Engineering Structures 112 (2016) 23–46

Contents lists available at ScienceDirect

Engineering Structures
journal homepage: www.elsevier.com/locate/engstruct

Review article

Review on dynamic and quasi-static buffeting response of transmission

lines under synoptic and non-synoptic winds
Haitham Aboshosha a, Amal Elawady b, Ayman El Ansary b, Ashraf El Damatty b,⇑
a
Boundary Layer Wind Tunnel Laboratory (BLWTL), Western University, London, ON, Canada
b
Civil and Environmental Engineering Department, Faculty of Engineering, Western University, London, ON, Canada

a r t i c l e i n f o a b s t r a c t

Article history: This study reviews the literature on the dynamic response of a Transmission Line (TL) system under syn-
Received 22 July 2015 optic wind (conventional atmospheric boundary layer) as well as non-synoptic wind loading (down-
Revised 23 December 2015 bursts). Gust-induced response for the conductors and the towers are covered and the limitations in
Accepted 4 January 2016
the current structural design codes for wind loading are identified. Three main sections are considered
Available online 19 January 2016
in this study covering synoptic wind loading, downburst, and main conclusions and recommendations.
For the case of synoptic wind events, four design codes (ASCE 74 2010, AS/NZS 2010, BS 2001, IEC
Keywords:
2003) specialized in TLs are considered for comparison. Using the ASCE 74 as a datum for normalization,
Dynamic response
Transmission line
a code ratio (CR) is evaluated for various parameters to assess the discrepancy between the codes. The
Conductors code ratio for conductor forces CRFc is found to be ranging between 0.81 and 1.44. For tower forces code
Synoptic winds ratio CRFt, a discrepancy range of 0.68 and 1.85 is noticed. The study highlights the main reasons behind
Non-synoptic downburst these discrepancies. For the case of downbursts, the study reveals that the event’s size and its relative
Gust factor location to the tower lead to a number of critical load cases that need to be considered. The study pro-
Buffeting vides important design considerations for both synoptic and non-synoptic winds. At the end of the study,
a list of the main gaps existing in current design codes and recommendations to fill out these gaps is
provided.
Ó 2016 Elsevier Ltd. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2. Dynamic and quasi-static responses of TLs under synoptic winds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.1. Literature on the dynamic buffeting response of transmission lines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.2. Approach used by various codes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.2.1. Design velocity Vd . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.2.2. Design wind pressure qd . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.2.3. Conductor force Fc . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.2.4. Conductor gust response factor GC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.2.5. Tower force FT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.2.6. Tower gust response factor GT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.3. Quantitative comparison between codes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.3.1. Design wind velocity and pressure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.3.2. Conductor forces and the gust factor GC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2.3.3. Tower forces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
3. Response of TLs under downburst wind loading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
3.1. Field measurements and numerical modeling of downburst . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
3.1.1. Field measurements during downbursts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
3.1.2. Downburst numerical, analytical and empirical modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

⇑ Corresponding author.
E-mail addresses: haboshos@uwo.ca (H. Aboshosha), damatty@uwo.ca (A. El Damatty).

http://dx.doi.org/10.1016/j.engstruct.2016.01.003
0141-0296/Ó 2016 Elsevier Ltd. All rights reserved.
24 H. Aboshosha et al. / Engineering Structures 112 (2016) 23–46

3.1.3. Downburst wind field characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36

3.2. Structural modeling of a transmission line under downburst loading. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
3.2.1. Lattice tower and guy wires . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
3.2.2. Conductor lines. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
3.3. Behavior of transmission lines subjected to downbursts and corresponding critical load cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
4. Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
Acknowledgment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
Appendix A. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44

