You are on page 1of 11

Journal of Building Engineering 6 (2016) 225235

Contents lists available at ScienceDirect

Journal of Building Engineering


journal homepage: www.elsevier.com/locate/jobe

Aerodynamic Mitigation and Shape Optimization of Buildings: Review


Maryam Asghari Mooneghi a, Ramtin Kargarmoakhar b,I
a
PhD, Advanced Technology and Research, Arup, San Francisco, CA, USA
b
PhD, T.Y. Lin International, San Francisco, CA, USA

art ic l e i nf o a b s t r a c t

Article history: Usually the design of the external shape and orientation of buildings is driven by architectural con-
Received 3 October 2015 siderations, functional requirements and site limitations, rather than by aerodynamic considerations.
Received in revised form This, most of the times, results in structures becoming bluff bodies characterized by high wind-structure
28 January 2016
interaction induced loads. These effects can be signicantly reduced through novel aerodynamic miti-
Accepted 29 January 2016
gation techniques and optimal aerodynamic shape design procedures. This paper reviews the past/recent
Available online 4 February 2016
work on various aerodynamic mitigation techniques developed for reducing wind loads on buildings by
Keywords: modifying their shapes and/or adding simple architectural elements. Aerodynamic mitigation techniques
Buildings applicable to low-rise and high-rise buildings have been reviewed. In addition, aerodynamic shape op-
Wind
timization techniques for reducing wind loads on tall buildings are presented and the suitability and
Aerodynamic Mitigation
challenges of using Computational Fluids Dynamics (CFD) for this application are discussed. An overview
Shape Optimization
CFD of the optimization techniques namely gradient based methods and non-gradient based methods are
presented. It is expected that this research can ignite an interest in using aerodynamic shapes and
consideration of the structures shape, in terms of wind performance, early in the design process. This
paper also serves as a source for various techniques that can be used for reducing wind loads on
buildings.
& 2016 Elsevier Ltd. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225
2. Aerodynamic mitigation techniques for buildings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
2.1. High-rise buildings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
2.1.1. Minor modications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
2.1.2. Major modications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228
2.2. Low-rise buildings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228
2.2.1. Parapets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
2.2.2. Passive aerodynamic devices. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
3. Aerodynamic shape optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230
3.1. An overview of optimization algorithms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230
3.2. Aerodynamic shape optimization of tall buildings using CFD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
4. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
Acknowledgement. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
Appendix Optimization algorithms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234

E-mail address: maryam.asghari@arup.com (M. Asghari Mooneghi).

http://dx.doi.org/10.1016/j.jobe.2016.01.009
2352-7102/& 2016 Elsevier Ltd. All rights reserved.
226 M. Asghari Mooneghi, R. Kargarmoakhar / Journal of Building Engineering 6 (2016) 225235

1. Introduction aerodynamic shape optimization approach is presented and the


suitability and challenges of using computational uid dynamics
Wind loads are one of the most critical parameters for design of (CFD) for aerodynamic shape optimization of tall buildings are
buildings. Buildings typically have sharp corners which can cause discussed. Different optimization algorithms which can be used for
wind ow separation resulting in strong wind-structure interac- aerodynamic shape optimization problems are briey presented. It
tion induced loads (e.g. high roof suctions in case of low-rise is shown that the use of aerodynamic shapes and consideration of
buildings and acrosswind loads and wind-induced vibrations the structures shape, in terms of wind performance, early in the
produced by unfavorable aerodynamic corners in case of high-rise design can result in signicant improvements in the response of
buildings). buildings under wind loads.
One approach for reducing the wind loads on buildings is to use
Aerodynamic Mitigation techniques. These methods effectively
use simple and innovative architectural features to modify the 2. Aerodynamic mitigation techniques for buildings
aerodynamic shape of the buildings in order to reduce the wind
loads. Aerodynamic modications assist either by disrupting the 2.1. High-rise buildings
formation of strong corner vortices; or by breaking the coherent
formation of vortices; or by diverting ows in the separation zone The world is currently undergoing the biggest wave of tall
over the roof edge or away from the weak members. building construction in history. The use of higher-strength ma-
A second strategy to achieve reductions in wind-induced loads terials, lightweight oors, and curtain wall system in the con-
on buildings is to use Aerodynamic Shape Optimization techni- struction of tall buildings has reduced building weight, stiffness,
ques. In optimal shape design problems, the optimization of a and damping values. Therefore, tall buildings are more susceptible
performance criterion depends on the shape of a boundary. The to wind loads and wind-induced excitations which have the po-
creativity and insight of an experienced designer is required to tential to reduce their structural safety and cause discomfort to the
reduce the design problem to a well-posed optimization problem. occupants. Also, these excessive motions can create high base
This involves the denition of objective functions that specify the loads which increase the cost of the structure.
goals of the optimization, design variables that determine the For typical tall buildings, aerodynamic forces are drag force
aerodynamic shape, as well as constraints that dene a feasible (alongwind), lift force (acrosswind) and torsional moment
region of the design space. The optimization algorithm nds the (Fig. 1a).
values of the geometric parameters that optimize the objective The wind-induced response of tall buildings is usually domi-
function while satisfying the constraints. For optimizing the shape nated by dynamic acrosswind loading resulting from wind vortex
of civil structures like long-span bridges and tall slender towers shedding [3] as shown in Fig. 1b. When wind blows over a bluff
different objective functions can be taken into account e.g. redu- structure, ow separates and causes periodic shedding of vortices.
cing the drag force and/or vortex-induced forces. Aerodynamic This periodic vortex shedding exerts acrosswind forces on the
shape optimization allows the designer to explore more alter- body by creating uctuating pressures. Strouhal number is a non-
natives for the design of new aerodynamic shapes compared to the dimensional parameter that denes the dominant frequency of the
traditional methods which are limited to a certain number of pre- uctuations in the acrosswind forces and is expressed as (Eq. (1)):
selected geometries by the designer.
S = fB/U (1)
Shape optimization has progressed signicantly over the years
and has been used in many areas of engineering. Examples include where, f is the frequency of vortex shedding, S is the Strouhal
but are not limited to the design of external surfaces of aircrafts, number, U is the wind speed and B is the building width. Strouhal
especially wings, design of cars and civil structures, such as tall number is a function of the shape of the building with values
buildings and long-span bridges [1]. Several algorithms have been between 0.1 to 0.3, e.g. about 0.14 for a square cross section and
developed that can efciently perform aerodynamic shape opti- 0.2 for a roughly circular cylinder [4]. Vortex-induced vibrations
mization. Dulikravich [2] presented a comprehensive review on (VIV) occurs when the frequency of vortex shedding, f , ap-
the aerodynamic shape optimization methods with applications in proaches one of the natural frequencies of the building. This leads
aircraft airfoil design. to amplied acrosswind response. Vortex-induced vibrations are
The intention of the current paper is to review the past/recent the prime problem in self-excited vibration of tall exible
work on aerodynamic mitigation and aerodynamic shape optimi- buildings.
zation methods developed for reducing wind loads on buildings. In Aerodynamic mitigation techniques which utilize modications
the rst part of this paper different aerodynamic mitigation to the external shape of a building (e.g. corner modications,
techniques which aim at reducing wind loads on high-rise and variation of the cross section shape and size along the height of
low-rise buildings are reviewed. In the second part, an overview of the building, etc.) can signicantly reduce building response under

