A Dilatant Two-Fluid Debris Flow Model for Hazard Analysis in Changing Mountain Environments
Open access
Author
Date
2023Type
- Doctoral Thesis
ETH Bibliography
yes
Altmetrics
Abstract
Gravitationally driven flows of mud and sediment debris are causing a growing threat to mountain populations. Rock/ice avalanches, glacier lake outburst floods (GLOFs) and debris flows are in- creasingly a result of a global temperature rise, which is leading directly to the thawing of mountain permafrost and melting of glaciers. When coupled with extreme precipitation events, the mobiliza- tion of loose sediments leads to dangerous water-saturated flows that can cause human fatalities and severe infrastructure damage. Understanding the dynamics of debris flows is essential to develop land planning and technical measures to protect mountain communities.
Numerical modeling of debris flows provides hazard engineers with a predictive tool to help plan and construct mitigation measures, including developing real-time warning systems. With the recent increase in computer power, it is now possible to simulate debris flow motion from initiation to run-out. Numerical modeling therefore links initial conditions, including precipitation and sediment availability, to flow conditions in the torrent and run-out fan. Despite recent progress, the application of numerical models is limited by the lack of understanding of the general kinematical behavior of two-phase flows, including the frictional interaction of fluid-solid mixtures with the basal surface, as well as the shearing interaction between the solid and fluid phases.
The rheological problem is compounded by the complex interaction of the debris with the basal surface leading to bed erosion. Modeling debris flow motion with entrainment involves accurately predicting the interplay between debris flow composition (time evolution of the solid-fluid components) coupled with the geological setting. The lack of understanding of these complex physical processes and geo-mechanical feedbacks is preventing a reliable, and predictive, application of debris flow models in engineering practice. In this thesis, we develop, test and calibrate a depth-averaged, two-fluid debris flow model with erosion and dilatancy. The model consists of six partial differential equations governing mass, mo- mentum, and granular temperature. The key feature of the model is to include dilatant effects that are associated with the shearing of granular debris. We assume that the expansive and contracting action of the solid volume under shearing governs the volume of the pore space and therefore the behavior of the interstitial fluid. Moreover, the dilatant response of the solid matrix governs mass and momentum exchanges between the two layers. The model is validated using actual, field-scale debris flow measurements from the Swiss Illgraben test site (Wallis, Switzerland).
The comparison between Illgraben data and the simulations reveals that the model reproduces the space-time evolution of the solid/fluid flow composition in the streamwise direction, meaning from the leading edge of the debris flow to the debris flow tail. Following the model development, we investigate the connection between the flow composition and the flow rheology. As the well-known debris flow scientist Iverson argued [7], it is impossible to accurately simulate a debris flow from initiation to runout using a fixed and uniform rheology.
Based on Illgraben data, as well as on careful laboratory experiments, we find a physical formulation where the flow composition evolution governs the rheological changes. We propose two different methods to compute the frictional resistance from the solid/fluid composition of the flow. They are conceptually different but mathematically equivalent to large degrees. We test and compare these two approaches using data from the torrent of Ritigraben (Wallis, Switzerland). Additionally, we i introduce an erosion model adapted for two-layer debris flow, in which erosion is also a function of the flow composition. Drone flights performed before and after a specific debris flow event in Ritigraben provide us with reliable and rare erosion data, which allows us to test and calibrate the rheological and erosion models. The numerical results are in good agreement with the field measurements. In the final theme presented in this thesis, we validate the proposed two-layer debris flow model using several well-documented case events. We study three GLOFs, two that occurred in Peru (Lake 513, 2010 and Lake Palcacocha, 1941) and one in Kyrgyzstan (Lake Uchitel in the Aksay Valley). The event documentation contains precise field observations, making the direct comparison and analysis of the model results consistent and insightful. A characteristic of long-running GLOFs is the transformation in the flow composition from initiation to run-out. Directly after initiation, usually from a moraine collapse, the flow consists primarily of fluid. However, as the flow acquires solid material via bed erosion, the flow transforms from a mud flow to a type of granular debris flow. Conversely, as the slope flattens, the solid matrix collapses (contracts) and therefore stops while the fluid which is washed-out of the interstitial pore space and continues to flow downstream.
As the proposed frictional rheology depends on the internal solid-fluid composition, modeling these transitions is crucial for a model’s accuracy. We demonstrate that the model can reproduce the entire flowing behavior of all three events, for which many of flowing transitions and phase separa- tions occurred. The back-calculation of the three GLOF events is performed using the same set of rheological parameters for all three case studies. In summary, we develop a dilatancy-based, two-fluid debris flow model that captures flow regime transitions based on the evolution of the solid-fluid composition. This composition depends directly on erosion and deposition processes which are driven in turn by the rheological behavior and the terrain topography. We test the proposed model on real-scale field data obtained from two debris flow sites in Switzerland. Finally, we apply the model to simulate well-documented GLOF events. Future applications will help to determine if the proposed model can be used for practical hazard analysis in changing mountain environments. Show more
Permanent link
https://doi.org/10.3929/ethz-b-000623278Publication status
publishedExternal links
Search print copy at ETH Library
Contributors
Examiner: Müller, Christoph
Examiner: Kaitna, R.
Examiner: Bartelt, P.
Examiner: McArdell, B.
Publisher
ETH ZurichSubject
Debris flow modelingOrganisational unit
03865 - Müller, Christoph R. / Müller, Christoph R.
More
Show all metadata
ETH Bibliography
yes
Altmetrics