ABSTRACT
We investigate the wave-optical light scattering properties of deformed thin circular films of constant thickness using the discrete-dipole approximation. Effects on the intensity distribution of the scattered light due to different statistical roughness models, model dependent roughness parameters, and uncorrelated, random, small-scale porosity of the inhomogeneous medium are studied. The suitability of the discrete-dipole approximation for rough-surface scattering problems is evaluated by considering thin films as computationally feasible rough-surface analogs. The effects due to small-scale inhomogeneity of the scattering medium are compared with the analytic approximation by Maxwell Garnett, and the results are found to agree with the approximation.
ABSTRACT
We study theoretically the light-scattering efficiency of paper coatings made of starch acetate pigments. For the light-scattering code we use a discrete dipole approximation method. The coating layer is assumed to consists of roughly equal-sized spherical pigments packed either at a packing density of 50% (large cylindrical slabs) or at 37% or 57% (large spheres). Because the scanning electron microscope images of starch acetate samples show either a particulate or a porous structure, we model the coatings in two complementary ways. The material can be either inside the constituent spheres (particulate case) or outside of those (cheeselike, porous medium). For the packing of our spheres we use either a simulated annealing or a dropping code. We can estimate, among other things, that the ideal sphere diameter is in the range 0.25-0.4 microm.
ABSTRACT
The Foldy-Lax multiple-scattering equations in matrix form were examined for their suitability for solving the phase function and degree of linear polarization of clusters of particles as a complementary approach to the superposition T-matrix scheme. The constituent particles used for verification against the latter method were spheres with size parameters close to unity, with their aggregation governed by generalized Brownian motion. The number of constituents can be enlarged beyond the limitations of the benchmark method by a relatively small sacrifice in accuracy, but increasing constituent size and introducing nonsphericity easily result in numerical noise and overwhelming computational effort unless a large amount of the accuracy is going to be traded off.
ABSTRACT
We develop a second-order ray-tracing model that includes the constructive interference of reciprocal rays and Fresnel reflections from an inverse gamma-type distribution function of path length to predict the negative polarization branch seen in some astronomical bodies. We expect that such a path-length distribution might resemble the path lengths undertaken by rays incident upon some astronomical bodies. The resulting negative polarization is largely wavelength independent and depends primarily on the power law in the path-length distribution, which coincides with some observations.
