OpenLB makes simulation software accessible to SMEs
If you’re a simulation engineer, then the following scenario may well be familiar to you. You spend weeks meshing a geometry, configuring the setup and wrestling with grid convergence issues – and then end up starting the whole process from the beginning again due to a grid cell error that either causes instabilities or necessitates the next manual iteration of the optimization process. How many times have you wished that the entire process was more robust, faster and more automated? Large corporations can afford to hire a service provider to deliver a ready-to-use simulation process. But what are small and medium-sized enterprises or development engineers in large companies supposed to do if they want to use simulations to test a new product or process idea? OpenLB is a simple and efficient to use simulation software solution supported for this target group by the Steinbeis Consulting Center for Computational Engineering (CE).
wo separate multiphase simulations in the geometry of a real porous rock. In the simulation on the left, liquid water (blue) displaces hydrogen gas (white), while in the simulation on the right, the water is displaced by the gas. The whole system is driven by a pressure difference of 50 kPa across an area 1.2 mm wide. The fluid properties correspond to compressed hydrogen (7.1 kg/m³) and liquid water (992 kg/m³).
Reference simulation of a turbulent reacting micromixer. Two chemicals (blue and yellow) react when brought together in a mixing chamber. The turbulent carrier fluid is modeled as LES, while the reactants and products are modeled as separately coupled advection-diffusion equations.
OpenLB is based on the Lattice Boltzmann Method (LBM). At first, this method was largely overlooked because it required too much computing power and was so different to established approaches. But the world has changed, and today it has become a much more attractive proposition. Even kids playing the latest computer game on their graphics card (GPU) late at night now have the necessary computing power under their desk. Commercial and open-source LBM software is now widely available, and companies like DLR, Airbus and Porsche have already been using it for computational fluid dynamics (CFD) for over ten years.
But as wide as the range of available software is, training people to use it and developing your own setup can be very costly. This is where the OpenLB software supported by the Steinbeis experts in Wörth am Rhein comes in. “OpenLB is a powerful open-source CFD solution that provides a lattice Boltzmann alternative to OpenFOAM and includes several ready-made tutorials and example applications”, says Steinbeis Entrepreneur PD Dr. Mathias Joachim Krause. What sets it apart is that it can perform calculations up to 32 times faster on standard CPUs, and many more times faster than this on GPUs. Users can thus easily perform simulations with several hundred million cells even on desktop computers. Moreover, OpenLB can be adapted to your company’s specific requirements.
OpenLB does the meshing for you
The nodes are generated fully automatically within the specified geometry in a matter of seconds, making manual meshing a thing of the past. A range of tutorials corroborated by established benchmarks are available. Examples include virtual wind tunnels, continuous stirred-tank reactors, pumps, reactors, filters and even blood vessels – with fluids simply flowing through or being mixed, separated, reacted, heated and cooled. Several ready-to-use models are also available for particulate and non-Newtonian flows, including in dissolved and undissolved porous media. The scalable software makes it possible to perform accurate large eddy or direct numerical simulations for industrial applications, with or without mesh refinement.
Goodbye manual parameter studies
As Mathias Joachim Krause emphasizes, “The integrated sensitivity analysis and optimization functions dispense with the need for trial-and-error methods and laborious parameter setting”. This means that the simulation setup can be automatically checked for relevant dependencies during runtime and the appropriate solution identified directly from the simulation, saving time and money. What previously took several weeks can now be done in a matter of hours. The Steinbeis experts are rightly proud of this quantum leap in efficiency that helps to take product development to the next level.