High-mobility of unconstrained rock avalanches: Numerical simulations of a laboratory experiment and an Argentinian event
High-mobility of unconstrained rock avalanches: Numerical simulations of a laboratory experiment and an Argentinian event
复制标题
无约束岩石雪崩的高流动性：实验室实验和阿根廷事件的数值模拟
DOI：
--
复制
发表时间：
2016
期刊：
影响因子：
--
通讯作者：
M. Jaboyedoff
中科院分区：
文献类型：
--
作者：
I. Manzella;I. Penna;K. Kelfoun;M. Jaboyedoff
岩崩是最危险的滑坡现象之一。它们蕴藏着巨大且极速流动的物质，携带着强大的破坏力，并且达到了极高的机动性。本文展示了阿根廷安第斯山脉 Potrero de Leyes 案例的回溯分析结果。这种岩石雪崩在不受限制的地形上传播，即使没有任何渠道化效应，也显示出极大的流动性。首先使用连续模型 VolcFlow 进行模拟，然后在 Potrero de Leyes 上进行无约束颗粒流实验。结果表明，摩擦模型在实验中效果良好，但它只能在现场重现纵向跳动，并且前提是使用的基础摩擦角接近相应的测量法。在这里，岩石雪崩开始时是一个连贯的质量，仅在传播过程中分解和破碎，横向扩散被高估，粘性行为会产生更好的结果。其特征为近似椭圆形、梯形或舌形（见图1b）。 1903 年在加拿大落基山脉南部发生的 Frank 滑坡（图 1b）就是这种岩石雪崩的一个例子（Cruden 和 Martin，2007；Charrière 等，2016）。最后，低迁移率的特征是倒T形（见图1c），这是由具有相反斜率的高能量耗散冲击引起的。一个典型的例子是 1987 年在意大利发生的 Valpola 岩石雪崩（见图 1c）（Azzoni 等，1992）。图 1. 根据 Nicoletti 和 Sorriso-Valvo (1991) 以及 Corominas (1996) 的分类，岩石崩塌的典型形状和相关实例的草图： a) 沙漏形状和 Pandemonia 溪岩石崩塌（来源：不列颠哥伦比亚省）； b) 梯形形状和弗兰克滑动岩石雪崩（来源：阿尔伯塔社区发展）； c) 倒 T 形和 Valpola 岩石雪崩（来源：Azzoni 等人，1992）。改编自曼泽拉 (2008)。已经开发了几种模型来尝试模拟这些危险和高流动性现象的传播。特别是，由于它们的类似流动的行为，基于非定常流动的圣维南方程和 Savage 和 Hutter (1989) 工作的连续介质力学模型经常被用来模拟它们的传播，例如饥饿（1995）；曼格尼-卡斯特尔诺等人。 （2005）；皮鲁利和曼格尼 (2008)；普达赛尼和哈特 (2007)。为了能够从危险预测的角度使用它们，实验室实验的模型验证和案例历史的回溯分析变得至关重要，例如（康斯坦丁内斯库等人，2011）。它们提高了我们对最合适的流变学以及传播和就位过程的理解。它们帮助我们发现不同规范的优点和局限性，改进其开发和使用，从而改进风险区域的建模和评估。在本研究中，考虑了在日内瓦大学进行的无约束颗粒流实验和 Potrero de Leyes 岩石崩落的情况。选择这个历史案例是因为，即使可用数据有限且准确性不是很高，它也代表了保存完好的无约束岩石雪崩的罕见案例。这个事件的研究非常有趣，因为即使没有通道化效应并且它属于上述中间迁移流类别，它也恰好具有非常低的 Fahrböschung。此外，由于在缺乏地形约束的情况下难以模拟横向扩散，因此不受约束的颗粒流经常给数值建模者带来问题。为了对这些数据进行回溯分析，使用了岩浆与火山实验室（法国克莱蒙费朗）开发的连续介质力学模型 VolcFLow（Kelfoun 和 Druitt，2005 年；Kelfoun 等人，2009 年）。改变了主要参数并测试了不同的流变性，以尝试不仅适应质量的跳动和扩散，而且适应最终沉积物的厚度。这使我们能够提高对无约束流动类型的理解，并了解可用于感兴趣区域的第一近似危险评估的流变学类型。
Rock avalanches are one of the most hazardous landslide phenomena. They involve huge, extremely rapid flowing masses, carrying a strong power of destruction and reaching high mobility. The present article shows results of back-analysis of the Potrero de Leyes case in the Argentinian Andes. This rock avalanche has travelled on an unconstrained topography showing a great mobility, even without any channelizing effects. Simulations are carried out with VolcFlow, a continuum model, first on an unconstrained granular flow experiment and then on Potrero de Leyes. Results show that frictional model works well for the experiment, but that it can only reproduce the longitudinal runout in the field, and this provided that the basal friction angle used is close to the corresponding measured Fahrböschung. Here, where the rock avalanche starts as a coherent mass that disaggregates and shatters only during its propagation, lateral spreading is overestimated and a viscous behavior produces better results. characterized by a nearly oval, trapezoidal or tongue shape (see figure 1b). The Frank slide (fig. 1b), which took place in 1903 in the southern Rocky Mountains of Canada, is an example of this kind of rock avalanche (Cruden