The FDTD computation of electromagnetic wave scattering from surfaces

Abstract

The radiative properties of engineering surfaces with microscale surface textures (patterned or random roughness and coating) are of fundamental and practical importance. In the rapid thermal processing or arc/flash-assisted heating of silicon wafers, the control of energy deposition through radiation and the surface temperature measurement using optical pyrometry require in-depth knowledge of the surface radiative properties. These properties are temperature, wavelength, and surface texture dependent. It is important that these properties can be modeled and predicted with reasonable accuracy. This study builds the foundation by solving the Maxwell equations that describe the electromagnetic wave reflection from the one-dimensional random roughness surfaces. The surface height conforms to the normal distribution, i.e., a Gaussian probability density function distribution. The models produce very accurate bi-directional reflectivity with its accuracy limited by the numerical scheme. The numerical algorithm of Maxwell equations' solution is based on the well-developed finite difference time domain (FDTD) scheme and near-to-far-field transformation. Various computational modeling issues that affect the accuracy of the predicted properties are quantified and discussed. The predicted properties were compared and found in good agreement with the published work