Interferometric Reconstruction of Three-Dimensional Flow Fields through NaturalPixel Decomposition

摘要

Reconstruction of three-dimensional refractive-index fields from multidirectionalholographic interferograms can be powerful flow diagnostic tool in many engineering application. It is noninvasive and can capture gross fields instantaneously. Interferometric tomography for reconstructing flow fields, however, confronts ill-posed problems, that is, use of insufficient data. Being sampled under the presence with opaque models and enclosures, the data are sparse, nonuniform, and sometimes incomplete in projection and scanning. Here a new method, termed the variable grid method, is developed and tested to incorporate the distinct features of interferometric data for flow field reconstruction. The method is based on natural pixel decomposition. It utilizes variable rectangular grid elements in representing a field with a finite series. When the complementary field concept, which is iterative in nature, is coupled with the method, a priori information on the field to be reconstructed can be efficiently incorporated. This can further enhance reconstruction accuracy. The variable grid method allows better reconstruction resolution with greater accuracy in the region where sampled rays are more closely crossed. This region is usually of more engineering interest. Keywords: Flow visualization, Interferometry, Tomography, Three-dimensional, High speed aerodynamics, Reprints. (jhd)