PETER IRWIN-wind and Highrise Buildings - Negatives and Positives

5/22/2018 PETER IRWIN-wind and Highrise Buildings - Negatives and Positives

1/9

1

Wind and Tall Buildings Negatives and Positives

Peter Irwin+, John Kilpatrick*, Jamieson Robinson+

and Andrea Frisque#

+ Rowan Williams Davies & Irwin Inc., 650 Woodlawn Road West, Guelph, Ontario, Canada N1K 1B8* RWDI Anemos Limited, Unit 4, Lawrence Way Industrial Estate, Dunstable, Bedfordshire, UK LU6 1BD

# RWDI AIR Inc., 830 999 West Broadway, Vancouver, British Columbia, Canada V5Z 1K5

Abstract

Wind is often regarded as the foe of tall buildings since it tends to be the governing lateral load. Carefulaerodynamic design of tall buildings through wind tunnel testing can greatly reduce wind loads and theiraffect on building motions. Various shaping strategies are discussed, aimed particularly at suppression ofvortex shedding since it is frequently the cause of crosswind excitation. The use of supplementary dampingsystems is another approach that takes the energy out of building motions and reduces loads. Differentapplications of damping systems are described on several buildings, and an example of material savingsand reduced carbon emissions is given. Wind also has some potential beneficial effects particular to tallbuildings. One is that, since wind speeds are higher at the heights of tall buildings, the potential forextracting wind energy using wind turbines is significantly improved compared with ground level. Thepaper explores how much energy might be generated in this way relative to the building's energy usage.

Other benefits are to be found in judicious use of natural ventilation, sometimes involving double layer wallsystems, and, in hot climates, the combination of tailored wind and shade conditions to improve outdoorcomfort near tall buildings and on balconies and terraces.

Keywords: Wind loads, tall buildings, building motions, wind tunnel testing, sustainability, human comfort

IntroductionThe economics of constructing tall buildings

are greatly affected by wind as their heightincreases. To counteract wind loads and keepbuilding motions within comfortable limits canrequire robust structural systems that drive up

costs. Both the loads and motions are oftensubject to dynamic amplification in both thealong wind and crosswind directions. Theseeffects are heavily dependent on shape. Hencethe current trend towards considering theaerodynamics of the shape very early in thedesign of the very tall towers. The curtain wallloads also tend to increase with height primarilydue to the fact that wind speeds in generalincrease with height, and the winds at groundlevel and on terraces or balconies are increased.All these effects are familiar to experienceddevelopers and designers of tall towers and can

be categorized as potential problems to be solvedthrough the use of wind tunnel testing. Wind isthe foe.

However, there are also some effects ofwind that can make it a friend. One example isthe possible use of the building as the platformfor wind turbines for generating energy. Withthe available wind power increasing as windvelocity cubed, the higher winds at the upper

levels of tall towers present a potentialopportunity to access greater wind energy than isavailable at ground level. There is also inprinciple an opportunity to use the amplifiedwinds around tall towers to naturally ventilatethe building. In hot climates the increased winds

around the towers base can, in conjunction withshading, provide opportunities to improve thethermal comfort around the building.

This paper will examine situations wherewind is the foe and measures that can be taken tofrustrate the foe. It will also look at situationswhere it is a friend and look at how far one cango in enlisting the help of this friend.

Wind as the Foe

Influence of Shape and How to Confuse the

FoeOne of the critical phenomena that effect tall

slender towers is vortex excitation, which causesstrong fluctuation forces in the crosswinddirection. This is probably the main behaviourthat distinguishes tall towers from mid-risebuildings. The well-known expression ofStrouhal gives the frequency Nat which vorticesare shed from the side of the building, causingoscillatory across-wind forces at this frequency.


2/9

2

b

USN = (1)

whereS= Strouhal number, U= wind speed, b=building width.