and Martin, 2007, Charrière et al., 2016). Finally, the low mobility ones are characterized by a reverse T-shape (see figure 1c), caused by the high energy dissipating impact with the opposite slope. A typical example is the Valpola rock avalanche (see figure 1c), which took place in Italy in 1987 (Azzoni et al., 1992). Figure 1. Sketches of typical shapes and picture of related examples of rock avalanches according to the classification of Nicoletti and Sorriso-Valvo (1991) and Corominas (1996): a) Hour glass shape and Pandemonium creek rock avalanche (source: Province of British Columbia); b) trapezoidal shape and Frank slide rock avalanche (source: Alberta Community Development); c) inverse T shape and Valpola rock avalanche (source: Azzoni et al, 1992). Adapted from Manzella (2008). Several models have been developed to try to simulate the propagation of these dangerous and high-mobile phenomena. In particular, because of their flow-like behaviour, continuum mechanics models based on the St Venant equations of unsteady flow and on the work of Savage and Hutter (1989) are often used to model their propagation, e.g. Hungr (1995); Mangeney-Castelnau et al. (2005); Pirulli and Mangeney (2008); Pudasaini and Hutter (2007). In order to be able to use them in a hazard predictive perspective, models validation on laboratory experiments and back-analysis of cases history become crucial, e.g. (Constantinescu et al., 2011). They improve our understanding of the most suitable rheology and of the propagation and emplacement processes. They help us detecting advantages and limits of the different codes, refining their development and employment and consequently improving modelling and assessment of area at risk. In the present study an unconstrained granular flow experiment carried out at the University of Geneva and the case of Potrero de Leyes rock avalanche were considered. This case history was chosen because, even if data available are limited and the accuracy is not very high, it represents a rare case of a well-preserved unconstrained rock avalanche. This event is very interesting to be studied since even if there is not a channelizing effect and it belongs to the aforementioned category of intermediate mobility flows, it happened to have a very low Fahrböschung. In addition unconstrained granular flows pose often problems to numerical modelers because of the difficulty in simulating lateral spreading with the lack of topographical constrains. To back-analyse these data a continuum mechanics model developed at the Laboratoire Magmas et Volcans (Clermont-Ferrand, France), VolcFLow (Kelfoun and Druitt, 2005, Kelfoun et al., 2009), was used. Main parameters were changed and different rheologies were tested to try to fit not only the runout and spreading of the mass but also the thickness of the final deposit. This has allowed improving our understanding of unconstrained type of flow and to have an idea of the kind of rheology that could be used for a first approximation hazard assessment in the region of interest.