Head of the Group
Dr. Mohaddeseh Mousavi Nezhad
Dr Mousavi Nezhad is an Associate Professor in Computational Mechanics and Data Science in the School of Engineering of the University of Warwick. She received her PhD from the University of Exeter in 2010. Her doctoral research project was on stochastic modelling of transport process in heterogeneous porous media. After her PhD, she joined the computational mechanics group in the University of Glasgow as an associate research fellow for a project funded by EDF Energy, to investigate crack propagation in nuclear graphite under gas flow and thermal shocks. She is member of Warwick Centre for Predictive Modelling (WCPM) and leads PMPM research group. Her research focuses on modelling flow and transport in heterogeneous deformable porous materials.
Elisa’s research project aims at investigating the solute transport and mixing in the hyporheic zone. It represents the interface between the aquifer and the stream where the flow exchange and mixing between the surface water and groundwater occur. This region provides a significant contribution to the attenuation of pollutants and the self-purification of the river water. Diffusion and mixing are challenging to predict within this critical area hence the present research. The solute transport and mixing in a porous medium under turbulent flow conditions are investigated. A numerical model is developed meant to capture the mentioned phenomena. Our setup is representative of mixing processes taking place within the hyporheic zone and considers the transport of dissolved chemicals close to the interface between a free fluid system and a porous medium. The study is grounded on the use of the random walking technique and an appropriate Lagrangian mixing model. The first one, consisting of the discretisation of the solute in particles, is employed to track the motion of solute particles due to diffusion processes. The second one enables the prediction of the temporal evolution of chemical concentration in the hyporheic zone, as well as the mixing of the solute mass within the porous domain. The theoretical and numerical results are benchmarked against experimental data quantifying the vertical variation of the effective dispersion coefficient with depth below the sediment-water interface.
In fractured reservoirs, constructive interaction with the natural fracture system is critical to the stimulation treatment. To be most effective, hydraulic fractures should cross and connect natural fracture system, but it is possible that arrest, diversion, or offset could occur to hydraulic fractures in the intersections. Nima’s research aims to identify the controls on fracture propagation in deep rocks in order to carry out sealing capacity of faults in fractured reservoirs. The study contains developing analytical and numerical models for fracture propagation in anisotropic and heterogeneous media. Conducting laboratory experiments and outcrop observations to provide primary input data for the models and to validate the numerical results will be also included.
Derek’s research is to develop a new accurate and practical numerical model for deep analyzing the failure propagation in order allow geophysical disasters risk assessment. An advanced computational framework based on Material Point Method, considering random field, into slope failures would be constructed, so as to describe the physical and mechanical behaviours of geophysical and geotechnical hazards such as general slope failures and landslides, from the initiation state (internal stresses concentration) to the post-failure stage (final equilibrium state), to overcome the limitations about modelling and geological mechanisms which is simulated by Finite Element Method. The proposed model will be further developed based on a basic MPM code framework. By doing this, a whole new accurate and practical numerical model for deep analyzing the failure propagation will be established, which give a unique insight for geophysical disasters risk assessment.
Mathematical Groundwater flow models (i.e. subsurface flow, mass transport) are used to investigate a wide variety of hydrogeological conditions such as managing groundwater resources assessment, prediction, and remediation for contamination. To construct a groundwater model Choosing and identifying the appropriate parameters is an imperative step. Inverse methods are broadly applied in the field of hydrology to solve problems regarding groundwater modelling. In general sense, inverse modelling refers to the process of using the results of actual observations/measurements to conclude the values of the model and/or its parameters characterizing the system being modelled. The evaluation of estimated parameter uncertainty is the most focused on in the hydrology/groundwater inverse modelling literature as it is the most relevant factor affecting mass transport prediction. It is of paramount importance to quantify contribution of uncertain parameters in uncertainty of process modelling and this requires developing a reliable theory for rigorous sensitivity analyses which is the aim of this project. The project relies on Global Sensitivity Analysis methods and aims to propose new metrics to quantify influence of uncertainty parameters describing non-linear and turbulent flow, on mixing process modelling.
A better understanding of flow of chemical substances through porous media is essential in the development of many engineering applications. For example, in Carbon Capture and Storage, compressed Liquid CO2 is stored below ground in porous geologic features. Through better understanding, the efficiency of this technology could be enhanced. A key research problem is the requirement of accurate modelling despite uncertainty in both flow behaviour and porous media structural parameters – information on these two key components is often scarcely available. The ultimate goal of Matthew’s research is to produce a model which provides probabilistic information on local tracer concentration evolutions through a porous media sample. To do so a probability density function method will be utilized which accounts for advective transport, pore-scale dispersion and chemical fluid phase reactions. A multi-level stochastic method will be used which allows a combination of expensive and cheap approximate solvers to achieve accurate outputs more efficiently.
Due to significant variability of rock types, an extensive amount of information about the micro-structure of the rocks are required for their characterisation. My research aims to investigate and characterise different rock types and develop models for predicting hydromechanical properties of the rocks. In this work, image analysis methods are used to classify the rock types. Rock features are extracted from 2D X-ray images which include the spatial distribution of mineralogy, homogeneity and pore size distribution of samples. Machine Learning algorithms are used to train and test classification data and develop permeability predictive models. The permeability models will be implemented in an existing finite element algorithm to assess their performance for simulation of hydromechanical behaviour of the rocks. The results of this research are expected to develop an improved measure for evaluating structural integrity of earth structures.