Interface Roughness Measurement Of Coatings With Ultrasonic Method

Interface roughening is an effective measure to improve bonding strength. However, the residual stress increases as the roughness increases in the process of service, which leads to a high degree of stress concentration and causes the coating to peel off. Characterization of interface roughness can provide an important basis for in-service quality inspection and life monitoring of the coating parts. Almost all existing methods confined to surface roughness measurement, and it’s very difficult to characterize interface roughness at the end of coating preparation. In this paper, a novel method has been proposed for characterizing interface roughness which is based on the ultrasonic reflection coefficient amplitude spectrum (URCAS). The principle of obtaining roughness information from overlapped signals is given and the idea of modification for measuring inhomogeneous coatings is also presented. On this basis, the influence on measurement results by testing conditions and material heterogeneity is analyzed through numerical simulation. Finally, the validity of this method is verified by experimental tests. The main contents are as follows:(1) Starting with the ultrasonic propagation in the model of isotropic homogeneous layered medium, the URCAS of layer structure, which contains rough interface, is derived based on phase screen approximation theory. Besides, the theoretical basis on interface roughness measurement is established combining with correlation coefficient matching algorithm.(2) The construction method of random rough surface model is introduced, which meets the Gaussian height distribution function and specified autocorrelation function. Furthermore, finite-difference time domain model used for ultrasonic numerical simulation is set up and the set of boundary conditions and numerical stability conditions are analyzed.(3) The influence of roughness measurements by ultrasonic wave length λ, beam coverage, shape errors and inhomogeneity of coating is analyzed through numerical simulation. The results are as follows:roughness ranging from 20～65μm,5～45μm and 2～40μm can be measured by different sound sources whose center frequencies are 15MHz, 20MHz and 25MHz respectively; in order to obtain reliable ultrasonic measurements, it’s necessary to make the beam coverage D and the contour single peak average distance S meet the relationship of D≥5S in one dimensional case; the smaller the roughness, the greater the impact on measurement results by shape errors, and they’re adverse factors in ultrasonic inspection which are difficult to be eliminated; when the porosity is small, it’s possible to weaken the influence of coating inhomogeneity and improve the ultrasonic measurement accuracy through determining the attenuation of the coating and using it to correct the URCAS.(4) Interface roughness of standard roughness specimens and tungsten carbide (WC) coating specimen is measured in experiments. It is shown that the absolute error is less than 3.3μm and the relative error is less than 17% between ultrasonic and laser confocal method within standard roughness specimens whose nominal roughness Ra are 3.2μm,6.3μm,12.5μm and 25.0μm respectively, except Ra=3.2μm. For WC coating specimen, the absolute error is less than 2.5μm and the relative error is less than 20% between ultrasonic and metallographic method. However, measurement errors are bigger on both sides of the region by the influence of shape errors.