What they have done and what I can do

Last two years, I spent quite a lot of time on reading about what people have done on the research about oxide film on light alloy.  In the early study on my final year research project, my literature review mainly focus on how oxide film affects the mechanical properties of the casting component, and how modelling work has been done to track the films, both during the filling process and solidification.  After I submitted my research report last year, my co-supervisor C R sent me several copies of the work had been done in Japan. The study in Prof. Ohnaka's group using the theory established by Prof. Campbell in Birmingham to model the porosity formation in castings. Then they developed a software to model the filling and solidification in casting. The application is modelling of high pressure die casting  produced in China and Japan. They published loads on this topic, but personally, I feel their prediction is not that accurate. There are two possible reasons for this: a)the fundamental theory is not correct or only correct in some circumstances; b) their models have some limitation.

When I started my PhD, I felt what I should do was the work they did as viewed in the first part of the literature, i.e. modelling the formation of oxide film in light alloy, but developed their work. However, later NRG said we should try to model the porosity inside the casting, since this is more damaging than oxide film. According to Prof. Campbell's theory, oxide films can be a nuclei of porosity, so modelling oxide film formation and distribution can help the understanding of porosity distribution in castings. And this is the basic theory of Ohnaka's work.

Then what I can do?

I feel I can develop the model built by CR to a more accurate way and validate it.

I could develop two kinds of approaches, one on the double oxide film formation, the other on the bubble entrainment. Both two models  could use particle module to track. But to achieve this, I need to talk with Flow Science developers.

After accurate modelling of the distribution of double oxide films and micro-bubbles, I think I can use the package developed by PDL's group to model the growth and morphology of those porosity.

This sounds fantastic, doesn't it?