Fig. 1. (a) Aerodynamic forces on a tall building; (b) Vortex shedding (plan view).
M. Asghari Mooneghi, R. Kargarmoakhar / Journal of Building Engineering 6 (2016) 225235 227

wind loads by altering the wind ow pattern around the building methods that can be utilized to improve the wind performance of
and can lead to a more economic and comfortable design [46]. tall buildings (Fig. 2). These methods can result in reductions in
Shape effects from a wind engineering perspective have been in- both alongwind and acrosswind responses compared to plain
vestigated by Davenport [7] through aerodynamic model tests. rectangular shape buildings. The modication of windward cor-
Many researches were performed to study the relationship be- ners can change the characteristics of the separated shear layers
tween the aerodynamic characteristics of a structure and the re- leading to narrower downstream wake and thus can be very ef-
sulting wind-induced excitation level [817]. Aerodynamic mod- fective in reducing the drag and uctuating lift forces.
ication techniques aim particularly at suppression of vortex Fig. 3 shows a graphical representation of the ow eld around
shedding and can generally be classied into two groups [18]: six different cross section alternatives obtained using CFD simu-
lations. The analysis was performed using ANSYS Fluent Software.
2.1.1. Minor modications All the models were run transiently with 20 m/s uniform inlet
These modications have negligible effects on the overall velocity using k- SST turbulence model. The k- SST model was
structural and architectural design of the building. Common selected because of its proved ability for modeling detached ows
building shapes are square or rectangle which cause the building with periodic vortex shedding [19]. The outlet was considered to
to experience relatively strong vortex-induced forces. These ex- be a pressure outlet. Top, bottom, and the two sides of the com-
putational domain were assigned as symmetric plane. All the
citation forces can be reduced through minor modications to the
building faces were assigned to be No-slip walls. It can be seen
aerodynamic shape of the tower. For instance, modications in the
form Fig. 3 that corner modications (in this case modications
corners of the cross sectional shape of the building such as slotted
were made to a building with square plan shape) can change the
corners, chamfered corners, corner recession, roundness of corners
ow pattern around the building. The effectiveness of each of
and changing the orientation of the building relative to the most
these strategies does depend on the type and size of the corner
frequent strong wind direction are among the minor modication
modications relative to the size of the plan. For example, it can be
seen that the plan shapes with Recessed Corners or Double
Chamfered caused smaller separation zones and narrower wakes
as compared to a simple square shape. These techniques can be
effective in reducing the alongwind (drag) as well as acrosswind
(lift) forces as compared to a sharped edged square cross section.
Many researchers studied the effect of corner modications
and their impact on the aerodynamic forces on tall buildings in-
cluding chamfered, recessed and slotted corners [3,4,13,2022].
While many of these studies demonstrated the benets of corner
geometry modications, there are some cases showing that
modications to building corners can be ineffective and even
sometimes result in adverse effects [14,23]. As a matter of fact,
these methods should be utilized with care to avoid any un-
favorable effects. It should also be noted that the effectiveness of
corner modications depends on the oncoming wind direction
[20] and the size of the modications relative to the size of the
Fig. 2. Minor aerodynamic modications. plan [3,11]. Irwin et al [24] suggested that the corner modications

Fig. 3. Flow eld around different cross sections.


228 M. Asghari Mooneghi, R. Kargarmoakhar / Journal of Building Engineering 6 (2016) 225235

should extend about 10% of the building width to be effective. becomes weakened and disrupted by the ow of air through the
Taipei 101 skyscraper in Taiwan and Mitsubishi Heavy In- structure. This aerodynamic modication method was in-
dustries Yokohama Building in Japan are examples of tall buildings vestigated by several researchers [12,14,29,30]. These studies
that successfully utilized the corner modication techniques to showed that openings in the upper half of the buildings
reduce the alongwind and acrosswind responses. Corner mod- (especially openings in the alongwind direction) can be very
ications in Taipei 101 skyscraper resulted in 25% reductions in the effective for reducing the acrosswind response of tall buildings.
base moment [24]. Creating rounded corners is another effective However, the efcacy of this approach does depend on the
approach for improving the aerodynamic behavior of tall buildings location and size of the openings. Dutton and Isyumov [12] also
showed that the effectiveness of these gaps might be inuenced
against the wind excitations [5,16,25].
by the level of turbulence in the approaching ow.
Twisting or rotating the buildings: Rotating the building can be
2.1.2. Major modications
very effective because the designer can assign the orientation of
These modications have signicant effects on the structural
the building such that its least favorable aspect does not coin-
and architectural design of the building. For instance, varying the
cide with the strongest wind direction. Also, twisted building
shape of building and setbacks along the height, tapering, inclu-
forms are effective in reducing vortex-induced vibrations by
sion of openings at top and twisting the building are among the
avoiding simultaneous vortex shedding along the building
major modication methods that can be utilized to come up with height. The availability of studies investigating the effects of
an aerodynamically favorable building shape (Fig. 4). Some of twisting the building on the wind loads is limited [31]. The
these approaches are discussed in the following. twisted form can be found in todays tall building designs such
as Turning Torso, in Malmo, Sweden and Shanghai Centre, in
Tapering and setbacks: Through tapering or setbacks, the width Shanghai, China.
of a building can be varied along its height causing the vortices
to become incoherent and shed at different frequencies at dif- 2.2. Low-rise buildings
ferent heights. This is mainly due to the fact that the frequency
of vortex shedding depends on the width of the building (Eq. 1). Wind-related disasters are among the most costly natural ha-
This results in dramatic reduction of the associated uctuating zards to occur in the US each year in addition to loss of many lives
forces. Kim et al [26], Kim and You [17], and You et al [27] [32]. Low-rise buildings such as residential houses, commercial
investigated the effects of tapering for reducing the wind- and industrial structures, constitute more than 70 percent of the
induced response of tall buildings. Tapering has a more sig- building stuck in the United States and account for the majority of
nicant effect in acrosswind direction than that in the along- losses due to wind storms [33].
wind direction [17,26]. This method has been in part used in Roof systems are exposed to higher loading than any other
order to optimize the shape of Burj Khalifa, Dubai. The building building element. The worst wind suctions on roofs normally oc-
shape is extremely efcient from a wind loading point of view so cur for cornering or oblique wind directions. These extreme suc-
that the tower does not need any supplementary damping tions are the result of conical vortices which form along the roof
system [28]. The Millennium Tower in Tokyo, Japan and Transa- edges (Fig. 5) and are the main reason for most of the wind-in-
merica pyramid in San Francisco also exploit the use of tapering duced damages. Shingles, tiles, or pavers placed on roofs are most
effects along their heights for mitigating the wind forces. vulnerable to being dislodged and becoming wind-borne debris
Varying cross-section shape: Varying the cross-section shape with [34]. Loose roong components could lead to rain water intrusion
height, e.g. going from square to round has a similar effect. In and losses to interior appliances and building contents [35]. The
this case, Strouhal number varies with height (Eq. 1) which need to reduce roof damages due to wind effects has recently
causes the vortices to shed over a broad range of frequencies. become one of the most important challenges for designers,
This method can be especially effective in reducing the across- manufacturers, and building code ofcials [34,3639].
wind forces [24]. Simple modications to the shape of the roof edge and/or
Porosity or openings: The addition of openings to a building is utilizing roof accessory structures can change the ow pattern on
another method to improve the aerodynamic behavior of a the roof and can lead to reductions in the wind loads and conse-
structure. By allowing air to bleed through the building via quently damage to low-rise buildings. These methods are appro-
openings or porous sections, the formation of the vortices priately entitled vortex suppression techniques since they
mainly aim at disrupting and deecting the conical vortices from
the roof which are the main causes of extreme roof suctions. These
methods are mainly in form of using parapets with various shapes
and congurations or passive aerodynamic edges that suppress
vortex generation on the roof. Vortex suppression techniques can
be classied in the following four categories based on their aero-
dynamic mechanism [40]:

