Evaluation of Nondestructive Methods for Ceramic Heat Exchanger Applications

摘要

For both seeded defects and naturally occurring defects, there is little correlation between test results for small defects. This is because X-Ray is sensitive to density variations and ultrasonics and scanning laser acoustic microscopy (SLAM) are sensitive to acoustic impedance variations. The grain boundaries and microporosity of the material produce a high scattering background for the acoustic methods, masking small inclusions. X-Ray is insensitive to grain boundary effects and the uniform microporosity averages out over the material thickness. If minor inclusions are to be detected, X-Ray must be used. Ultrasonics and SLAM are sensitive only to the presence of an open crack, not its width. If cracks are to be detected, ultrasonics or SLAM must be used. SLAM is by far the fastest method of scanning for OD cracks, but only ultrasonics will find ID cracks. Results of this first phase study give four guidelines to tube design. First, surface irregularities are a major limitation to testing. The eventual tests will be much more sensitive if surface finish can be improved. Second, porosity is another major limit to testing. Pores are not generally strength limiting in this material, but pores scatter ultrasound very badly and produce irregular densities on X-Ray film, in both cases obscuring more important defect indications. Third, the dimple shape of the closed end is essentially untestable. The fourth design guideline is provided by the fractography results. In some cases, failure was initiated in or near clusters of large grains. Defect sizing depends upon the detection methods. 5 figs. (ERA citation 13:026555)