Numerical investigation of the effect of road structures on ambient air quality—for their better design

Two-dimensional numerical simulation for investigating wind and concentration field around a double-decked road structure was performed using a standard k-ε turbulence model. The main objective of this paper is to study how road fences installed at a double-decked road affect ambient air quality, especially, pollutant concentration at some downstream locations. For model validation, calculated results were compared with available field experiment. Performance of the standard k-ε model was also compared with that of the renormalization group k-ε model for the double-decked road. Obtained results clarified how and how much pollutant concentration distribution is influenced by road structures with and without fences: the fences on the upper deck have generally positive effect on decreasing of air pollution near ground level while those on the ground always not. The computer code we used is CFX4.     

On the numerical study of contaminant particle concentration in indoor airflow

The application of three turbulence models—standard k–ε, re-normalization group (RNG) k–ε and RNG-based large eddy simulation (LES) model—to simulate indoor contaminant particle dispersion and concentration distribution in a model room has been investigated. The measured air phase velocity data obtained by Posner et al. [Energy and Buildings 2003;35:515–26], are used to validate the simulation results. All the three turbulence model predictions have shown to be in good agreement with the experimental data. The RNG-based LES model has shown to yield the best agreement. The flow of contaminant particles (with diameters of 1 and 10 μm) is simulated within the indoor airflow environment of the model room. Comparing the three turbulence models for particle flow predictions, the RNG-based LES model through better accommodating unsteady low-Reynolds-number (LRN) turbulent flow structure has shown to provide more realistic particle dispersion and concentration distribution than the other two conventional turbulence models. As the experimental approach to access indoor contaminant particle concentration can be rather expensive and unable to provide the required detailed information, the LES prediction can be effectively employed to validate the widely used k–ε models that are commonly applied in many building simulation investigations.     

Dispersion of gas pollutant in a fan-filter-unit (FFU) cleanroom

Numerical simulation and experimental measurement of flow and concentration fields in a working fan-filter-unit (FFU) cleanroom have been conducted in this study. The purpose of the study is to find out the unsteady concentration distribution of a leaking gas pollutant. The standard K–ε model was used for the simulation of the flow field. To obtain the gas concentration field, SF₆ gas with a certain concentration was released as a simulated leaking source from a valve manifold box (VMB) for 5 or 10 min, respectively. Three Fourier transform infrared spectrometers (FTIRs) were simultaneously used to measure the spatial and temporal distributions of SF₆ concentrations. The measured data were then compared with the numerical results and the agreement is seen to be quite good. From the numerical results, the pollutant hot spots, peak pollutant concentration at the end of leaking, and time taken for the concentration to reduce to near background level are obtained.     

Evaluating the influence of the width of inlet slot on the prediction of indoor airflow: Comparison with experimental data

The aim of this work is to investigate the influence of two values of inlet slot width on the velocity characteristics and turbulent intensity of the airflow inside a rectangular room. The experimental data used to check the numerical results concerns a rectangular room where the air is supplied horizontally on the upper left and is exhausted through an opening on the lower right on the opposite side. The performance of three turbulence models, standard k-?, RNG k-?, and k-ω, in predicting the three-dimensional airflow in that room has also been investigated. The results for Reynolds number of 5000 are presented for dimensionless horizontal velocities and turbulent kinetic energy for two planes of the room and two inlet arrangements, one opening as large as the room and another with half of the width of the room. The results have indicated that the main features of the flow were captured by the three turbulence models investigated. On the whole, the performance of the standard k-? model was better than those of the other two turbulence models. In particular, the k-ω model performed better in the configuration with the largest air opening than in that with the smallest one, while the RNG k-? model presented the opposite behavior. The comparative study between both geometries demonstrated that for slots much smaller than the width of the room, three-dimensional effects become important.     

New inflow boundary conditions for modelling the neutral equilibrium atmospheric boundary layer in computational wind engineering

Modelling neutral equilibrium atmospheric boundary layers (ABL) in CFD is an important aspect in computational wind engineering (CWE) applications. In this paper, new inflow boundary conditions are introduced from the viewpoint that these boundary conditions should satisfy the turbulence model employed. The new set of inflow turbulence boundary conditions is an approximate solution to the standard k-ε model transport equations. The capability of these boundary conditions to produce an equilibrium ABL is demonstrated by performing numerical simulations in an empty domain. The new inflow turbulence boundary conditions in this paper support future practical applications in CWE and future research in modelling equilibrium ABLs.     