1. Methods aiming at eliminating straight sharp edges which


create the vortices
2. Methods aiming at disrupting the vortex formation (like partial
or porous parapets)
3. Methods aiming at disturbing the vortices (like porous fence or
screen, rooftop cylinders and splitters)
4. Methods aiming at displacing the formed vortices (like high
parapets)

Fig. 6 shows a schematic of some types of aerodynamic miti-


Fig. 4. Major aerodynamic modications. gation techniques that were proposed in the literature to alleviate
M. Asghari Mooneghi, R. Kargarmoakhar / Journal of Building Engineering 6 (2016) 225235 229

the corner suction pressures on roofs of low-rise buildings. Please permanent architectural elements or can be attached to the roof
note that the drawings in Fig. 6 are not in their actual scale for during preparations for high-wind events.
better presentation of each approach. The effectiveness of roof edge conguration on mitigating wind
loadings was investigated by Blackmore [51] and Savory et al [52]
2.2.1. Parapets using wind tunnel testing or eld measurements performed on
Parapets have been used as a standard architectural feature for aerodynamic curved and chamfered roof eave edges. Although
many years mainly for buildings with at roofs. Many wind tunnel these methods were shown to be effective, they are not practical
testing and eld studies have been performed on the effectiveness sometimes e.g. for buildings with eaves.
of parapets for reducing wind loads on building roofs. Basically, A variety of roof edge devices have been developed that can be
the parapets lift the separated shear layers clear of the roof surface attached to roof eaves (e.g. semicircular gutters, cantilevered spoi-
and thus dissipate the high local corner or edge suctions over a lers, etc.). These devices were shown to facilitate the reduction of
larger area. However, this can result in increased loads on interior extreme wind pressures on the roof with varying degrees of aero-
regions, thus inuencing the overall effectiveness of parapets [44]. dynamic effectiveness and architectural practicality [5356]. For
One of the rst studies that specically considered the role of instance, Huang et al [57] investigated the effectiveness of eight
these building components in vortex suppression was carried out typical softs that are commonly used in the residential houses in
by Baskaran and Stathopoulos [41]. Their results showed that high China for alleviating rooftop extreme wind pressures using wind
parapets generally reduce the high suctions on roof corners while tunnel testing. Results showed that the presence of these elements
low parapets may increase the roof suctions at the edges and signicantly reduced the negative peak wind pressures at edges and
corners. A similar effect was observed by Asghari Mooneghi et al corners of the roof while not changing the wind loads on the other
[49] by performing large-scale testing on a at roof building with roof regions. Most of these studies have been based on small-scale
solid parapets of different heights in Wall of Wind open jet facility model testing (scales between 1:50 and 1:200) in wind tunnels.
at Florida International University. Baskaran and Stathopoulos [41] However, large-scale and full-scale testing is more favorable for as-
also showed that generally the perimetric parapets have a more sessing the performance of aerodynamic devices [58] because they
signicant effect on reducing corner pressure coefcients than a allow for accurate modeling of architectural details, avoiding adverse
parapet present on only one side of the roof. The effectiveness of scale effect by testing at higher Reynolds numbers (which is more
parapets is also a function of the parapet conguration. Surry and pronounced for curved surfaces that might be used in aerodynamic
Lin [40] investigated the effectiveness of various parapet cong- devices [59] and using greater spatial resolution of the pressure taps
urations, including saw-tooth partial parapets, semi-cylindrical in critical regions on the roof (e.g. corners and edges) to capture the
parapets, solid and porous roof corner splitters and isolated porous high localized suctions [49]. For this reason large- and full-scale tests
parapets on a scaled model of Texas Tech University (TTU) research have been performed recently for developing roof suction mitigation
building. The isolated porous parapets were shown to be the most techniques. Blessing et al [60] exploited roof gravel scour testing and
effective alternative in reducing suctions near the roof corner. pressure testing to assess the effectiveness of several aerodynamic
Kopp et al [43] studied the effects of various parapets in de- devices in reducing high uplift pressures at corners and edges of at
creasing area-averaged loads under the corner vortices. The spoi- roofs through full-scale testing. Huang et al [47] carried out eld
lers and porous continuous parapets performed best with 44% and measurements to examine the efcacy of three types of aero-
56% maximum reductions in peak pressure coefcients near roof dynamic mitigation plates, including full-length roof-edge plate,
corners. Wu [44] investigated a specic parapet called a Conical roof-corner plate, and discrete roof-edge plates with different spaces
Vortex Disrupter through full-scale tests. Results showed sig- on pitched roofs. Results showed that these plates could signicantly
nicant reductions on both local and area-averaged wind loads reduce the mean and uctuating pressure coefcients in the wind-
near the roof corners on a at roof building. Other studies on the ward corner. Bitsuamlak et al [48] investigated the effectiveness of
effect of parapet conguration on mitigating peak suctions on the simple architectural elements including gable end and ridgeline
roof of low-rise buildings include those of Pindado and Meseguer extensions, trellis, and wall extensions for the reduction of roof and
[42] and Suaris and Irwin [50]. wall corner suctions using 6-fan Wall of Wind facility at Florida in-
ternational University. The wind loading mechanism of aerodynamic
2.2.2. Passive aerodynamic devices mitigation devices were also investigated using CFD. Aly [61] used
Beyond investigating the efcacy of parapets in reducing high CFD simulations to study the efcacy of aerodynamic roof mitigation
wind suctions on the roofs, recent studies have focused on mod- devices in reducing wind suctions on the roof of low-rise buildings.
ifying the roof corners or edges and/or adding various aero- Mitigation devices including barriers, circular edges, inclined edges
dynamic edge elements to mitigate the extreme negative pres- and airfoil edges were investigated. The objective was to propose an
sures caused by conical vortices. These elements can be used as aerodynamic mitigation technique which could not only reduce the

Fig. 5. (a) Conical vortices on a at roof at cornering winds; (b) Suction variation under conical vortices.
230 M. Asghari Mooneghi, R. Kargarmoakhar / Journal of Building Engineering 6 (2016) 225235

Solid parapet [41] Porous parapet [42] Discontinuous parapet [43]

Partial parapet [25] Perimetric spoiler [44] Aerodynamic edge [45]

Discontinuous perforated parapets [46] Full length roof-edge plate [47] Trellis (pergola) [48]
Fig. 6. Various aerodynamic mitigation techniques for reducing wind loads on roof of low-rise buildings [25,4148].

