ultraFluidX - Background

Printer-friendly version

What is ultraFluidX?

ultraFluidX is a simulation software for computing physical flows for external aerodynamics. Employing the Lattice Boltzmann Method, the software is based on a particle approach for the description of fluid flows. This method is optimally suited for calculations on massively parallel GPU hardware, leading to significantly reduced turnaround times.
What is ultraFluidX' aim?

The aim of ultraFluidX is to maximize the simulation's performance while ensuring a high level of result accuracy. Thus, applying ultraFluidX in industry ensures a significant decrease both in overall costs and computing times for flow simulations, resulting in the emergence of new CFD markets.
How does ultraFluidX work and what kind of potential does it offer?

In order to increase computing power and to improve results quality, algorithmic, infrastructural and model-based approaches are adopted. Methodical improvements, such as the possibility to locally refine the computational mesh in order to resolve boundary layers or strong flow gradients, help to increase the overall efficiency of the computations without loss of results quality.
Additionally, the implemented turbulence models also reduce the required number of cells and at the same time enable higher accuracy levels through acknowledged model-like assumptions on flow behaviour.
In its current state of development, ultraFluidX is capable of employing multiple GPUs within the same host system for performing a simulation. Hereby, ultraFluidX achieves almost ideal linear scalability.
The software’s capabilities are continuously developed. A major focus of current developments is on obtaining approximately the same scalability when running simulations within a network of workstations.
For this purpose, ultraFluidX offers three-dimensional flow visualization capabilities for realtime flow analysis and monitoring while the simulation is still running. An importer for geometries saved as STL-files is available for data exchange with external tools. Simulation results can be exported employing the widely used EnSight format. This allows using highly efficient post-processing tools for detailed data analysis.


Meet us at the next exhibition:

MOC Munich //
24. Jun. 2019