19–22 May 2026
Europe/Paris timezone

PINNs enhanced multi-resolution modeling of laminar vortex dynamic process in pore-scale MICP

20 May 2026, 15:05
15m
Oral Presentation (MS04) Biological Processes in Porous Media MS04

Speaker

Dianlei Feng (Tongji University)

Description

Microbial mineralization is a novel bioremediation and consolidation technology. However, its mineralization process is influenced by a variety of complex factors (such as microbial species, urea concentration, and the evolution and distribution of pore vortex structures) at the pore scale, presenting highly nonlinear characteristics and a certain degree of uncertainty in distribution and evolution. Due to the difficulty in real-time observation of the reaction-flow coupling process and the dynamic changes of pore structure in the pore space, experimental studies are hard to deeply explore the micro-mineralization mechanism at the pore scale. Based on the lattice Boltzmann method (LBM), Eulerian finite element method (FEM), and cellular automata (CA), this study constructed a multi-physics coupling numerical model for pore-scale microbial mineralization. High-resolution numerical simulation in space was achieved by using the Physical Informed Neural Network (PINNs) method. Combining GPU parallel acceleration technology, a three-dimensional complex pore flow-reaction coupled universal multi-physics field solver was independently developed. The full-scale mineralization process simulation of three-dimensional microfluidic chips was successfully realized, and the experimental phenomena of the microfluidic chips were quantitatively reproduced. Based on the verified model, the distribution law of calcium carbonate precipitation and the influence of the initial pore structure on its evolution process were quantitatively analyzed, providing quantitative suggestions and prediction tools for optimizing the biological grouting strategy. The mechanism of the vortex phenomenon caused by the dynamic evolution of pore structure and its influence on the uniformity of mineralization were further explored. Through quantitative analysis of the vortex evolution distribution based on the Liutex vortex identification method, the correlation between vortices and the generation amount of calcium carbonate as well as the degree of solute mixing was studied. The dynamic coupling influence mechanism of vortices and reaction processes in the microbial mineralization evolution system of porous media was preliminarily and quantitatively revealed. This research provides predictive analysis methods and models for the refined application design and process control of microbial mineralization technology.

Country China
Acceptance of the Terms & Conditions Click here to agree

Author

Dianlei Feng (Tongji University)

Presentation materials

There are no materials yet.