loads on the roofs but also had minimum drag and lift forces exerted In principle, they attempt to mimic nature in order to nd the op-
on it. Aerodynamic mitigation devices with relatively high lift and timum of the objective function. One of the key features of these
drag forces can become wind born debris impacting other structures algorithms is that they search from multiple points in the design
downwind or they may introduce excessive loads to the main space, instead of moving from a single point like gradient based
structure. It was shown that slope-in aerodynamic mitigation de- methods. Thus, although in general there is no proof that these
vices which could be replaced by solar panels [62] were relatively methods converge to global optima, experience entails that they
effective in reducing roof induced suctions. converge to global optima in most cases (Fig. 7b).
The main drawback of non-gradient based methods is that
these algorithms are generally computationally slower than the
3. Aerodynamic shape optimization gradient based ones. Indeed, the non-gradient based algorithms
can require thousands of function evaluations and, in some cases,
3.1. An overview of optimization algorithms become non-practical. In order to overcome these difculties, the
so-called Hybrid algorithms, which take advantage of the robust-
The search for an optimal state is one of the most fundamental ness of the non-gradient based methods and the fast convergence
principles in our world. Optimization is to nd the best solution to a of the gradient based methods, have been proposed by different
certain designated problem. Numerical methods for optimizing the scholars. A set of analytically formulated rules and switching cri-
performance of engineering problems have been studied for many teria can be coded into the program to automatically switch back
years. For optimization of an objective, different categories of opti- and forth among the different algorithms as the iterative process
mization techniques namely Gradient based methods and Non- advances. Each technique provides a unique approach with vary-
gradient based methods can be used. The basic idea behind the ing degrees of convergence, reliability and robustness at different
gradient based methods is that a given function reaches its extremes stages during the iterative optimization process.
(minimum or maximum) in the direction of its gradient. Gradient It is also interesting to introduce to the reader the multi-ob-
based methods are in general computationally faster (they require jective optimization problems vs. single-objective optimization
fewer objective function evaluations in case of problems with low problems. In a large number of problems, there exists a need to
number of design variables) than non-gradient based methods. Their nd optimal solutions due to more than one objective. In this case,
main drawback is that they might converge to local minima, and a multi-objective approach must be employed. Engineering pro-
their convergence to global minima depends mainly on the chosen blems demanding low cost, high performance and low losses are
starting point (initial guess) by the user (Fig. 7a). On the other hand, an example of applications where this approach is needed. For the
non-gradient based methods, do not rely, most of the time, on single-objective optimization problems, a unique optimal solution
strong mathematical basis and make use of neither the gradient or exists. However, for multi-objective optimization problems, there
the second derivative of the objective function as a direction of exist a set of compromised solutions, known as the Pareto-optimal
descent. These methods work based on function evaluations alone. solutions or non-dominated solutions, which are based on the
M. Asghari Mooneghi, R. Kargarmoakhar / Journal of Building Engineering 6 (2016) 225235 231

competing objectives. The goal of multi-objective optimization is design, the ultimate goal of computational simulation methods
to nd such set of solutions. Once the solutions are obtained, the should not just be the analysis of prescribed shapes, but the auto-
designer can choose a nal design with further considerations. matic determination of the true optimum shape for the intended
Non-gradient based methods are more suitable for multi-objective application. This is the underlying motivation for the combination of
optimization problems since they are able to nd the entire range computational uid dynamics with numerical optimization methods
of Pareto-optimal solutions. Gradient based methods typically use for aerodynamic shape optimization problems. Some of the earliest
a weighting method to combine different objectives into one for studies of such an approach were made by Hicks et al [65] and Hicks
handling multi-objective optimization problems. In this case, re- and Henne [66] for aircraft wing design. Aerodynamic shape opti-
sults would be dependent on the chosen weighting coefcients. mization using CFD has been used for many years in aerospace
An overview of different gradient based and non-gradient based [67,68] and automotive [69,70] industries and has recently become
optimization algorithms are provided in Appendix A. Clearly, the subject of increasing interest in civil structures especially for
knowledge about the nature of the problem is a requirement for aerodynamic design of the shape of tall buildings [71].
choosing the most suitable optimization tool for an application. A general approach for aerodynamic shape optimization using
CFD is shown in Fig. 8. The essential components of an aero-
3.2. Aerodynamic shape optimization of tall buildings using CFD dynamic shape optimization problem are objective functions,
constraints, design variables that dene the possible geometries, a
The goal of aerodynamic shape optimization is to accurately ow solver, and a numerical optimization method. Examples of
and efciently determine surface shapes that attain optimal objective functions for aerodynamic shape optimization of tall
aerodynamic performance [1]. While the effects of geometric buildings can be reducing the drag force and/or vortex-induced
modications to the shape of tall buildings, such as utilizing re- vibrations. The optimization algorithm nds the values of the
cessed or chamfered corners, etc. (discussed in Section 2.1) can geometric parameters that optimize the objective function while
signicantly improve the aerodynamic response of tall buildings, a satisfying the constraints. These components should be selected
systematic approach for taking full advantage of aerodynamic carefully as they have a direct impact on the accuracy and ef-
shape optimization for buildings is not fully explored yet. ciency of solution [1]. Samareh [72] provides summaries of shape
Experimental method using wind tunnel testing, provide the basis parameterization techniques that can be used to dene design
of the traditional cut and try approach for the design of new aero- variables. In general, it is important that the selected para-
dynamic shapes. In this approach, several congurations are in- meterization technique provides sufcient exibility in order to
vestigated in a wind tunnel and the one that yields the best aero- realize truly optimal designs. It is also desirable that the number of
dynamic performance is identied. The reason behind using wind parameters necessary to dene the shape to be small so that a
tunnel testing for designing the new aerodynamic shapes is that the reasonable convergence rate of the optimization can be obtained.
relation between the external shape of a building and the resulting As discussed in Section 3.1, various numerical optimization meth-
intensity of the wind excitations is complicated and the improvements odologies can be used for aerodynamic shape optimization applica-
in wind effects that can be obtained by specic geometric modica- tions. Hicks et al [65] were the rst to apply gradient based methods
tions is difcult to predict without performing experiments [63]. to aerodynamic shape optimization problems. They used the method
Merrick and Bitsuamlak [64] examined the effects of building shape on of feasible directions, which is based on conjugate gradients, to opti-
wind loading patterns for high-rise buildings by analyzing various mize airfoil shapes in transonic ow governed by the small-dis-
buildings with foot prints of square, circular, triangular, rectangular and turbance equation. Since this pioneering work, the application of
elliptical shapes using wind tunnel database. Results of this research gradient based methods to aerodynamic shape optimization problems
outlined the general wind loading characteristics of simple building remained an active area of research. However, as the aerodynamic
shapes. The sensitivity of each building shape to vortex-shedding was shape optimization problem is a complex one with possibly many local
determined. While it is possible to use wind tunnel testing for aero- optima, non-gradient based methods (also called Evolutionary Algo-
dynamic shape optimization, this approach is highly demanding due to rithms (EA)) such as Genetic Algorithm (GA) are more suitable for this
the time and cost limitations for performing each test. As a matter of application to ensure reaching the global optimum. Evolutionary Al-
fact, only a limited number of possible congurations which are chosen gorithms have the advantages such as robustness, suitability to parallel
based on engineering experience and judgement can be examined. computing and simplicity in coupling CFD codes. Owing to these ad-
The use of computational simulation to scan many alternative vantages over the non-gradient based methods, EAs have become
designs has proved extremely valuable in practice. With the ad- increasingly popular in a broad class of design problems [73]. How-
vances in computational uid dynamics and computing power of ever, the implementation of aerodynamic shape optimization using
modern computers, CFD has contributed to cut aerodynamic design CFD is intrinsically difcult for bluff civil structures due mainly to the
cost and time scales by reducing the number of required wind turbulent ow eld in which the structures are immersed, the high
tunnel tests. Currently, CFD is mainly used for estimating aero- Reynolds number values, and the multi-objective nature of the design
dynamic performance of a given structure conguration and it still problem [74]. So, while there are clear advantageous for using non-
suffers the limitation that it does not guarantee the identication of gradient based optimization theories for aerodynamic shape optimi-
the best possible design. To ensure the realization of the true best zation applications, the use of these algorithms for shape optimization

Fig. 7. Optimization methods: (a) Gradient based (b) Non-gradient based.


