Wind for Dense and Hi Rise Cities - Tokyo Institute of...
Transcript of Wind for Dense and Hi Rise Cities - Tokyo Institute of...
The 7th International Conference on Urban Climate, Yokohama, Japan (June 29‐July 3, 2009)
1
External Wind Modeling for Dense and Hi‐Rise Cities
Validation of turbulence modeling against wind tunnel tests and on‐site measurements
Raymond M. H. Yau1 Arup Fellow, Wilson Jiang1, Sui‐Hang Yan1, Trevor S. K. Ng1, Rumin Yin1, & Edward Y.Y. Ng2 1 Ove Arup and Partners Hong Kong Ltd,
2 Department of Architecture, Chinese University of Hong Kong
Abstract
Hong Kong and many other urban areas are densely populated city; hi‐rise build form has advantages in land‐use, transportation efficiency, but places severe strains on ventilation environment. With increasing public concern, guideline on Air‐Ventilation‐Assessment (AVA) has been introduced to encourage good practices in building design through the use of Computational Fluid Dynamics (CFD).
Despite the prevalence of using CFD in modeling the urban ventilation environment, there remain uncertainties on what is the optimal combination of turbulence models and simulation parameters (inflow condition, wall functions, model domain) that gives acceptable accuracy while commanding a reasonable level of computational resources. This work addresses the uncertainties through, firstly, studying each modeling parameter in an idealized representation of a city‐scape, comparing wind‐tunnel results against CFD; and secondly, the findings are verified through comparing on‐site measurements taken in Shinjuku to CFD data.
Within our framework, the Detached‐Eddy‐Simulation (DES) combines the computational efficiency of RANS models and the accuracy of LES models, and is best able to represent the interactions between detached flows of closely packed buildings typical of Asian cities. Recommendations on other simulation parameters are also discussed.
This work builds on our previous research [Yau et al Arup 2007] and forms the core of our recommendations towards formulation of CFD guidelines on Urban Wind Performance Analysis of Hong Kong.
1 Introduction
Our study is broadly divided into two stages. For the first stage, a simplified model representing typical scenarios was constructed and the CFD result was verified against results from the wind tunnel. The objective of this stage is to utilize a relatively simple situation to highlight the different features of the flow to facilitate comparison. Results from the averaged DES turbulence model is compared against conventional RANS models. In the second stage, we apply our findings to a real‐life test case at the Shinjuku District of Japan through a comprehensive Air Ventilation Assessment (AVA).
2 Idealized Model
In the first part of this paper, we present selected results form a combined wind tunnel scale and real scale CFD study of wind effects on urban wind climate. The wind tunnel test of an idealized city was performed by Yoshie et al (Yoshie et al 2005), this was used as comparison against CFD performed within Arup. The comparisons focused on accuracy and reliability for different simulation conditions commonly used in practical applications were presented and discussed. An in‐depth was also conducted to investigate the influence of boundary inflow condition, wall function for buildings, surrounding building range. We discuss the merits and draw‐backs of different turbulence models with particular focus on wind effects at pedestrian level.
Turulence Model Comparison with Wind Exp.
0.00
1.00
2.00
3.00
4.00
5.00
6.00
0.00 1.00 2.00 3.00 4.00 5.00 6.00
Experimental Data
CFD
Res
ult
DES k-e RNG
Wind Tunnel Test
Computational Simulation
Single high‐rise model with surrounding rows of buildings, testing perimeters located around the centre pedestrian region.
The seventh International Conference on Urban Climate, 29 June - 3 July 2009, Yokohama, Japan
The 7th International Conference on Urban Climate, Yokohama, Japan (June 29‐July 3, 2009)
2
Figure 1 Methodology of Study – An idealized city profile: A 1/400th scale model is constructed and tested in the wind tunnel. While the equivalent computational geometry in both 1/400th and full‐scale are constructed and tested using computational fluid dynamics techniques.
2.1 Analysis
2.1.1 Turbulence Models
One of the most important objectives of our analysis is to assess the performance of the different turbulence model currently in use. The particular areas of focus are:
1. Qualitative accuracy in capturing the important flow features
2. Quantitative accuracy in predicting pedestrian level wind velocities
3. Efficient use of computational resources
In addition to the standard and modified k‐ε discussed by Yoshie et al (Yoshie et al 2005), we also performed simulations using the following models:
Reynolds Average Navier‐Stokes
• Spallart‐Allmaras (SA)
• Reynolds Normalized Group (RNG)
Eddy Models
• Detached Eddy Simulation (DES)
• Large Eddy Simulation (LES) – subject of future work.
By comparing the results generated from the CFD model against data collected through wind tunnel anemometry, we can see that DES typically outperforms the other models.
Figure 2 Comparison of overall CFD results against Wind tunnel data. In terms of velocity magnitude, DES outperforms conventional RANS models.
DES is inherently an unsteady simulation, and a time averaging is performed to calculate the averaged velocity over approximately 18s (estimated to be significantly larger than the (Strouhal frequency)‐1. In our upcoming work, we will study the optimum period required for the DES model to reach steady state.
3 Air Ventilation Assessment of the Shinjuku Test Case
The findings from the first part of this study were applied to a realistic built area at the Shinjuku District in Japan using the Air Ventilation Assessment Methodology (AVA).