1. Introduction dynamic behavior of TLs is discussed in detail. The procedures used

Electricity is carried by Transmission Lines (TLs) from the
source of power generation to the distribution system. Overhead
transmission lines consist mainly of support towers, conductors,
insulators and ground wires. Conductors are responsible for trans-
mitting the electricity and they are attached to the towers using
insulators. Ground wires protect the line from lightning strike.
Optimal TL design is particularly important when the site of power
generation is geographically remote from population centers, such
as in the case of many hydroelectric dams in North America. The
ubiquity and full integration of electronics into modern life means
that power outages due to TL failure are unacceptable from the
standpoint of both social and economic losses. Past reports of TL
failure due to weather conditions [37,79], including High Intensity
Wind (HIW) events (downbursts and Tornadoes), emphasize the
importance of accurate design wind loads.
The current study summarizes the previously conducted work
related to the buffeting dynamic behavior of TL under synoptic
and non-synoptic winds in order to identify the limitations and
gaps in the current codes and to suggest how these gaps can be
closed. The study is structured in the form of three main sections.
Dynamic response of transmission line structures under synoptic
winds is discussed in Section 2. In this section, four design codes
[12,13,20,69] specialized in TLs are considered to compare
between the approaches utilized to estimate the forces acting on
the conductors and the towers under synoptic winds. This compar-
ison is provided in a tabulated format. For each quantity, a code
ratio (CR) is obtained by normalizing the values predicted from
the three codes [13,20,69] by the value obtained from ASCE 74
[12]. Section 3 covers the response of TL structures under down-
burst winds. This section provides a description of the downburst
wind field and a literature review showing previous field measure-
ments and numerical modeling of downbursts. The last section
(Section 4) identifies gaps in design codes and recommendations
to fill out these gaps for both synoptic and non-synoptic wind cases
studied in Sections 2 and 3. This section starts by presenting the
gaps existing in the current design codes related to the gust
response of TLs under synoptic wind and recommendation for fill-
ing these gaps. The second part of Section 4 provides the main con-
clusions and recommendations for the downburst case.
2. Dynamic and quasi-static responses of TLs under synoptic
winds
This section discusses the response of TLs to fluctuating synop-
tic winds and is divided into two main parts. In the first part (Sec-
tion 2.1), a literature review on the buffeting dynamic behavior of
TLs is presented emphasizing the main parameters affecting the
response. In addition, a literature review on field tests conducted
during the last few decades to measure loads on conductors and
towers has been provided. In the second part (Section 2.2), the
quasi-static method used by design codes to account for the
by four major design codes to obtain the design forces on the con-
ductors and the towers are compared to identify the sources of dis-
crepancy between different codes’ approaches.
2.1. Literature on the dynamic buffeting response of transmission lines
Dynamic responses of transmission lines can be classified into
the following groups: (i) buffeting due to the incoming turbulence,
(ii) vortex shedding and (iii) galloping. The current study focuses
on the dynamic buffeting response of transmission lines which
affects both the lines and the towers. Vortex shedding and gallop-
ing affect mainly the lines and can lead to adverse effects such as
clashing of wires, excessive conductor sags, wire fatigue, and gen-
erally collapse of line components. For deeper explanation of such
phenomena, authors recommend reading studies by Den Hartog
[40], Davison et al. [36], Havard and Pohlman [59], Rawlins [109]
and more recent studies such as Macdonald et al. [90], Ohkuma
and Marukawa [107], Gurunga et al. [54], Chabart and Lilien [24],
Dyke and Laneville [44], Boddy and Rice [21], Jones [72], Lilien
et al. [89], Yan et al. [142], and Ma et al. [91].
