Abstract

Large Eddy Simulation (LES) has for a long time been considered a promising technique for predicting complex turbulent flows. Industrial use of LES is, however, still fairly limited due to the cost of predicting time dependent flow with the required mesh resolution. With the aid of parallel computers, this state of affairs is about to change, as computational fluid dynamics (CFD) computations with more than a million points now can be done at reasonable cost. For long flexible suspension bridges, wind loads, and in particular the fluctuating part of the load, is critical for the design and construction and extensive wind tunnel experiments are routinely carried out prior to design and construction. The principle of LES is simple and attractive. Unlike traditional turbulence modeling based on time-averaging of all turbulent quantities, LES consists of simulating the large scale turbulent fluctuations and modeling only the quantities too small to be captured by the grid. This introduces the subgrid model, which models the effect of the small scale turbulence. The subgrid model is however not the only requirement for LES. In addition, the mesh must be fine enough to resolve sufficiently small “large” eddies and the numerical viscosity of the space discretization scheme should be significantly smaller than the subgrid viscosity.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call