K. Kosaka, “Wave-front curvature in geometrical optics”, Phys.
Rev. E, vol. 57, no. 5, pp. 6127–6134, 1998.

14. P.C.Y. Chang, J.G. Walker, and K.I. Hopcraft, “Ray tracing in absorbing media”,
J. Quant. Spectrosc. Radiat., vol. 96, pp. 327–341, 2005.

15. W.A. Shurcliff, Polarized Light: Production and Use, Harvard, 1962.

16. Y.M. Ahmed, M. Jongewaard, M. Li, and E.R. Blatchley III, “Ray Tracing for
Fluence Rate Simulations in Ultraviolet Photoreactors”, Environ. Sci. Technol.,
vol. 52, no. 8, pp. 4738-4745, 2018.

17. O. S. Heavens, Optical Properties of Thin Solid Films, Dover, 1991.

18. H. Anders, Thin Films in Optics, Focal Press, 1967.

19. Y. Shuai, X. Xin-Lin, T. He-Ping, “Radiation performance of dish solar

concentrator/cavity receiver systems”, Sol. Energy, vol. 82, pp. 13–21, 2008.

20. H. C. van de Hulst, Light Scattering by Small Particles, Dover, New York, 1981.

248 | CHAPTER 3: RAY OPTICS INTERFACES

21. C. F. Bohren and D. R. Huffman, Absorption and Scattering of Light by Small
Particles, Wiley, Weinheim, 2004.

22. D. Hestroffer and C. Magnan, “Wavelength dependency of the Solar limb

darkening”, Astron. Astrophys., vol. 333, pp. 338–342, 1998.

23. ISO 24157:2008: Ophthalmic optics and instruments — Reporting aberrations

of the human eye, International Organization for Standardization, Geneva,
Switzerland. Amendment 1, ibid., 2019.

24. ANSI Z80.28-2017: American National Standard for Ophthalmics — Methods

of Reporting Optical Aberrations of Eyes. American National Standards Institute,
Alexandria, VA.

THEORY FOR THE GEOMETRICAL OPTICS INTERFACE | 249

250 | CHAPTER 3: RAY OPTICS INTERFACES
 4

Multiphysics Interfaces and Couplings

The Ray Optics Module contains predefined multiphysics interfaces and couplings
to facilitate easy setup of models of ray propagation with significant attenuation and
heating. The Ray Heating physics interface described in this chapter is found under
the Optics>Ray Optics branch ( ) when adding a physics interface.

In this chapter:

• The Ray Heating Interface

• Theory for the Ray Heating Interface

251
 The Ray Heating Interface
The Ray Heating interface ( ) combines the Geometrical Optics interface with the
Heat Transfer in Solids interface. The Ray Heat Source multiphysics feature is added
automatically. The Ray Heating interface is used to model electromagnetic wave
propagation in optically large systems in which the refractive index is temperature
dependent. The energy lost due to the attenuation of rays in an absorbing medium
creates a heat source that is included in the temperature computation.

When a predefined Ray Heating interface is added from the Optics>Ray Optics
branch ( ) of the Model Wizard or Add Physics windows, Geometrical Optics and Heat
Transfer in Solids interfaces are added to the Model Builder. A Multiphysics node is also
added, which automatically includes the multiphysics coupling feature Ray Heat Source.

On the Constituent Physics Interfaces

The Geometrical Optics interface is used to model electromagnetic wave propagation
when the wavelength is much smaller than the smallest geometric entity in the model.
It includes built-in tools for computing the intensity, phase, and optical path length of
rays. The rays can be polarized, unpolarized, or partially polarized. Frequency
distributions and ray propagation in absorbing media are also supported.

The Heat Transfer in Solids interface is used to model heat transfer by conduction,
convection, and radiation. A Heat Transfer in Solids model is active by default on all
domains and a Heat Transfer in Fluids model is added with an empty selection. All
functionality for including other domain types, such as a fluid domain, is also available.
The temperature equation defined in solid domains corresponds to the differential
form of the Fourier’s law that may contain additional contributions like heat sources.