Buffeting dynamic response of the conductors is governed by
the fluctuating incoming wind speeds and the properties of the
conductors including their damping. Damping of the conductors
results from a structural contribution and an aerodynamic contri-
bution. Structural contribution of the damping is typically minor
in the order of 0.05% as indicated by Bachmann et al. [15] and thus
can be neglected. Aerodynamic contribution to damping, which is
typically referred to as aerodynamic damping fa , is usually domi-
nant and can be calculated using the expression proposed by
Davenport [31] and is shown in Eq. (1) [101].
pC d qD2 V
fa ¼ ð1Þ
4 m f D
where Cd is the drag coefficient of the conductor and can be taken
equal to 1.0 according to the ASCE 74 [12] and AS/NZS [13]; q: air
density; m: mass per unit length; D: Conductor diameter; V:  mean
wind velocity; f : conductor frequency.
As indicated from the equation above, the aerodynamic damp-
 and
ing, fa , is directly proportional to the mean wind velocity, V,
is inversely proportional to the line mass, m. This contribution
can rise up to 60% of the critical damping according to Loredo-
Souza and Davenport [83] for light weight conductors subjected
to high velocities (i.e. a critically damped system returns to its
equilibrium position without oscillating).
The study conducted by Loredo-Souza and Davenport [83] con-
cluded that the background response is, indeed, the main contrib-
utor to the total fluctuating response. However, the resonant
component can be also important if the line characteristics and
wind velocities lessen the aerodynamic damping. There were only
minor resonant effects for the cases of high aerodynamic damping.
For low aerodynamic damping, dynamic effects should be
considered but confirmation of a threshold value of aerodynamic
 H. Aboshosha et al. / Engineering Structures 112 (2016) 23–46
damping, below which dynamic behavior is of significant concern,
requires further research.
Later on, Loredo-Souza and Davenport [84,85] conducted vari-
ous wind tunnel tests on conductors subjected to turbulent synop-
tic wind and compared the conductor responses with those
obtained theoretically using the statistical method employing the
influence lines [35]. The study indicated a very good agreement
between the experimental and the theoretical results. This statisti-
cal method is general for any structure. It accounts for both the
background and the resonant components and considers the effect
of higher modes on the resonant component.
Also several full-scale conductor wind loading experiments
have been conducted during the last few decades to acquire char-
acteristics of wind and corresponding conductor response. Shan
[116] conducted a subsequent investigation to evaluate several
past full-scale conductor wind loading experiments, particularly
the experiments conducted by the Bonneville Power Administra-
tion [133], Ontario Hydro [75], Hydro Quebec [67], and the EPRI
Transmission Line Mechanical Research Center (TLMRC) [115].
For example, Bonneville Power Administration [133] conducted
field measurements on a 500 kV lattice tower subjected to extreme
wind conditions. The tower was attached to conductors with spans
252 m and 450 m on the sides. The longer span was passing over a
valley while the shorter span was passing over a flat terrain. A total
number of 23 records of the wind speed, wind direction and struc-
tural response of a transmission line were recorded. The obtained
records were extensively analyzed by Mehta et al. [98] and Mehta
and Kadaba [99,100] aiming to improve the analytical model
developed by Davenport [34] to predict conductor response in
extreme wind. Based on their studies the following points were
found: (1) Winds traversing over the valley showed a wide varia-
tion in mean profile and turbulence; (2) Effective drag coefficient
of the conductor was found to vary between 0.48 and 0.75; (3)
Noticeable resonant peaks occurred in the frequency range from
0.1 to 0.4 Hz in the conductor response spectra; (4) Most of the
fluctuating conductor responses (75%) appeared to be due to the
background component and a minor contribution of the resonant
component was found. This was due to the high conductor aerody-
namic damping in the order of 40%.
-
Influence Line of F
+
F Lin et al. [81] tested an aeroelastic model of a guyed mast, at
