This past week, one focus was looking at partial pair correlation functions for each atom type as the cut-off for how near atoms could be to each other was altered. Another goal was to learn how to run the programs required to do Honeycutt-Anderson (HA) analysis of the files generated by the Rever Monte Carlo (RMC) simulation. It is still unclear what may be the best cut-off to use for the RMCs, but so far the HA analysis is looking very interesting. Hopefully it concurs with results from the Voronoi analysis.
This week Voronoi analysis was done on Ni24Zr76. A trend that was observed was that the counts overall decrease as temperature increases. An interesting peak in counts was observed at 1200C for some indices as well.
Below is a powerpoint descibing the apparatus (the code) used to look at the atomic structure of the metallic liquids and glasses using Voronoi analysis.
The first week I focused on analyzing the experimental S(q) and g(r) data from the x-ray diffraction done at Washington University, St. Louis. The trends observed for Ni24Zr76 were that as temperature decreases, the height of the first peak (S(q)1) increases. Contrary to predictions, the position of the peak, q, only shifted by a total of about 0.03 1/A between highest and lowest temperatures. Another observation was that the percent composition of Nickel has a strange effect on the g(r) function in the glassy data in that another smaller peak appears farther in from the tallest peak.
A reverse-Monte Carlo program was tested on S(q) data at 1200C, where the chi squared vale was 2.9364 and the first peak was within .05 the height of the experimental S(q).