Browsing by Author "Veveakis, Manolis"

This dissertation aims to understand the physics of geomaterial under different loading rates, employing a combination of experimental analysis and theoretical modeling. After obtaining a comprehensive understanding of geophysics, the objective shifts to exploring methods for enhancing the mechanical properties of materials to mitigate or prevent structural failures.

One particular focus in experimental studies is placed on understanding the phenomenon of desiccation, where the volumetric shrinkage rate of the geomaterial can be manipulated by adjusting atmospheric conditions. The experimental results from desiccation tests, supported by triaxial tests, reveal that the behavior of geomaterials is dependent on the loading rate, indicating a rate-dependent response. This observation highlights the need to consider viscoplasticity in the mechanical analysis of these geomaterials. Subsequently, a theoretical mathematical model incorporating viscoplasticity is utilized to describe the stress distribution within the geomaterial. By comparing the predicted locations and the number of stress singularities obtained from the model with the observed locations and the number of cracks in desiccation tests conducted under controlled atmospheric conditions, the effectiveness of the model in capturing the mechanical behavior of the geomaterial is assessed.

Once the mechanical behavior is understood and the corresponding theoretical model is validated, modifications in soil properties can be achieved through adjustments to viscosity. Initially, increasing the viscosity results in the formation of more cracks with narrower spacing. However, as viscosity continues to increase, it eventually leads to the complete prevention of failure. Desiccation experiments containing fluids with varying viscosities were conducted to validate the predicted failure pattern. The experimental results align with the theoretical predictions, providing confirmation of the anticipated behavior.

During desiccation tests conducted on amended soil samples, it was observed that crack development was mitigated, indicating that cracks initially appeared but remained suspended during the dehydration process. Due to the complexity of solving time-dependent partial differential equations with shifting boundary conditions, capillary experiments were introduced to provide insights into the force development within soil particles with the loss of water. The morphology and force development from capillary tests revealed distinct outcomes during dehydration: in the capillary system with distilled water and low viscosity fluid, a rapid force reduction (drop to zero) occurred as the capillary bridges broke, while the presence of high viscosity fluid resulted in a rebound followed by a high attraction force due to bonding formation in the capillary system.

In all, this dissertation offers a novel perspective on describing soil behavior and provides a micro-scale explanation of force development in soil dehydration.

Deep-seated landslides are known as large slides involving millions of cubic meters that move as a rigid block on top of a deep (below the roots of the trees and the groundwater level) basal layer of heavily deformed minerals. This kind of landslides geometrically shears as translational/rotational (depending on the stratigraphy of the area), with very low velocities (cm/year) during long periods (years to tens of years). However, their collapse is usually very sudden, happening within minutes and without previous warning, reaching high velocities up to 120m/s (as the 1963 Vaiont landslide in Italy, \cite{Muller1964}). The catastrophic and fast collapse of this kind of landslides makes the evacuation of the area that is going to be affected almost impossible, thereby possibly causing fatalities and infrastructure damages. Moreover, the lack of understanding of the physical processes behind the mechanisms of failure of this kind of landslides makes the development of reliable early warning systems (or tools/protocols to stop the acceleration of the landslide) challenging, therefore potentially causing significant damages to civil infrastructures. The landslide-prone areas are widespread around the world, having a detrimental fatality rate of tens of thousands. Hence, landslides are a globally threatening natural hazard with disproportional consequences.

This thesis focuses on the understanding of the mechanism of the fast collapse of large deep-seated landslides and provides the first-stage tool for the early warning system. First, is described the Vardoulakis Forecasting Model (VFM), which is a heat energy-based mathematical model that considers that the temperature of the shear band material is critical in the behavior and stability of the landslide. The model contemplates the external and internal factors of this kind of landslide. The external factors of a landslide are considered as the loading conditions, such as groundwater level. And the internal factors of the landslide are focused on the thin shear band, such as the reduction of the friction coefficient of the material, hence, the loss of resistance of the basal material due to continuous friction and cycles of loading-unloading of external forces such as the groundwater. Moreover, the constitutive law used in the VFM theoretically implies that the material of the shear band (usually clay or clay-like material) is rate (velocity) hardening and thermal softening \cite{Vardoulakis2002}, but this assumption has never been tested experimentally. This model has been applied previously by \cite{Veveakis2007} for the case of the famous Vaiont landslide, which collapsed catastrophically in 1963 causing over 2000 fatalities. However, the study did not consider a time-dependence of the loading conditions, and the parameters of the basal material were taken from the literature. This thesis thus presents an extension of the work that the late Professor Vardoulakis and Professor Veveakis started from 2002 until 2007, by implementing the VFM to other case studies with time-dependent loading conditions. Moreover, the present thesis proves the theory that the temperature plays a critical role in the behavior of deep-seated landslides by instrumenting an active deep-seated landslide for the first time, called El Forn landslide (Andorra), with a thermometer in the shear band. The log-samples of this landslide have been studied in the laboratory in different ways, firstly in the triaxial machine to test the theoretical constitutive law of Vardoulakis that the clay material inside the shear band is rate hardening and thermal softening. The tests performed in the triaxial machine have validated for the first time that, indeed, the basal material (as a clay-like material) behaves as Vardoulakis postulated. Furthermore, micro-scale tests, such as X-Ray diffraction, SEM-EDS, MicroCT, and Plasticity Index have been performed to understand the effect of this behavior. Hence, mineralogical, textural, porosity, and plasticity results have been obtained for the samples, and, indeed exists a correlation of why the basal material is velocity and thermal sensitive.