Fig. 1. Influence line for the member force that has a changing sign (modified from
Cook [27]).
25
For a freestanding tower, Holmes [64] studied the dynamic
response of single tower (without conductors) to address the gust
response factors. These factors were derived assuming a linearly
tapered tower. The gust factors considered the height effect, mean
velocity profile, mode shape of the tower, and the type of the load.
Later, Holmes [65] continued what he reported in part ‘‘I” to
include an expression of the aerodynamic damping of a freestand-
ing tower. In addition, Holmes [65] provided an expression to eval-
uate the top deflection of the tower. In addition, Holmes [66]
extended his comprehensive study of part I and II and reported sta-
tic load distribution of mean, background fluctuating, and resonant
components of the wind loads on freestanding lattice tower.
The natural frequency of a typical tower, with a 60 m height or
less, is typically higher than 1 Hz [69,135,81]. That is significantly
higher than the frequencies corresponding to the maximum turbu-
lence energy, which leads to a negligible resonant response compo-
nent for the towers. Harikrishna et al. [58] studied the peak base
moment of a typical transmission tower, which was mainly due
to the background component, and found that the peak base
moment predicted by the Indian code [68] and the Australian code
[13] agreed with field measurements, whereas the British code
[19] overestimated the base moment by approximately 30%. Full-
scale measurements of axial forces in tower legs by Savory et al.
[110] agreed with predictions from the British code [19], however
a higher sampling rate than available was needed for the predic-
tion of the dynamic structural response.
On the other hand, taller towers with lower natural frequencies
are expected to be dynamically excited and the literature on tall
guyed masts for telecommunications applications describes find-
ings from studies concerning isolated lattice-frame support struc-
tures. For example, analytical and full-scale vibration
measurements up to 341 m height on a guyed mast subjected to
strong winds showed the importance of including the nonlineari-
ties in the guys [108]. Frequency domain analysis [127] has yielded
reliable estimates of peak responses when the effective stiffness of
guys, obtained from a static analysis under mean wind loads, was
considered. Aeroelastic model experiments by Lou et al. [88] in a
wind tunnel found that a tall tower (180 m full-scale height) was
sufficiently flexible to be susceptible to dynamic excitation by
the wind and the resonant component could not be neglected.
However, the effects from the conductors were neglected, which
is a problematic assumption since field measurements indicated
that the vibration characteristics of towers were greatly affected
by the conductors [101]. Yasui et al. [143] conducted an analytical
study of wind-induced vibration of conductors supported by either
a suspension or a tension tower. The difference in the way the tow-
ers supported the conductors had a noticeable influence on the
response characteristics. Furthermore, from a three-dimensional
FEM model, Battista et al. [16] found that the first mode of vibra-
tion involves both tower and conductor motions; although
increases in conductor tension due to the wind load were not
included in the analysis of the natural frequency and a 2% damping
ratio was assumed across all vibration modes, which appears to be
insufficient damping especially for the modes which are mainly
due to conductor motion.
1:100 length-scaling, in a wind tunnel. A half-span of lines on
either side of the mast was modeled, thus full- and multi-span
motions were neglected. Loredo-Souza and Davenport [86] com-
pared the peak forces developing in different tower members
obtained using various codes with those obtained using the statis-
tical methods employing the influence line [35]. Their results indi-
cated a possible 20% underestimation of the peak forces obtained
by the codes than those obtained by the statistical method. This
underestimation is a result of neglecting the contribution of higher
modes that happens in the members having an influence line with
26 H. Aboshosha et al. / Engineering Structures 112 (2016) 23–46

