Study Of Spatial-Average-Based Surface Roughness Measurement Techniques Using Static And Dynamic Speckle Patterns

Surface roughness measurement has always been an important research subject in many engineering fields. As one of the indicators of surface quality, surface roughness can be used for the prediction of surface performance. So far many techniques, developed from the earliest subjective judgment (e.g., visual inspection, fingernail test) to the objective comparators (e.g., capacitance method, ultrasound method), from the contact mechanical stylus instruments to the noncontact optical profilers, and from the low power optical microscopy to the revolutionary scanning tunneling microscopy (STM) and atomic force microscopy (AFM), have been presented for measuring surface roughness. But more attentions should be paid to the other two techniques, the light scattering techniques and the speckle contrast techniques, which can offer real time assessments of weak scattering surfaces and have the most promising potential for on-line or in-process use.On the basis of a review of the existing measuring methods constructed on different principles, with the intention to resolve the problems remained in the light scattering techniques and the speckle contrast techniques, a method SAM based on the static speckle statistics and DSAM on the dynamic speckle statistics are presented for the surface rough measurements of weak scattering objects.It will be more convenient for observers when letting the oblique incident light illuminate the specimen to view and record the speckle patterns in the near-field. But the mostly used Beckmann's theory only gives the far-field solutions while the Fresnel diffraction equation is merely applicable for describing the wave propagation between two parallel planes with a normal incidence light, or a geometrical-optical approximation must be introduced to calculate the emergent field of the rough surface if an oblique incident light is used. To solve these problems, an inclination factor of the reflective surface is obtained and the original model for the transmission aperture diffraction is extended to the one for the diffraction of reflective surface. Started from the basic Fresnel-Kirchhoff diffraction equation, the diffraction equation under the case of oblique incidence is derived. The light scattering methods (LSM's) are sensitive to the effects of speckles appearing in the distribution of scattered light, which leads to the fluctuations of measurement results. To avoid this, a speckle method is proposed, where the speckles are regarded as the useful carrier of the information of surface microirregularities but not the noise. Another problem for LSM's is the influence of the diffuse-light component on the received light intensity signal, which cannot be eliminated in the experiment. Treating the specular field as the coherent superposition of the uniform-deterministic specular component and the irregular-random diffuse component, the contribution from the diffuse light component is ridded of the total signal.A main disadvantage of the speckle contrast methods (SCM's) is the sensitivity of the measured results to the variations of the surface correlation length, which decreases the reliability of SCM's. Also an obstacle encountered in the practical use is originating from the theoretical basis of SCM's, the ensemble average, which cannot be realized in engineering application. To solve these problems, the local ergodicity assumption made by Goodman is introduced, a whole field spatial averaging is performed, and then the theory of surface roughness measurement from the static speckle pattern is established on the basis of spatial average.Dynamic speckle theories are usually employed in velocity measurements or vibration analysis of the strong scattering surface, but rarely applied in surface roughness measurements of weak scattering surface. As the continuation of SAM, by analyzing the statistical properties of dynamic speckle patterns and the imaging process, a technique DSAM using dynamic speckle properties for measuring the roughness of a moving surface is given in this study. The theoretical foundation of DSAM is derived in detail.The computer simulations are conducted for verifications of SAM and DSAM, by which the comparisons are also made with LSM and SCM. As the theories expect, SAM and DSAM show good one-to-one relationship between the characteristic parameters and the rms roughness, eliminate the influence of the diffuse light component and the surface correlation length on the measuring results. The vibration tests are made for SAM and DSAM, and the results demonstrate that both methods are insensitive to the displacement or vibration of the specimen surface. The results of DSAM are similar to the ones of LSM, which indicates that DSAM has the potential in in-process use as LSM does. Experiments are presented for further illumination.