Session Overview
NOMTM3: Computational Fluid Dynamics models
Thursday, 23/Jul/2015:
9:30am - 10:30am

Session Chair: Manabu Kanda, Tokyo Institute of Technology
Location: Caravelle Room


Thermal stratification and vegetation effects on the urban microclimate – a CFD case study

Bharathi Boppana1, Yushi Liu1, Hee Joo Poh1, Matthias Roth2

1Institute of High Performance Computing, Singapore; 2National University of Singapore, Singapore

Increasing urbanization renders both challenges and opportunities in myriad ways. While the job prospects, good living standards etc. belong to the latter category, the former include land scarcity, urban heat island phenomenon, meeting the needs such as infrastructure, resources etc. of increasing population. Of these, understanding urban microclimate, especially in densely populated cities is very crucial as it provides the basis for creating a highly liveable residential town. Although, several urban studies have been done in the past, most were conducted in mid-latitude regions and only very few in the tropics and sub-tropic regions.

Therefore, we aim to develop an urban microclimate – multi-physics integrated simulation tool (UM-MIST) that will enable town planners, architects, policy makers etc. to develop more sustainable cities, and which will provide better insights into the flow physics in a densely populated tropical city like Singapore. This multi-physics tool will incorporate the effects of thermal stratification, vegetation and solar irradiance on urban flows, and enable the whole master planning, urban design & environmental modelling to be carried out in a single urban digital platform.

The preliminary study will consist of CFD simulations for a small urban site (less than 0.5 km2) in Singapore. This urban site comprises of various complex features such as high-rise residential buildings of approximately 50m height, low-rise commercial buildings of approximately 10m height, courtyards, vegetation, roads etc. The objectives of this preliminary study are to (i) represent all the complex features of the geometry to the required detail, (ii) perform robust meshing, and (iii) analyse the effects of the presence and absence of vegetation in both neutral and unstably stratified flows. The steady-state RANS simulations will be carried out using Star-CCM and OpenFOAM and preliminary results will be presented in detail at the conference.


High resolution numerical study of wind, thermal effects and pollution dispersion in urban neighborhoods in Toulouse and Marseille


1MFEE, R&D, EDF; 2CEREA, Ecole des Ponts ParisTech

Detailed, high resolution, 3D numerical simulation are used to study short episodes of local urban climate and pollution dispersion in the neighborhood of Bordelongue in Toulouse and Saint Marcel in Marseille in the framework of the French ANR project EUREQUA. These urban areas consist of various types of buildings and obstacles: house districts, tower blocks, highway, some local streets, vegetation areas, etc. The 3D geometry of this urban area was constructed with an in house tool developed around the open-source geometry and mesh generator SALOME, based on the available geophysical data SIG from the French geographical institute (IGN). The open-source computational fluid dynamic (CFD) code Code_Saturne, with the atmospheric option developed at CEREA, was used to carry out the simulations. The atmospheric model already implemented by solving the Navier-Stokes equation with a k-epsilon turbulent model is used to compute the air flow in the urban area. It is coupled with a newly implemented heat transfer and radiative model, which is able to take into account the effects of solar and atmospheric radiation, local traffic, as well as the heat transfer at the solid surfaces (building walls, ground, roads, etc). The vegetation area, principally composed with trees, is considered as a porous layer which induces a drag force to the air flowing through it. The pollutants of the local traffic emissions are considered as passive scalars, which are solved by the transport equations, transported by local winds and dispersed by local turbulence. The global meteorology condition is taken into account by using the measurement data at the meteorological station installed in the area during the different campaigns of 72-hour measurements in the framework of ANR project EUREQUA or by using results from mesoscale simulations performed over the region using the Meso-NH code. The simulation results of the air flow, heat transfer, and pollution dispersion are compared with the various measurements obtained both with a mobile station circuiting along a trajectory through several regularly spaced observation points in the neighborhood and fixed stations especially set up in the area during the campaigns. These validation results will also be compared with sociologically analyzed results based on the subjective perceptions of the environment by the local residents during the campaigns. The numerical tool and collected results can be used for the future studies of urban renewal scenarios.