232 M. Asghari Mooneghi, R. Kargarmoakhar / Journal of Building Engineering 6 (2016) 225235

of civil structures is often impractical since the objective functions and optimum might be missed due to the uncertainty at the predicted
constraints are evaluated using CFD and an extremely high number of point which may mislead the optimization process in a wrong
function calls which can easily reach the order of thousands, is re- way. As a matter of fact, the selected DoE strategy is of paramount
quired for obtaining the optimum solution. importance for achieving a satisfactory accuracy of the surrogate
One way of tackling the efciency issue of evolutionary search model. Generally, sampling plans that more evenly ll the design
methods is to use a Surrogate-Based Optimization (SBO) metho- space can reduce the bias error of the surrogate. A plenty of sur-
dology in CPU-intensive aerodynamic shape optimization appli- rogate methods, search algorithms and updating algorithms have
cations. The main idea in SBO is to utilize data sampling and been proposed in the literature for a variety of applications in-
surface tting strategies to parameterize the space of possible cluding CFD-based aerodynamic shape optimization [67,74,76,78].
solutions via a simple, computationally inexpensive model to be It should also be noted that the repeated evaluation of the ob-
used for the purpose of numerical optimization. So, the whole jective functions required during the optimization demands fast ow
optimization process is managed by the surrogate model outputs. solutions [1]. As a matter of fact, the ow solver has a signicant
This is often referred to as optimizing the response surface of the inuence on the efciency of the optimization. Buildings are im-
system. The basic process for a surrogate-based optimization mersed in very turbulent ows due to ground roughness effect
consists of the following steps [75]: which after interacting with buildings represent very wide temporal
and spatial scales governed by highly non-linear Navier-Stokes dif-
1. Dening a sampling plan for the design space: This contains ferential equations which are computationally expensive to solve
both the samples required for constructing the surrogate model [79]. Detailed reviews of effective ow solvers for the Navier-Stokes
and some additional samples needed for verifying the surrogate equations are given by Pueyo [80]. Reynolds-Averaged Navier-Stokes
model. (RANS) models can be used for obtaining fast and reasonably reliable
2. Performing numerical simulations at design points selected simulations even for complex shapes and high Reynolds numbers,
from the sampling plan. and can be used for obtaining both steady and unsteady aerodynamic
3. Constructing the surrogate model: Different surrogate models measures (URANS). On the other hand, Large Eddie Simulation (LES)
can be used such as polynomial regression, Kriging, radial basis methods and/or hybrid RANS/LES methods are more accurate than
functions, neural networks, and support vector regression [76]. RANS models for the prediction of the unsteady forces at the cost of
4. Validating the model: This is to nd out the predictive cap- higher computational burden. Note that the Navier-Stokes equations
abilities of the surrogate model. should be solved repeatedly for each updated geometry as the op-
5. Updating the model: Repeating the last four steps until the timization progresses. This issue makes the problem further com-
desired model validation accuracy is obtained. plicated. Therefore, an efcient approach can be to use URANS
6. Optimization: The search for the optimum is carried out on the models for narrowing down the solution domain and then use the
surrogate model. higher accuracy LES models when a greater level of detail is needed
for example when comparing two nal competing shapes [63].
The approximation efciency and its accuracy are major issues One challenge in the application of numerical optimization to a
in SBO. If the problem has a high number of design variables, the building aerodynamic shape design problem is regeneration of the
construction of surrogate model may cause extremely high com- mesh for geometry variations from the initial base geometry. So, one
putational cost, which makes the approximation inefcient. De- very important feature of the CFD-based shape optimization tech-
sign of Experiment (DoE) can be used to reduce the number of nique should be to have the possibility of updating the CFD meshes
design points [77]. However, by using this approach the global automatically. Without this ability, the CFD model may need to be
reprocessed before a suitable mesh with the required quality can be
obtained. Also, the boundary conditions may need to be reapplied
before the analysis can be launched. Mesh morphing techniques al-
low performing mesh modications without changing the mesh to-
pology and avoid a costly recalculation of the physics equations with
every new CFD model. In these methods, the mesh is rst imported
into to the mesh morpher tool. The mesh is parameterized and then
based on design constraints and realistic assumptions, deformation
vectors are added to the model. After each optimization iteration the
mesh morpher modies the mesh with respect to the original to-
pology, thus allowing a more robust and faster design optimization
environment using CFD. Quality controls for meshing can be set to
guarantee the quality of the CFD simulation results. This ensures that
the algorithm is independent from external inputs during the opti-
mization and makes mesh generation process faster. Mesh morphers
have been used in many applications of shape optimization using
CFD [8183] mainly for aerospace and automotive structures. Wei et
al [84] and Wei et al [85] developed a mesh morphing algorithm
based on a Laplacian smoothing approach which was very robust
even for marked geometric modications which are expected in
aerodynamic shape optimization of tall buildings.
Civil engineering structures which often have square or rectan-
gular shapes provide a wide variety of applications for bluff body
aerodynamics. The optimization of a trapezoidal bluff body was
conducted by Burman et al [86]. This study focused on the response
surface optimization of Navier-Stokes ow over a bluff body. The
optimization objectives were to minimize the mean drag coefcient
Fig. 8. Aerodynamic shape optimization utilizing CFD. and to maximize the measure of mixing dened as the time average
M. Asghari Mooneghi, R. Kargarmoakhar / Journal of Building Engineering 6 (2016) 225235 233

