The measurement of materials properties and the control of processing parameters is important for both materials development and quality control. Determination of these parameters is sometimes advantageous if done by means of microstructural characterisation as here additional information may be gained from the sample. These values, e.g. the grain size or the volume fraction of the present microconstituents, may allow the correlation with mechanical properties or processing properties. Unfortunately, the diffusional and displacive transformation products, martensite and bainite, exhibit very fine microstructures with a low contrast, so that conventional microstructural image analysis cannot be applied readily to distinguish these structures. As an alternative microstructural characterisation technique texture analysis based on Haralick parameters calculated from second order grey value statistics was successfully applied. It has been shown that the analysis of large sample areas can be done automatically enabling the correlation of the texture data with the respective local microhardness using a neural network. The analysis is limited due to the dependence of the texture parameters on preparation and imaging conditions. A more detailed understanding of the individual Haralick parameters will be the basis to extend the method to a correlation with other properties of the sample that may be not easily accessible by physical testing, such as toughness.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:489757 |
| Date | January 2005 |
| Creators | Fuchs, Alexander |
| Publisher | University of Birmingham |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://etheses.bham.ac.uk//id/eprint/212/ |
Page generated in 0.002 seconds