Mixers
Multiphase mixing processes in stirred vessels where hydrodynamics, turbulence, and phase interaction govern efficiency, separation performance, and scale-up in chemical and process industries
Technical capabilities in MPflow
MPflow provides advanced CFD tools for simulating mixing processes in mechanically agitated systems, including complex multiphase interactions and turbulence-driven transport.
- Dedicated solvers for rotating machinery and impeller-driven flows
- Support for multiphase systems (liquid–liquid, gas–liquid) and phase separation
- Accurate modelling of turbulence, recirculation zones, and vortex structures
- Capability to simulate baffles, impeller geometries, and vessel internals
- Fully compatible with OpenFOAM libraries for flexible workflows
- Prediction of mixing time, phase distribution, and separation efficiency


Our CFD approach
We use high-fidelity CFD to resolve the complex hydrodynamics governing mixing and separation in industrial vessels.
- Detailed resolution of flow patterns, circulation loops, and dead zones
- Quantification of turbulence intensity and energy dissipation rates
- Analysis of baffle configurations and impeller design impact
- Prediction of phase distribution and separation characteristics
- Simulation across operating conditions (rotation speed, inlet velocity, fluid properties)
Physics-based ML acceleration
Physics-aware machine learning enhances CFD by accelerating predictions and enabling rapid exploration of mixer configurations.
- ML models trained on CFD-generated hydrodynamic fields and experimental data
- Fast prediction of flow structures, mixing efficiency, and phase separation
- On-the-fly deployment for turbulence modelling and source-term approximation
- Rapid evaluation of design variations (impeller type, baffle number, geometry)
- Generalisation across operating conditions and fluid systems
Why ML-CFD matters for Mixers
- Up to 7000× faster predictions → rapid evaluation of mixer configurations and operating conditions
- 4–7× acceleration with ML-enhanced turbulence modelling, supporting all standard RANS and LES models
- <7% deviation from CFD in flow patterns, mixing time, and phase distribution in full 3D systems
- Instant design exploration across impeller types, baffle configurations, and vessel geometries
- Optimises mixing efficiency and separation performance
- Reduces experimental effort and scale-up costs in industrial process design


