LBRG

The Lattice Boltzmann Research Group (LBRG) at Karlsruher Institut für Technologie (KIT) is proud to share our latest paper: “N-Component Free Energy Lattice Boltzmann Method with Reduction Consistency and Global Momentum Conservation”.

This work is the result of a collaboration by Michael Rennick, Xitong Zhang, Tim Niklas Bingert, Matthias J. Krause and Halim Kusumaatmaja.

This paper introduces a novel N-component free Energy Lattice Boltzmann Method, which adresses the long-standing issue of non-physical artifacts due to scale-up (N>=4) : absent fluid phases can spontaneously nucleate out of nowhere and numerical discretisation of errors in surface tension forces create artificial whole-domain momentum shifts. Our method manages to scale reliably to an arbitrary number of fluid phases by implementing an independent flux correction and a divergence-consistent force discretization. 

Key Highlights from the Paper:

 Reduction Consistency: Our framework ensures that absent fluid components do not unphysically appear, utilizing a flux correction in the Cahn-Hilliard equations that operates entirely independently of fluid mobility.

 Machine Precision Conservation: By modifying the surface tension force implementation to align with a numerical divergence-consistent form, we eliminate unphysical whole-domain velocity drifts down to machine precision.

 Arbitrary Fluid Phases: The model allows for the simulation of an arbitrary number of immiscible components (N>=4) in principle, with completely independent selection of all interfacial tensions.

 Broad Impact: This robust framework unlocks high-fidelity engineering simulations for complex real-world applications, from targeted drug delivery in nested microfluidic emulsions to liquid-liquid phase separations in cell biology and smart anti-fouling coatings.

This paper is published open access and can be found here https://arxiv.org/pdf/2605.22214.

The picture shows the formation of an emulsion droplet from the injected fluids over time. The surface tensions are configured to encourage cloaking of fluids 3 (yellow) and 4 (blue) by fluid 2 (pink).

LBRG

As part of this year's Computational Fluid Dynamics and Simulation (CFDS) Lab Course, the LBRG organized an excursion to the HoreKa supercomputer at SCC, KIT, featuring an engaging tour graciously guided by René Caspart.

Taking place shortly after the launch of the project phase, this excursion provided 10 students from Mathematics, Engineering, and Computer Science programs with vital firsthand insight into the high-performance computing environment where their simulations of a fluid dynamics cases, all inspired by real-world industry problems, will be running.

The CFDS Lab is offered by a funded collaboration between the Lattice Boltzmann Research Group and the Scientific Computing Center: Fedor Bukreev, Stephan Simonis, Mathias J. Krause, Gudrun Thaeter, Jasmin Hörter, and Martin Frank.
In the course, we enable students to efficiently use the large-scale computing infrastructures at KIT such as HoreKa for scientific simulations with highly efficient codes such as OpenLB.

This course is funded by the Federal Ministry of Education and Research (BMBF) and the Baden-Württemberg Ministry of Science as part of the Excellence Strategy of the German Federal and State Governments. More information is available at https://www.lbrg.kit.edu/cfdslab.php

We gratefully acknowledge the computing time for our course provided on the high-performance computer HoreKa by the National High-Performance Computing Center at KIT (NHR-Verein, NHR-Alliance Karlsruhe Institute of Technology (KIT)). This center is jointly supported by the Federal Ministry of Education and Research and the Ministry of Science, Research and the Arts of Baden-Württemberg, as part of the National High-Performance Computing joint funding program. HoreKa is partly funded by the German Research Foundation.
 

LBRG

The LBRG at KIT is continuing to expand what is possible with LBM in OpenLB and reached a new milestone with compressible magnetohydrodynamics.
In our new preprint we validated this approach on established magnetohydrodynamics (MHD) benchmarks and applied it to a challenging astrophysical showcase.

Key Highlights & Performance

• Efficiency: Implemented within the open-source framework OpenLB, our code achieves up to 98.9% of the GPU hardware roofline performance.

• Versatility: Discretizes both ideal compressible and resistive incompressible MHD systems (encompassing Euler and Navier-Stokes limits).

• Validation: Benchmarked against established compressible and incompressible MHD cases.

From Theory to Space:

To demonstrate our framework's support for dynamic solid geometries, shifting magnetic fields, and fluid-structure interaction, we simulated a moving, surface-resolved magnetized asteroid modeled after 16 Psyche in supersonic early solar wind flow.

Methodology, Implementation and Visualization: Fedor Bukreev, Adrian Kummerländer, Mathias J. Krause.

YouTube: https://www.youtube.com/watch?v=itsxjEagKmQ
Preprint: https://arxiv.org/abs/2606.00641