Welcome to the Lattice Boltzmann Research Group
The Lattice Boltzmann Research Group (LBRG) is an interdisciplinary research group that aims to take advantage of novel mathematical modeling strategies and numerical methods to enable large-scale simulations and optimal control of fluid flows for applications in process engineering. The LBRG aims at a better fundamental understanding of suspensions in general and for the improvement of mechanical processes and medical treatments. In particular the LBRG designs and uses models, algorithms, and open source simulation tools such as OpenLB, always taking advantage of modern high performance computers for the simulation of, for example:
- Particulate fluid flows
-
Thermal flows
-
Turbulent flows
-
Material transport and chemical reactive flows
-
Light transport
-
Fluid-structure interaction
-
Flows in porous media and complex geometries
The LBRG’s teaching and education concept is project- and research-oriented, offering for example basic programming courses, lectures on parallel computing, software tutorials, and advanced seminars on particular fluid flow simulations as well as optimal control theory.
Read more:
Latest News
The Lattice Boltzmann Research Group (LBRG) at Karlsruher Institut für Technologie (KIT) is excited to share our latest paper: “An Integrated Open-Source Software System for the Generation and Analysis of Subject-Specific Blood Flow Simulation Ensembles.”
This work is the result of a collaboration by Simon Leistikow*, Thomas Miro*, Adrian Kummerländer, Ali Nahardani, Katja Grün, Marcus Franz, Verena Hoerr, Mathias J. Krause, and Lars Linsen (*joint first authors).
Modeling patient-specific blood flow is notoriously complex, often relying on fragmented and inefficient workflows. To truly understand how blood flow shifts under different physiological conditions, researchers need to generate entire "simulation ensembles"—yet most existing tools are only built to handle one simulation at a time.
We introduced an integrated software system that bridges Voreen and OpenLB. By merging medical image processing, high-performance fluid simulation, and interactive visual analysis into a single interface, this new setup brings together the best of both CFD and MRI.
Key Highlights from the Paper:
- Seamless Integration: OpenLB embeds smoothly into Voreen's interactive volume-rendering framework, enabling true in-situ visualization.
- A Smoother Workflow: Moving to a single graphical user interface (GUI) drastically cuts down the manual workload and eliminates the tedious, error-prone process of file conversions.
- Accessible to Non-Experts: You no longer need an advanced background in computing to configure, run, and evaluate complex Lattice Boltzmann simulations.
- Deeper Clinical Insights: Combining MRI data directly with CFD gives us a much clearer, comprehensive view of critical hemodynamic biomarkers, such as wall shear stress and aneurysm parameters.
The article is published Open Access and available here.
An overview of the different Voreen-OpenLB use cases can be found on youtube:
- Use Case A
- Use Case B
- Use Case C
Last week, the Lattice Boltzmann Research Group (LBRG) at Karlsruher Institut für Technologie (KIT) met with project partners Willi Mayer Holzbau GmbH, UBP-Consulting GmbH, and CeMOS - Research and Transfer Center of Technische Hochschule Mannheim at Willi Mayer Holzbau in Bisingen. The meeting marked the kick-off for the InvestBW-funded project INFRA (Innovative nachhaltige Wärmeeffizienz-Analyse im realen Altbau / Innovative sustainable thermal efficiency analysis in real-world existing buildings).
Heating and cooling buildings accounted for 25.6% of Germany’s energy consumption in 2023. While improving the insulation of existing buildings offers significant energy savings, current energy analyses are technically complex, largely manual, and limited by a shortage of qualified personnel.
The INFRA project aims to standardize and digitalize the energy analysis and retrofit planning process for existing buildings. The project focuses on two main objectives:
Analysis Tool: Automates the creation of digital twins using drone-based geometry capture, exterior thermography, and indoor temperature data. It combines simulations and optimization methods to inversely calculate the thermophysical properties of the building envelope, and subsequently applies AI techniques to identify materials based on a comprehensive database.
Planning Tool: A system that utilizes the digital twin data to automatically generate fully digital plans for energy-efficiency improvements.
By automating these processes, the project aims to significantly reduce the documentation and calculation workload for energy consultants and make building retrofits more scalable.
That’s a wrap on an intensive training on Algorithmic Differentiation (AD)! Last week, Shota Ito, Johannes Grafen and Leonardo Dorneles participated in the AD training at the Karlsruher Institut fur Technologie (KIT).
It was a great opportunity to dive into the theory and practical application of AD during the lectures and hands-on coding sessions.
The two-day training on AD covered everything from the foundational theory to scalable AD tools for high-performance computing:
Day 1: Introduction to AD, manual differentiation basics, and building an operator overloading tool from scratch.
Day 2: Practical applications using CoDiPack (forward, reverse, and vector modes), learning about Jacobian and higher order derivative computations
Thank you to the organizers Max Sagebaum (Rheinland-Pfalzische Technische Universitat Kaiserslautern-Landau (RPTU), NHR@SW), Nicolas R. Gauger (Rheinland-Pfalzische Technische Universitat Kaiserslautern-Landau (RPTU), NHR@SW), Jan Kieseler (Karlsruher Institut fur Technologie (KIT)), and Mathias J. Krause (Lattice Boltzmann Research Group (LBRG)) for hosting such an insightful, hands-on event!