Table 1
Summary of previous studies’ scope and main finding.

Reference Scope Main finding/conclusion

Davenport [31]
Loredo-Souza and Davenport [83–85]
Wang et al. [135]
Lin et al. [81]
Loredo-Souza and Davenport [86]
Harikrishna et al. [58]
Peil and Nolle [108]
Sparling and Wegner [127]
Lou et al. [88]
Yasui et al. [143]
Battista et al. [16]
Behncke and Ho [17]
Solari [125,126]
Simiu and Scanlan [121]
Holmes [61]
Aboshosha and El Damatty [6]
Studied the dynamic behavior of suspension
bridge cables
Conducted aeroelastic wind tunnel tests for
transmission line wires
Studied the dynamic behavior of
transmission lines
Studied the peak responses of transmission
line and compared the results with different
codes
Assessed the importance of conducting
dynamic analysis on transmission towers
Conducted numerical dynamic analyses of
TL subjected to turbulent wind
Reviewed code approaches to evaluate the
peak responses
Assessed the gust response factor developed
by Davenport [32]
Studied the responses of the line under
downbursts
Proposed an expression for cable aerodynamic damping
Background response was the main contributor to the total fluctuation response
Transmission towers can be considered non-dynamically excited
Peak response was mainly due to the background component
Indian Standard [68], Australian Standard/New Zealand Standard [14] agreed
with the field measurements while British Standard [20] overestimated the
base moment by 30%
Tall towers were dynamically excited
Natural frequency of the line was affected by the method of supporting the
wires. Coupled tower-conductor vibration modes were found
Span Reduction Factor and gust velocity factors presented in the codes were
suitable for synoptic wind and not for non-synoptic wind
Good agreement between the responses obtained from wind tunnel tests and
using the gust factor expression was found
Span reduction factor for synoptic wind cannot be used for downburst wind

