Yee lattice

From AbInitio

(Difference between revisions)
Jump to: navigation, search
Revision as of 18:48, 22 October 2005 (edit)
Stevenj (Talk | contribs)

← Previous diff
Current revision (14:41, 18 April 2017) (edit)
Ardavan (Talk | contribs)

 
Line 1: Line 1:
{{Meep}} {{Meep}}
-[[Image:Yee-cube.png|250px|center|thumb|Illustration of Yee lattice in 3d for a single grid voxel.]]+[[Image:Yee-cube.png|400px|center|thumb|Illustration of Yee lattice in 3d for a single grid voxel.]]
-In order to discretize Maxwell's equations with second-order accuracy (for homogeneous regions), FDTD methods ''store different field components at different grid locations''. This discretization is known as a '''Yee lattice'''.+In order to discretize Maxwell's equations with second-order accuracy (for homogeneous regions where there no discontinuous material boundaries), FDTD methods ''store different field components for different grid locations''. This discretization is known as a '''Yee lattice'''.
-The form of the Yee lattice in 3d is shown in the illustration here for a single cubic grid voxel (<math>\Delta x \times \Delta x \times \Delta x</math>). The basic idea is that the three components of '''E''' are stored at the ''edges'' of the cube in the corresponding directions, while the components of '''H''' are stored at the ''faces'' of the cube.+The form of the Yee lattice in 3d is shown in the illustration above for a single cubic grid voxel (<math>\Delta x \times \Delta x \times \Delta x</math>). The basic idea is that the three components of '''E''' are stored for the ''edges'' of the cube in the corresponding directions, while the components of '''H''' are stored for the ''faces'' of the cube.
More precisely, let a coordinate <math>(i,j,k)</math> in the grid correspond to: More precisely, let a coordinate <math>(i,j,k)</math> in the grid correspond to:
-:<math>\mathbf{x} = (i \hat\mathbf{e}_1 + j \hat\mathbf{e}_2 + k \hat\mathbf{e}_3) \Delta x</math>.+:<math>\mathbf{x} = (i \hat\mathbf{e}_1 + j \hat\mathbf{e}_2 + k \hat\mathbf{e}_3) \Delta x</math>,
-Then, the <math>\ell</math><sup>th</sup> component of <math>\mathbf{E}</math> or <math>\mathbf{D}</math> (or <math>\mathbf{P}</math>) is stored at the locations +where <math>\hat\mathbf{e}_k</math> denotes the unit vector in the ''k''-th coordinate direction. Then, the <math>\ell</math><sup>th</sup> component of <math>\mathbf{E}</math> or <math>\mathbf{D}</math> (or <math>\mathbf{P}</math>) is stored for the locations
:<math>(i,j,k)+ \frac{1}{2} \hat\mathbf{e}_\ell \Delta x</math>. :<math>(i,j,k)+ \frac{1}{2} \hat\mathbf{e}_\ell \Delta x</math>.
-The <math>\ell</math><sup>th</sup> component of <math>\mathbf{H}</math>, on the other hand, is stored at the locations+The <math>\ell</math><sup>th</sup> component of <math>\mathbf{H}</math>, on the other hand, is stored for the locations
:<math>(i+\frac{1}{2},j+\frac{1}{2},k+\frac{1}{2})-\frac{1}{2} \hat\mathbf{e}_\ell \Delta x</math>. :<math>(i+\frac{1}{2},j+\frac{1}{2},k+\frac{1}{2})-\frac{1}{2} \hat\mathbf{e}_\ell \Delta x</math>.
-In two dimensions, the idea is similar except that we set <math>\hat\mathbf{e}_3=0</math>. The 2d Yee lattices for the TE polarizations ('''E''' in the ''xy'' plane and '''H''' in the ''z'' direction) is shown in the figure at right.+In two dimensions, the idea is similar except that we set <math>\hat\mathbf{e}_3=0</math>. The 2d Yee lattice for the <i>P</i>-polarization ('''E''' in the ''xy'' plane and '''H''' in the ''z'' direction) is shown in the figure below.
-[[Image:Yee-te.png|thumb|right|150px|Yee lattice in 2d for the TE polarization.]]+[[Image:Yee-te.png|thumb|center|250px|Yee lattice in 2d for the TE polarization.]]
-The consequence of the Yee lattice is that, whenever you need to compare or combine different field components, e.g. to find the energy density <math>(\mathbf{E}^* \cdot \mathbf{D} + |\mathbf{H}|^2)/2</math> or the flux <math>\textrm{Re}\, \mathbf{E}^* \times \mathbf{H}</math>, then the components need to be '''interpolated''' to some common point. Meep does this interpolation for you&mdash;in particular, whenever you compute energy density or flux, or whenever you output a field to a file, it is stored at the locations <math>(i+0.5,j+0.5,k+0.5)</math>: the centers of each grid voxel.+The consequence of the Yee lattice is that, whenever you need to compare or combine different field components, e.g. to find the energy density <math>(\mathbf{E}^* \cdot \mathbf{D} + |\mathbf{H}|^2)/2</math> or the flux <math>\textrm{Re}\, \mathbf{E}^* \times \mathbf{H}</math>, then the components need to be '''interpolated''' to some common point (in order to remain second-order accurate). Meep automatically does this interpolation for you wherever necessary&mdash;in particular, whenever you compute energy density or flux, or whenever you output a field to a file, it is stored for the locations <math>(i+0.5,j+0.5,k+0.5)</math>: the centers of each grid voxel.
[[Category:Meep]] [[Category:Meep]]