SETTINGS FOR PHYSICS INTERFACES AND COUPLING FEATURES

When physics interfaces are added using the predefined couplings, for example Ray
Heating, specific settings are included with the physics interfaces and the coupling
features.

However, if physics interfaces are added one at a time, followed by the coupling
features, these modified settings are not automatically included.

252 | CHAPTER 4: MULTIPHYSICS INTERFACES AND COUPLINGS

For example, if single Geometrical Optics and Heat Transfer in Solids interfaces are
added, COMSOL adds an empty Multiphysics node. You can choose from the available
coupling features but the modified settings are not included.

Coupling features are available from the context menu (right-click the
Multiphysics node) or from the Physics toolbar, Multiphysics menu.

TABLE 4-1: MODIFIED SETTINGS FOR A RAY HEATING INTERFACE

PHYSICS INTERFACE OR MODIFIED SETTINGS (IF ANY)

COUPLING FEATURE

Geometrical Optics Compute intensity and power is selected from the

Intensity computation list in the physics interface Ray
Properties section.
Heat Transfer in Solids No changes.
Ray Heat Source The Domain Selection is the same as that of the
participating physics interfaces.
The corresponding Geometrical Optics and Heat
Transfer in Solids interfaces are preselected in the Ray
Heat Source section.

PHYSICS INTERFACES AND COUPLING FEATURES

Use the online help in COMSOL Multiphysics to locate and search all the
documentation. All these links also work directly in COMSOL
Multiphysics when using the Help system.

Coupling Features
The Ray Heat Source coupling feature node is described in this section.

Physics Interface Features

Physics nodes are available from the Physics ribbon toolbar (Windows users), Physics
context menu (Mac or Linux users), or right-click to access the context menu (all
users).

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

THE RAY HEATING INTERFACE | 253

 • The available physics features for The Geometrical Optics Interface are listed in the
section List of Geometrical Optics Interface Physics Features.
• The available physics features for The Heat Transfer Interfaces are listed in the
sections Domain Features, Boundary Features, and Edge and Point Features in the
COMSOL Multiphysics Reference Manual.

Ray Heat Source

Use the Ray Heat Source multiphysics coupling ( ) to compute the heat source
resulting from the attenuation of rays in an absorbing medium. This heat source is
defined to a dependent variable that is defined on all domain mesh elements in the
selection of the Ray Heat Source node. The ray power should be solved for in the
coupled Geometrical Optics interface.

SETTINGS
The Label is the default multiphysics coupling name.

The Name is used primarily as a scope prefix for variables defined by the coupling node.
Refer to such variables in expressions using the pattern <name>.<variable_name>. In
order to distinguish between variables belonging to different coupling nodes or physics
interfaces, the name string must be unique. Only letters, numbers, and underscores (_)
are permitted in the Name field. The first character must be a letter.

The default Name (for the first multiphysics coupling in the model) is rhs1.

RAY HEAT SOURCE

This section defines the physics involved in the multiphysics coupling. By default, the
applicable physics interface is selected in the Source and Destination lists.

You can also select None from either list to uncouple the node from a physics interface.
If the physics interface is removed from the Model Builder then the applicable list
defaults to None as there is nothing to couple to.

If a physics interface is deleted and then added to the model again, and in
order to reestablish the coupling, you need to choose the physics interface
again from the Source or Destination lists. This is applicable to all
multiphysics coupling nodes that would normally default to the once
present physics interface. See Multiphysics Modeling Workflow in the
COMSOL Multiphysics Reference Manual.

254 | CHAPTER 4: MULTIPHYSICS INTERFACES AND COUPLINGS

Theory for the Ray Heating Interface
The Ray Heating Interface theory is described in this section:

• Unidirectional and Bidirectional Couplings

• Coupled Heat Transfer and Ray Tracing Equations
• Heat Source Calculation

Unidirectional and Bidirectional Couplings

The Ray Heating multiphysics interface combines geometrical optics with heat transfer
to compute the heat source that is generated as rays propagate through an absorbing
medium. The Ray Heating interface can either be used to set up a unidirectional or
bidirectional coupling between the constituent physics interfaces.