Numerical studies on air flow around a cube

In this paper, air approach flow moving towards a cube will be studied using computational fluid dynamics (CFD). The Reynolds Averaging of Navier-Stokes (RANS) equation types of k-ε turbulence model are used. Some RANS predicted results are compared with different upstream air speeds. Flow separation at the corner above the top of the cube, level of separation and reattachment are investigated. Reference is made to the experimental data on wind tunnels reported in the literature.A method similar to ‘recirculation bubble promoter’ is used for different approach flow speed distributions. Problems encountered in numerical simulations due to the sharp corner are discussed with a view to obtaining better prediction on recirculation flow in regions above the top of the cube. Correlations between the turbulent kinetic energy above the cube and the recirculation bubble size are derived for different distributions of approach flow speed.By limiting the longitudinal velocities in the first cell adjacent to the sharp edge of the cube or rib, and making good use of the wall functions at the intersection cells of the velocity components, positions of maximum turbulent kinetic energy and the flow separation and reattachment can be predicted by a standard k-ε model. The results agree with those obtained in the experiments.     

Applicability of linear type revised k-ε models to flow over topographic features

Accurate prediction of the wind energy distribution over terrains is essential for the appropriate selection of a suitable site for a wind power plant. This paper presents two-dimensional numerical simulations of flow over three common types of topographic features, i.e., a hill and two types of slopes (up-slope and down-slope). In a previous investigation by the present authors [Lun, Y.F., Mochida, A., Murakami, S., Yoshino, H., Shirasawa, T., 2003. Numerical simulation of flow over topographic features by revised k-ε models. J. Wind Eng. Ind. Aerodyn. 91(1-2), 231-245], the revised k-ε model proposed by P.A. Durbin [1996. Technical note: on the k-ε stagnation point anomaly. Int. J. Heat Fluid Flow 17, 89-90] was applied to flow prediction over a hill. Although, this model works well for flow around bluff bodies, a limitation was revealed in the area downstream of the hill. In this study, two new revised k-ε models proposed by Y. Nagano, H. Hattori and T. Irikado [2001. Prediction of flow over a complex terrain using turbulence model. In: Proceedings of the TED-Conference’01, JSME, in Japanese] and by Y. Nagano and H. Hattori [2003. A new low-Reynolds-number turbulence model with hybrid time-scales of mean flow and turbulence for complex wall flows. In: Proceedings of the Fourth International Symposium on Turbulence, Heat and Mass Transfer, Antalya, Turkey, October 12-17, 2003], i.e., the Ω and S-Ω models, were employed. These models are based on a mixed-time-scale concept. Their performance in predicting flow over various topographic features, namely a hill, up-slope and down-slope, was investigated. The problem of the Durbin model was corrected by the Ω model. However, a drawback of the Ω model was found in the upstream region. A new model, the S-Ω model, was introduced and was found to correct this problem. The S-Ω model showed best agreement with experiments for the hill case and the slope cases.     

Experimental and numerical study of a full scale ventilated enclosure: Comparison of four two equations closure turbulence models

Full-scale experimental and computational fluid dynamics (CFD) methods are used to investigate the velocity and temperature fields in a mechanically ventilated enclosure. Detailed airflow fields are measured in three cases of ventilation air temperature: an isothermal case, a hot case and a cold case. The ventilation system creates an axisymmetric jet which is developing near the ceiling. The experimental data are used to test four two equations turbulence models: a k–ε realizable model, a k–ε RNG model, a k–ω model and a k–ω SST model. It is found that, even if the models can predict reasonably the hot and isothermal cases global values of temperature and velocity, none of the models is reliable concerning the cold case. Moreover, a detailed analysis of the jet shows that none of the models is able to predict the exact experimental velocity and temperature fields.     

