Customarily, a material has been sectioned to study its internal grain microstructure and thus in the process is destroyed. Using x-rays, however, there are two nondestructive methods of determining the sources of diffraction spots and hence the internal grain microstructure of a sample. One technique consists of placing a wire in the path of a diffracted ray so that its image is prevented from appearing on the detector screen. Ray-tracing is then done to locate the source within the sample from whence the rays emanate. In this experiment, we investigate the other technique of determining source location by recording diffraction patterns at ten equally-spaced detector distances and then graphing the data with reasonable-fit lines using the least-squares fitting routine. We then perform a ray-tracing triangulation technique to pinpoint the location of the source from which the rays are coming. Cluster analyses are employed and plots of ray number versus pixel position of certain points at some particular detector distances are created. An error propagation analysis is then carried out as a check to the cluster analyses and graphs of error deviation along the detector path versus ray number are constructed. With statistical error analyses and construction of error boxes using chosen pixel error deviations and delta z error values, the best error measurement using the detector method was found to be plus/minus 100 microns. In this study, it was found that the detector method provided a much poorer resolution than the traditional wire technique of which there is a source size precision of within 1-5 microns. The detector method, though, is sufficient for large-grain material studies.
PDF
download
878987797
028-85220240
