The morphological characterisation of grains and grain boundaries.

It is well reported that the grain size of polycrystalline materials is important in determining mechanical properties. Within this thesis investigations are reported from an inter-comparison of grain size methods, the impact of an incomplete network of grain boundaries after chemical etching (missing boundaries) on grain size measurements and finally the relationship between grain size and the misorientation of grains. The experimental techniques used are manual grain size measurement methods, automatic image analysis and Electron Backscatter Diffraction (EBSD). The materials used are a ferritic, single phased, equi-axed steel (mild steel) and a ferrite/pearlite steel. The relative simplicity of the mild steel microstructure effectively removes complicating factors such as multi-phases or grain elongation allowing a more focused investigation into grain size methods. The ferrite/pearlite steel can be found more readily in industrial applications where impact toughness is an important property. Manual measurement methods and automatic image analysis are used for the inter-comparison of grain size methods. The manual methods are described in the standard ASTM E112, and are the lineal, planimetric, single circle and three circle methods. It is shown that there is a difference across the methods in the number of measurements required to obtain a specified accuracy. Also that sampling is critical to ensure that the measurements made are representative of the microstructure in that it is more important to measure more specimens than more fields of view within one specimen or many grains within one field of view. Automatic image analysis can provide a substantial database and thus is a useful grain sizing method. It is demonstrated that the number of pixels in a digital image forming a grain boundary, determined by the pixel resolution, will influence the measurements. Also from using digital images a specific number of grain boundaries are removed thus artificially creating missing boundaries making their impact on grain size measurements quantifiable. Manual and automatic measurements are conducted and it is shown that the mean grain size is relatively less affected by missing boundaries than might be perceived from a visual inspection. An EBSD map provides a complete network of grain boundaries since a boundary is formed from the orientation of one grain to another (misorientation), compared to standard metallographic techniques, e.g. chemical etching. EBSD maps are compared with optical images of the same fields of view and missing boundaries are then located. From this it is shown that there is a significant difference between two operators in determining the location of boundaries. EBSD is also used to investigate the correlation between the location of missing boundaries and the misorientation of the grains at that boundary and shows a trend of higher misorientation for boundaries not chemically etched up. The final section of this thesis is concerned with the relationship between the misorientation of small/small, small/large and large/large grains and the size of grains of the ferrite phase from the ferrite/pearlite steel. EBSD maps provide the misorientation data and are also used for grain sizing conducted on the automatic image analysis system. From the misorientation data the boundaries can be categorised as low or high angle and it is found that there are more low angle boundaries between small/small grains and small/large grains than between large/large grains.