DOI: 10.1615/TSFP10
STATE DETECTION AND HYBRID SIMULATION OF BIOMEDICAL FLOWS
SINOPSIS
Hybrid simulations might be very interesting to save computational resources for a variety of applications involving different flow regimes. Combining proper models for laminar, transitional, or turbulent flows, described either by (U)RANS, LES or DNS depending on the needs, an accurate solution could then be obtained for complex configurations on existing computers. To do so, it is necessary to decide which kind of model should be used in which part of the numerical domain in space and time. As shown in this work, the spectral entropy Sd obtained from solving the eigenvalue problem for the temporal autocorrelation function, can be used in order to uniquely and automatically quantify the flow state and differentiate between laminar, transitional, or turbulent regime; as such, it delivers a direct measure of turbulence intensity. Using Sd, two hybrid simulations of biomedical flows have been carried out. The statistically-steady blood nozzle benchmark proposed by the FDA is simulated by URANS/LES, while the pulsating flow in a cerebral aneurysm is solved by LES/DNS. In both cases, savings in computational time and disk storage are observed, while keeping a very high accuracy.