Simulation of mean flow and turbulence over a 2D building array using high-resolution CFD and a distributed drag force approach

A numerical simulation has been performed of the disturbed flow through and over a two-dimensional array of rectangular buildings immersed in a neutrally stratified deep rough-walled turbulent boundary-layer flow. The model used for the simulation was the steady-state Reynolds-averaged Navier-Stokes equations with linear and non-linear eddy viscosity formulations for the Reynolds stresses. The eddy viscosity was determined using a high-Reynolds number form of the k-ε turbulence-closure model with the boundary conditions at the wall obtained with a standard wall-function approach. The resulting system of partial differential equations was solved using the SIMPLE algorithm in conjunction with a non-orthogonal, colocated, cell-centered, finite volume procedure. The predictive capabilities of the high-resolution computational fluid dynamics (CFD) simulations of urban flow are validated against a very detailed and comprehensive wind tunnel data set. Vertical profiles of the mean streamwise velocity and the turbulence kinetic energy are presented and compared to those measured in the wind tunnel simulation.It is found that the performance of all the turbulence models investigated is generally good—most of the qualitative features in the disturbed turbulent flow field through and over the building array are correctly reproduced. The quantitative agreement is also fairly good (especially for the mean velocity field). Overall, the non-linear k-ε model gave the best performance among four different turbulence closure models examined. The turbulence energy levels within the street canyons and in the exit region downstream of the last building were underestimated by all four turbulence closure models. This appears to contradict the ‘stagnation point anomaly’ associated with the standard k-ε model which is a result of the excessive turbulence energy production due to normal straining. A possible explanation for this is the inability of the present models to account properly for the effects of secondary strains on the turbulence and/or for the effects of large-scale flapping of the strong shear layer at the canopy top.The results of the high-resolution CFD simulations have been used to diagnose values of the drag coefficient to be used in a distributed drag force representation of the obstacles in the array. Comparisons of the measured spatially-averaged time-mean mean velocity and turbulence kinetic energy in the array with predictions of the disturbed flow using the distributed drag force approach have been made.     

Evaluation of RANS turbulence models for simulating wind-induced mean pressures and dispersions around a complex-shaped high-rise building

Re-ingestion of the contaminated exhaust air from the same building is a concern in high-rise residential buildings, and can be serious depending on wind conditions and contaminant source locations. In this paper, we aim to assess the prediction accuracy of three k-? turbulence models, in numerically simulating the wind-induced pressure and indoor-originated air pollutant dispersion around a complex-shaped high-rise building, by comparing with our earlier wind tunnel test results. The building modeled is a typical, 33-story tower-like building consisting of 8-household units on each floor, and 4 semi-open, vertical re-entrant spaces are formed, with opposite household units facing each other in very close proximity. It was found that the predicted surface pressure distributions by the two revised k-? models, namely the renormalized and realizable k-? models agree reasonably with experimental data. However, with regard to the vertical pollutant concentration distribution in the windward re-entrance space, obvious differences were found between the three turbulence models, and the simulation result using the realizable k-? model agreed the best with the experiment. On the other hand, with regard to the vertical pollutant concentration distribution in the re-entrant space oblique to the wind, all the three models gave acceptable predictions at the concentration level above the source location, but severely underestimated the downward dispersion. The effects of modifying the value of the turbulent Schmidt number in the realizable k-? model were also examined for oblique-wind case. It was confirmed that the numerical results, especially the downward dispersion, are quite sensitive to the value of turbulent Schmidt number.     

Effects of plume buoyancy and momentum on the near-wake flow structure and dispersion behind an idealized building

Dispersion simulations of buoyant and neutral plume releases within the recirculation cavity behind a cubical building were performed using a commercially available CFD code and the RNG k-ε turbulence model. Plume buoyancy was observed to affect the size and shape of the cavity region and the flow structure and concentration profiles within. Source momentum of a neutral plume release had similar effects on the flow structure and the cavity region to that caused by plume buoyancy. However, the effects of momentum on the concentration profiles were noticeably different from that caused by plume buoyancy. Plumes released immediately downwind of a cubical building appear to alter the flow field and dispersion characteristics of the cavity recirculation region due to their inherent momentum and buoyancy. A greater fraction of a plume was captured inside the wake as the plume became increasingly buoyant. Contrarily, greater plume momentum resulted in smaller plume fractions captured inside the wake. Inclusion of these effects in the downwash algorithms would improve the accuracy of modeling results for far-field concentration distributions and would be mandatory in accident assessments where accurate predictions of short-term, near-field concentration fluctuations near source releases are required.     