If the medium properties are not temperature dependent and the domains do not
undergo significant thermal expansion, then a unidirectional coupling is appropriate.
The heat source generated by the attenuation of rays can be included in the heat
transfer simulation by first computing the ray trajectories using a Ray Tracing study
step, then computing the temperature using a Stationary study step. Both study steps
are described in the COMSOL Multiphysics Reference Manual.

If the refractive index of the medium is temperature dependent, a more accurate

approach is to set up a bidirectional coupling in which the attenuation of ray power
affects the temperature distribution, which in turn affects the ray trajectories. The two
physics can be solved for using an iterative solver loop consisting of a time-dependent
solver for the ray trajectories and a stationary solver for the temperature. Such a solver
sequence can be set up automatically using the Bidirectionally Coupled Ray Tracing
study step, also described in the COMSOL Multiphysics Reference Manual.

Coupled Heat Transfer and Ray Tracing Equations

Under static conditions, the heat equation can be written as

ρC p u ⋅ ∇T + ∇ ⋅ q = Q

q = – k∇T

THEORY FOR THE RAY HEATING INTERFACE | 255

 where

• ρ (SI unit: kg/m3) is the density,

• Cp (SI unit: J/(kg·K)) is the heat capacity at constant pressure,
• T (SI unit: K) is the temperature,
• Q (SI unit: W/m3) is the heat source, and
• k (SI unit: W/(m·K)) is the thermal conductivity.

The propagation of each ray can be modeled using a set of coupled first-order
equations:

------- = – ∂ω
dk dq- ∂ω
------- ------ = -------
dt ∂q dt ∂k

where k (SI unit: rad/m) is the wave vector, ω (SI unit: rad/s) is the angular
frequency, t (SI unit: s) is time, and q (SI unit: m) is the ray position vector. In an
isotropic medium, the frequency and wave vector are related by the expression

ck
ω = ------------
n( q)

where c = 299,792,458 m/s is the speed of light in a vacuum and n (dimensionless) is

the refractive index. If the medium is absorbing, then the intensity of the ray decreases
and it deposits some power in the domain as it passes through. This is described in
greater detail in the section Attenuation Within Domains in Theory for the
Geometrical Optics Interface.

A bidirectional coupling between the Geometrical Optics and Heat Transfer interfaces
may be required for any of the following reasons:

• The refractive index is temperature dependent.

• The medium undergoes thermal expansion and the refractive index is strain
dependent.
• The medium undergoes thermal expansion, causing rays be reflected or refracted at
different angles when entering or leaving the medium.

When none of these conditions is met, the Geometrical Optics interface affects the
Heat Transfer interface, and a unidirectional coupling is sufficient.

256 | CHAPTER 4: MULTIPHYSICS INTERFACES AND COUPLINGS

Heat Source Calculation
Consider a ray j transferring an amount of power Qj through a domain. During a small
time interval Δt, the ray propagates a short distance Δs from initial position qj to
position qj+Δqj and its power decreases by an amount ΔQ. Then conservation of
energy dictates that the deposited ray power in the domain should increase by ΔQ:

N
∂Q j
Q src ( r, t + Δt ) ≈ Q src ( r, t ) –  --------
∂t
- Δtδ ( r – q j )
j=1

where δ is the Dirac delta function. In the limit as this time derivative becomes
arbitrarily small, this expression can be rewritten as

N
dQ src ∂Q j
--------------
dt
- = –
 --------
∂t
- δ ( r – qj )
j=1

In this expression, every ray is treated as an infinitesimally small point source. This can
lead to numerical issues because the heat source becomes infinitely large at the location
of each point source. In the following section an alternative expression for the domain
heat source is given.

