Granular flow rheology; the key to understanding the exceptional mobility of pyroclastic density currents

火山碎屑密度流（PDC）是火山岩、浮石、火山灰和气体的热雪崩，沿着火山的侧翼下降。他们可以摧毁和掩埋100平方公里的地形。其高温、固有的流动性和不可预测的性质使其成为最危险的火山现象之一。自公元1600年以来，火山碎屑流已导致超过90，000人死亡，占所有火山死亡人数的33%，使其成为火山死亡的最大原因。预测火山碎屑密度流在给定火山的流动路径和淹没程度取决于我们对以下方面的理解：（i）所涉及的流动机制;（ii）开发能够忠实地捕捉这些流动的动态性质并准确模拟过去事件的模型;以及（iii）概率地应用这些模型，这样，就可以考虑到某一火山未来所有可能的情况，以便绘制概率危险图。在这里，我们将处理（i）和（ii），但我们在（iii）中的跟踪历史表明了我们的长期意图。因此，这项研究的基本原理源于强烈的最终用户定义的需求，以及推动这些复杂的多相（颗粒和气体）自然流动科学的动机。因此，本研究的目的是提高预测火山周围的火山碎屑密度流淹没区的能力，在理解流动行为和流变性质的变化，因为它propagations.During流动之间的相互作用取得突破，火山碎屑密度流逐步发展的地区，不同的物理性质和流动机制。通常，流动在底部形成高颗粒浓度，颗粒之间存在摩擦或碰撞接触。一个压倒一切的灰云在此之上发展，颗粒浓度低，大多数颗粒由热气体的湍流对流支撑。当流动在地形上传播时，这些上部和下部区域对坡度和山谷限制的变化有不同的响应。上、下单元的加速、减速和扩展发生在不同的点上，并且可以诱导流动分离。这些上层灰云倾向于从母基流中分离出来，并向意想不到的方向移动，这往往会导致致命的后果。这项研究将侧重于了解流变变化的基础颗粒流，并将考虑如何，反过来，调制质量通量到压倒性的灰云。我们将测试的假设，在基础的暗流流变学的变化，部分地形引起的，导致孔隙流体压力波动，饲料的产生和分离的上层湍流灰云从他们的父母暗流。我们将通过整合从互补领域，地貌，实验和计算研究中获得的数据，特别是利用为工程应用开发的尖端建模工具来实现这一目标。我们将建立在对工业颗粒流的理解的重要新进展的基础上，以了解流动流变学如何变化（通过时间和空间），以及是什么控制了这些变化。我们的研究结果将形成一个新的本构流变学描述的基础上，提供了一个基本的一步，允许从目前用于危险量化的流动模拟工具中采用的流动平均流变学定律的进步。这项工作的扩展，特别是新一代模拟工具的应用将产生具有较低相关不确定性的危险图。使用我们已经开发的概率风险映射方法，我们将量化改进的程度。该项目是及时的，并将受益于与一个主要的基于汉堡的工业颗粒流倡议的协同作用，以及项目合作伙伴正在进行的研究。

Pyroclastic density currents (PDCs) are hot avalanches of volcanic rock, pumice, ash and gas that descend the flanks of volcanoes. They can destroy and bury 100's km2 of terrain. Their high temperatures, inherent mobility and unpredictable nature render them one of the most hazardous volcanic phenomena. Since 1600AD, pyroclastic flows have resulted in over 90,000 deaths, 33% of all volcanic fatalities recorded, making them the single biggest cause of death at volcanoes. Forecasting the flow paths and the extent of inundation by pyroclastic density currents at a given volcano depends on our understanding of (i) the flow mechanisms involved (ii) developing models that can faithfully capture the dynamic nature of those flows and accurately simulate past events, and (iii) applying those models probabilistically, so that all possible future scenarios at a given volcano can be considered in order to generate probabilistic hazard maps. Here we will tackle (i) and (ii), but our track history in (iii) demonstrates our longer-term intention. The rationale for this research therefore stems from both a strong end-user defined need, as well as motivation to advance the science of these complex multiphase (particle and gas) natural flows. The aim of this research is therefore to improve the capability of forecasting pyroclastic density current inundation zones around volcanoes by making breakthroughs in understanding the interplay between flow behaviour and how the rheological nature of the flow changes as it propagates.During flow, pyroclastic density currents progressively develop regions that vary in their physical nature and flow mechanisms. Typically, the flows develop high particle concentrations at the base, with frictional or collisional contacts between the particles. An overriding ash cloud develops above this, where particle concentration is low and most particles are supported by turbulent convection of hot gases. As the flows propagate over topography, these upper and lower regions respond differently to changes in slope and valley confinement. Acceleration, deceleration and spreading of the upper and lower units occur at different points, and flow separation can be induced. The propensity for these upper ash clouds to separate from the parent basal flow and travel in unexpected directions often results in lethal consequences. This research will focus on understanding the rheological variations in the basal granular flow and will consider how it may, in turn, modulate mass flux into the overriding ash cloud. We will test the hypothesis that variations in the basal undercurrent rheology, in part induced by topography, result in pore fluid pressure fluctuations that feed the generation and separation of upper turbulent ash clouds from their parent undercurrents. We will achieve this by integrating data obtained from complementary field, geomorphological, experimental and computational studies, in particular utilising cutting-edge modelling tools developed for engineering applications. We will build on important new advances in the understanding of industrial granular flows to characterise how flow rheology varies (through time and space), and what controls those variations. Our results will form the basis for a new constitutive rheology description, providing a fundamental step forward by allowing advance from flow-averaged rheology laws currently employed in flow simulation tools used for hazard quantification. Extensions of this work, in particular the application of the new generation simulation tools will produce hazard maps that have lower associated uncertainties. Using methods we have already developed for probabilistic hazard mapping, we will quantify that degree of improvement. The project is timely and will benefit from synergy with a major Edinburgh-based initiative on industrial granular flows, as well as ongoing research by project partners.