A comparative study of the Mellor-Yamada and k-ε two-equation turbulence models in atmospheric application

The present study systematically compares the Mellor-Yamada (MY) model and the k−ε algebraic stress model in order to verify the possibility of using the k−ε algebraic stress model in atmospheric applications. The results of the parameterization process and atmospheric application of both models confirmed that the MY model neglected the pressure redistribution effect of buoyancy due to 〈u_iu_j〉 and 〈u_iθ〉 and that of shear due to 〈u_iθ〉. In addition, the MY model overestimated the turbulent energy dissipation. Based on the formulation of the k−ε algebraic stress model, we modified the constant value C_μ(=0.09) in the standard k−ε model to obtain the variables C_μM and C_μH to account for atmospheric stability. Finally, the results of the simulation obtained from the Wangara experiment verify the possibility of using the k−ε algebraic stress model in atmospheric application.     

The numerical computation of near-wall turbulent flow over a steep hill

The present work performs a detailed comparison between numerical computations for the flow over a two-dimensional steep hill and some newly obtained laboratory data. Six turbulence models were tested: four eddy-viscosity models (κ-ε, RNG-ε, κ-ω, SST) and two second-moment models (SSG-RSM-ε, BSL-RSM-ω). The experiments were conducted in a water channel and were specially planned such that the large separated flow region that is formed on the lee side of the hill could be well scrutinized. The experimental results include complete profiles of the mean velocity components and of the two-dimensional Reynolds stress tensor and were obtained through the laser Doppler anemometry. A particular concern of this work has been to achieve a detailed experimental and numerical characterization of the near-wall flow region. As such, for most of the measuring stations, at least eight points were located in the viscous sublayer. The work also shows the distribution of wall-shear stress in detail. The ω-equation-based models were observed to perform much better than the ε-equation-based models. The length of separated flow region, mean velocity profiles and wall-shear stress were all reasonably well predicted. The flow properties on the hill top were particularly difficult to describe. The turbulence properties in the reversed flow region were best simulated by the BSL-RSM model.     

The effect of ceiling configurations on indoor air motion and ventilation flow rates

The purpose of this paper is to evaluate the effects of a building parameter, namely ceiling configuration, on indoor natural ventilation. The computational fluid dynamics (CFD) code Phoenics was used with the RNG k–? turbulence model to study wind motion and ventilation flow rates inside the building. All the CFD boundary conditions were described. The simulation results were first validated by wind tunnel experiment results in detail, and then used to compare rooms with various ceiling configurations in different cases. The simulation results generated matched the experimental results confirming the accuracy of the RNG k–? turbulence model to successfully predict indoor wind motion for this study. Our main results reveal that ceiling configurations have certain effects on indoor airflow and ventilation flow rates although these effects are fairly minor.     

Numerical modeling of passive scalar dispersion in an urban canopy layer

A turbulent dispersion model describing the dispersion of a passive scalar from a localized source released in a built-up environment (urban area) is presented. The proposed model simulates both the flow field in the urban complex using the ensemble-averaged, three-dimensional Navier-Stokes equations with a standard k-ε turbulence closure model and the turbulent diffusion using transport equations for the mean concentration and concentration variance of the scalar. Two models for the scalar dissipation rate, required to close the transport equation for concentration variance, are investigated. Results of a detailed comparison of the flow and turbulent dispersion between a comprehensive water channel experiment and the model predictions are presented. The water channel experiment is unique in the sense that it includes data obtained from the dispersion of both continuous and nearly instantaneous releases of a tracer from a point source located within a regular array of building-like obstacles, and this data include measurements of both the mean concentration and concentration variance.     