CREATING A BOUNDED HEAT SOURCE TERM

The Ray Heat Source node defines a variable for the contribution to the heat source
by rays in each mesh element. This variable uses constant shape functions, so it may be
discontinuous across boundaries between elements. For a mesh element j with volume
Vj the average heat source Qsrc,j changes according to the expression

N
dQ src 1 ∂Q j
- = – -----
--------------
dt Vj  --------
∂t 
- δ ( r – q j ) dV
j=1

The integral on the right-hand side is a volume integral over element j. The resulting
time derivative of the heat source is the average value over the mesh element, which
may be written more concisely as

Nj
dQ src, j 1 ∂Q j
------------------
dt
- = – -----
Vj  --------
∂t
-
i=1

where the sum is taken over all rays that are within mesh element j.

THEORY FOR THE RAY HEATING INTERFACE | 257

258 | CHAPTER 4: MULTIPHYSICS INTERFACES AND COUPLINGS
 5

Glossary

This Glossary of Terms contains modeling terms in a ray optics context. For
mathematical terms as well as geometry and CAD terms specific to the COMSOL
Multiphysics® software and documentation, see the glossary in the COMSOL
Multiphysics Reference Manual. For references to more information about a term,
see the index.

259
 Glossary of Terms
actinometric quantity A physical quantity that describes radiative transfer in terms of
the number of photons, rather than the total mount of energy.

blackbody, ideal A theoretical surface that absorbs all radiation that is incident upon
it, regardless of wavelength or surface temperature.

blackbody radiation Diffuse emission of radiation from the surface of an ideal

blackbody.

coherence The degree to which two rays interfere with each other, or to which a ray
may interfere with itself when subjected to a time delay.

cross grating A type of diffraction grating with periodicity in two directions.

diffraction grating A periodic structure from which transmitted and reflected radiation
can only propagate in specific directions governed by the frequency of radiation and
the size of the unit cell.

diffraction order In a diffraction grating, an integer indicating the number of

wavelengths between waves emanating from adjacent unit cells before they
constructively interfere.

diffuse emission Emission of radiation from a surface in a random direction, following

a probability distribution based on Lambert’s cosine law.

diffuse reflection Reflection at a surface in a random direction, following a probability

distribution based on Lambert's cosine law.

emissivity, spectral The ratio of the amount of radiation per unit surface area, per unit
wavelength or frequency, released by an object, compared to an ideal blackbody
radiation source of the same size and shape.

Fresnel equations A set of equations for computing the reflection coefficients and
transmission coefficients for rays as they cross a boundary between two media with
different refractive indices.

260 | CHAPTER 5: GLOSSARY

Gaussian beam A beam of electromagnetic radiation in which the electric field
amplitude follows a Gaussian distribution. In the ideal case, this is the TEM00 mode
solution to the electromagnetic wave equation in free space.

geometrical optics The analysis of electromagnetic wave propagation through

electrically large domains by treating the electromagnetic waves as rays.

grating See diffraction grating and cross grating.

graybody radiation Diffuse emission from a surface at which the spectral emissivity is
constant, independent of wavelength or temperature. Blackbody radiation is the
special case for which this constant value of the spectral emissivity is unity.

intensity A measure of the energy density of an electromagnetic wave; for a plane

polarized wave, the intensity is equal to the magnitude of the time-averaged Poynting
vector.

internal transmittance A material property indicating the fraction of electromagnetic

energy that is transmitted through a material sample of a given thickness, neglecting
Fresnel losses at the surfaces of the sample. Can be used as an alternative means of
specifying the imaginary part of the refractive index.

marginal ray The ray that is released from the center of the object at the maximum
aperture. This ray will pass through the edge of the aperture stop.

monochromatic light Light of uniform frequency or wavelength.

Mueller matrix A 4-by-4 matrix which can be multiplied by the Stokes vector to apply
the effects of various optical components to a ray.

optical dispersion model A function giving the refractive index as a function of

wavelength. The coefficients for any given model will usually be determined
empirically.

optical axis For simply rotationally symmetric optics, this is the axis of rotational
symmetry. The optical axis will pass through the center of curvature.

optical path length The integral of refractive index along the ray trajectory. Over a
given time interval, the optical path length of a ray is equal to the distance the ray
would have propagated in a vacuum.

