Research / Flow and Transport in Porous Media
System Application to Non-destructive Study of Coupled Flow in Porous Media
Files marked with a lock are password-protected. Students: use the username and password given in class.
Mesut Pervizpour's Dissertation (PDF)
Table of Contents
Dissertation Committee Members:
Abstract
The problem of coupled flow in porous media under combined thermal and hydraulic gradients is re-captured/stated in terms of transient responses and a non-destructive testing approach under closed boundaries. The analytical models developed, numerical simulation engines programmed and optimization routines incorporated for a system approach in non-destructive determination of coupled flow phenomenological coefficients (Fig 1).
The study is unique due a list and combination of factors such as:
Experimental, Numerical and Verification System Responses
Overview
A synopsis of the research will be included soon, please refer to the PDF copy of the dissertation: Mesut Pervizpour's Dissertation in PDF
To Summarizes The Efforts for This Study:
The "meaningful" requirement is that the system provides a reasonably good representation of the actual phenomena. An oversimplified model may provide results and conclusions that do not apply to the real phenomenon being modeled. An overcomplicated one may constrain potential applications, render theory too difficult to be useful, and strain available computational resources. Perhaps the most distinguishing characteristic between an average development and an outstanding one is the ability to provide such a system view encapsulating all and providing a good balance between complexity and accuracy.
Table of Contents
Dissertation Committee Members:
| » Prof. Sibel Pamukcu | (Dissertation Advisor), Civil & Environ. Eng. Dept., Lehigh University, Bethlehem, PA, 18015 |
| » Prof. G. P. Lennon | (Committee Chairman), Civil & Environ. Eng. Dept., Lehigh University, Bethlehem, PA, 18015 |
| » Prof. Richard Sause | Civil & Environ. Eng. Dept., Lehigh University, Bethlehem, PA, 18015 |
| » Prof. Fazil Erdogan | Mech. Eng. & Mechanics Dept., Lehigh University, Bethlehem, PA, 18015 |
| » Prof. Hugo S. Caram | Chemical Eng. Dept., Lehigh University, Bethlehem, PA, 18015 |
| » Dr. Scott A. Raschke |
Abstract
The problem of coupled flow in porous media under combined thermal and hydraulic gradients is re-captured/stated in terms of transient responses and a non-destructive testing approach under closed boundaries. The analytical models developed, numerical simulation engines programmed and optimization routines incorporated for a system approach in non-destructive determination of coupled flow phenomenological coefficients (Fig 1).
The study is unique due a list and combination of factors such as:
- Non-destructive test approach to preserve soil properties and constancy
- Controled application of hydraulic and thermal graidents at boundaries
- Measurement and use of transient state parameters
- Analytical representation of coupled transient fluxes
- Numerical simulation agent representing the non-destructive model
- Parameter estimation in collocation of experimental & simulated responses for critical parameter determination
- a) Non-destructive testing method and tool development:
A transient testing method and setup (Fig 1) (one-dimensional (Fig 2) and triaxial) is developed for application and response measurement of transient coupled hydraulic and thermal fields in porous media. The transient pressure and temperature measurements under various types of applied gradients (check here for responses) are measured without disturbing or changing the constituents (Fig 3) of the original sample repeatedly.(Fig 4,Fig 5,Fig 6)
- b) Mathematical modeling and representation of transient non-destructive phenomena:
Mathematical modeling of the generalized governing transient equations is based on non-equilibrium thermodynamics, deformable porous media and conservation equations. Generalized forms of governing equations are developed (check chapter 3 of dissertation for details) that are capable of expressing the non-destructive test and applied boundary conditions.
- c) Numerical modeling and simulations of the transient phenomena by programming:
The developed generalized transient coupled governing partial differential equations (PDEs) (Fig 1, Fig 2) are used in numerical simulations with boundary conditions identical to experimental conditions. The formulation of the numerical model (Method of Lines Fig 3) was based on converting the governing PDE's in to ODE's (discretization Fig 4) that were solved by implicit integration (Fig 5) in the simulations (Fig 6) by the codes developed in FORTRAN and C++.
- d) Numerical optimization and programming for parameter identification:
A parameter estimation (Fig 1) process based on the optimization of an objective function was used to determine the phenomenological coefficients of the porous media. The objective function was based on the collocated transient pressure and temperature responses from non-destructive tests and numerical simulations (Fig 2). The Quasi-Newton gradient-based search method (Fig 3, Fig 4) was implemented numerically. The collocation of the transient pressure and temperature responses from non-destructive tests and the repeated simulations of the numerical model generating updated simulation responses satisfied the required level of convergence and helped determine the coupled flow parameters of the soil specimen (Fig 5). However, the search process was not conducted on all parameters at once as in a curve-fit approach. Instead a system approach was used based on the non-destructive mind-set by employing different gradient combinations and test results in uniquely identifying each set of phenomenological coefficient. The approach permitted verification of the non-destructive process developed by uniquely identifying each parameter.
Experimental, Numerical and Verification System Responses
Overview
A synopsis of the research will be included soon, please refer to the PDF copy of the dissertation: Mesut Pervizpour's Dissertation in PDF
To Summarizes The Efforts for This Study:
The "meaningful" requirement is that the system provides a reasonably good representation of the actual phenomena. An oversimplified model may provide results and conclusions that do not apply to the real phenomenon being modeled. An overcomplicated one may constrain potential applications, render theory too difficult to be useful, and strain available computational resources. Perhaps the most distinguishing characteristic between an average development and an outstanding one is the ability to provide such a system view encapsulating all and providing a good balance between complexity and accuracy.