A generalized multiphase modelling approach for multiscale flows

2021 
Abstract Multiphase flows are ubiquitous both in nature and industry. A broad range of interfacial scales, ranging from fine dispersions to large segregated interfaces, is often observed in such flows. Standard multiphase models rely on either the interface-averaging approach, which is suitable for the modelling of dispersed flows, or on the interface-resolving approach, which is ideal for large segregated interfaces. This results in the inability of such models to deal with complex multiscale flows, and different generalized hybrid modelling approaches having been proposed to overcome this shortcoming. This work presents a novel generalized multifluid modelling approach where large segregated interfaces are identified in the multifluid field from the local interface topology and resolution, avoiding the need for a-priori thresholds of the local volume fraction used in the majority of the models available in the literature. Interface compression and suitable modelling closures for drag and surface tension are activated in the large interfaces regions, whilst the model reverts to a standard multifluid formulation in the regions of small/dispersed interfaces. An assessment against different benchmark cases shows that the approach is as accurate as one-fluid interface-resolving techniques for large/segregated interfaces, while successfully recovering the expected multifluid behaviour for fully dispersed flows. Further, a prototypical multiscale flow has been simulated to demonstrate that the model can effectively switch between large-interface and dispersed-interface mode based on the local flow conditions and mesh size. It is concluded that the present approach represents a promising step towards the development of a comprehensive multiphase model capable of simulating complex multiscale flows of industrial interest.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    53
    References
    2
    Citations
    NaN
    KQI
    []