| dc.description.abstract | The present PhD thesis is concerned with electron backscatter diffraction (EBSD) based in-situ observations of fine-scaled polycrystalline microstructures during thermomechanical processing in the scanning electron microscope (SEM). The main aim of the work was to demonstrate the capabilities of the equipment available for such examinations at the SEM lab at NTNU. The thesis is divided into five separate parts. Part I covers a general introduction to the EBSD technique and a brief description of the main experimental equipment used in the PhD work.
In Part II the effects of varying the SEM parameters, the operating conditions of the charged-coupled device (CCD) camera and the EBSD indexing parameters were examined by measuring different types of stainless steels. It was shown that the spatial and angular resolution is strongly affected by nature of the examined sample and the operating conditions of the microscope and the CCD camera, indicating that these parameters must be carefully considered prior to an EBSD measurement. At the same time, it was demonstrated that the recently developed EBSD system for high speed pattern collection is a very promising tool for microstructural characterization of polycrystalline materials. By using this system, measurements may be carried out at much higher speeds than traditional on-line EBSD with minimal penalty on the indexing reliability. Moreover, it was confirmed that the number of diffraction bands detected in the Kikuchi patterns may have significant impact on the confidence in indexing during an automated EBSD analysis. A minimum of six bands was required to accurately measure the volume fraction of retained austenite for the examined twophase material.
The first section of Part III involves studies of deformation-induced phase transformations in a supermartensitic steel (SMSS) containing about 40 vol. % retained austenite in the as-received (i.e. intercritically annealed) condition, using in-situ tensile testing in combination with EBSD analyses in the SEM. It was demonstrated that the equipment design could be used to unravel crystallographic and morphological features of the microstructure constituents in the steel during plastic deformation at room temperature. The martensite formation was initiated already at low strains, and increased gradually with increasing plastic strains up to about 10%. The transformed martensite maintained an orientation relationship with its parent austenite that satisfied the Kurdjomov-Sachs (K-S) criteria. Moreover, it was seen that the martensite formed homogeneously within the microstructure, independent of the crystallographic orientations of the retained austenite. But no new martensite variants, besides those already present in the as-received condition, did form during the deformation.
In the second section of Part III the deformation-induced martensite variant selection in the SMSS was examined in more detail in the temperature range from -60 °C to 150 °C. At each testing temperature, this austenite transformed back to martensite during plastic deformation at a rate which was controlled by the accumulated plastic strain in the material. On the other hand, the applied strain rate did not affect the overall transformation rate. Moreover, the subsequent Schmid factor analysis revealed that the martensite variant selection was independent of the local slip activity within the austenite. Therefore, only the martensite variants present in the parent steel developed during the phase transformation, which is consistent with the observations from the previous section of Part III. In addition, their individual intensities remained approximately constant within each prior austenite grain throughout the deformation process. This means that the deformation-induced martensite variants nucleated from the same sites as those being operative in the intercritically annealed base material. Thus, the observed variant selection is just another example of the inherent reversible nature of the martensite transformation.
In Part IV the ferroelastic domain reorientation in polycrystalline rhombohedral LaCoO3 was demonstrated by backscatter electron (BSE) imaging and electron backscatter diffraction (EBSD) combined with in-situ compression testing. Four different domain orientations were confirmed within each examined grain, in line with expectation from crystallography. During in-situ compression reorientation of domains were observed favoring domains with the c-axis close to perpendicular to the stress field. In-situ compression combined with BSE/EBSD was shown to be an excellent tool for studying the influence of mechanical stress on grain/domain structure in polycrystalline ceramics.
Finally, Part V involves high speed EBSD measurements in conjunction with hot-stage annealing in the SEM of an Al-Mg-Si alloy after equal-channel angular pressing (ECAP). The microstructure evolution was in this investigation monitored at different temperature intervals up to 300°C, using pattern collection speeds in the range from 100 to 300 patterns per second (pps). It was confirmed that the high speed EBSD system is a very powerful tool for observing the recrystallization behaviour of the examined alloy during hot-stage annealing. However, the present equipment design was restricted by the simple heating stage applied. Therefore, a more sophisticated hot stage is required to conduct EBSD measurements in parallel with the heating of the sample. Still, it was observed that continuous recrystallization occurred within the alloy up to the annealing temperature of 270°C, with no marked change in texture or the fraction of high angle grain boundaries. Following annealing at 300°C, on the other hand, a marked change in grain size and texture evolution was observed, indicating a change from continuous recrystallization to discontinuous grain coarsening. | nb_NO |