Numerical study of the influence of albedo on the microclimate of Bergpolder Zuid, Rotterdam

Yasin Toparlar1,2, Bert Blocken1,3, Bino Maiheu2, Gert-Jan van Heijst4

1Building Physics and Services, Eindhoven University of Technology, Eindhoven, the Netherlands; 2Environmental Modeling, Flemish Institute for Technological Research, Mol, Belgium; 3Building Physics Section, University of Leuven, Heverlee, Belgium; 4Turbulence and Vortex Dynamics, Eindhoven University of Technology, Eindhoven, the Netherlands

The rapid trend towards urbanization has increased the popularity of studies investigating urban microclimates. Factors such as the Urban Heat Island (UHI) effect and summer-time heat waves can have a significant influence on outdoor thermal comfort levels and building energy demand. For the investigation of urban microclimates, historically, observational approaches (such as field measurements) have been employed and lately, with the advancements in computing power, computational methods are becoming more popular because of the possibility to conduct scenario analysis. Computational Fluid Dynamics (CFD) is one of these computational methods which has proven to have the potential to accurately predict urban microclimate. Compared to other computational methods employed to investigate urban microclimate (i.e. Energy Balance Models), CFD has the capability to couple flow field with the temperature field at the cost of more computational resources. In this study, CFD simulations are performed to predict urban temperatures in the Bergpolder Zuid region, located in Rotterdam. The area is planned to be renovated and one of the aims is to increase its climate resilience. Based on the municipal drawings of the Bergpolder region, a high-resolution computational domain is generated. The computational domain has a hexagonal shape with an edge length of 2.4 kilometer and a height of 400 meter. In total, the grid consists of 6,610,456 hexahedral cells. The simulations are performed using the 3D unsteady Reynolds-averaged Navier-Stokes (URANS) equations in combination with the realizable k-epsilon turbulence model. Several physical phenomena influencing urban microclimate such as wind flow and heat transfer (conduction, convection and radiation) are considered in the simulations. As for evapotranspiration, during morning (6:00 – 11:00 h) and afternoon (15:00 – 18:00 h) a constant and uniform sink value of 80W/m2 is imposed and during noon time (11:00 – 15:00 h) the sink value is specified as 130 W/m2 (at the ground level). Meteorological data used in this study is obtained from the hourly averaged dataset of the Royal Dutch Meteorological Institute (KNMI) by the Rotterdam weather station, which is located near Rotterdam airport, 4 km northwest of Bergpolder Zuid. In the first part of this research, resulting surface temperatures on the region were validated using experimental data from high-resolution thermal infrared satellite imagery, which were recorded during the heat wave of July 2006. The resulting average surface temperatures showed a deviation of 7.9% with the satellite imagery data. In the second part, the investigation focused on the effect of different physical parameters, such as thermal diffusivity and albedo on outdoor thermal comfort, considering the validated time range. Preliminary results show that changing the thermal diffusivity of the building materials and albedo value of the ground cover can influence the average surface temperatures in an urban environment significantly. Complete results regarding the effect of physical parameters on outdoor thermal comfort are still being evaluated and the results will be available in the full paper.


Advanced numerical analysis on sensible heat flux from building external surfaces to the surrounding atmosphere using a heat balance simulation and CFD

Kan CHEN, Takashi ASAWA

Tokyo Institute of Technology, Japan

Sensible heat flux from building external surfaces is one of the essential factors warming the atmosphere around buildings and causing urban heat island effect. Field measurements are difficult to cover the whole sensible heat fluxes from the building external surfaces with complex geometries and the scales of wind tunnel experiments are different from that of reality. Therefore numerical analysis is considered to be effective method for grasping the sensible heat flux from the whole building surfaces. Although surface temperature, air temperature and air flow near the building surfaces, all of which dominate the sensible heat flux should be considered in the prediction, previous studies did not deal with these mutual and detailed relations.

This study applies the 3D CAD-based heat balance simulation, which was developed by the authors’ group and can calculate the detailed surface temperature distribution taking into account the building geometry, and a CFD simulation. By using Low Reynolds number k-ε model (Low-Re) for the CFD simulation, we can simulate air temperature and wind speed distribution inside the wall boundary layers. The purpose of this study is to examine the method to predict the sensible heat flux from the building external surfaces by using coupled analysis of the heat balance simulation and the CFD (Low-Re) simulation. And this study is also implemented to figure out the appropriate prediction formula for CHTC (Convective heat transfer coefficient) that can be used for the prediction of heat fluxes by high Reynolds number k-ε model (High-Re), in order to apply this method to the analysis of urban block scales.

Firstly, we used High-Re and Low-Re to predict CHTC and compared these results with the result of a wind tunnel experiment. It is shown that the Low-Re showed the similar result as the wind tunnel experiment. Then we used coupled analysis of the heat balance simulation and CFD (Low-Re) to predict the sensible heat flux from the building external surfaces of a house complex. We confirmed that with considering the building geometry, wind speed attenuation inside veranda spaces and undeveloped boundary layer on the edge affect the prediction of sensible heat flux. Finally we found out that the correlation between CHTC and the effective wind speed near the building surface was high in forced convection dominant cases. The experimental approximation formula obtained from the analysis is considered to be appropriate for the heat flux prediction using High-Re for urban block scales.