The Strouhal number is a constant with avalue typically in the range 0.1 to 0.3. For asquare cross-section it is around 0.14 and for arough circular cylinder it is about 0.20. WhenN

matches one of the natural frequencies rNof the

building, resonance occurs which results inamplified crosswind response, as illustrated inFigure 1. From Equation 1 this will happenwhen the wind speed is given by

S

bNU r= (2)

Figure 1. Effect of Vortex Shedding on Response

The peak in the response in Figure 1 can bemoved to the right on this plot if the buildingnatural frequency is increased and if it can bemoved far enough to the right the wind speedwhere the peak occurs will be high enough that itis not a concern. This is the traditional approachof adding stiffness but this approach can becomeextremely expensive if the peak has to be moved

a long way to the right. However, the height ofthe peak is sensitive to the building shape and,with astute aerodynamic shaping, the peak canbe substantially reduced or even eliminated.There are several directions that one can go indeveloping an aerodynamically favourableshape, Figure 2.

Figure 2.Shaping Strategies

Softened Corners: - Square or rectangularshapes are very common for buildings and

experience relatively strong vortex sheddingforces. However, it is found that if thecorners can be softened throughchamfering, rounding or stepping theminwards, the excitation forces can besubstantially reduced. The softening shouldextend about 10% of the building width infrom the corner. The corners on Taipei 101where stepped in order to reduce crosswindrespond and drag, resulting in a 25%reduction in base moment, Irwin (2005).

Tapering and Setbacks: -As indicated in

Equation 1, at a given wind speed, thevortex shedding frequency varies dependingon the Strouhal number Sand width b. Ifthe width b can be varied up the height ofthe building, through tapering or setbacks,then the vortices will try to shed at differentfrequencies at different heights. Theybecome confused and incoherent, whichcan dramatically reduce the associatedfluctuating forces. Burj Dubai, Figure 3, is aclassic example of this strategy.

Varying Cross-Section Shape- A similar

effect can be achieved by varying the cross-section shape with height, e.g. going fromsquare to round. In this case the StrouhalnumberSvaries with height, which again, inaccordance with Equation 1 causes theshedding frequency to be different atdifferent heights. This again results inconfused vortices.

Wind velocity

Crosswind

Response

Vortex shedding

No vortex shedding

Softened corners

Tapering and setbacks

Varyingcross-sectionshape

Spoilers

Porosity or openings


3/9

3

Spoilers: - One can also reduce vortexshedding by adding spoilers to the outside ofthe building. The most well known form ofspoilers is the spiral Scruton strake used oncircular chimneystacks. Architecturally and

practically, the Scruton strake leavessomething to be desired for circularbuildings, but other types of spoiler could beused that might be more acceptable, such asvertical fins at intervals up the height.

Porosity or Openings: -Another approachis to allow air to bleed through the buildingvia openings or porous sections. Theformation of the vortices becomes weakenedand disrupted by the flow of air through thestructure.

Figure 3. Burj Dubai spiralling set backs(photo courtesy Skidmore Owings & Merrill)

While vortex shedding is the principalculprit causing undesirably high crosswindmotions, another cause is buffeting by turbulence

cast off from upstream buildings. This is lesseasy to deal with through the building shapesince the origin of the turbulence is not thebuilding itself. However, some cross-sectionalshapes, e.g. a lens shape, are more prone toacross-wind buffeting because their streamlinedshape causes them to act somewhat like avertical aerofoil, generating high crosswind forcevariations for relatively small changes in angle ofattack of the wind. Shape changes that make

them less like an aerofoil can help in thissituation. The best tool for assessing shapeeffects is the wind tunnel. Irwin et al (1998)describe wind tunnel experiments with differentshapes. However, though hundreds of tallbuildings are tested annually in commercial windtunnel laboratories around the world, there isonly limited feedback received about the actualperformance of these tall buildings.

Recent comparisons of wind tunnelpredictions with the long-term measured wind-induced accelerations of tall buildings inChicago (Kilpatrick 2007) reaffirm the validityand effectiveness of wind tunnel model methodsto predict the response of tall buildings to wind.A comparison of the predicted accelerations of a344m tower with measured data acquired duringa winter 2003 wind storm is given in Figure 4.On average, the measured data compare

favourably with the wind tunnel predictions.Linear regression curves fitted to themeasurements indicate that total damping ratioincreases with wind speed, likely including anamplitude-dependant structural dampingcomponent, in addition to a positiveaerodynamic damping component in the along-wind direction. Some data suggested negativeaerodynamic damping in the across-winddirection, re-inforcing the need for aeroelasticmodel tests for super-tall buildings.