Current revision

Meep
Download
Release notes
FAQ
Meep manual
Introduction
Installation
Tutorial
Reference
C++ Tutorial
C++ Reference
Acknowledgements
License and Copyright
Illustration of Yee lattice in 3d for a single grid voxel.
Enlarge
Illustration of Yee lattice in 3d for a single grid voxel.

In order to discretize Maxwell's equations with second-order accuracy (for homogeneous regions where there no discontinuous material boundaries), FDTD methods store different field components for different grid locations. This discretization is known as a Yee lattice.

The form of the Yee lattice in 3d is shown in the illustration above for a single cubic grid voxel (\Delta x \times \Delta x \times \Delta x). The basic idea is that the three components of E are stored for the edges of the cube in the corresponding directions, while the components of H are stored for the faces of the cube.

More precisely, let a coordinate (i,j,k) in the grid correspond to:

\mathbf{x} = (i \hat\mathbf{e}_1 + j \hat\mathbf{e}_2 + k \hat\mathbf{e}_3) \Delta x,

where \hat\mathbf{e}_k denotes the unit vector in the k-th coordinate direction. Then, the \ellth component of \mathbf{E} or \mathbf{D} (or \mathbf{P}) is stored for the locations

(i,j,k)+ \frac{1}{2} \hat\mathbf{e}_\ell  \Delta x.

The \ellth component of \mathbf{H}, on the other hand, is stored for the locations

(i+\frac{1}{2},j+\frac{1}{2},k+\frac{1}{2})-\frac{1}{2} \hat\mathbf{e}_\ell  \Delta x.

In two dimensions, the idea is similar except that we set \hat\mathbf{e}_3=0. The 2d Yee lattice for the P-polarization (E in the xy plane and H in the z direction) is shown in the figure below.

Yee lattice in 2d for the TE polarization.
Enlarge
Yee lattice in 2d for the TE polarization.

The consequence of the Yee lattice is that, whenever you need to compare or combine different field components, e.g. to find the energy density (\mathbf{E}^* \cdot \mathbf{D} + |\mathbf{H}|^2)/2 or the flux \textrm{Re}\, \mathbf{E}^* \times \mathbf{H}, then the components need to be interpolated to some common point (in order to remain second-order accurate). Meep automatically does this interpolation for you wherever necessary—in particular, whenever you compute energy density or flux, or whenever you output a field to a file, it is stored for the locations (i + 0.5,j + 0.5,k + 0.5): the centers of each grid voxel.

Personal tools