Large-eddy simulations of turbulent flow over complex terrain using modified static eddy viscosity models

The performances of the modified static eddy viscosity model proposed by Inagaki et al. [2005. A mixed-time-scale SGS model with fixed model-parameters for practical LES. Transactions of the American Society of Mechanical Engineers: Journal of Fluids Engineering 127, 1–13] and of its variations were examined in the large-eddy simulation (LES) of a turbulent flow over a two-dimensional steep hill under a neutral condition. Unlike the standard Smagorinsky model, the modified static eddy viscosity models require no explicit wall damping function. Moreover, the modified static models showed excellent computational stability. The accuracy of the modified static models was assessed by comparing the results with those obtained from a wind tunnel experiment conducted by Ishihara and Hibi [1998. An experimental study of turbulent boundary layer over a steep hill. Proceedings of the 15th National Symposium on Wind Engineering, 61–66 (in Japanese)] and Ishihara et al. [2001. A wind tunnel study of separated flow over a two-dimensional ridge and a circular hill. Journal of Wind Engineering 89, Proceedings of the Fifth Asia-Pacific Conference on Wind Engineering, Kyoto, Japan, pp. 573–576] and the standard Smagorinsky model. The original model of Inagaki et al. (2005) produced unacceptable predictions, mainly due to the inaccurate estimation of the turbulence characteristics near the ground surface of the flow approaching the hill. On the other hand, the modification of the original model of Inagaki et al. (2005), in which the modified Leonard term was adopted for the parameterization of the square of the velocity scale in the expression of the subgrid-scale (SGS) eddy viscosity, gave more accurate predictions and produced the best results among the modified static models introduced in this study.