MoST-DFG Collaboration - Theoretical, numerical and experimental investigations of gravity-driven fluid-granular mixture flows

流体-颗粒混合流在工业过程中的各种应用和对泥石流等自然灾害的预测都是推动流动的因素。泥石流是一种重力驱动的泥沙颗粒和水的流动，它部分或过度地填充颗粒物质的空隙。尽管在地球物理质量流的模拟、计算和实验方面取得了进展，但此类流体-颗粒多相流的预测仍然是一个最具挑战性的课题，主要体现在两个方面：(1)理论和数值模拟；(2)实验研究。一方面，在大多数连续介质模型中，这种混合流通常被视为单相介质，尽管它们显然是两相混合物。在这种单相方法中，碎屑混合物要么被视为包含某些塑性行为的非牛顿流体，要么被视为库仑连续体，其中间隙流体的影响被参数地结合在一起，或者被视为具有相同组成速度的混合物。很少有多组分混合物模型被构建并应用于泥石流。一些现有的模型考虑了这样的多相概念，即流体饱和的颗粒混合物，尽管自然颗粒流通常不是完全饱和的，或者被水过饱和。另一方面，同时测量流体-颗粒混合流中所有组分的速度和体积分数也是一项困难的任务。虽然许多实验研究讨论了间隙流体对颗粒流动的影响，但由于测量颗粒-流体混合物中流体相的动力学是困难的，所以只测量颗粒介质的动力学。在本项目中，我们将尝试建立一个流体-颗粒两相模型，该模型能够通过两层方法来描述这种欠饱和和过饱和混合物及其转变，其中流体饱和的颗粒下层被纯颗粒上层覆盖，对于欠饱和情况，或者被纯流体覆盖，对于过饱和情况。在实验研究方面，将发展PIV和PTV耦合的间接图像测量技术，用于测量动态混合流中的流体速度、颗粒位置、速度和体积分数。在这种耦合方法中，PIV用于同时测量流体速度，PTV用于测量颗粒速度。其优点是，PIV方法对识别流体相的播撒颗粒的识别精度较高，而PTV方法比PIV方法能提供更详细的颗粒速度信息。理论模型将用任意高阶精度的间断伽辽金(DG)方法对各种欠饱和或过饱和颗粒流体混合物的重力驱动流动进行数值检验，并通过实验结果进行验证。

Fluid-granular mixture flows are motivated by various applications in industrial processes and predictions of natural hazards such as debris flows. A debris flow represents a gravity-driven flow of sediment particles and water, which fills the interstices of granular material partially or excessively. Despite the past developments in modeling, computing and experimenting geophysical mass flows, the prediction of such fluid-granular multiphase flows is still a most challenging topic mainly in two aspects (i) theoretical and numerical modeling and (ii) experimental investigation. On the one hand, in most continuum models, such mixture flows are often treated as a single-phase medium, even though they are clearly two-phase mixture. In such single-phase approaches, the debris mixture is considered either as a non-Newtonian fluid including some plastic behavior or as a Coulomb continuum where the effect of the interstitial fluid is incorporated parametrically or as a mixture with the same constituent velocities. Very rarely multi-constituent mixture models are constructed and applied to debris flows. A few existing models consider such multiphase concepts as fluid-saturated granular mixtures, although natural granular flows are often not fully saturated or else over-saturated by water. On the other hand, the simultaneous measurement of velocities and volume fractions of all constituents in a fluid-granular mixture flow is also a difficult task. Although many experimental studies discuss the effect of the interstitial fluid on the granular flow, only the dynamics of the granular medium is measured, because of the difficulty in measuring the dynamics of the fluid phase in a granular-fluid mixture.With this project, we will attempt to develop a fluid-granular two-phase model, which is able to describe such under-saturated and over-saturated mixtures and their transition by means of a two-layer approach, in which the fluid-saturated granular lower layer is overlain by either the pure granular upper layer, for the under-saturated case, or the pure fluid, for the over-saturated case. For experimental investigations, the indirect image measurement technique will be developed, by coupling PIV and PTV, to measure the fluid velocities, the granular positions, velocities and volume fractions during the dynamic mixture flow. In this coupling method, PIV will be used for simultaneously measuring the fluid velocities, and PTV for measuring the velocities of granules. The benefits are that the accuracy of the PIV method is high for discriminating the seeding particles that identifying the fluid phase, while the PTV can offer more detailed information about granular velocities than PIV. The theoretical model will be examined numerically by the discontinuous Galerkin (DG) method of an arbitrary high-order accuracy in terms of gravity-driven flows of various under- or over-saturated granular-fluid mixtures and validated by experimental results.