Development Of An Optical Surface Roughness Gauge Based On The Laser Speckle Contrast Method

This work represents a noncontact nondestructive technique for surface roughness measurement. The technique is based on investigating the statistical properties of the speckle patterns that are developed whenever a surface is illuminated with a highly coherent light source such as a laser. The main objective is to correlate the sizes of the speckles with their light intensities to the surface roughness and profile. Different statistical approaches and correlation analysis have been implemented. Spatial data analysis, or variogram analysis, has been used to develop the relationship between observations taken at various locations. Among the other advantages of such an approach, this technique is simple to implement and suitable for operation in a manufacturing environment.

New material on computerized optical processes, computerized ray tracing, and the fast Fourier transform, Bibre-Bragg sensors, and temporal phase unwrapping. * New introductory sections to all chapters. * Detailed discussion on lasers and laser principles, including an introduction to radiometry and photometry. * Thorough coverage of the CCD camera.

The available statistical analyses of laser speckle from smooth surfaces have been extended to account for the effect of anisotropic surfaces on the contrast and on the autocorrelation function. Experimental results are presented for comparison with the theoretical results. For rough surfaces, an experiment is proposed for remotely determining surface roughness by using two different illuminating angles. A significant aspect of the procedure is that photographic processing has been removed from the experiment so that the technique is of interest in radar applications. An extensive literature survey of methods for measuring correlation length is reported. The various techniques are classified into three categories, profilometric, optical, and photomicrographic. The most promising procedures are discussed and two techniques are recommended for laser radar research. (Author).

Three methods for determining the statistics of a surface's roughness are proposed. All three use some variation of the phase contrast method, and all appear to have advantage over a recently suggested method. (Author).