Investigation of microstructural evolution by real-time SEM of high-temperature specimens

• Main Author: Fielden, Iain Michael
• Other Authors: Cawley, Jess ; Rodenburg, John ; Smith, Alan
• Published: Sheffield Hallam University 2005
• Subjects: 620.11299
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.435340

This thesis presents the results of a project to investigate the growth of grains and movement of grain boundaries in face centred cubic metals, using Environmental Scanning Electron Microscopy (ESEM).The original aim proved impractical without considerable modification to the microscope technique. The result of this was an imaging technique suitable for "real-time" . characterisation of dynamic microstructures, evolving as materials are heated, cooled or held at high temperatures in the SEM. The technique is adaptable to both conventional "high-vacuum" SEM and environmental SEM. The development of the technique is described, and its application to hot metal specimens. The technique has been applied to various metals, but most notably to steel. The project has yielded probably the first "real-time" images of grain growth with time in steel, the first images of Austenite decomposition and phase change occurring in steel, the first images of grain growth in a bulk gold alloy and images of grain growth in an aluminium alloy. It is shown that the motion of grain boundaries in polycrystalline metal bulks is discontinuous ("jerky") and that this jerky motion occurs independent of grain boundary grooving. It is also shown that the first manifestation of austenite decomposition is an as-yet unexplained micron-scale "cellular" sub-structure within the austenite grain. It is further shown that in cooling of steel at slow-to-moderate speeds, the first appearance of permanent non-austenite structure is the precipitation of relatively large carbides at surfaces. Unexpectedly, this observation is in a slightly hypo-eutectoid steel, in which a slight excess of ferrite would be expected, leading to the logical but erroneous expectation that pro-eutectoid ferrite should be the first phase to precipitate. In slow-to-moderate cooling of near-eutectoid steel, it is shown that the number of nuclei initiating the austenite-to-pearlite transformation is small by comparison to the number of austenite grains present and that the austenite-to-pearlite transformation front sweeps from grain to grain with relative ease.