Figure 4.Comparison of RMS Accelerations

0.01

0.1

1

10 100

Estimated Gradient Wind Speed (mph)

NormalizedN-SRMSAcceleration

Measured Linear Regression

Likely accelerations

estimated from wind

tunnel data


4/9

4

Inherent and Supplementary Damping

A tall tower under the action of wind tendsto act as an energy storage device. Thefluctuating wind forces cause the tower to moveand the motions gradually build up, alternatelyexchanging kinetic energy with stored elasticenergy as the building sways to and fro. Thelimit on these motions is set by the ability of thetower to dissipate energy faster than the rate atwhich the wind feeds energy in. A measure ofthe towers energy dissipation ability is itsdamping ratio: the higher the damping ratio, thebetter. For the purpose of determining buildingaccelerations the inherent damping ratio in tallbuildings is usually assumed to be in the range of0.01 to about 0.02. However, there isconsiderable uncertainty as to its exact value andit appears possible that it might be significantlylower than 0.01 in some buildings. The achieved

total damping is a combination of twocomponents: inherent damping and designedsupplementary damping.

The approach of adding supplementarydamping to the structure therefore has twobenefits: a) it clearly can reduce motions byincreasing the towers ability to dissipate energy;and b) it greatly reduces the uncertainty as towhat the building damping actually will be.

The importance of the damping parameter towind loading and serviceability issues is often

overlooked. In many cases, a change of 0.5% inthe inherent damping assumption can have agreater effect on loading and comfort than can beachieved by all reasonable structural measures(see Figure 5).

Figure 5. Generally-observed relative effects on buildingmotions of changing mass, stiffness, and damping.

Once the idea of adding supplementarydamping has been adopted, typically the exactvalue of the inherent damping becomes muchless critical since it only contributes part of theoverall damping (often a minor part).

The opportunity here is to start the designprocess with a known, achievable, high level oftotal damping. If the design relies on anuncertain amount of inherent damping alone, andthe assumption is inaccurate, then it is possiblethat the structural materials will not be efficientlyutilized.

When incorporating even a simplesupplementary damping system, it is typical tobe assured of a damping value of 3% of criticalor more. With this as a reliable baseline, anoptimal structural design can quickly beachieved with minimal iteration, since the mass,

stiffness, and damping can all be carefullycontrolled and optimized together. The potentialfor cost savings, enhancing sustainability, andcreating more useable floor space throughreduced structure are substantial (Gamble andRobinson, 2007).

Supplementary Damping Systems

There is a wide range of possibilities,generally categorized as Distributed DampingSystems and Mass Damping Systems.Distributed damping needs to be located in areas

of greatest relative movement between structuralelements. Therefore, these locations dependhighly upon the structural system. The mosteffective location for a Mass Damping System(Tuned Sloshing Damper, Tuned Liquid ColumnDamper, Tuned Mass Damper, or Active MassDamper) is close to the area of peak amplitude ofvibration, which is typically near the top of thebuilding.

The optimum choice for any given projectwill depend on a variety of factors: the amount ofdamping needed, space available, the number of

modes of vibration involved, accessibility formaintenance etc.

The Tuned Mass Damper has been quitepopular. The principle is illustrated in Figure 6,it being that a small mass m is attached to themain mass of the tower M via a spring andviscous damping system, the natural frequencyof the small mass-spring system being tuned toclose to that of the main system. The Taipei 101TMD, Figure 7, is a prominent example of this


5/9

5

type, the spring in this case being based on thegravity stiffness provide by the pendulumprinciple. The mass m is a 660 tonne steel ballsuspended high in the tower. The Taipei 101application is unusual in that the TMD has beenused as a special feature and attraction of thetower, being visible to visitors as illustrated inFigure 8. As the building period becomes longerthe simple pendulum type TMD does requireplenty of vertical space. Where vertical space isnot available, more complex pendulum devicescan be used such as the compact compoundpendulum TMD used in the Bloomberg tower inNew York, Figure 9. AMD systems can providemore damping for the same mass, being drivenby active control systems, and can handle morethan one mode of vibration. However, they aremore complex and, although the mass may besmaller, some of the saved space is used up bythe larger stroke requirements.