The seventh International Conference on Urban Climate, 29 June - 3 July 2009, Yokohama, Japan
The 7th International Conference on Urban Climate, Yokohama, Japan (June 29‐July 3, 2009)
3
36
10
98
7
6
5
3
1418
1920
2
1
21
13
22
23243031
32
17 16 1533
34
29
25
26
27
28
11
D
C
Œ “́ _
Figure 3 Tests points at the Shinjuku Area
Figure 4 shows some extracted point results under 16 wind directions. In general, CFD could is able to give good quantitative prediction over 16 wind directions if the point is located at the centre of domain and surrounded by certain layers of buildings. However, there are a number of exceptions, such as point 11 and 28, which are located at the edge of the model. In particular, the VRs are always over‐predicted for those locations where wind arrives from the non‐building model area (i.e. northern part of point 11 & 28). This is consistent with our findings from stage 1 where we found that a “wind pre‐conditioning” layers of buildings is important for providing an accurate simulation result. This finding is consistent the 2R surrounding area stipulated in the AVA guidelines. i.e. there would have 1H area for the “wind pre‐conditioning” layers of buildings on ensuring modeling accuracy.
Figure 4 Wind rose of selected test points
4 Conclusions
In this study, the underlying assumptions and physics of air ventilation in a high‐rise and densely built environment is first analysed in a simplified model. In the second stage, the finds were applied and verified in a real‐life situation at the Shinjuku district of Japan.
RANS models are the industrial standards in engineering practices due to its simplicity and relativity low computational demand. However, this type of model has its limitations on prediction accuracy. Large Eddy Simulation Models (LES) requires significant computational resources and are impractical for commercial applications for urban environment flows. Implementation of Detached Eddy Simulation model is an optimum compromise that combines the accuracy of LES in the regions of interest while using a practical level of computational resources. The simulation results shown that DES model outperforms the others at both
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
N
NNE
NE
ENE
E
E SE
SE
SSE
S
SSW
SW
WSW
W
WNW
NW
NNW
Exp
SA
RNG
V2F
DES
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
N
NNE
NE
ENE
E
E SE
SE
SSE
S
SSW
SW
WSW
W
WNW
NW
NNW
Exp
SA
RNG
V2F
DES
0
0.1
0.2
0.3
0.4
0.5
0.6
N
NNE
NE
ENE
E
ESE
SE
SSE
S
SSW
SW
WSW
W
WNW
NW
NNW
Exp
SA
RNG
V2F
DES
Point 6
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
N
NNE
NE
E NE
E
E SE
SE
SSE
S
SSW
SW
WSW
W
WNW
NW
NNW
Exp
SA
RNG
V2F
DES
Point 2
Point 28
Point 11
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
N
NNE
NE
E NE
E
ESE
SE
SSE
S
SSW
SW
WSW
W
WNW
NW
NNW
Exp
SA
RNG
V2F
DES
Point 15
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
N
NNE
NE
ENE
E
ESE
SE
SSE
S
SSW
SW
WSW
W
WNW
NW
NNW
Exp
SA
RNG
V2F
DES
Point 17
The seventh International Conference on Urban Climate, 29 June - 3 July 2009, Yokohama, Japan
The 7th International Conference on Urban Climate, Yokohama, Japan (June 29‐July 3, 2009)
4
windward, leeward and congested locations although DES model may not give a good prediction at open space due to grid resolution issue. DES model is recommended for studying external wind modeling.
5 Acknowledgement
The Authors will like to Thank Prof. Mochida, Prof Yoshie, Dr Kataoka, and Prof Tominaga of AIJ CFD Group for providing the experimental data and invaluable comments. In addition, we will like to show our appreciation for Arup Design and Technology Executive Committee for providing funding to this enabling technology research project (#77314‐13).
6 Bibilography
Architectural Institute of Japan. Guidebook for practical applications of CFD to pedestrian wind environment around buildings. Architectural Institute of Japan, Japan, 2007
CD adapco Group. STAR‐CD – Methodology. CD adapco Group.UK, 2003
Franke, J., (2006). Recommendations of the COST action C14 on the use of CFD in predicting pedestrian wind environment. In: The Fourth International Symposium on Computational Wind Engineering, Yokohama, Japan, July 2006.
Hu C‐H., Wang, F., (2005), ‘Using a CFD approach for the study of street‐level windsin a built‐uparea’, Building and Environment, 40(5):617‐631
Kataoka H., T. Tamura, Y Okuda and M. Ohashi. Numerical evaluation of the wake field behind high‐rise building by RANS and LES. 12th International Conference on Wind Engineering. Australia, 2007.
Li X.X., C.H. Liu, D.Y.C. Leung and K.M. Lam, Recent progress in CFD modeling of wind field and pollutant transport in street canyons. Atmospheric environment. Vol. 40 (2006) 5640‐5658
Planning Department of HKSAR. Feasibility study for establishment of air ventilation assessment system. Hong Kong, 2005
Sahm P., P. Louka,M. Ketzel, E. Guilloteau and J. Sini. Numerical and experimental modeling of pollutant dispersion in a street canyon. Journal of Wind Engineering and Industrial Aerodynamics. Vol. 90 (2002) 321‐339
Yau R., S. H. Yan, R. Yin. Air Ventilation Assessment by External Wind Modelling Using Different Turbulence Models. Arup Research Review (2008) 32‐35
Yoshie, R., Akashi, M., Tominaga, Y., Hiroto, K., Yoshikawa, M., (2005) ‘Cross Comparisons of CFD Prediction for Wind Environment at Pedestrian Level around Buildings’, The Sixth Asia‐Pacific Conference on Wind Engineering (APCWE‐VI) Seoul Korea, September 12‐14, 2005
Walton A., A.Y.S. Cheng and W.C. Yeung. Large‐eddy simulation of pollution dispersion in an urban street canyon –part I: comparison with field data. Atmospheric Environment. Vol. 36 (2002) 3601‐3613
The seventh International Conference on Urban Climate, 29 June - 3 July 2009, Yokohama, Japan