Interference effects around two model high rise buildings in asimulated non-synoptic event

Matthew Haines1, Mark Sterling2, and Andrew Quinn2

1School of Civil Engineering, University of Birmingham, Birmingham,B15 2TT, UKmrh046@bham.ac.uk

2School of Civil Engineering, University of Birmingham, Birmingham,B15 2TT, UK

Abstract

This study will investigate the interference effects between two model high rise CAARC build-ings at a scale of1 : 1000 in non-synoptic outflow. The University of Birmingham non-synopticeffects simulator will be used as well as a numerical model designed to simulate the simulator. Thetwo building’s yaw angle will be altered around a central axis to allow for various wind directionsto be compared. The approach of using a physical and numerical model will allow a faster andmore thorough understanding of the physical processes which produce the pressure coefficientsaround the model buildings.

1 Introduction

ABI (2005) notes that the loss of life, disruption and economic losses arising due to extreme stormsare increasing rapidly in recent years largely due to an increase in densely populated urban centres andan increase in wealth across much of Europe, USA and China. There is also evidence to suggest thatthese storms are increasing in frequency and magnitude although such evidence is far from conclusive(Kasperki, 1998). Any increases in magnitude or frequency of extremestorms are likely to result inserious damage to the urban infrastructure, the world economy and societyas a whole. In Europe itis predicted that by 2080, there will be an increase in "wind-related insured losses from extreme Eu-ropean storms by at least....e25-30bn" (ABI, 2005), although this estimate does not take into accountpopulation and wealth increase.

Within the last few years, thunderstorm downburst type events have received considerable inter-est from the wind engineering community with a mixture of measurement campaigns(Holmeset al.(2008)), physical experiments (Chay & Letchford (2002), Lin & Savory (2006), McConvilleet al.(2009)) and numerical modelling (Linet al. (2007) and Masonet al. (2008)) amongst others beingcarried out. Figure 3a illustrates that severe thunderstorms can producea streamwise velocity distri-bution which differs from the typical boundary layer flow.

1.1 Experimental Setup

1.1.1 Non-synoptic effects simulator

The basic University of Birmingham simulator set up is outlined in McConvilleet al. (2009) andis illustrated in figure 1. Some changes have since been made with a large raised platform beingconstructed to allow pressure taps to be placed on the floor if needed and the fans now being switchedoff after 1s after the flaps opening mechanism has released. This allows a study of the vortex which

1

6th European and African Wind Engineering Conference 2

is commonly found in thunderstorm downbursts rather than a study of the steady impinging jet whichfollows after this initial vortex.

Figure 1: The University of Birmingham downburst simulator

1.1.2 Numerical Simulation

The numerical domain is set up using OpenFoam, a10m × 10m × 2.5m box has a cylindrical .stlfile positioned in the centre with the inlet to the domain positioned on one surface of this .stl file ata height of2m. The inlet is set up to match the conditions found in the downburst simulator at thisheight, with an inlet velocity of13.7ms−1 and a turbulence intensity of13%. The roof and floor weresimulated as walls and the sides were modelled as outlet conditions. The numerical domain mesh andinitial flow for the DES case after0.155s is illustrated in figure 2.

Figure 2: The numerical domain of the simulation of the downburst simulator showing the mesh andthe initial flow of the DES simulation after 0.155s.

1.2 Previous single building studies

The potential wind loading arising as a result of downbursts or non-synoptic wind events has seenrelatively little research, especially within the experimental community. Previouswork by Chay &Letchford (2002) observed that for a stationary impinging jet on a model cube there was little dif-ference in the trend when compared to synoptic wind loading. However there were differences largedifferences in turbulence, approximately20% turbulence intensity in wall jet flow compared to0.5%in a synoptic wind tunnel. This led to greater variation in flow separation, in addition the wall jettended to have higher windward face pressures epecially when the cubewas close to the jet(C

D= 1).

For a translating impinging jet they found that the trend in wind loads again compared reasonably wellwith synoptic wind events except that the position of the jet caused the windward and leeward faces tochange and thus the wind loads changed accordingly.

6th European and African Wind Engineering Conference 3

The wind loading around a model high rise building has been previously examined by Haineset al. (2012) who examined the loading caused by the primary vortex formed on thestart up of thesimulator. It was felt that this was a more realistic simulation of a downburst event since in many casesthe maximum wind loading will be associated with the initial gust(s) front(s). In thisarrangement thewind loading differed with a positive pressure being found on the leewardface of the building atcertain times which is illustrated in figure 4a. The cause of this was a vortex forming on the rear ofthe building which is illustrated in figure 4b.Cp on this figure was defined as:

Cp =p− pref

1

2ρrefVref

2(1)

Wherep is the pressure at the location being measured,pref is a reference pressure taken outside ofthe flow,ρref is the density of the air andVref is typically taken to be the mean velocity at a fixedheight of10m. It is acknowledge that choosing a different value ofVref results in different values ofCp, althuogh the sign of the coefficients remain the same.

Figure 5 shows the vortex on the rear of the building rolling back and strikingthe rear of thebuilding in the same way that the flow visualisation in figure 4b does.

(a) Streamwise vertical profile

0

5

10

15

20

25

30

35

40

-500 0 500 1000

Ve

loci

ty (

m/s

)

Time (s)

Synoptic wind and downburst velocity- time

comparison

Synoptic wind

Downburst wind

(b) Synoptic and downburst velocity time history

Figure 3: (a) A schematic illustration of the mean streamwise velocity profile corresponding to a’typical’ downburst and a typical boundary layer wind. (Lin & Savory,2006). (b) A comparison of asynoptic and downburst wind velocity time history.

