In the present study, numerical simulations were carried out on an areaway-attached basement model with a single-sided opening. The attention was focused on wind-driven natural ventilation performances and airflow characteristics in the basement space. Two CFD approaches: Reynolds averaged Navier-Stokes (RANS) and large-eddy simulation (LES) were used and the influence of inflow turbulent fluctuations in LES modeling was also examined. First, the simulation results were validated against the previous experimental data in terms of effective air change rate in the basement space. As a comparison result, a good agreement was found between the measurement and LES with inflow turbulent fluctuations while RANS tends to underestimate the ventilation rate by 40% on the average. Furthermore, based on LES modeling, the mean airflow patterns within and around the areaway-attached basement model was investigated in detail for different wind incidence angles to examine the influence of wind direction on ventilation performance in the basement space.
The contribution ratio of indoor climate (CRI) based on CFD was developed to estimate the individual contribution of heat factors to any location inside a room. However most of the studies related to CRI involve forced convection airflow fields. Thus the purpose of this study is to make the CRI effective also in natural convection air flow fields. In this study, a new calculation method was proposed by means of setting a uniform heat sink. A CFD simulation coupled with convection and radiation has been performed in an atrium and the CRI of all heat sources are calculated by the new definition of CRI. Also an approximation method is proposed to discuss the sensitivity of each heat source.
This paper describes a study of the feasibility about the thermal insulation efficiency of porous material on the proposed system. First, the window frame was designed using a porous material such as the packed bed. Then, to verify its thermal insulation efficiency, the temperature contribution of the window frame was evaluated using CFD simulation with different coupled conditions. In addition, to verify the condensation effect, the relative humidity was also calculated. The results show the thermal load was inversely proportional to the indoor/outdoor pressure differences. The condensation in the insulation depends on the outdoor temperature, humidity ratio, and porosity.
When hazardous substance is diffused, it is necessary to identify pollutant source immediately. Reverse simulation, which is the solution of the transport equation in negative time, is useful method for identifying pollutant source with CFD. Then the negative diffusion simplifies dispersive errors and reverse simulation cannot be carried out. We introduced a method to improving numerical instability. In this paper, we introduce reverse simulation result in cavity flow and survey the dependence of reverse simulation accuracy on grid resolution and filter width.
The comfortable environment in the truck cabin is important for a driver of trucks. Apart from the issue of comfort and health for driver it may cause a serious automobile accident because of fatigue and concentration loss. To know the flow field and air quality controlling the ventilation behavior in a cabin there are two technique of CFD and the experiment. In this report, by the comparison of CFD and the PIV experiment the flow behavior in the truck cabin is estimated. The six scale of Ventilation Efficiencies as SVE3 and SVE4 are examined by CFD. The diffusion behavior of the pollutant inside the car will be examined in the future.
LES coupled with a bi-molecular reaction model are conducted on the reactive air pollutants dispersion in urban street canyons which have two different roof-height variations. Nitrogen monoxides (NO) emitted from a line-source set on the bottom of the street canyon disperse and react with Ozone (O3) included in free streams. The reactivity of air pollutants has a considerable influence on their concentrations of a city block scale. In addition, the roof-height change also has a significance impact on the urban ventilation, making concentrations in the canyon space of product the highest in the even roof height case, meanwhile O3 in the non-uniform height case.
Aerodynamic sound radiate from a flow around has been numerically predicted by two methods. The first one is DNS which solve the compressible Navier-Stokes equations directly, and the second one is based on Lighthill Tensor which is the right side of Lighthill's equation. Recently, analyzing Lighthill Tensor is more used because of high cost of DNS. However, this method still has several problems especially about the termination of the computational domain. To solve computational domain problem, we use DNS in this study. The sound field radiated from the two-dimensional square cylinder in a uniform flow is computed by change the number of vortices captured by the computation. In order to clarify vortices radiating sound, we investigate sound mechanism. In addition, we calculate Powell's and Lighthill's acoustic sound sources.[This abstract is not included in the PDF]
Cavitating flows often have unsteady large 3D structures. We therefore believe that Large Eddy Simulation(LES) should be applied to the numerical simulation of cavitating flows. We used Dynamic Smagorinsky model for the subgrid scale model of LES and source/sink model for cavitation model. Weak compressibility is introduced by the low Mach assumption. As a basic test case, we calculated cavitating flows around 2D hydrofoil, Clark-Y11.7%, NACA0015 and validated accuracy by comparison with experimental data. The breakdown characteristics of the hydrofoils were not simulated accurately. The discrepancy between the simulation and measured breakdown characteristics is most likely to be related to that the present simulation cannot accurately predict the cavity length or large unsteady structures.[This abstract is not included in the PDF]
Aerodynamic sound generated from a propeller fan is simulated by a decoupling approach of flow and sound fields. The incompressible Large Eddy Simulation (LES) is performed by finite element method with a point dipole source, which is obtained by calculating the fluid force acting on the blades and struts of the fan. The predicted sound pressure spectrum shows good agreement with the experimental result.[This abstract is not included in the PDF]
The effects of global warming on the water quality and ecosystem of Lake Biwa were predicted by carrying out numerical simulations in the cases when atmospheric temperature increases by 5 oC and does not change in the following 100 years. As a result, the overturn of waters occurred every year since the increasing rate of water temperature at water bottom during the stratified seasons was larger than that at water surface. The decrease in dissolved oxygen concentration due to decreasing solubility caused the release of nutrients from sediments and eutrophication, resulting in the further decrease in dissolved oxygen concentration.[This abstract is not included in the PDF]
One-dimensional diffusion equation is investigated as a fundamental research for the reverse simulation. The diffusion equation of a concentration emitted from a point source is numerically solved to examine filtering methods for preventing numerical instability associated with the reverse simulation. Spatial oscillation of the concentration profile appears because the high-wavenumber Fourier components of the concentration increase owing to the numerical error and the components become as large as the low-wavenumber components representing the concentration profile. The method of filtering the concentration flux gave a better result because the fourth-order derivative term reduces the high-wavenumber components selectively.[This abstract is not included in the PDF]
The subgrid-scale model (SGS) in large-eddy simulation is studied. In this work, a novel one-equation-type SGS model is designed including no length scale. The objective of this study is to propose the new SGS model and check its performance compared with other models. For this purpose, we apply these models to a turbulent channel flow at Reτ=395 and Rayleigh-Bénard convection at Ra =3.81×105 and then show the major results of them.[This abstract is not included in the PDF]
The authors have constructed a statistical theory of inhomogeneous turbulent flow, clearly satisfying the covariance under the coordinate-transformations. This theory has been developed mainly based on TSDIA; 2-sacale direct interaction approximation, much more attention paid on the using of the covariant or invariant properties under the coordinate-transformations.With this new method, various kinds of turbulent-statistical properties can be modeled in a manner always satisfying the covariance. In this paper the authors produced an algebraic model for the Reynolds stress and found the model satisfying the covariance clearly.[This abstract is not included in the PDF]