GLOSSARY OF TERMS | 261

 part IA subsequence of geometry instructions, usually comprising several geometry
primitives, that can be added repeatedly to a COMSOL geometry sequence to avoid
repetitive steps in geometry setup.

phase The quantity that indicates the variation of the instantaneous electric field with
respect to time and position.

plane of incidence For nonnormal incidence of a ray at a surface, the unique plane
containing the surface normal and the incident ray direction vector. For normal
incidence, any plane containing the surface normal may be selected.

photon The quantum of the electromagnetic field.

polarization A description of the orientation of the electric field for a polarized or

partially polarized ray, and for the way in which this electric field varies as the ray
propagates.

polarization, circular The state of polarization in which the electric field amplitude has
two orthogonal components of equal magnitude, with a phase shift of π/2 between
them, such that the electric field vector traces a helical pattern as the ray propagates.

polarization, linear The state of polarization in which the electric field only oscillates
within a single plane.

power The energy transferred per unit time.

Poynting vector The vector representing the directional energy flux of an

electromagnetic field.

primary ray A ray that is released into a model independently of the trajectories of
other rays. Primary rays are created by release features.

principal curvature direction A unit vector which, together with the normal to a
surface, defines a plane in which one of the principal radii of curvature of a surface is
defined.

polychromatic light Light with a distribution of frequencies or wavelengths.

principal curvature The reciprocal of one of the principal radii of curvature of a

surface.

262 | CHAPTER 5: GLOSSARY

principal radii of curvature The maximum and minimum values of the radius of
curvature of curves that can be obtained via the intersection of a surface with a plane.

quasi-monochromatic light Light with a distribution of frequencies in which the

changes in free-space wavelength are much smaller than the average free-space
wavelength.

radiant exitance, spectral The amount of electromagnetic energy emitted from a

surface per unit time, per unit surface area, per unit wavelength or frequency.

radiant exitance, total The amount of electromagnetic energy emitted from a surface
per unit time, per unit surface area.

radiometric quantity A physical quantity that describes radiative transfer in terms of

the amount of energy, including such units as watts, watts per square meter, watts per
steradian, and watts per square meter per unit wavelength.

reflection coefficient A dimensionless coefficient that is used to compute the intensity

of reflected rays. At boundaries between media, reflection coefficients are computed
using the Fresnel equations.

refractive index The ratio of the speed of light in a vacuum to the speed of light in a
medium. This is also known as the absolute refractive index.

refractive index, relative The ratio of the absolute refractive index to the refractive
index of air.

release feature A feature that releases a predetermined number of rays and assigns
initial values to dependent variables associated with the rays.

secondary ray A ray that is produced by the interaction of existing rays with a domain
or boundary feature. The degrees of freedom for all secondary rays are preallocated
when beginning to run a study.

Snell’s law The set of equations that indicates the directions in which reflected and
refracted rays will propagate when a ray interacts with a boundary between two media.

specular reflection Idealized reflection at a smooth surface in which the reflected ray
lies within the plane of incidence and the angles of incidence and reflection are equal.

GLOSSARY OF TERMS | 263

 spot diagram A type of plot for visualizing the focusing capability of an optical system.
It displays the intersection points of rays with a surface, which is often the image plane
of the optical system.

Stokes parameters The four Stokes parameters completely characterize the intensity
and polarization of a fully polarized, partially polarized, or unpolarized ray. The Stokes
parameters can be interpreted as indicators of the ray intensity that would be measured
when sending a ray through various arrangements of polarizers and wave retarders.

thermo-optic dispersion model A function giving the change in refractive index as a

function of wavelength and temperature.

transmission coefficient A dimensionless coefficient that is used to compute the

intensity of refracted rays at boundaries between media. Transmission coefficients are
computed using the Fresnel equations.

wave vector A vector in the direction of ray propagation, which indicates the spatial
dependence of the instantaneous electric field for coherent radiation.

wave number The magnitude of the wave vector, equal to ω/c, where ω is the angular
frequency and c is the speed of light in the medium.

wave retarder An optical component that applies a phase shift to radiation of a specific
polarization with respect to radiation of an orthogonal polarization.

