IDENTIFICATION OF STEEL PHASES AND CONSTITUENTS BY COMBINED EBSD AND HIGH RESOLUTION EPMA CARBON MEASUREMENTS

摘要

The mechanical properties of multiphase steels depend on various factors such as the composition, fraction of the different phases/constituents and internal strain level. Precise measurements are required to understand steel behaviours as well as to establish reliable prediction models. In this project, EBSD and EPMA measurements are combined to obtain a complete characterisation of dual phase and TRIP steel samples as well as to overcome some ambiguities encountered when only one of these two techniques is used. SEM images after chemical etching are also performed to validate the results. EBSD can easily differentiate ferrite and austenite grains based on their crystallographic difference. However, the lattice deviations between ferrite, bainite and martensite are too small to be indexed as different phases. As reported by many authors including Ryde, ferrite and martensite can be differentiated based on the quality of the diffraction patterns. Despite its dependence on the crystal orientation, the diffraction quality is predominantly reduced by the crystallographic defects in martensite. This segmentation is further confirmed by EPMA where martensite regions have a carbon concentration of between 0.8 - 1.2 wt% in the measured samples. The discrimination between ferrite and bainite using the diffraction quality method is, however, more ambiguous. Zaefferer et al. utilized the higher concentration of dislocations in bainite, measured using the kernel average misorientation (KAM), to segment out bainitic regions from ferrite grains. However, in dual phase steels, potential ferritic regions with high density of transformation induced dislocations near martensite islands complicate the identification of bainite. Both have high local misorientation. High resolution carbon mapping performed with a FEG microprobe can resolve this ambiguity. Bainite can be identified as regions of high misorientation values containing carbon, whereas dislocation regions will have a carbon content similar to that in ferrite grains. A further complication is the presence of pearlite, as this constituent also exhibits higher KAM values and carbon level than ferrite. As the internal structure of pearlite cannot be distinguished in a microprobe, the measured carbon concentration is a mixture of the one of ferrite and cementite based on the lamellae spacing. This illustrates the need of sensitive EPMA measurements to distinguish small carbon variation between pearlite and bainite structures. Strategies must be developed to maximize counting statistics while reducing the effects of contamination and topography, and maintaining a good spatial resolution.