Figure 6. TMD Principle

Figure 7. Schematic of Taipei 101 Simple Pendulum TMD

Figure 8. Taipei 101 TMD as an Attraction

Figure 9. Compact Compound Pendulum TMDBloomberg Tower, New York

Tuned Sloshing Dampers and Tuned LiquidColumn Dampers are economical alternatives toTMD systems. In place of a solid mass, there isa mass of liquid (usually water) and the

dimensions of the tank are designed so that thewave action or liquid column motion havenatural frequencies close to those of the building.Baffles are added to dissipate the energy of theliquid motion. Figures 10 and 11 illustrateapplications of these systems. Although they aremore economical to build, typically thesesystems use more space than a TMD or AMD,and this must be factored into the final choice ofsystem.

M

m

kd,c d

k,c


6/9

6

Figure 10. Tuned Sloshing Damper

Figure 11. Tuned Liquid Column DamperRandom House, New York

Distributed Damping (using viscous orviscoelastic dampers) can be effective but thereare a number of practical issues to solve inintegrating them into the structural system.These types of systems typically take thefollowing forms: (a) Viscoelastic materials todissipate energy that are placed within multiplestructural connections, (b) cross-bracing thatincorporates fluid or viscoelastic elements todissipate energy in multiple frame locations, or(c) outrigger, column, or shear-wall systems thatincorporate hydraulic or viscoelastic elements todissipate energy. In these systems specialattention needs to be given to ensuring that theviscous or viscoelastic elements perform

satisfactorily at the very small operationalamplitudes experienced under wind loading.Also, if the local forces induced by the dampersbecome too high, the structure tends to deformaround them making them less effective.

Example of CO2 Emission Reductions

Achieved by incorporating a Damping System

The Residences at College Park Phase III,located in downtown Toronto, Canada is aproposed tall tower being built by CanderelStoneridge. In the course of the wind studies byRWDI a range of heights for the tower from 75to 80+ floors was considered. The floors fromthe fifth story up are planned to containresidential condominium units. More details ofthe studies on this tower can be found inRobinson et al (2008).

The lateral system had been initiallydeveloped for the 75-storey tower and theincrease in height could only be achieved withthis system if the damping of the tower could beincreased. Therefore studies of supplementarydamping systems were initiated. However, once

the idea of a damping system had beenintroduced, and as planning for the projectdeveloped further, another interesting scenariothat was explored was to keep the tower at 75stories and to exploit the benefits of the dampingsystem by removing structure.

For this latter scenario, structural engineersHalcrow Yolles examined reductions in re-inforced concrete wall thicknesses and in steelthat could be achieved through use of thedamping system. A total damping ratio of 3.0%was targeted with 2% assumed to come from the

inherent structural damping and the extra 1%from a Tuned Sloshing Damping (TSD) systemmade up of water tanks. The TSD system wasidentified as the most cost effective systembearing in mind the space required and buildingprogram. The net material savings were assessed

at 1,400 m3

of concrete, 88,000 kg of mild

reinforcing steel and 9,300 kg of post-tensionsteel strand.

The damping system was optimized to the50-year return period event, with a goal for thewave behavior to remain predictable during large

amplitude motions. To provide a damping ratioof approximately 3.0% for the three fundamentalmodes of vibration (X-sway, Y-sway andtorsion), two dual-axis TSD tanks wereconfigured. The water in these tanks could alsobe utilized for the fire-suppression system, thusremoving the necessity of installing a specialtank dedicated to fire-suppression.


7/9

7

A cost/benefit analysis was performedwhich indicated that significant savings instructural costs (between $400,000 and$500,000) could be expected. These costs actedas an offset to the expense of designing andconstructing the TSD system. Theenvironmental benefits of reducing the need forconcrete and steel were also assessed. Theytranslated into reductions in greenhouse gasemissions (CO2) of about 670 tonnes, which isequivalent to removing about 143,000 cars fromthe road for one typical day in North America(Robinson et al, 2008).

