06 Fakultät Luft- und Raumfahrttechnik und Geodäsie

Permanent URI for this collectionhttps://elib.uni-stuttgart.de/handle/11682/7

Browse

Publication Type: search.filters.UBSTyp.ZeitschriftenartikelInstitute: search.filters.UBSInstitut.Institut für Statik und Dynamik der Luft- und RaumfahrtkonstruktionenSubject: search.filters.subject.624

Search Results

Now showing 1 - 4 of 4
  • Thumbnail Image
    ItemOpen Access
    Optimization of the groundwater remediation process using a coupled genetic algorithm-finite difference method
    (2021) Seyedpour, Seyed Morteza; Valizadeh, Iman; Kirmizakis, Panagiotis; Doherty, Rory; Ricken, Tim
    In situ chemical oxidation using permanganate as an oxidant is a remediation technique often used to treat contaminated groundwater. In this paper, groundwater flow with a full hydraulic conductivity tensor and remediation process through in situ chemical oxidation are simulated. The numerical approach was verified with a physical sandbox experiment and analytical solution for 2D advection-diffusion with a first-order decay rate constant. The numerical results were in good agreement with the results of physical sandbox model and the analytical solution. The developed model was applied to two different studies, using multi-objective genetic algorithm to optimise remediation design. In order to reach the optimised design, three objectives considering three constraints were defined. The time to reach the desired concentration and remediation cost regarding the number of required oxidant sources in the optimised design was less than any arbitrary design.
  • Thumbnail Image
    ItemOpen Access
    Affine transformations accelerate the training of physics-informed neural networks of a one-dimensional consolidation problem
    (2023) Mandl, Luis; Mielke, André; Seyedpour, Seyed Morteza; Ricken, Tim
    Physics-informed neural networks (PINNs) leverage data and knowledge about a problem. They provide a nonnumerical pathway to solving partial differential equations by expressing the field solution as an artificial neural network. This approach has been applied successfully to various types of differential equations. A major area of research on PINNs is the application to coupled partial differential equations in particular, and a general breakthrough is still lacking. In coupled equations, the optimization operates in a critical conflict between boundary conditions and the underlying equations, which often requires either many iterations or complex schemes to avoid trivial solutions and to achieve convergence. We provide empirical evidence for the mitigation of bad initial conditioning in PINNs for solving one-dimensional consolidation problems of porous media through the introduction of affine transformations after the classical output layer of artificial neural network architectures, effectively accelerating the training process. These affine physics-informed neural networks (AfPINNs) then produce nontrivial and accurate field solutions even in parameter spaces with diverging orders of magnitude. On average, AfPINNs show the ability to improve the L2relative error by 64.84%after 25,000 epochs for a one-dimensional consolidation problem based on Biot’s theory, and an average improvement by 58.80%with a transfer approach to the theory of porous media.
  • Thumbnail Image
    ItemOpen Access
    Uncertainty with varying subsurface permeabilities reduced using coupled Random Field and extended Theory of Porous Media contaminant transport models
    (2022) Seyedpour, Seyed Morteza; Henning, Carla; Kirmizakis, Panagiotis; Herbrandt, Swetlana; Ickstadt, Katja; Doherty, Rory; Ricken, Tim
    To maximize the usefulness of groundwater flow models for the protection of aquifers and abstraction wells, it is necessary to identify and decrease the uncertainty associated with the major parameters such as permeability. To do this, there is a need to develop set of estimates representing subsurface heterogeneity or representative soil permeability estimates. Here, we use a coupled Random Field and extended Theory of Porous Media (eTPM) simulation to develop a robust model with a good predictive ability that reduces uncertainty. The coupled model is then validated with a physical sandbox experiment. Uncertainty is reduced by using 500 realisations of the permeability parameter using the eTPM approach. A multi-layer contaminant transport scenario with varying permeabilities, similar to what could be expected with shallow alluvial sediments, is simulated. The results show that the contaminant arrival time could be strongly affected by random field realizations of permeability compared with a modelled homogenous permeability parameter. The breakthrough time for heterogeneous permeabilities is shorter than the homogeneous condition. Using the 75% confidence interval (CI), the average contaminant concentration shows 4.4% variation from the average values of the considered area and 8.9% variation in the case of a 95% confidence interval.
  • Thumbnail Image
    ItemOpen Access