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1−0.1

00.10.20.30.40.50.60.70.80.9

Windwardface

Windwardroof edge

Roofface

Leewardroof edge

Leewardface

Cp values at time 8.066 s

Cp

Normalised Building Height

(a) Pressure coefficients across thebuilding as the vortex passes over-head

(b) Vortex on leeward side of building

Figure 4: The pressure on the model building and the associated rear vortex

6th European and African Wind Engineering Conference 4

(a) Vortex approaching the building (b) Vortex striking the building

Figure 5: The interaction of the vortex and the building in a URANS simulation

2 Previous interference effects studies in synoptic winds

Whilst a study on an individual building has merits, a modern urban environment typically has a vari-ety of buildings which are often densely spaced. For this reason interference effects around buildingsof many shapes, sizes and separation are now being considered including but not limited to work byKhanduriet al. (1997), Lam & Leung (J.G. Zhao) and Orlando (2001). Kimet al. (2011) found a pos-itive correlation between peak suctions for local peak pressures and the height ratio of the interferingbuilding. Huiet al. (2012) found that the building shape and direction of oncoming wind greatlyinflu-enced local peak pressures and that the vertical edges and cornerswere most affected. In some casesthe minimum pressures on a building face were found to be40% higher with an interfering buildingpresent.

The full paper will examine what happens when a second building is introduced behind the initialbuilding. It is anticipated that the introduction of a second building will interfere with the formation ofthe vortex to the rear of the building changing the wind loading from the one building case. It is alsoanticipated that the vertical wind loading profile of a thunderstorm downburst will produce differentresults than previous interference effects studies which are discussedfurther in section 2.

Two CAARC dimension buildings (Melbourne, 1980) will be used positioned at a distance ofXD

= 1.5 where X is the distance from the jet centre andD is the diameter of the simulator jet, whichis 1m. This is to maintain consistency with the previous work by Haineset al. (2012). The buildingswill be at a scale of1 : 1000 which will correspond to a physical size of30 × 46 × 183 mm. Thesebuildings will be separated by between10mm and200mm, which will correspond to a separationdistance of between10m and200m. The buildings will be rotated about at an axis centered betweenthe two buildings and yaw angles of0◦, 30◦, 60◦ and90◦ will be initially tested. Each building will betapped on each face as well as the roof with and based on the findings of Hui et al. (2012) additionaltaps being placed at the corners of the building so that the effect of flow separation and vortices inthese regions can be well measured as well as the expected increase in local peak pressures.

In addition to the experimental setup an identical set of experiments will be runin a numericalsimulation of the downburst simulator and the results compared. In order to ensure that the flowfeatures are adequately resolved the snappyHexMesh feature of OpenFoam will be used to ensure thatsufficient Y+ values are obtained to resolve a boundary layer on the floor/ building region and aroundthe buildings themselves.

References

ABI. 2005. Financial risks of climate change. [Online; downloaded 19th March 2010].

6th European and African Wind Engineering Conference 5

Chay, M.T., & Letchford, C.W. 2002. Pressure distributions on a cube in asimulated thunderstormdownburst, Part B: moving downburst observations.Journal of wind engineering and industrialAerodynamics, 90, 733–753.

Haines, M., Sterling, M., & Quinn, A. 2012. The simulation of non-synoptic effects for wind damagestudies.In: BBAA7. BBAA7, Jinjiang Tower hotel, Shanghai, China.

Holmes, J.D., Hangan, H.M., Schroeder, J.L., Letchford, C.W., & Orwig, K.D. 2008. A forensic studyof the Lubbock-Reese downdraft of 2002.Wind and structures, 11(2), 137–152.

Hui, Y., Tamura, Y., & Yoshida, A. 2012. Mutual interference effects between two high-rise buildingmodels with different shapes on local peak pressure coefficients.Journal of wind engineering andindustrial dynamics, 104-106, 98–108.

Kasperki, M. 1998. Climate change and design wind load concepts.Wind and structures, 1(2), 145–160.

Khanduri, A.C., Bèdard, C., & Stathopoulos, T. 1997. Modelling wind-induced interference effectsusing backpropagation neural networks.Journal of wind engineering and industrial dynamics, 72,71–79.

Kim, W., Tamura, Y., & Yoshida, A. 2011. Interference effects on local peak pressures between twobuildings.Journal of wind engineering and industrial dynamics, 99, 584–600.

Lam, K.M., & Leung, M.Y. H. J.G. Zhao. Interference effects on wind loadings of a row of closelyspaced tall buildings.Journal of wind engineering and industrial dynamics, 96, 562–583.

Lin, W.E., & Savory, E. 2006. Large-scale quasi steady modelling of a downburst outflow using a slotjet. Wind and structures, 9, 419–440.

Lin, W.E., Orf, L.G., Savory, E., & Novacco, C. 2007. Proposed large-scale modelling of the transientfeatures of a downburst outflow.Wind and structures, 10, 315–346.

Mason, M.S., Wood, G.S., & Fletcher, D.F. 2008. Numerical simulation of downburst winds.Journalof Wind Engineering and Industrial Aerodynamics, 97, 523–529.

McConville, A.C., Sterling, A.C., & Baker, C.J. 2009. The physical simulation of thunderstormdowndrafts using an impinging jet.Wind and structures, 12(2), 133–149.

Melbourne, W.H. 1980. Comparison of measurements on the CAARC standard tall building model insimulated model wind flows.Journal of Wind Engineering and Industrial Aerodynamics, 6, 73–88.

Orlando, M. 2001. Wind-induced interference effects on two adjacentcooling towers.Engineeringstructures, 23, 979–992.