About this Abstract |
| Meeting |
2026 TMS Annual Meeting & Exhibition
|
| Symposium
|
Artificial Intelligence Applications in Integrated Computational Materials Engineering (AI-ICME)
|
| Presentation Title |
Machine Learning-Augmented Finite Volume Modeling and Inverse Optimization of Velocity and Pressure Distribution of Bentonite Slurry in Slurry Shield Tunneling Applications |
| Author(s) |
Somnath Somadder |
| On-Site Speaker (Planned) |
Somnath Somadder |
| Abstract Scope |
Bentonite slurry, a non-Newtonian fluid with yield stress and shear-thinning behavior, is widely used in slurry shield tunneling to maintain face stability and control infiltration. This study presents an integrated framework combining high-fidelity Finite Volume Method (FVM) simulations with supervised Machine Learning (ML) for both forward prediction and inverse design of slurry flow behavior. A parametric study using ANSYS Fluent generated 48 simulations across varying pipe diameters, inlet velocities, and CMC concentrations. Trained ML models (ANN, decision trees, ensembles) accurately predicted pressure drop and centerline velocity (Rē > 0.98). Feature analysis identified pipe diameter and velocity as key factors. Inverse optimization using ML surrogates enabled rapid estimation of input conditions for desired outputs, eliminating costly iterative CFD routines. This ML-augmented approach supports real-time rheological monitoring, process optimization, and adaptive control in tunneling operations, offering a scalable solution for intelligent automation in non-Newtonian fluid systems. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Machine Learning, Modeling and Simulation, Computational Materials Science & Engineering |