Adding polymers to a Newtonian solvent alters the turbulence, ultimately resulting in the macroscopic effect of drag reduction (DR) in a turbulent wall-bounded flow (Toms, 1946). Due to the important technological implication, the phenomenon gained large research interest in the last decades, but it remains poorly understood given the complex multi-scale coupling between polymers and turbulence.
Polymer DR exhibits a peculiar feature, the Maximum Drag Reduction (MDR) asymptote, a universal state independent of the polymers and solvent. To understand the complex phenomenon of DR, the comprehension of its universal states as MDR is essential. Direct numerical simulation is the best candidate to investigate such phenomena.
We use a physically consistent model for polymer chains that has no restrictions in spanning the entire parameter space. This is crucial to reproduce realistic conditions and fully characterise the polymers/turbulence coupling without any a priori assumptions. The motion of every single polymer chain (modelled as a FENE dumbbell) is addressed by means of a hybrid Eulerian-Lagrangian approach that couples polymer dynamics and Navier-Stokes equations.
Such high-fidelity simulations require the integration of billions of dumbbells and Eulerian collocation points and are only possible by exploiting the computing capabilities of HPC super-computers equipped with GPUs.
Sapienza Università di Roma, Italy.