Wind as the Friend

Wind Turbines

The simplest place conceptually to put awind turbine on a tall building is right on top,high enough above the roof to be free of thedisturbed airflow in the separation zone close tothe roof. Provided there are no other tall towersof similar height nearby this gives the wind anuninterrupted path to the turbine for all winddirections. It is interesting to examine howmuch power one can obtain from such a turbine,assuming it can turn to face any wind direction,in comparison to the typical power consumptionof the building. As a reference extreme case wewill examine the situation where the turbinediameter is the same as the maximum width ofthe building.

The building selected is one tested by RWDIin its wind tunnels. It is a 41-floor, 137 m high,rectangular tower with side dimensions of 52 mby 23 m. It is assumed to be in a climate similarto that of Rochester, New York, where theaverage wind speed at 10 m height in openterrain is 4.5 m/s.

Assume the turbine center is 30 m above theroof with a diameter equal to the larger buildingdimension of 52 m. This is a massive turbine butwe are looking to gain an impression of what ittakes to obtain significant wind power.

Assuming the building is located in a suburbanenvironment the average wind speed Vat the

turbine height is estimated to be =V 5.79 m/s.The available power per square meter ofapproaching wind may be calculated from

TVP 3

21r= (3)

where =r air density (about 1.2 kg/m3), 3V =

average of the cube of wind velocity, and T=numbers of hours per year. Assuming the windspeed distribution is approximately Rayleigh in

form, then 3V may be estimated using

33 91.1 VV . Also 8766T hours.Therefore the available power in the wind atturbine height is calculated to be

=P 1,945 kWh/(yr.m2) (4)

To calculate the energy extracted by theturbine the available wind power P is multipliedby the turbine area and the turbine efficiency.Even an ideal turbine cannot exceed about 59%efficiency (the so-called Betz efficiency) becausemuch of the airflow tends to deflect around theturbine due to the turbine drag. Real turbines

typically do not achieve overall efficienciesabove about 40%. Assuming 40% efficiency the52 m diameter turbine will therefore extract thefollowing energyEfrom the wind.

65.11000/4.021231945 ==EMWh/yr (5)

Assuming the building uses approximately300 kWh/(yr.m2) and that the used floor area isabout 45,000 m2, its energy consumption will be13.5 MWh/yr. Comparing this figure with theenergy from the wind turbine it is estimated that

the turbine can produce about 12% of thebuildings energy requirements. However, this iswith a very large 52 m diameter turbine, whichcould pose many design problems to mount ontop of a building. Smaller turbines will clearlyproduce much less power. For example a moremodest 10 m diameter turbine would produceless than % of the buildings requirements inthe wind zone we have selected. Clearly windpower needs to be considered as just onecomponent in a series of measures to reduce atowers dependence on external sources ofenergy, since the incremental improvement it

provides is unlikely to come close to meeting thetotal demand.

Schemes that involve locating the turbineslower in the building than on the roof have beenused in some designs. There are additionalcomplexities to be examined for these schemes,primarily caused by the disturbed airflowscreated by the surrounding structure of the tower.In one of the designs considered for the Freedom


8/9

8

Tower in New York, the approach was taken ofhaving a very open cable structure in the upperportions to reduce these disturbances. Suchstructures at altitude can attract significant iceaccumulations under certain meteorologicalconditions (in cloud icing), and the potentialhazards of falling ice need to be dealt with.Another example of turbines located lower in thetower is the Pearl River Tower, designed bySkidmore Owings Merrill, Figures 12a and 12b.In this case the wind speeds at the turbinelocations were increased through aerodynamicshaping of the entrance channels. The shaping ofthe channels also increased the ranges of winddirections where the turbines were able tooperate. RWDIs experience on several projectsis that, because of the wind speed cubedrelationship for wind power, it is very importantto evaluate the aerodynamic interference of thewind caused by surrounding structures since they

can have a large impact on available windenergy.

