Dispersed Phase and Separated Flows

Technical capabilities in MPflow

• Multiphase modelling of dispersed flows (droplets, bubbles, particles)
• Accurate simulation of separated flows with sharp phase interfaces
• Interface tracking and capturing using advanced VOF/level-set techniques
• Support for annular, stratified, slug, and bubbly flow regimes
• Prediction of phase distribution, coalescence, breakup, and transport
• High-resolution modelling of interfacial forces (surface tension, drag, lift)

Our CFD approach

• Eulerian and hybrid Eulerian–Lagrangian frameworks for dispersed phases
• Direct resolution of interface dynamics in separated flows
• Modelling of droplet/bubble breakup and coalescence mechanisms
• Accurate treatment of phase interactions and momentum exchange
• Simulation of flow regime transitions under varying operating conditions
• Capability to handle complex geometries and industrial-scale systems

Physics-based ML acceleration

• ML-enhanced prediction of interfacial dynamics and phase distribution
• Acceleration of multiphase simulations across flow regimes
• Surrogate modelling for breakup, coalescence, and drag closures
• Hybrid CFD–ML framework preserving physical consistency
• Rapid evaluation of operating conditions and design parameters

Why ML-CFD matters for Dispersed & Separated Flows

• Up to 1000–3000× faster predictions: rapid evaluation across multiple flow regimes
• 7× acceleration with ML-enhanced multiphase and turbulence modelling
• High accuracy (<5% deviation) in phase distribution and interface dynamics
• Instant exploration of flow regimes (bubbly, slug, annular, stratified)
• Scalable across industries: energy, chemical, oil & gas, and process engineering
• Reduces reliance on empirical correlations and costly experimental campaigns