CFD simulations of gas dispersion around high-rise building in non-isothermal boundary layer

Urban heat island phenomena and air pollution become serious problems in weak wind regions such as behind buildings and within street canyons, where buoyancy effect cannot be neglected. In order to apply CFD techniques for estimation of ventilation and thermal and pollutant dispersion in urban areas, it is important to assess the performance of turbulence models adopted to simulate these phenomena. As the first step of this study, we carried out wind tunnel experiments and CFD simulations of gas and thermal dispersion behind a high-rise building in an unstable non-isothermal turbulent flow. The standard k-ε model and a two-equation heat-transfer model as RANS models, and LES, were used for the CFD simulation. One of the important purposes of this study was to clarify the effect of inflow turbulence (both velocity and temperature) on flow field and gas/thermal dispersion for the LES calculation. Thus, LES calculations with/without inflow turbulence were conducted. The inflow turbulence was generated through a separate precursor simulation. The calculated results showed that both RANS models overestimated the size of the recirculation region behind the building and underestimated the lateral dispersion of the gas. Turbulent flow structures of LES with and without inflow turbulence were completely different. The LES result with inflow turbulence achieved better agreement with the experiment.     

A computational model to calculate the flow-induced pressure fluctuations on buildings

A computational model to predict the flow-induced pressure fluctuation on bluff bodies is presented. Unlike direct and large-eddy simulation, the present model employs a stochastic model to generate plausible velocity fluctuations (synthetic turbulence) that satisfy the mean turbulent quantities such as turbulent kinetic energy (k) and dissipation energy rate (ε). This model has three main components: (1) prediction of mean flow quantities by solving the 3D Navier-Stokes equations using the standard k-ε model with Kato and Launder modifications; (2) generating a synthetic turbulent velocity field using a stochastic model and finally (3) solving the Poisson equation that governs the pressure fluctuations field. Flow around the low-rise building at Texas Tech was analyzed using the developed model. Two different wind angles of attack are considered for the analysis. Results obtained using the developed model are compared with wind tunnel and field measurements. The computed rms values for pressure fluctuations show good agreement with the experimental results.     

CFD investigation of turbulence models for mechanical agitation of non-Newtonian fluids in anaerobic digesters

This study evaluates six turbulence models for mechanical agitation of non-Newtonian fluids in a lab-scale anaerobic digestion tank with a pitched blade turbine (PBT) impeller. The models studied are: (1) the standard k-? model, (2) the RNG k-? model, (3) the realizable k-? model, (4) the standard k-ω model, (5) the SST k-ω model, and (6) the Reynolds stress model. Through comparing power and flow numbers for the PBT impeller obtained from computational fluid dynamics (CFD) with those from the lab specifications, the realizable k-? and the standard k-ω models are found to be more appropriate than the other turbulence models. An alternative method to calculate the Reynolds number for the moving zone that characterizes the impeller rotation is proposed to judge the flow regime. To check the effect of the model setup on the predictive accuracy, both discretization scheme and numerical approach are investigated. The model validation is conducted by comparing the simulated velocities with experimental data in a lab-scale digester from literature. Moreover, CFD simulation of mixing in a full-scale digester with two side-entry impellers is performed to optimize the installation.     

Improved k-ε model and wall function formulation for the RANS simulation of ABL flows

The simulation of Atmospheric Boundary Layer (ABL) flows is usually performed using the commercial CFD codes with RANS turbulence modelling and standard sand-grain rough wall functions. Such approach generally results in the undesired decay of the velocity and turbulent profiles specified at the domain inlet, before they reach the section of interest within the computational domain. This behaviour is a direct consequence of the inconsistency between the fully developed ABL inlet profiles and the wall function formulation.The present paper addresses the aforementioned issue and proposes a solution to it. A modified formulation of the Richards and Hoxey wall function for turbulence production is presented to avoid the well-documented over-prediction of the turbulent kinetic energy at the wall. Moreover, a modification of the standard k-ε turbulence model is proposed to allow specific arbitrary sets of fully developed profiles at the inlet section of the computational domain.The methodology is implemented and tested in the commercial code FLUENT v6.3 by means of the User Defined Functions (UDF). Results are presented for two neutral boundary layers over flat terrain, at wind tunnel and full scale, and for the flow around a bluff-body immersed into a wind-tunnel ABL. The potential of the proposed methodology in ensuring the homogeneity of velocity and turbulence quantities throughout the computational domain is demonstrated.