Figure 12a. Pearl River Tower(photo courtesy of Skidmore Owings & Merrill)

Figure 12b. Wind Tunnel Model of Pearl River Tower

Natural Ventilation and Connection to the

Outdoors

Sustainable high-rise buildings that providecomfortable, healthy and efficient work-lifeenvironments are clearly desirable as thedensification of living space for the worldsincreasing population proceeds. One componentof this is the provision of fresh air and aconnection to the outdoors, which can beachieved through natural ventilation, usingoperable windows, double facades, ventilationstacks, balconies, patios, terraces and gardens.There can be multiple benefits to these types ofbuilding components. Large amounts of fresh aircan be provided to the spaces without the needfor power-consuming ventilation equipment.Occupants can control and increase the fresh airprovision as needed. Depending on the climate,cooling loads may be removed without the need

for mechanical conditioning, or at leastventilation loads can often be reduced. Buildingelements such as balconies can also help toreduce solar gains through shading. High-riseresidential units that provide access to theoutdoors combine the feel and the advantages ofliving in smaller, sub-urban communities withthe advantages of central, high-rise living.

Natural ventilation concepts as well asoutdoor spaces at higher elevations need toaddress the potentially high winds at higherelevations. Unlike wind power potential, which

typically falls short of the buildings needs, thewind can provide ventilation far in excess of thebuildings needs. This can be controlled byusing smaller openings or operable windowsthan on lower structures. As well, notuncommon in moderate-height residentialtowers, balconies, patios and similar structures,frequently experience winds that are too strongfrom a comfort perspective. Though there issome research into providing improved natural


9/9

9

ventilation to residential towers, for example seePriyadarsini et al. (2004), most of the efforts tonaturally ventilate tall buildings has beenfocused on office buildings. Over the last years anumber of naturally ventilated towers have beenbuilt and evaluated (see Pasquay 2004) using theconcept of double facades. Double shells protectoperable windows at higher elevations from highwind speeds and reduce acoustical problemsfrom operable windows at the same time. Oneearly example of such a building is theCommerzbank building in Frankfurt. Thebuilding features a double faade, with operablewindows in the interior shell. To control stackeffect, the building is subdivided intoindependent segments which also include 4-storey atria with gardens. The building and windtunnel model are shown in Figures 13 and 14.

Figure 13. Commerzbank Building in Frankfurt, Germany

Figure 14. Wind Tunnel Model used for Natural VentilationStudies of Commerzbank

For future buildings, careful windengineering can be used to go beyond doublefacades and create semi-enclosed spaces at highelevations, providing building occupants withsheltered outdoor spaces. Lattices, screens,glazing and wall elements could be used tocreate desirable, wind-protected and shadedpatios and terraces in a high-rise environment.

References

Gamble, S.L., and Robinson, J.K., (2007),Supplementary Damping of Buildings A

Comfort Level, Published in Structural EngineerMagazine, December 2007 Issue.Irwin, P.A., Breukelman, B., Williams, C.J., andHunter, M.A. (1998), Shaping and OrientingTall Buildings for Wind, ASCE StructuresCongress, San Francisco.

Irwin, P.A. (2005). Developing WindEngineering Techniques to Optimize Designand Reduce Risk, 2005 Scruton Lecture,published by UK Wind Engineering Society,Institute of Civil Engineers, One Great GeorgeSt, London, UK.Kilpatrick, J. (2007). Validation of model-scalepredictions of tall building performance usingfull-scale measurements, Ph.D. thesis, Facultyof Graduate Studies, University of WesternOntario.Pasquay, T., Natural ventilation in high-risebuildings with double facades, saving or wasteof energy, Energy and Buildings, Volume 36,Issue 4, April 2004, Pages 381-389, Proceedingsof the International Conference on Solar Energyin Buildings CISBAT 2001Priyadarsini, R., Cheong, K. W., and Wong, N.H. Enhancement of natural ventilation in high-rise residential buildings using stack system,Energy and Buildings, Volume 36, Issue 1,January 2004, Pages 61-71Robinson, J., Wesolowsky, M., and Fitzgerald,G., The Supplementary Damping Advantage Flexibility for a High-Rise Building, Proceedingsof the 17th Congress of IABSE, Chicago, Sept.17-19, 2008

Documents

PETER IRWIN-wind and Highrise Buildings - Negatives and Positives