In situ SEM-EBSD observations of the hcp to bcc phase ...

More documents

Recommendations

Info

830 G.G.E. Seward et al. / Acta Materialia 52 (2004) 821–832Fig. 11. Schematic of the competitive growth between intragranularplates and grain boundary allotriomorphs. The dark grey regionsrepresent the growing b allotriomorphs. The arrows indicate the directionof the moving a–b interfaces between the b allotriomorphs anda matrix, and the b–b interfaces between the b allotriomorphs andintragranular plates.4.4. Comparison of in situ SEM and vacuum furnacetransformed samplesIn situ heating in the SEM inherently involves observationof a free surface, and thus it is important toconsider their influence on the phenomena studied. Oneapproach is to section a larger volume sample that hasbeen similarly heat-treated ex situ, and subsequentlycompare its internal and surface microstructures usingSEM. Fig. 12 shows EBSD map data highlighting theresults of applying this method. A sample of slab materialwith dimensions 10 10 8 mm was used; this sizeenables the examination of an internal region that is atleast 2a grain diameters distant from any surface. Thesurface and internal section in Figs. 12(a) and (c) showsimilar microstructures exhibiting large domains, eachdomain having a particular crystallographic orientation.Two types of domains are observed in Figs. 12(b) and(d): (i) regions like those marked A showing no clearsubstructure, and (ii) regions marked P showing parallelarrays of low-angle grain boundaries. A-domains correspondto allotriomorphic a regions, and a P-domainarises by coalescence of a stack of single-variant a plateswhich grew within a particular b grain. We infer that thelarge domains of similar orientation represent a minimumestimate of the grain size for the b phase microstructureprior to transformation. These observationsshow that the process of allotriomorphic growth is notrestricted to the free surface. The deviations of dihedralangles from 120° observed in Fig. 12 (cf. Fig. 6) occurduring b–a transformation, firstly because several crystallographicallyequivalent a plate variants can form ineach b grain, and secondly because a allotriomorphsgrow along b–b grain boundaries and expand into theneighbouring grains. It is this process that makes thedetermination of the prior b phase microstructure fromthe final alpha phase microstructure problematic, andalso the reason why only a minimal estimate of the bFig. 12. EBSD map data from (a) and (b) surface and (c) and (d) internal section of ex situ transformed Ti slab material, with areas of plates (P) andallotriomorphs (A). (a) and (c) Colour coded crystallographic orientation data, highlighting the similarity in microstructure, with domains showingcomplicated grain boundary geometry and grain sizes frequently >1 mm diam. (b) and (d) Maps of variation in diffraction pattern quality. Grainboundaries or internally strained areas often show less intense diffraction and thus this parameter is used to highlight subtle substructure withindomains, such as packets of plates with similar orientation (P).
G.G.E. Seward et al. / Acta Materialia 52 (2004) 821–832 831phase grain-size can be inferred. The transient growthprocess observed in situ at the initial stage of the a–btransformation leaves no record in the a–b–a samplestransformed ex situ, and thus the effect of the free surfaceon this phenomenon is difficult to assess. However,it can be deduced from Figs. 12(b) and (d) that a platesand allotriomorphs in these samples, nucleate not onlyat the free surface but also within the volume. In the insitu experiments it is possible that nucleation andgrowth occur at lower driving forces because the freesurface assists these processes; however, the microstructuralobservations appear to be at least qualitativelysimilar at the surface and within the samplevolume.A comparison of the microtexture of the two microstructuresin Fig. 12 is shown in the contoured pole figuresof Fig. 13. Visual inspection of the pole figureindicates that the distribution of orientations in the twosections is very similar, indicating that there is not asignificant effect from the free surface. A rigorous calculationof the b phase microtexture is not possible dueto variant selection in the b–a transformation. However,assuming the Burgers OR is obeyed, it is possible to showthat the dominant elements of the a texture are consistentwith a combination of (0 0 1)h100i and (1 1 2)h111i texturecomponents observed in this contribution and reportedby others [9,10], these b phase orientations aresuperimposed on the a phase pole figure.The unique advantage of the in situ approach is thatit provides direct observation of the transient processesinvolved in the a–b transformation. Ultimately, theseobservations should improve our ability to predict, andFig. 13. Contoured pole figures showing the crystallographic texture of (a) surface and (b) internal section of ex situ transformed Ti slab material.Visual inspection of the distribution of orientations shows that a similar texture is developed at the surface and internally. Reconstruction of the btexture from the inherited a texture is problematic (due to variant selection in the b–a transformation); however, the overlay of b orientationsindicates that the a textures are consistent (assuming Burgers OR) with the (0 0 1)h100i and (1 1 2)h111i components observed in the in situexperiments.
Page 1 and 2: Acta Materialia 52 (2004) 821-832ww
Page 3 and 4: ture component showed a minor inten
Page 5 and 6: G.G.E. Seward et al. / Acta Materia
Page 7 and 8: G.G.E. Seward et al. / Acta Materia
Page 9: G.G.E. Seward et al. / Acta Materia

In situ SEM-EBSD observations of the hcp to bcc phase ...

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?