Zernike polynomial A type of polynomial basis that is orthonormal on the unit circle
in 2D. Each term comprises a radial component and an azimuthal component, which
are multiplied together. The Zernike polynomial basis is used to report optical path
differences when interpreting monochromatic optical aberrations.

264 | CHAPTER 5: GLOSSARY

I n d e x
A accumulator (node), boundaries 154 intensity computation 37, 72, 193–194
accumulator (node), domains 153 interference 73
accuracy order, of time stepping 75 internet resources 11
Application Libraries window 13
K knowledge base, COMSOL 15
application library examples 31, 33, 35–
36, 38–39, 44–46, 52, 71, 85, 126, 137, L latitude 127

140 linear polarizer (node) 149

auxiliary dependent variable 88, 94 linear wave retarder (node) 150

auxiliary dependent variable (node) 157 longitude 127

B blackbody radiation (node) 129 M material discontinuity (node) 84

medium properties (node) 79
C caustic surface 197
mirror (node) 95
circular wave retarder (node) 151
MPH-files 13
common settings 11
Mueller matrix (node) 151
count reflections 49, 74
multiphysics
cross grating (node) 146
ray heat source 254
cross sections (optical scattering) 234
multiscale electromagnetics modeling 18,
D deposited ray power (node), boundaries 35, 135, 137
153
N nodes, common settings 11
deposited ray power (node), domains
nonlocal accumulator (node) 156
152
nonlocal couplings 60
diffraction gratings 33, 39, 140, 225
diffraction order (node) 145, 148 O optical dispersion model

documentation 12 Cauchy 80, 167

Conrady 80, 167
E Edlén air model 71, 81, 166
Herzberger 80, 167
efficiency factor (optical scattering) 235
Schott 80, 167
emailing COMSOL 14
Schott extended 80, 167
extinction paradox 238
Sellmeier 80, 167
F fluence rate calculation (node) 152 Sellmeier modified, type 1 80, 167
fluence rate calculation, theory 211 Sellmeier modified, type 2 80, 168

G gaussian beam (node) 120 temperature dependent Sellmeier 43,

graded-index media 18–19, 28, 72 80, 168

grating (node) 140 optical path length 51, 74, 95, 135

I ideal depolarizer (node) 149 P phase 73

illuminated surface (node) 124 photometric data import (node) 119

INDEX| 265
 physics interfaces, common settings 11 total internal reflection 30
plane of incidence 197 total number of rays 58
polarization 73, 149, 193 transmission probability 58
of released rays 103, 123
W wall (node) 89
principal radii of curvature 196
wall condition
R random number generation 76, 95, 135 diffuse scattering 76, 90
ray continuity (node) 157 disappear 90
ray heat source (node), multiphysics 254 freeze 90
ray index 57 general reflection 90
ray position 77 isotropic scattering 90
ray properties (node) 78 mixed diffuse and specular reflection
ray termination (node) 130 90
ray termination, theory 212 pass through 90
refractive index of exterior domains 28, specular reflection 90
71 stick 90
release (node) 97 wave vector 77
release from boundary (node) wavelength distribution of released rays
ray optics 106 68
release from data file (node) 116 websites, COMSOL 15
release from edge (node) 111
Z Zernike polynomials 51, 241
release from electric field (node) 135
release from far-field radiation pattern
(node) 137
release from grid (node) 112
release from grid on axis (node) 116
release from point (node) 112
release from point on axis (node) 112
release from symmetry axis (node) 110

S secondary rays 69
solar radiation (node) 126
spot diagram (plot) 48
standard settings 11
store ray status data 74
sun position 128

T technical support, COMSOL 14

thin dielectric film (node) 89
time zone 127

266 | I N D E X