The measurement of properties is very important for both material design and quality control. As materials’ properties are determined by the microstructure of the materials such as grain size or the volume fraction of the present phases, microstructural characterisation is a powerful tool for property prediction. Unfortunately, microstructural characterisation has not been widely applied with all steels such as pearlitic steels or complex multi-phase steels due to their complex microstructures. These microstructures may contain features that cannot be resolved by optical microscopy, and in which important information is contained in their texture rather than simply their grey level. These microstructures were investigated in this study using image texture analysis. Fourier transform-based analysis was applied to pearlitic microstructures to extract the image orientation information. The orientation information as well as the grey value of low pass filtered image was used as predicates in a split-merge algorithm to segment the pearlitic colonies. A supervised classification method based on various statistical measures including a number of 2-point statistics (Grey Level Co-occurrence Matrix measures) was developed to distinguish the bainite (upper bainite and lower bainite), martensite and ferrite phases in steels. The influence of etching on the analysis results was also investigated.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:600266 |
| Date | January 2014 |
| Creators | Liu, Xi |
| Publisher | University of Birmingham |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://etheses.bham.ac.uk//id/eprint/4842/ |
Page generated in 0.0695 seconds