(114b) The Next-Generation Predictive Tools for Multiphase Flows

The ability to predict the behaviour of multiphase flows accurately, reliably, and efficiently addresses a major challenge of tremendous economic, scientific, and societal benefit. These flows are central to micro-fluidics, virtually every processing and manufacturing technology, oil-and-gas and nuclear applications, and biomedical applications such as lithotripsy and laser-surgery cavitation. Significant advances have been made in developing numerical procedures to simulate these complex flows, however, there remains a large gap between what is achievable computationally and 'real-life' systems; the latter are beyond what can be addressed with current methods. As a result, the use of empirical correlations to bridge this gap remains the norm. We will outline the framework that we are currently developing as part of the Multi-scale Examination of MultiPHase physIcs in flowS (MEMPHIS, http://www.memphis-multiphase.org/) programme, funded by the Engineering and Physical Sciences Research Council, UK, to minimise the use of correlations and shift towards the use of numerical simulations as a truly predictive tool that can be used as a sound basis for design. This involves a transparent linkage between input and prediction, allowing systematic error-source identification, and optimal, model-driven experimentation, to maximise prediction accuracy; optimal selection of massively-parallelisable numerical methods, capable of running efficiently on 10⁵-10⁶ core supercomputers, optimally-adaptive, three-dimensional resolution, and sophisticated multi-scale physical models. The talk will discuss progress made on several of the above areas using as examples bubbly flow, and slug flows.