maximum negative velocity. Response surface methodology was Various aerodynamic modications to the shape of tall buildings
utilized in the optimization process. Mack et al [87] used multiple for mitigating the wind induced loads were reviewed. The
surrogates for the shape optimization of a 2D trapezoidal bluff body aerodynamic modications of a buildings cross-sectional shape
model. Polynomial response surface and radial basis neural net- (e.g. corner cut, corner recession, slotted corner, etc.), variation
works were used as surrogates. Objective functions were minimizing of the cross section shape and/or its size along the height of
the total pressure loss coefcient and maximizing the mixing cap- building, twisting the building, porosity and openings, etc. can
ability of time dependent ows over a 2-D bluff body. signicantly reduce building response in alongwind (drag force)
Very recently, attempts have been made to use aerodynamic as well as acrosswind (acrosswind force due to vortex shedding)
shape optimization in the eld of tall building design. Bobby et al directions by altering the wind ow characteristics around the
[71] proposed a framework for the aerodynamic shape optimiza- building.
tion of tall buildings using CFD based on the concept of low-di- Aerodynamic mitigation techniques for low-rise buildings were
mensional models for describing the global aerodynamic perfor- discussed. Various methods for the mitigation of wind high
mance of tall buildings. A low-dimensional model was introduced suctions on the roof of low-rise buildings were described.
which allows an efcient way for reducing the computational Parapets and aerodynamic edge devices are simple, economic,
demand of the CFD simulations, thus expediting the optimization. and non-intrusive elements and can be utilized for reducing
A mesh morphing tool was used for automatically updating the wind loads and mitigating damage risk at the source. In addi-
mesh. The objective function was to minimize the drag force. Two- tion, they can be used as a comparatively cost-effective solution
dimensional CFD simulation was used for each optimization to the roof edge damage problem which is widely experienced
iteration using Reynolds-Averaged Navier-Stokes (RANS) model. during high wind events. These elements can either be used for
The low-dimensional model allowed extension of the results to retrotting of existing buildings or they can be incorporated into
the overall aerodynamic performance of tall buildings. Elshaer et the design of new buildings for reducing peak suctions at critical
al [88] proposed an optimization framework for aerodynamic locations of the building envelope.
shape optimization of tall buildings. The method couples the op- An overview of optimization techniques including gradient
timization algorithm using genetic algorithm (GA), computational based and non-gradient based was briey presented. Ad-
uid dynamics (CFD) solver, and the neural networks (NN) model. vantages and disadvantageous of each category were discussed.
The objective was to reduce the drag force acting on a tall building Non-gradient based optimization methods can converge to
by changing the shape of its corners. Large eddy simulation (LES) global optimum at the expense of more computational cost
models were used for numerical simulation of the wind behavior. compared to gradient based algorithms. Although these meth-
ods seem to be more attractive for aerodynamic shape optimi-
Bernardini et al [74] performed multi-objective aerodynamic
zation problems which have possibly a lot of local optimum
shape optimization of tall buildings using a surrogate-based op-
points, they require hundreds or thousands of analysis code
timization method. The multiple objectives were minimizing the
implementation. This is extremely costly, and often impossible,
mean drag coefcient as well as the standard deviation of the lift
for CFD-based aerodynamic shape optimization of civil struc-
coefcient. Ordinary Kriging surrogate model was used. A speci-
tures in which each evaluation of the objective/constrained
cally developed strategy was adopted to update the Kriging
functions requires a full CFD analysis to be carried out. Surro-
models in order to perform additional CFD runs efciently. Shell
gate-based optimization can be used to tackle the efciency is-
scripting, parallelized computations and mesh morphing algo-
sue of non-gradient based optimization methods.
rithms were used to improve the efciency and consistency of the
The suitability and challenges of aerodynamic shape optimiza-
framework.
tion using CFD for reducing wind loads on tall buildings were
These researches showed that in spite of recent advancements,
discussed and the recent researches on this topic were pre-
there are still many challenges for the CFD-based aerodynamic
sented. Although published information on how to design
shape optimization of tall buildings. These limited studies mainly
structures for optimal aerodynamic performance is limited, full
focused on optimizing the dimensions of some predened local
utilization of aerodynamic optimal design could enable properly
corner modications in tall buildings. To the authors knowledge,
designed buildings to withstand extreme wind loads cost-ef-
there is no reported work till date on global shape optimization of
fectively. Therefore, the aerodynamics of a tower shape needs to
tall buildings against wind load effects. Although this eld of re-
be considered as a critical design parameter from the very out-
search is fairly new and is constantly improving, it is believed that set. In this way, designers could optimally control building
full utilization of aerodynamic shape optimization could enable geometric parameters to favorably impact the wind loads and
design of tall buildings that withstand wind load effects more other adverse wind-induced responses.
efciently. With the rapidly increasing use of the distributed par-
allel computing and adequate computing hardware, aerodynamic
shape optimization is becoming more affordable than before. Acknowledgement

We would like to thank Florida International University (FIU)


4. Conclusions for funding this research. The helpful comments received from
Professor Girma Bitsuamlak from Western University are greatly
The shape and orientation of most buildings are driven mainly by appreciated. Authors would like to acknowledge Dimple Rana and
architectural considerations, functional requirements and site lim-
Alexej Goehring of Arup for their inputs in the ow simulation
itations, rather than by aerodynamic considerations. Consequently,
using CFD.
these structures are characterized by high wind-structure interac-
tion induced loads. Signicant reduction in wind loads can be
achieved by various types of aerodynamic modications to the
shape of the structure and/or aerodynamic shape optimization Appendix Optimization algorithms
techniques. In this paper the past/recent work on various techniques
developed by many researchers to reduce wind loads on buildings Fig. A1 shows an overview of different gradient based and non-
were reviewed which are summarized in the following: gradient based optimization algorithms.
234 M. Asghari Mooneghi, R. Kargarmoakhar / Journal of Building Engineering 6 (2016) 225235

Fig. A1. Figure A: Summary of optimization methods.

References Wind Engineering and Industrial Aerodynamics. 97 (2009) 455467.