Field data of the El Forn landslide has been obtained, such as the shear band's temperature, groundwater pressure, and displacement of the landslide. The data has demonstrated that, indeed, the temperature of the material of the shear band varies when the pressure changes, and then the landslide accelerates. The field data has shown that for this case study, the material is thermal sensitive when the water pressure varies, not when the landslide accelerates and, due to friction, the material heats.

The VFM model has been applied to four different cases, Vaiont (Italy), Shuping (Three Gorges Dam, China), Mud Creek (California, USA), and the El Forn (Andorra) landslides. The first three landslides have been implemented in the model by using literature data, and the model has reproduced with accuracy the behavior of the three landslides. Finally, the El Forn landslide has been applied to the VFM by implementing field and experimental data, thus reducing the uncertainty of the mathematical model, which accurately reproduces its behavior as well.

The VFM allows to forecast and control deep-seated landslides by using the heat-energy based mathematical model, and the constitutive law. This model works in a dimensionless form of the parameters, to avoid complications in the model by working with so many parameters. Furthermore, this unique model allows accounting in it the external loading and several parameters of the material of the shear band. By taking the heat-diffusion equation in dimensionless form, allows working with only a single dimensionless parameter, that includes the material parameters and the external loading. The single dimensionless parameter is then plotted against the temperature of the shear band (calculated by the model) and is, thus, mapped in the phase space. The phase-space is a curve calculated by the heat equation in the dimensionless form at a steady-state. It is a generic curve for all materials and allows to map the behavior of the landslide with the single dimensionless parameter against the temperature. This mapping allows to locate the creeping stage of the landslide and see if the landslide is close to collapse. Hence, the VFM can become a very useful tool to control and forecast the behavior of a deep-seated landslide and take remediation measures in time.

Knowledge of the drying process of geomaterials is meaningful and helpful in the field of geotechnical and geo-environmental engineering. This experimental study focuses on drying tests on geomaterial samples with monitored surface moisture contents and controlled environmental conditions. Using the digital image correlation (DIC) method to analyze the sample’s displacement, the 3D displacement plots and volumetric strain maps are obtained after calculation. By combining the monitored moisture content with analyzed displacement and volumetric strain plots, the phenomenon and characteristics of a geomaterial’s drying process are discussed and concluded. This study offers a better understanding of the deformation of geomaterials during the drying process in 3D.

Understanding and predicting the behavior of porous media holds unexpected potential for technological advances toward resilience and sustainability. Indeed, these materials are ubiquitous and exhibit a rich palette of processes, both multiphysics and multiscales, which are potential sources of inspiration for engineering design. Along these lines, the intended outcomes of this dissertation are twofold: 1) predicting the resilience of porous media and 2) enhancing behaviors of interest in these materials that could inspire sustainable metamaterials design. Geomaterials, a particularly complex subclass of porous media, will be the primary focus.

This program starts by laying down a general theoretical framework, based on non-equilibrium thermodynamics and differential geometry. A generalized relaxation equation is derived to ensure systematic satisfaction of the second law of thermodynamics. This is associated with a variational framework, based on Fermat's principle, that generalizes that of Onsager, in order to reckon with gyroscopic forces - that is, nondissipative but nonconservative forces.

This framework is then applied to modeling the microstructure of porous media, upon which the behavior of these materials largely depends. To that aim, phase-field modeling is employed to capturing the exact microstructural geometry, in association with digital rock physics based on microtomographic imaging. This effort is required to model processes too complex to be described by a unique constitutive law, such as pressure solution, as studied first in this dissertation. Therein, a microstructural viscosity is derived to capture the kinetics of processes, which is crucial for modeling geomaterials, since the associated timescales span from the engineering to the geological times.

Upon narrowing down the complexity of porous media processes, it is possible to extract the necessary and sufficient microstructural information through morphometry. From running phase-field simulations on a large variety of synthetic microstructures, a general morphometric strength law is inferred, which builds upon seminal works on metals and ceramics. This morphometric framework is applied to predicting the strength of various porous materials, including rocks and bones, from their microstructural geometry.

