Multi-scale modeling and finite element simulation of diffusion in porous media
Doctoral thesis, 2019
Porous media comprise a large range of natural and industrial materials and are typically complex on multiple length scales. At lower (micro-scopic) length scales, such media consist of a solid skeleton and fluid-filled pores in between. At higher (macro-scopic) length scales, transport of a migrating pore fluid can be observed in addition to the mechanical stress-strain response. The interaction between the pore fluid and the solid skeleton determines, in addition to the properties of the individual phases, the fully coupled response of the medium. Therefore, the investigation of macro-scale properties needs to take into account the processes between micro- and meso-scopic heterogeneities.
This thesis investigates transport processes in porous media on multiple scales by applying computational homogenization and finite element simulation. Biot’s equations
of (linear) consolidation are introduced and combined with sharp and diffuse interface formulations that are established to investigate the effect of meso-scale heterogeneities, e.g. in form of fractures, on the overall material behavior. The scale transition of the heterogeneous porous meso-scale towards an homogeneous macro-scale problem is derived via Variationally Consistent Computational Homogenization.
Taking into account this modeling framework, the thesis investigates:
1) How the numerical modeling of heterogeneous porous media can be used to calibrate laboratory experiments.
2) How the fluid transport in fractured rock can be modeled by applying sharp and diffuse interface formulations.
3) How the effective diffusivity of porous media can be derived for the special case of three-phase concrete, where diffusion takes place preferably on interfaces in the structure. In this industrial porous material, fluid transport is not a relevant process but rather the diffusion of chloride ions in the fluid phase is of interest.
All in all, the thesis reveals limits and establishes possibilities by the numerical modeling of heterogeneities in porous media with the aim to provide a deeper understanding of the transport processes in porous media on multiple scales.
This thesis investigates transport processes in porous media on multiple scales by applying computational homogenization and finite element simulation. Biot’s equations
of (linear) consolidation are introduced and combined with sharp and diffuse interface formulations that are established to investigate the effect of meso-scale heterogeneities, e.g. in form of fractures, on the overall material behavior. The scale transition of the heterogeneous porous meso-scale towards an homogeneous macro-scale problem is derived via Variationally Consistent Computational Homogenization.
Taking into account this modeling framework, the thesis investigates:
1) How the numerical modeling of heterogeneous porous media can be used to calibrate laboratory experiments.
2) How the fluid transport in fractured rock can be modeled by applying sharp and diffuse interface formulations.
3) How the effective diffusivity of porous media can be derived for the special case of three-phase concrete, where diffusion takes place preferably on interfaces in the structure. In this industrial porous material, fluid transport is not a relevant process but rather the diffusion of chloride ions in the fluid phase is of interest.
All in all, the thesis reveals limits and establishes possibilities by the numerical modeling of heterogeneities in porous media with the aim to provide a deeper understanding of the transport processes in porous media on multiple scales.
Hydro-Mechanical Coupling
Fractured Rock
Poroelasticity
Three-Phase Concrete
Computational Homogenization
VDL, Tvärgata 4C, Chalmers
Opponent: Professor Marc-André Keip, Institute of Applied Mechanics (CE), University of Stuttgart, Germany
Author
Nele Pollmann
Chalmers, Industrial and Materials Science, Material and Computational Mechanics
N. Pollmann, R. Jänicke, J. Renner, H. Steeb; Numerical Investigation of the Effective Skempton Coefficient in Porous Rock Containing Fluid-Filled Fracture Networks
N. Pollmann, R. Jänicke, J. Renner, H. Steeb; Experimental determination of the Skempton coefficient: Challenges and opportunities
Diffuse interface modeling and variationally consistent homogenization of fluid transport in fractured porous media
European Journal of Mechanics, A/Solids,; Vol. 84(2020)
Journal article
N. Pollmann, F. Larsson, K. Runesson, K. Lundgren, K. Zandi, R. Jänicke; Modeling and computational homogenization of chloride diffusion in three-phase meso-scale concrete
Porous media comprise a large range of natural or industrial materials with a broad spectrum of applications. It is part of our everyday life e.g. in form of bread, a simple rinsing sponge or a brick used for building. Thus, it has always been part of our environment.
Porous media are highly complex on different scales which gets clearer with the aid of an example such as sandstone. When hiking in the Elbe Sandstone Mountains one could take a closer look at the rock and get aware of small and large fractures as a form of heterogeneities. Taking an even closer look, e.g. with a magnifier, small grains are visible such as in sand.
Of special interest is the behavior of this material if it is saturated with fluid (e.g. water, oil, gas). If the porous media is filled with fluid, transport processes occur, e.g in form of fluid flow. This is, among others, important for geothermal energy production, environmental remediation or the corrosion process of reinforced concrete buildings (due to de-icing fluid).
During the production of deep geothermal energy e.g., the rock is hydraulically stimulated, i.e. fractures are induced by pumping water under high pressure in the ground. These fractures enhance the conductivity of the rock, which is used for energy production. The hydraulic stimulation of rock causes seismic attenuation, i.e. small to large earthquakes in the reservoir. Therefore, the ability to detect, to understand and to simulate seismic attenuation helps decision makers to forecast whether or not a rock region is suitable for hydraulic stimulation.
This thesis numerical examines the processes in porous media, to gain results that can e.g. be used to interpret field data such as of seismic exploration. Therefore different numerical approaches are investigated on the basis of the Theory of Porous Media and validated against suitable benchmarks. In summary the thesis provides a deeper understanding of diffusion processes in porous media and investigates a suitable toolbox to compute the overall material behavior of porous media, taking into account heterogeneities such as fractures in rock or aggregate content in concrete, on different scales.
Porous media are highly complex on different scales which gets clearer with the aid of an example such as sandstone. When hiking in the Elbe Sandstone Mountains one could take a closer look at the rock and get aware of small and large fractures as a form of heterogeneities. Taking an even closer look, e.g. with a magnifier, small grains are visible such as in sand.
Of special interest is the behavior of this material if it is saturated with fluid (e.g. water, oil, gas). If the porous media is filled with fluid, transport processes occur, e.g in form of fluid flow. This is, among others, important for geothermal energy production, environmental remediation or the corrosion process of reinforced concrete buildings (due to de-icing fluid).
During the production of deep geothermal energy e.g., the rock is hydraulically stimulated, i.e. fractures are induced by pumping water under high pressure in the ground. These fractures enhance the conductivity of the rock, which is used for energy production. The hydraulic stimulation of rock causes seismic attenuation, i.e. small to large earthquakes in the reservoir. Therefore, the ability to detect, to understand and to simulate seismic attenuation helps decision makers to forecast whether or not a rock region is suitable for hydraulic stimulation.
This thesis numerical examines the processes in porous media, to gain results that can e.g. be used to interpret field data such as of seismic exploration. Therefore different numerical approaches are investigated on the basis of the Theory of Porous Media and validated against suitable benchmarks. In summary the thesis provides a deeper understanding of diffusion processes in porous media and investigates a suitable toolbox to compute the overall material behavior of porous media, taking into account heterogeneities such as fractures in rock or aggregate content in concrete, on different scales.
Subject Categories
Mechanical Engineering
Areas of Advance
Building Futures (2010-2018)
ISBN
978-91-7905-170-9
Doktorsavhandlingar vid Chalmers tekniska högskola. Ny serie: 4637
Publisher
Chalmers
VDL, Tvärgata 4C, Chalmers
Opponent: Professor Marc-André Keip, Institute of Applied Mechanics (CE), University of Stuttgart, Germany