[21] Kumar KS, Irwin P, Davies A. Design of tall building for wind: wind tunnel vs.
codes/standards. Third National Conference on Wind Engineering. Calcutta,
[1] M. Nemec, Optimal Shape Design of Aerodynamic Congurations: A Newton- India, 2006, 318-325.
Krylov Approach, University of Toronto, Canada, 2003. [22] T. Tamura, T. Miyagi, The effect of turbulence on aerodynamic forces on a
[2] G. Dulikravich, Aerodynamic shape design and optimization: Status and square cylinder with various corner shapes, Journal of Wind Engineering and
trends, Journal of Aircraft. 29 (1992) 10201026. Industrial Aerodynamics. 83 (1999) 135145.
[3] M. Gu, Y. Quan, Across-wind loads of typical tall buildings, Journal of Wind [23] K.C.S. Kwok, N. Isyumov, Aerodynamic measures to reduce the wind-induced
Engineering and Industrial Aerodynamics. 92 (2004) 11471165. response of buildings and structures, Structural Engineers World Congress,
[4] P.A. Irwin, Bluff body aerodynamics in wind engineering, Journal of Wind San Francisco, CA, 1998.
Engineering and Industrial Aerodynamics. 96 (2008) 701712. [24] P. Irwin, J. Kilpatrick, A. Frisque, Friend or Foe, Wind at Height., CTBUH 8th
[5] A. Kareem, T. Kijewski, Y. Tamura, Mitigation of motions of tall buildings with World Congress, Dubai. Dubai, 2008.
specic examples of recent applications, Wind and Structures. 2 (1999) [25] Kwok KCS. Aerodynamics of tall buildings. Ninth International Conference on
201251. Wind Engineering. New Delhi, 1995, 180-205.
[6] P. Irwin, B. Breukelman, C. Williams, Hunter M. Shaping and Orienting Tall [26] Y.-M. Kim, K.-P. You, N.-H. Ko, Across-wind responses of an aeroelastic tapered
Buildings for Wind., Structural Engineers World Congress, San Francisco, 1998. tall building, Journal of Wind Engineering and Industrial Aerodynamics. 96
[7] A. Davenport, The Response of Six Building Shapes to Turbulent Wind, Phi- (2008) 13071319.
losophical Transactions of the Royal Society of London 269 (1971) 385394. [27] K.-P. You, Y.-M. Kim, N.-H. Ko, The evaluation of wind-induced vibration re-
[8] K. Kwok, P. Bailey, Aerodynamic Devices for Tall Buildings and Structures, sponses to a tapered tall building, The Structural Design of Tall and Special
Journal of Engineering Mechanics 113 (1987) 349365. Buildings. 17 (2007) 655667.
[9] K.C.S. Kwok, Effect of building shape on wind-induced response of tall [28] P.A. Irwin, Wind engineering challenges of the new generation of super-tall
building, Journal of Wind Engineering and Industrial Aerodynamics. 28 (1988) buildings, Journal of Wind Engineering and Industrial Aerodynamics. 97
381390. (2009) 328334.
[10] Melbourne NH, Cheung JCK. Designing for serviceable accelerations in tall [29] H. Okada, L. Kong, The Effects of Open Passage on Reducing Wind Response of
buildings. 4th International Conference on Tall Buildings. Hong Kong and Tall Building., 29th Technical Report, Public Works Research institute. Japan
Hanghai, 1988, 148-155. (1999) 561566.
[11] Melbourne NH. Wind loading and environment of tall building. International [30] N. Isyumov, A.A. Fediw, J. Colaco, P.V. Banavalkar, Performance of a tall
Conference on Tall Buildings and City Development. Brisbane, 1989, 159-164. building under wind action, Journal of Wind Engineering and Industrial
[12] R. Dutton, N. Isyumov, Journal of Wind Engineering and Industrial Aero- Aerodynamics. 42 (1992) 10531064.
dynamics. 36 (1990) 739747. [31] D. Kelly, D. Poon, P. Irwin, J. Xie, Wind Engineering of the Shanghai Center
[13] H. Hayashida, Y. Iwasa, Aerodynamic shape effects of tall building for vortex Tower, Advances in Hurricane Engineering: American Society of Civil En-
induced vibration, Journal of Wind Engineering and Industrial Aerodynamics. gineers (2012) 426436.
33 (1990) 237242. [32] E. Simiu, R.H. Scanlan, Wind effects on structures, Third edition, John Wiley &
[14] K. Miyashita, J. Katagiri, O. Nakamura, T. Ohkuma, Y. Tamura, M. Itoh, et al., Sons, Inc, New York, N.Y., 1996.
Wind-induced response of high-rise buildings Effects of corner cuts or [33] J.D. Holmes, Wind Loading of Structures, Taylor & Francis, London and New
openings in square buildings, Journal of Wind Engineering and Industrial York, 2007.
Aerodynamics. 50 (1993) 319328. [34] Asghari Mooneghi M. Experimental and Analytical Methodologies for Pre-
[15] A. Kareem, Y. Tamura, Mitigation of wind-induced motions of tall buildings. dicting Peak Loads on Building Envelopes and Roong Systems. FIU Electronic
Tall Building Structures: A World View. Chicago, Illinois Council on Tall Theses and Dissertations. Paper 1846. http://digitalcommons.u.edu/etd/
Buildings and Urban Habitat, Lehigh University, 1996. 18462014.
[16] H. Kawai, Effect of corner modications on aeroelastic instabilities of tall [35] T. Baheru, A.G. Chowdhury, J.-P. Pinelli, G. Bitsuamlak, Distribution of wind-
buildings, Journal of Wind Engineering and Industrial Aerodynamics. 7476 driven rain deposition on low-rise buildings: Direct impinging raindrops
(1998) 719729. versus surface runoff, Journal of Wind Engineering and Industrial Aero-
[17] Y.-M. Kim, K.-P. You, Dynamic responses of a tapered tall building to wind dynamics. 133 (2014) 2738.
loads, Journal of Wind Engineering and Industrial Aerodynamics. 90 (2002) [36] M. Asghari Mooneghi, P. Irwin, A. Chowdhury, Design Guidelines for Roof
17711782. Pavers against Wind Uplift, Structures Congress 2015: American Society of
[18] J.A. Amin, A.K. Ahuja, Aerodynamic modications to the shape of the shape of Civil Engineers (2015) 26792688.
the buildings: A review of the state-of-the-art, Asian Journal of Civil En- [37] B. Mintz, A. Chowdhury, A. Mirmiran, N. Suksawang, Kargarmoakhar R. De-
gineering (Building and Housing) 11 (2010) 433450. sign, Development, and Testing of a Composite Roong System, Journal of
[19] P. Catalano, M. Amato, An evaluation of RANS turbulence modelling for Composites for Construction. (2015) 04015052.
aerodynamic applications, Aerospace Science and Technology. 7 (2003) [38] F. Habte, M. Asghari Mooneghi, A. Gan Chowdhury, P. Irwin, Full-scale testing
493509. to evaluate the performance of standing seam metal roofs under simulated
[20] K.T. Tse, P.A. Hitchcock, K.C.S. Kwok, S. Thepmongkorn, C.M. Chan, Economic wind loading, Engineering Structures. 105 (2015) 231248.
perspectives of aerodynamic treatments of square tall buildings, Journal of [39] M. Asghari Mooneghi, P. Irwin, A. Gan Chowdhury, Towards guidelines for
M. Asghari Mooneghi, R. Kargarmoakhar / Journal of Building Engineering 6 (2016) 225235 235

