A priori and a posteriori tests of inflow conditions for large-eddy simulation

Abstract

Comparisons of inflow conditions for large-eddy simulations of turbulent, wall-bounded flows are carried out. Consistent with previous investigations, it is found that the spectral content of the inflow velocity is important. Inflow conditions based on random-noise, or small-scale eddies only, dissipate quickly. Temporal and spatial filtering of a time series obtained from a separate calculation indicates that it is important to capture eddies of dimensions equal to or larger than the integral length scale of the flow. Three methods for generating inflow velocity fields are tested in a simulation of spatially developing turbulent channel flow. Synthetic turbulence generation methods that introduce realistic length scales are more suitable than uncorrelated random noise, but still require fairly long development lengths before realistic turbulence is established. A recycling method based on the use of turbulent data obtained from a separate calculation, in different flow conditions, was found to result in more rapid transition. A forcing method that includes a control loop also appears to be effective by generating turbulence with the correct Reynolds stresses and correlations within less than ten channel half heights.

Efficient Generation of Inflow Conditions for Large Eddy Simulation of Street-Scale Flows

“…Recall that, as noted earlier, Keating et al (2004) used Batten et al (2004)'s procedure (i.e. a synthetic turbulence generation method) to generate inflow data for a plane channel flow and found that the redevelopment was slow requiring at least 20d.…”

“…Alternatively, a time-evolving LES or direct numerical simulation (DNS) with a periodic (Garcia et al, 2004;Keating et al, 2004) or a 'modified periodic' (Lund et al, 1998) inlet-outlet boundary condition can be used to generate appropriate turbulent flow. The latter authors used a sort of 'precursor simulation', in which the velocity field at an appropriate downstream station is stored and imposed in suitably re-scaled form as inflow data for the primary computation.…”

“…The procedure is easy to code and the efficiency is high. However, Keating et al(2004) used this procedure to generate inflow data for a plane channel flow and found that the redevelopment was slow -after a distance of twenty channel half-heights the turbulence in the core was clearly not yet fully developed. Klein et al(2003) developed a technique generating artificial velocities as inflow data for jet flows, which reproduced first and second order one-point statistics as well as locally given correlations.…”

Large-eddy simulation of low-frequency unsteadiness in a turbulent shock-induced separation bubble

“…However, this is only true as long as the prescribed length scales are at least as large as the integral length scales of the real flow. Failing to meet that requirement can lead to laminarization issues [29,52] just as when only white noise is added to the flow. The prescribed mean-velocity profile was obtained from the semi-analytical method described in Li [31] and the prescribed Reynolds stresses were obtained from an earlier simulation under similar flow conditions as the one considered here.…”

Comment on “Diffusion by a random velocity field” [Phys. Fluids 13, 22 (1970)]

“…Kraichnan's turbulence generator was used in many subsequent studies to model aeroacoustic noise, [2][3][4] inlet boundary conditions, [5][6][7][8][9][10] verification and validation of turbulence solvers, 11,12 and a priori and a posteriori testing of turbulence models. 13,14 To enforce a divergence-free velocity field, Kraichnan's method requires that two vectors be perpendicular. In this comment we show that this condition is only valid in the infinitesimal limit and should be replaced with an otherwise simple modification based on the discrete divergence of the velocity field.…”