Friday 21 March 2025
Turbulent two-phase flows are a common occurrence in many industrial processes, such as boiling heat transfer and liquid-liquid extraction. These flows can be notoriously difficult to simulate, as they involve complex interactions between two or more phases that are often characterized by different physical properties.
A team of researchers has developed a new numerical solver called CaNS-Fizzy that is specifically designed to tackle these challenging simulations. The solver uses a combination of advanced algorithms and modern computing architectures to efficiently solve the equations governing turbulent two-phase flows.
At its core, CaNS-Fizzy employs a one-fluid formulation of the two-phase flow problem, which allows it to solve a single set of equations for both phases in the entire domain. This approach eliminates the need for explicit interface reconstruction, which can be computationally expensive and prone to errors.
The solver uses a finite difference method to discretize the governing equations, which are then solved using a pressure correction scheme and a fast direct solver for the pressure Poisson equation. The resulting solution is highly accurate and efficient, making it well-suited for large-scale simulations.
One of the key features of CaNS-Fizzy is its ability to efficiently solve complex interface problems. This is achieved through the use of a diffuse interface approach, which allows the solver to capture the intricate details of the interface without requiring explicit reconstruction.
The team has tested CaNS-Fizzy using a range of benchmark cases, including simulations of liquid-liquid emulsions and gas bubbles rising in a liquid. The results show that the solver is able to accurately capture the complex dynamics of these flows, even at large scales.
In addition to its accuracy and efficiency, CaNS-Fizzy is also designed to be highly scalable. This means that it can take advantage of modern computing architectures, such as graphics processing units (GPUs), to solve larger and more complex problems than would be possible on traditional CPU-based systems.
The team’s results have been tested on a range of GPU-based systems, including the Leonardo supercomputer at Cineca in Italy. The results show that CaNS-Fizzy is able to scale efficiently as the problem size increases, with only a modest increase in computational time required for larger simulations.
Overall, CaNS-Fizzy represents a significant advance in our ability to simulate turbulent two-phase flows. Its combination of accuracy, efficiency, and scalability make it an attractive tool for researchers and engineers working on a wide range of applications, from chemical engineering to aerospace.
Cite this article: “Simulating Turbulent Two-Phase Flows with CaNS-Fizzy”, The Science Archive, 2025.
Turbulent Two-Phase Flows, Numerical Solver, Cans-Fizzy, One-Fluid Formulation, Finite Difference Method, Pressure Correction Scheme, Diffuse Interface Approach, Gpu-Based Systems, Scalability, Chemical Engineering, Aerospace