design of loose-laid roof pavers for wind uplift, Wind and Structures 22 (2) [64] R. Merrick, G.T. Bitsuamlak, Shape effects on the wind-induced response of
(2016), http://dx.doi.org/10.12989/was.2016.22.2.000. high-rise buildings, Journal of Wind Engineering 6 (2009) 118.
[40] D. Surry, J.X. Lin, The effect of surroundings and roof corner geometric mod- [65] R.M. Hicks, E.M. Murman, G.N. Vanderplaats, An assessment of airfoil design
ications on roof pressures on low-rise buildings, Journal of Wind Engineering by numerical optimization., NASA Ames Research Center; Moffett Field, CA,
and Industrial Aerodynamics. 58 (1995) 113138. United States, 1974.
[41] A. Baskaran, T. Stathopoulos, Roof corner wind loads and parapet congura- [66] R.M. Hicks, P.A. Henne, Wing design by numerical optimization, Journal of
tions, Journal of Wind Engineering and Industrial Aerodynamics. 29 (1988) Aircraft. 15 (1978) 407412.
7988. [67] J.-E. Kim, V.N. Rao, R.P. Koomullil, D.H. Ross, B.K. Soni, A.M. Shih, Development
[42] S. Pindado, J. Meseguer, Wind tunnel study on the inuence of different of an efcient aerodynamic shape optimization framework, Mathematics and
parapets on the roof pressure distribution of low-rise buildings, Journal of Computers in Simulation. 79 (2009) 23732384.
Wind Engineering and Industrial Aerodynamics. 91 (2003) 11331139. [68] D.N. Srinath, S. Mittal, An adjoint method for shape optimization in unsteady
[43] G.A. Kopp, C. Mans, D. Surry, Wind effects of parapets on low buildings: Part 4. viscous ows, Journal of Computational Physics. 229 (2010) 19942008.
Mitigation of corner loads with alternative geometries, Journal of Wind En- [69] L. Dumas, CFD-based optimization in Automotive Aerodynamics, Optimization
gineering and Industrial Aerodynamics. 93 (2005) 873888. and Computational Fluid Dynamics: Springer (2008), pp. 191215.
[44] F. Wu, Full-scale study of conical vortices and their effects near corners, Texas [70] F. Muyl, L. Dumas, V. Herbert, Hybrid method for aerodynamic shape opti-
Tech University, . Lubbock, Texas, 2000. mization in automotive industry, Computers & Fluids. 33 (2004) 849858.
[45] Lin JX, Montpellier PR, Tillman CW, Riker WI. Aerodynamic devices for miti- [71] Bobby S, Spence SMJ, Bernardini E, Wei D, Kareem A. A complete perfor-
gation of wind damage risk The 4th International Conference on Advances in mance-based optimization framework for the design of tall buildings. 11th
Wind and Structures (AWAS'08). Jeju, Korea, 2008. International Conference on Structural Safety & Reliability. New York, NY, USA,
[46] Fu T-C. Development of Effective Approaches to the Large-Scale Aerodynamic 2013.
Testing of Low-Rise Building. FIU Electronic Theses and Dissertations. Paper [72] J.A. Samareh, Survey of shape parametrization techniques for high-delity
986. http://digitalcommons.u.edu/etd/9862013. multidisciplinary shape optimization, AIAA Journal. 39 (2001) 877883.
[47] Huang P, Jia C, Gu M. Field measurement for aerodynamic mitigation of wind [73] A. Oyama, Wing design using evolutionary algorithms., Tohoku University,
pressure on gable-roofed low-rise building The Eighth Asia-Pacic Conference Japan, 2000.
on Wind Engineering. Chennai, India, 2013. [74] E. Bernardini, S.M.J. Spence, D. Wei, A. Kareem, Aerodynamic shape optimi-
[48] G.T. Bitsuamlak, W. Warsido, E. Ledesma, A.G. Chowdhury, Aerodynamic mi- zation of civil structures: A CFD-enabled Kriging-based approach, Journal of
tigation of roof and wall corner suctions using simple architectural elements, Wind Engineering and Industrial Aerodynamics. 144 (2015) 154164.
Journal of Engineering Mechanics-ASCE. 139 (2013) 396408. [75] Y.V. Pehlivanoglu, B. Yagiz, Aerodynamic design prediction using surrogate-
[49] M. Asghari Mooneghi, P. Irwin, A. Gan Chowdhury, Large-scale testing on wind based modeling in genetic algorithm architecture, Aerospace Science and
uplift of roof pavers, Journal of Wind Engineering and Industrial Aero-
Technology. 23 (2012) 479491.
dynamics. 128 (2014) 2236. [76] Ahmed M, Qin N. Surrogate-based aerodynamic design optimization: use of
[50] W. Suaris, P. Irwin, Effect of roof-edge parapets on mitigating extreme roof
surrogates in aerodynamic design optimization. 13th International Conference
suctions, Journal of Wind Engineering and Industrial Aerodynamics. 98 (2010)
on Aerospace Science and Aviation Technology. Cairo, Egypt, 2009.
483491.
[77] R.H. Myers, D.C. Montgomery, C.M. Anderson-Cook, Response Surface Meth-
[51] P.A. Blackmore, Load reduction on at roofs - the effect of edge prole, Journal
odology: Process and Product Optimization Using Designed Experiments, 3rd
of Wind Engineering and Industrial Aerodynamics. 29 (1988) 8998.
ed, John Wiley & Sons, Inc, 2009.
[52] Savory E, Dalley S, Toy N. International Conference on Wind EngineeringThe
[78] E. Iuliano, D. Quagliarella, Proper Orthogonal Decomposition, surrogate
effects of eaves geometry, model scale and approach ow conditions on portal
modelling and evolutionary optimization in aerodynamic design, Computers
frame building wind loads. Journal of Wind Engineering and Industrial
& Fluids. 84 (2013) 327350.
Aerodynamics. 1992, 43: 1665-1676.
[79] A. Dagnew, G.T. Bitsuamlak, Computational evaluation of wind loads on
[53] Lin JX, Surry D. Suppressing extreme suction on low buildings by modifying
buildings: A review, Wind and Structures. 16 (2013) 629660.
the roof corner geometry. 7th US National Conference on Wind Engineering.
[80] A. Pueyo, An efcient NewtonKrylov mothod for the Euler and NavierStokes
University of California, USA, 1993.
equations, University of Toronto, Toronto, Canada, 1998.
[54] L.S. Cochran, E.C. English, Reduction of Roof Wind Loads by Architectural
[81] M.E. Biancolini, I.M. Viola, M. Riotte, Sails trim optimisation using CFD and RBF
Features, Journal of Architectural Science Review 40 (1997) 7987.
mesh morphing, Computers & Fluids. 93 (2014) 4660.
[55] Banks D, Sarkar P, Wu F. A device to mitigate vortex induced rooftop suction.
1st Americas Conference on Wind Engineering. Clemson University, Clemson, [82] R.G. Silva, A.P. Antunes, R.B. Flatschart, J.L.F. Azevedo, CFD based optimization
USA, 2001. of high-lift devices using a mesh morphing technique., The 29th Congress of
[56] M.K. Philips, Evaluation of Mitigation Measures for Reducing High Suction the International Council of the Aeronautical Sciences, St Petersburg, Russia,
Pressures on Roof Corners, Clemson University, Clemson, USA, 2003. 2014.
[57] P. Huang, X. Peng, M. Gu, Aerodynamic devices to mitigate rooftop suctions on [83] M.L. Staten, S.J. Owen, S.M. Shontz, A.G. Salinger, T.S. Coffey, A comparison of
a gable roof building, Journal of Wind Engineering and Industrial Aero- mesh morphing methods for 3D shape optimization, 20th International
dynamics. 135 (2014) 90104. Meshing Roundtable, Paris, France (2011), pp. 293311.
[58] Asghari Mooneghi M, Irwin P, Gan Chowdhury A. Partial Turbulence Simula- [84] Wei D, Spence SMJ, Kareem A. ynamic mesh for deformable or movable
tion Method for Small Structures. 14th International Conference on Wind boundaries in uid-structure interaction. 11th ASCE Joint Specialty Conference
Engineering. Porto Alegre, Brazil, 2015. on Probabilistic Mechanics and Structural Reliability. Notre Dame, IN, USA,
[59] R. Kargarmoakhar, A. Gan Chowdhury, P. Irwin, Reynolds number effects on 2012.
twin box girder long span bridge aerodynamics, Wind and Structures. 20 [85] D. Wei, S.M.J. Spence, A. Kareem, A. Jemcov, A structured mesh boundary
(2015) 327347. motion approach for simulating wind effects on bluff bodies with changing
[60] C. Blessing, A.G. Chowdhury, J. Lin, P. Huang, Full-scale validation of vortex boundaries, Journal of Wind Engineering and Industrial Aerodynamics. 126
suppression techniques for mitigation of roof uplift, Engineering Structures. (2014) 118131.
31 (2009) 29362946. [86] Burman J, Papila N, Shyy W, Gebart BR. Assessment of response surface-based
[61] A.M. Aly, Aerodynamic Mitigation of Wind-Induced Suction on Low-Rise optimization techniques for Unsteady ow around bluff bodies. th AIAA/ISS-
Buildings: A Comparative Study., Advances in Wind and Structures (AWAS'14), MO Symposium on Multidisciplinary Analysis and Optimization. Atlanta,
Busan, South Korea, 2014. Georgia, 2002.
[62] Chokwitthaya C, Aly AM. Retrotting Building Roofs with Aerodynamic Fea- [87] Mack Y, Goel T, Shyy W, Haftka R. Multiple surrogates for the shape optimi-
tures and Solar Panels to Reduce Hurricane Damage and Enhance Eco-Friendly zation of bluff body-facilitated mixing. 43rd AIAA Aerospace Sciences Meeting
Energy Production 14th International Conference on Wind Engineering and Exhibit. Reno, Nevada, 2005.
(ICWE). Porto Alegre, Brazil, 2015. [88] Elshaer A, Bitsuamlak G, Damatty AE. Aerodynamic shape optimization for
[63] A. Kareem, S. Spence, E. Bernardini, S. Bobby, D. Wei, Using computational corners of tall buildings using CFD. 14th International Conference on Wind
uid dynamics to optimize tall building design, CTBH journal (2013) 3843. Engineering (ICWE). Porto Alegre, Brazil, 2015.

You might also like