PARTICLES 2025

Simulating Turbulent Flows With Synthetic Inflow Turbulence Using Smoothed Particle Hydrodynamics

  • Goebel, Marco (Institute of Thermal Turbomachinery , KIT)
  • Buerkle, Niklas (Institute of Thermal Turbomachinery, KIT)
  • Chaussonnet, Geoffroy (Institute of Thermal Turbomachinery, KIT)
  • Koch, Rainer (Institute of Thermal Turbomachinery, KIT)
  • Bauer, Hans-Joerg (Institute of Thermal Turbomachinery , KIT)

Please login to view abstract download link

Smoothed Particle Hydrodynamics (SPH) has gained increasing attention due to its versatility in modeling complex fluid flows, with growing industrial applications. Many of these applications, such as spray atomization, are significantly affected by turbulence. Recent studies have shown that SPH is able to reproduce the turbulent energy cascade up to the kernel scale, enabling its integration into the framework of Large Eddy Simulation (LES). However, to effectively conduct LES-type simulations, turbulent inflow boundary conditions are required to maintain manageable computational domains. Currently, to the author’s knowledge, no such boundary conditions exist for SPH. In this work, a synthetic turbulent inflow generator used in grid-based CFD methods is adapted and implemented into the in-house SPH code turboSPH. Turbulent kinetic energy is injected into the flow by superimposing velocity fluctuations on top of the mean inlet velocity. The fluctuations are synthesized from a random velocity field constructed by the summation of Fourier harmonics with prescribed turbulent length and time scales. The development of turbulence downstream of the synthetic inflow is investigated through 2D and 3D test cases. When employing the previously used SPH scheme relying on background pressure to avoid particle disorder, the injected turbulent kinetic energy (TKE) is rapidly dissipated. However, when employing the more modern quasi-Lagrangian ALE-SPH approach, the injected TKE prevails and its rate of dissipation is significantly reduced. Using the ALE-SPH approach, key parameters, such as the ratio of the turbulent length scale to the kernel radius, are identified to ensure consistency between the prescribed and actual turbulence levels. This study demonstrates that turbulent inflow generators from grid-based CFD methods can be adapted to the ALE-SPH framework with only minor modifications. This enables the generation of adequate inflow conditions for LES-ALE-SPH simulations, which is a significant advancement towards enhancing their fidelity.