Boundary conditions
Boundary conditions are implemented in the adding-doubling method by
creating a layer which mimics the reflection and transmission at a boundary. This
layer is added to a slab to find the reflection and transmission for a slab including the
boundary conditions. If
is the Fresnel reflection for light incident from a
medium with index of refraction n0 on a medium with index of refraction n1 then
 |
(3.35) |
 |
(3.36) |
The square of the ratio of the indices of refraction is due to the n2-Law of radiance
(Appendix B) which accounts for the difference in radiances across across an index
of refraction mismatch. The factor of
is included to compensate for star
multiplication. If a glass slide is present then Equation (B.21) should be used to
calculate
with
,
and
.
Both operators are diagonal
because light is specularly reflected and the angle of incidence equals the angle of
reflection. The reflection and transmission operators for light travelling from medium
1 into medium 0 are
 |
(3.37) |
 |
(3.38) |
Since light is refracted at the boundary, care must be taken to ensure that the
incident and reflected fluxes are identified with the proper angles. For example, if
eleven Gaussian quadrature angles are used, all but three are totally internally
reflected at an interface with n1=1.5 and n2=1.0. This is caused by the uneven
distribution of quadrature points over the integration interval--in the Gaussian
quadrature scheme oblique angles are emphasized over angles near normal and,
therefore, most of the quadrature angles undergo total internal reflection.
If equal boundary conditions exist on both sides of the slab then, by
symmetry, the transmission and reflection operator for light travelling from the top to
the bottom are equal to those for light propagating from the bottom to the top.
Consequently only one set need be calculated. This leads to a faster method for
calculating the reflection and transmission for a slab with equal boundary conditions
on each side. Let the top boundary be layer 01, the medium layer 12, and the bottom
layer 23. The boundary conditions on each side are equal: R01=R32,
R10=R23, T01=T32, and T10=T23.
For example the light reflected from layer 01 (travelling
from boundary 0 to boundary 1) will equal the amount of light reflected from layer
32, since there is no physical difference between the two cases. The switch in the
numbering arises from the fact that light passes from the medium to the outside at the
top surface by going from 1 to 0, and from 2 to 3 on the bottom surface. The
reflection and transmission for the slab with boundary conditions are R30 and T03
respectively. These are given by
|
T02 = T12(E-R10R12)-1T01
|
(3.39) |
|
R20 = T12(E-R10R12)-1 R10T21+R21
|
(3.40) |
and
|
T03 = T10(E-R20R10)-1T02
|
(3.41) |
|
R30 = T10(E-R20R10)-1 R20T01+R01
|
(3.42) |
Further increases in efficiency may be made by exploiting the diagonal nature of the
reflection and transmission operators for an interface, since most matrix/matrix
multiplications above become vector/matrix multiplications.
|