ºÚÁϸ£ÀûÍø

Societal relevance

Multiphase flow processes – where gases, liquids, and solids interact – are central to many of today’s pressing technological challenges. They play a key role in the energy transition, the materials transition, and the drive for circular, sustainable production. For example, metal powders like iron can serve as carbon-free recyclable fuels for industrial heat and energy storage, alkaline water electrolysis is a main driver for green hydrogen generation, and the production of green steel (using hydrogen instead of coal) anticipate greatly reduced emissions.

Despite their importance, multiphase flows remain challenging to master; turbulence, chemical conversion, electrostatics, phase transitions and thermal effects all affect the flows. A lack of fundamental understanding, even of basic flow patterns and regimes, can hamper the design and optimization of novel processes and technologies. Researchers in the Multiphase Flow Systems cluster are addressing these challenges, relying on decades of expertise in fluid dynamics.

Research approach and methods

Our approach integrates advanced experiments and high-fidelity simulations to unravel multiphase flow physics (details are given below). We develop novel measurement techniques and custom experimental facilities to observe flows in unprecedented detail, and we create sophisticated numerical models (ranging from direct numerical simulation to reduced-order models) to predict and optimize reactor performance.

Crucially, we employ a multi-scale modeling strategy: phenomena from the smallest scale up to the equipment scale are connected in our models. This ensures that fine-scale insights improve the accuracy of large-scale reactor design. Fast computational models capable of predicting hydrodynamics, heat and mass transfer, chemical conversions and phase transitions in multiphase equipment are a key enabler for scaling up new processes.

Another hallmark of our research is the synergy between physical modeling and data-driven techniques. We leverage modern developments in machine learning (AI) alongside physics-based simulations. Data-driven algorithms are used to improve closure models for mass and heat transfer, for particle interactions, and to enable real-time reduced-order modeling. Additionally, AI supports advanced computer vision methods to segment and reconstruct dispersed entities like particles and bubbles from flow imagery.

Our researchers maintain close collaboration with industry partners, from large established companies to high-tech start-ups. This ensures our research addresses real-world needs, helping to scale up new technologies that make the Netherlands and the world more sustainable and future-proof.

Experimental research and facilities

We build and employ state-of-the-art experimental techniques to probe multiphase flows, to provide high-quality data needed to validate our models and reveal new physics. For example, we have developed advanced imaging methods like Magnetic particle tracking and Particle Image Velocimetry in combination with specialized digital image analysis, to help quantify granular flows. In flows opaque to visible light (e.g. inside industrial reactors or porous media), we use techniques such as Magnetic Resonance Imaging (MRI) to visualize phase distributions.

Our laboratories house a range of custom multiphase flow setups that replicate real process conditions. From bubble column reactors to fluidized beds with full optical access for studying gas–liquid and gas–solid flows, as well as high-temperature gas-solid experiments and specialized experiments for reactive flows. We also study electrochemical multiphase flows, e.g. by observing how microscopic gas bubbles nucleate, grow, and detach from electrode surfaces – data that helps to improve electrode designs and gas management in electrolyzers.

Computational and theoretical research

On the simulation side, we develop sophisticated computational models to capture multiphase flow behavior across scales. At the most detailed level, we perform Direct Numerical Simulations (DNS) of multiphase flows – resolving every eddy and interface in small domains to fundamentally understand interactions. Such simulations, for example, have illuminated how bubble swarms rise in liquids or how particles cluster in turbulent gas flows. We also use particle-resolved simulations (e.g. DNS with interface tracking or coupled CFD-DEM methods) to study phenomena like bubble–turbulence coupling and catalyst particle mixing at the grain scale. These high-fidelity results feed into our mesoscale and continuum models. We build Euler–Euler multi-fluid models and Euler–Lagrange models that can simulate industrial-scale systems (like reactors of several meters) by averaging or probabilistically representing the fine details. A key challenge in these models is closure laws – accurately representing effects of sub-grid interactions (collisions, coalescence, etc.).

Our team also works on model reduction and optimization techniques. We construct reduced-order models that capture the essence of complex multiphase dynamics in simplified form – speeding up simulations by orders of magnitude. This rigorous numerical research underpins the design of next-generation multiphase systems for fuels, chemicals, and materials – helping to make them smarter, smaller, and more efficient, in line with the goals of process intensification.