Statistical comparison of the geometry of second-phase particles
Source abstract
Abstract
In microscopic studies of materials, there is often a need to provide a statistical test as to whether two microstructures are different or not. Typically, there are some random objects (particles, grains, pores) and the comparison concerns their density, individual geometrical parameters and their spatial distribution. The problem is that neighbouring objects observed in a single window cannot be assumed to be stochastically independent, therefore classical statistical testing based on random sampling is not applicable. The aim of the present paper is to develop a test based on N-distances in probability theory. Using the measurements from a few independent windows, we consider a two-sample test, which involves a large amount of information collected from each window. An application is presented consisting in a comparison of metallographic samples of aluminium alloys, and the results are interpreted.
Source: openaire · Quality: candidate unverified · Source record ↗
Research context
OpenAlex topics: Optimal Experimental Design Methods · Soil Geostatistics and Mapping · Advanced Statistical Methods and Models
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
References (15)
Show source-provided references
1. Derr · 2000
2. Ohser · 2000
3. Nearest neighbor distributions in thin films, sheets, and plates10.1016/j.actamat.2005.12.017 ↗
4. Computations of contact distributions for representation of microstructural spatial clustering10.1016/j.commatsci.2006.01.008 ↗
5. Application of significance tests in quantitative metallographic analysis of a C-Mn-B steel10.1016/j.matchar.2006.01.006 ↗
6. Klebanov · 2005
7. Lehmann · 2005
8. R: A language for data analysis and graphics10.1080/10618600.1996.10474713 ↗
9. Buening · 1978
10. Mandl · 1975
11. Humphreys · 2004
12. Polmear · 2006
13. Hirsch · 2006
14. The influence of alloy composition on phase transformations and recrystallization in twin-roll cast Al–Mn–Fe alloys
15. Nonparametric tests based on N-distances10.1007/s10986-008-9028-2 ↗