From batteries to bones, rocks to concrete, porous materials are ubiquitous in the natural and engineered environment, yet remain elusive in their characterization. One of the fundamental challenges of studying porous materials comes down to a fundamental question of linking the microscale to the mesoscale. This work addresses two primary linkages for analysis--the chemical response and the mechanical response of the material. Minkowski functionals served as the primary vessel for understanding how material microstructural geometry ties to macroscale energetics. In the case of chemical systems, Minkowski functionals proved to be powerful predictive tools in both reaction steady states and reaction dynamics. These exponential linkage to morphometers serves as a basis for understanding how the interfacial geometry of system affects the non-mixed chemical behavior of said system over time.As a study on novel simulation frameworks for modeling discrete chemical behavior at the microstructural scale, this work also introduces a unique means for modeling interface chemistry--surface CRNs. Surface CRNs are asynchronous cellular automata models similar to Markov chain models. This class of simulator efficiently translates complex chemical behavior into relatively easy-to-follow reaction rules. This class of simulator has proven to be surprisingly accurate despite its simplicity, creating a strong basis for understanding chemical behavior at a discrete level. While one half of this work focused on the ability of Minkowski functionals to predict chemical behavior, the other half of this work focuses on their ability to link to the mechanics of a microstructure. To address the mechanics problem, Minkowski functionals were extracted from 3D x-ray tomographic scans and assessed mechanically via 3D printed and digitally modeled strength assessments. Ultimately, a deep learning model was trained that could accurately predict and recreate the mechanical response profile of a digitally simulated porous microstructure from just four Minkowski functionals. This extended further to 3D printed samples, allowing for the mechanical behavior of physical samples to be predicted just from its geometric descriptors.

Deep-seated landslides involve the movement of large, slow-moving, masses of soil creeping at often imperceivably slow speeds before catastrophically collapsing. Driven by a combination of natural and anthropogenic factors, these extreme slope failures events pose significant societal risks, manifesting in both economic losses and human casualties. These mass movements slip along a heavily-deformed shearing surface, known as the shear band. Because these sliding surfaces are typically made up of clay, they are especially sensitive to pressure and temperature changes. For several decades, these landslides were monitored via borehole instrumentation, which measured parameters such as pore pressure, displacement, and temperature. However, drilling and instrumenting these boreholes are invasive, labor-intensive, and expensive -- assuming that it is logistically feasible to access these sites, which are often remote and difficult to reach. This thesis offers a detailed exploration and deployment of a novel application of Interferometric Synthetic Aperture Radar (InSAR) data for monitoring and modeling the El Forn landslide in Andorra by merging a physics-based model for deep-seated landslide evolution, borehole data, and InSAR data, utilizing the Alaska Satellite Facility (ASF) Vertex platform and the European Ground Motion Service (EGMS) platform. This work begins by focusing on the performance of InSAR data from the EGMS platform with ASF's On Demand InSAR processing tools, highlighting the trade-off between precision and accuracy in the consideration of different resolution InSAR data. InSAR data is then intertwined with data-driven methodologies for a nuanced modeling of the El Forn landslide in Andorra through the calibration of a temperature-driven physics-based model using in situ data, optimized parameters were obtained to inform the creation of a risk map built through ordinary kriging and InSAR data. In order to create a more holistic metric for landslide stability, this research introduces a population risk mapping tool and consideration of a multi-objective optimization framework. This framework aims to guide communities and policymakers in understanding the risks associated with deep-seated landslides, facilitating informed decisions regarding disaster preparedness.

Landslides are a very common type of disaster. It happens in every state of the U.S and is defined as the movement of the mass of rock, debris, or earth down a slope. Debris flows, sometimes referred to as mudslides, mudflows, lahars, or debris avalanches, are common types of fast-moving landslides(Lynn et al. 1997). When a landslide takes place, it could bring down a large volume of mass which is enough to bury houses and buildings. Therefore, preventing or reducing the life and economic losses comes from landslides is an indispensable task for engineers. There are multiple factors that can cause landslides, including water level, stream erosion, changes in ground water. This paper will focus on examining the associations between rainfall and landslide displacement. The goal will be performing spatial and temporal estimation of landslide displacement in “Valle Febrraro” by using data during the slow-motion stage. We will analyze how we can possibly predict landslide dynamics in Valle Febrraro using precipitation data. We will adopt the concept of correlation coefficient to pinpoint at places where landslide dynamics might be sensitive to precipitation. In fact, we have identified dots with correlation coefficients close to negative 1 through calculation. Those dots are clustered in the lower left half of the selected region. After examining correlation coefficient for every single dot in the chosen space, we adopted kriging as a spatial estimation technique to predict the value of correlation coefficient at every place in the entire chosen space. Results indicate that for the regions where dots with correlation coefficient close to -1 are clustering have values close to -1 whereas those far away have values higher than -1. Besides kriging for the values of correlation coefficient for the entire chosen space, we also performed kriging for the displacement values everywhere inside the same space. We have seen that the results vary from one instant to another. The approach of kriging provides us with what are areas with high displacement values at each of the timestamp and therefore will provide useful insights for future landslide prediction. For temporal estimation, we use regression model to estimate landslide displacement values at one year. Besides, we will also modify the regression model in different ways to see how much better or worse the model will perform. In addition, we will also apply time series prediction technique auto-regression/auto-regressive models, make modifications to it and compare the results. The goal after creating and comparing these models is to perform some related error analysis. Although the results turn to be well in general, we have not noticed any obvious increase in the displacement when we increase the precipitation to several times of its original values when we are trying to optimize the model. Other ways for optimization do exist as we have found out that adjusting the model parameters or performing model for multiple times with each time predicting fewer values do help increase the accuracy.