論文

The subgrid-scale (SGS) models based on the coherent structure (CS) in grid-scale flow fields are proposed and are applied to (non-)rotating homogeneous turbulences and turbulent channel flows. The eddy viscosity is modeled by a coherent structure function (CSF) with a fixed model-parameter. The CSF is defined as the second invariant normalized by the magnitude of a velocity gradient tensor and plays a role of wall damping. The probability density function of the CSF is non-Gaussian showing an intermittency effect. The model parameter is locally determined, and it is always positive and has a small variance. These models satisfy a correct asymptotic behavior to a wall for incompressible flows. It is shown that the SGS models with an energy-decay suppression function which indicates also a pseudo-backscatter are consistent with the asymptotic material frame indifference in a rotating frame. Since the CSF characterizing turbulent flows has relation to the SGS energy dissipation, the present SGS models are applicable not only to (non-)rotating homogeneous and shear turbulences but also to laminar flows. The proposed models have almost the same performance as the dynamic Smagorinsky model for (non-)rotating homogeneous turbulences and turbulent channel flows, but these models do not need to average or clip the model parameter, use an explicit wall-damping function, or change the fixed-parameter, so that they are suitable for engineering applications of LES.

In a magnetohydrodynamics (MHD) power generation with a frozen inert gas plasma (FIP), the effect of the non-uniformity of the inlet plasma in the direction perpendicular to walls on the performance and the comparison of the performance with r-theta two-dimensional (2-D) numerical simulations are examined with r-z 2-D ones. The inlet non-uniformity of an ionization degree is kept throughout the channel and even in the boundary layer the state of the FIP is realized without an ionization instability. When the inlet ionization degree near the walls is higher than that in the main flow, the enthalpy extraction ratio decreases because the strong MHD interaction near the wall leads to the development of the boundary layer. On the other hand, it increases for the inverse distribution at the channel inlet, even if the introduced total inlet ionization degree is the same. In comparison with the results obtained from the r-theta simulation, the development of the boundary layer results in the deterioration of the performance, and the optimal condition shifts to the low inlet ionization degree and the high load resistance.

Thermal characteristics of a high temperature air combustion boiler is numerically examined. The high temperature air combustion is realized using a preheated air of about 800℃〜1000℃ under low oxygen content (〜5%). Numerical simulation reveals that more uniform temperature distribution in the boiler and higher heat transfer are obtainable using high temperature air in comparison with using normal temperature air owing to mild reactions in the high temperature field and heat recovery by regenerative burners. High speed injection of high temperature air into the boiler causes in-furnace exhaust gas recirculation, which results in low NOx emission. There is good agreement between experiment and numerical simulation in particular when using low calorific value gas, because more uniform temperature distribution is realized and average temperature in the boiler becomes lower.

The inapplicability of the dynamic Clark model to the large eddy simulation of incompressible turbulent channel flows is proved both analytically and numerically. The reason is neither a negative subgrid-scale eddy viscosity nor an incorrect near-wall scaling, but a negative effective viscosity in the viscous sublayer for the tensor diffusivity term that models the sum of the Leonard and the cross terms.