Laminar, Turbulent, and Transitional Simulations in Benchmark Cases with Cardiovascular Device Features

Abstract

Verification and validation of laminar, transitional, and fully turbulent flow simulations using URANS and hybrid RANS/LES (HRL) are performed for a capillary tube at Re = 500–5100 and for nozzle flow at Re = 500, 3500, and 6500. For the capillary tube case, good predictions for Re ≤ 1000 and ≥4000 were obtained using laminar and fully turbulent URANS models, respectively. A transition-sensitive URANS model performed well for the entire Re range, suggesting the ability to provide a universal model for laminar and turbulent URANS solutions, provided that the inlet turbulence intensity can be accurately prescribed. For the nozzle flow simulations, good predictions for Re = 500 and 6500 were obtained using laminar and fully turbulent URANS and HRL simulations, respectively. For the transitional case, Re = 3500, turbulent simulations performed well in the separated flow regions but were under-predictive in the throat region. The hybrid RANS/LES simulations performed better than URANS in the separated flow region. The most commonly used hybrid RANS/LES models failed either to trigger turbulence or transition to LES mode for low background turbulence. A new dynamic hybrid RANS/LES model [Bhushan and Walters, Physics of Fluids, 24, 015103, 2012] did not show such a deficiency and provided the best comparison with experimental data for the two higher Re cases.