Fluid, Solid and Radiation Modelling in Multiphysics Systems: Prof. Christopher Pain, Imperial College London
Date: Wed 5 Feb 2020, 15:00 - 16:00
Location: Graduate Centre GC114
Here we present an outline of the work by Applied Modelling & Computation Group on development of radiation transport/neutronics, thermal hydraulics/multi-phase flows, and solids mechanics modelling. We present the coupling of these models together with their applications which now cover a range of industrially important, multi-physics phenomena. Current applications include: criticality in fissile solutions, nuclear reactor analysis and accident scenarios, ocean and atmospheric prediction, pollution dispersion through urban environments, single and multi-phase flows and coastal defense structures. To maximise computational efficiency and accuracy, we employ adaptive mesh resolution to perform some of the multi-scale modelling in which the finite element / control volume meshes are optimized to represent the physics. It is shown how errors in the overall solution can be calculated and how this approach can also be used to generate error measures to guide mesh adaptivity/numerical resolution.As far as fluid-structure interaction modelling is concerned, the Finite Element Discrete Element Method (FEMDEM) is used for resolving the solids. FEMDEM is able to capture non-linear material behaviour and has multi-body and granular media capabilities so that sediment and scour can be modelled at the grain scale. The model Fluidity is used for the CFD. It is a multi-phase model based on arbitrary unstructured finite element meshes and has an anisotropic dynamic mesh adaptivity capability. This allows us to resolve highly complex geometries.We outline future research directions in multi-physics modelling, an approach to adaptive multi-scale modelling and how this may be linked to predictive modelling including uncertainty quantification, data assimilation and reduced order, or rapid, modelling.
Prof. Christopher Pain is head of the Applied Computation and Modelling Group (AMCG) of about 50 scientists and engineers at Imperial College and deputy lead of the Novel Reservoir Modelling and Simulation group (NORMS), at Imperial College. His main academic interests are in general Computational Fluid Dynamics (including multiphase flow in porous media, industrial multi-phase flows, oceanography and geophysical fluid dynamics), nuclear engineering (neutron and photon transport, nuclear criticality, nuclear reactor dynamics, nuclear waste repositories), optimisation, data assimilation, sensitivity analysis, reduced order modelling, parallel solution techniques. Professor Pain has supervised 50 successful PhDs and published 230 journal papers.