Search

Components which best utilise the properties of high temperature titanium alloys are characterised by thin sections of a few millimetres thickness and hundreds of millimetres length. These alloys however are difficult to work with, being highly reactive in a molten state, necessitating a low superheat during processing. Centrifugal casting is therefore utilised as a candidate production method, as under the centrifugal force, metal can rapidly fill thicknesses substantially less than a millimetre before it solidifies. However, due to the high liquid metal velocity developed there is a high risk of turbulent flow and of the trapping of any gas present within the liquid metal. This challenging application involves a combination of complex rotating geometries, significant centrifugal forces and high velocity transient free surface flows, coupled with simultaneous heat transfer and solidification. Capturing these interacting physical phenomena, free surface flows, trapped air and associated defects is a complex modelling task. Building upon earlier work on computational modelling the authors have previously described to capture and validate the fluid dynamics behaviour of rotating systems, this contribution considers the modelling and validation of such systems to capture the coupled flow and thermal solidification behaviour and associated defect development. A bench-mark test case is employed to validate the effect of solidification on the fluidity of an aluminium alloy. Validation is also performed against a series of casting experiments to establish the models ability to capture the filling process and predict defects due to air entrapment within the solidified metal.

Centrifugal casting offers one potential route through to high quality products in difficult to cast high temperature low superheat alloys. The coupling of free surface flows and complex rotating geometries, results in significant centrifugal forces; combined with rapid heat transfer and solidification this yields a significant computational modelling challenge. The ultimate objective of the work reported here is to develop a comprehensive computational model of centrifugal casting that can reliably predict the macro-defects that arise from the process. This contribution describes the development of the computational model, enabling capture of the free surface detail of the flow during the filling stage of centrifugal casting in what are inevitably complex three dimensional geometries. The work reported here concentrates on resolving the fluid film formation, air entrainment within the turbulent gas-fluid interface and transport of gas bubbles through the mould. Validation of the above phenomena is a key issue and a series of water model experiments has been performed and recorded using high-speed video image capture. The objective is to validate and refine the computational model predictions using repeatable high quality experimental data, before application of the model in analysing full-scale centrifugal casting process. A number of key observations arise from this work, such as accurate prediction of the initial pour dynamics, maintaining a sharp fluid–air interface on complex meshes in order to capture the fluid film and air bubble behaviour, and employing higher order schemes to capture the vortex formation. These phenomena are not trivial to capture within the computational modelling tools, however, for the models to be useful in the analysing the full scale casting process, such physics must be reflected within their predictive capability.