an alternating sign as indicated in Fig. 1. Loredo-Souza and Daven-
port [86] concluded that the dynamic response of transmission
towers depends strongly on the turbulence intensity and both
the structural and aerodynamic damping of the towers. So far,
there is no direct expression for the aerodynamic damping of the
towers.
Regarding field measurements of wind loads on lattice steel
towers, EPRI study at the TLMRC was conducted by Shan [114] to
collect data from a full-scale lattice transmission tower. This has
been done by erecting an experimental wind tower at the TLMRC
site. A number of anemometers were placed at different elevation
levels of the wind tower. Strain gages were also mounted on the leg
posts of the tower. The purposes of this experiment were to estab-
lish the wind characteristics at the TLMRC site and to gain experi-
ence using different types of instrumentation and data acquisition
techniques in field-wind loading experiments. The main results
that were drawn from this experiment can be summarized in the
following points: (1) the magnitude of wind velocity and direction
can vary considerably during a short period of time; (2) the mean
vertical wind profile does not hold constant as usually assumed;
(3) the turbulence intensity and the gust factor increase as the
height above the ground decreases; (4) the averaging time can
greatly influence the results of wind data analysis; (5) although
wind contains little energy beyond 1 Hz, structural responses
above 1 Hz can be excited; (6) strong relationships exist between
the wind velocity and the responses in the leg posts of the wind
tower. The study also provided some comments on how to conduct
the field-wind loading experiments as well as how to analyze the
wind and response data.
Table 1 summarizes the scope of previous studies focused on
the behavior of TL under synoptic wind and shows their main find-
ings. In the next subsection, the quasi-static approach utilized by
four different design codes to account for the dynamic effect on
the TLs components is discussed in detail. The procedures provided
by these design codes to obtain design forces on the conductors
and the towers are compared in order to assess the discrepancy
between different codes’ approaches. The main results of this com-
parison are utilized in the last section (Section 4) to identify the
gaps noticed in the four codes considered in the current study
and to provide possible recommendations to fill out these gaps.
2.2. Approach used by various codes
Design codes provide a simplified approach to obtain the peak
forces acting on the elements of TLs to be used for design. In this
subsection, four design codes [12,13,20,69] specialized in TLs are
considered to compare between the approaches provided by these
codes to estimate the forces acting on the conductors and the tow-
ers. A summary of this comparison is provided in Table 2. For a bet-
ter understanding of the comparison shown in this table, the main
terms and parameters presented in the table are first defined in
this subsection. This is followed by Section 2.3 which provides
the main steps required to apply design approaches specified in
the four codes and the results of a detailed comparison between
these codes.
2.2.1. Design velocity Vd
Wind velocity used in the design, Vd, of a TL represents the aver-
aged velocity over a small time interval, i.e. 3 s. This velocity can be
exceeded annually with a probability of 1/R, and through the n year
life span of the TL with a probability of 1  (1  1/R)n. Inverse of
the annual probability of exceedance, R, is commonly referred as
the return period of the storm. Typically, a 50 year return period
is used for the design of the TLs. Design codes allow for using dif-
ferent return periods, R, by introducing a scaling factor to the
50 year return period such as the ASCE 74 [12] and the AS/NZS
[13]. As mentioned above, reference wind velocities can be aver-
aged over time intervals of 3 s, 1 min, 10 min or 1 h. Conversion
between wind velocity averaged using different time intervals
can be obtained using a curve provided by Durst [43].
This curve was based on wind speeds not exceeding 40–50 mph
(18–22 m/s) and according to Sissenwine et al. [122] and Behncke
and Ho [17], the velocity ratio could be different at higher wind
speeds. The ratio between the 3 s averaged velocity to the 10 min
or the 1 h averaged velocity is referred as the gust velocity factor,
Gv.
2.2.2. Design wind pressure qd
Design wind pressure, or sometimes is referred to as the peak
wind pressure, qp, represents the dynamic pressure caused by the
design wind velocity Vd, and can be calculated using Eq. (2).
 H. Aboshosha et al. / Engineering Structures 112 (2016) 23–46
1
qp ðzÞ ¼ qV 2d ðzÞ ð2Þ
2 V3s Incoming Velocity
where q is the air density.
In general, the air density q depends on the temperature and
the atmospheric pressure. It is equal to 1.225 kg/m3 at 15 °C at
an atmospheric pressure 101.325 kPa. For other temperature and
pressures, the expression described by Eq. (3) can be used.
q 288 1:2104 H
¼ 0 e ð3Þ
q0 T
where q0 is the density at 0 elevation H at 15 °C (288 K) and equals
to 1.225 kg/m3, T0 is the temperature in K, H: the elevation above
the ground.
2.2.3. Conductor force Fc
Conductor force Fc represents an overall wind force transmitted
from the conductor to a tower. This force is usually in the trans-
verse direction of the conductor regardless of the wind direction
as shown in Fig. 2.
2.2.4. Conductor gust response factor GC
Conductor gust response factor, GC, is used to scale the conduc-
tor forces resulting from the design pressure (based on 3 s gust
speed) to account for the dynamic response of the conductor.
Accounting for the dynamic effect using the gust factor was origi-
nally proposed by Davenport [32,33] and examined by many
researchers such as Solari [125,126], Simiu and Scanlan [121] and
introduced in a new format by Zhou and Kareem [146]. The origi-
nal version of the gust factor proposed by Davenport [33] was
based on a 1 h-averaged wind speed. This means that it combines
the gust velocity, Gv, and the gust response factor. In general, the
gust response factor for a structure accounts for the background
and the resonant components of the responses. Fortunately, con-
ductors have a high aerodynamic damping at the range of the wind
speeds used in the design, which suppresses most of the resonant
component, and thus allows the design codes to neglect the reso-
nant component. It shall be mentioned that assuming a minor res-
onant component might not be true for lower wind speeds causing
smaller aerodynamic damping. Such a small aerodynamic damping
may not be enough to suppress the resonant motion and conse-
quently may cause fatigue problems to the cross arms.
2.2.5. Tower force FT
FT represents the total force acting on a tower panel in the direc-
tion of the wind. This force is a function of the design pressure, qp.
It is also a function of the drag coefficients, Cdx and Cdy, and the
panel projected areas, Ax and Ay, in the panel x and y directions,
respectively. Drag coefficients of the panel, Cdx and Cdy, are typi-
cally a function of the panel porosities /x and /y, in the x and y
directions, respectively.
2.2.6. Tower gust response factor GT
Similar to the conductor gust response factor Gc, the tower gust
response factor, GT, is used to scale the forces acting on the tower
panels resulting from the design pressure, qp, to account for the
dynamic response of the tower. As discussed earlier and indicated
by Holmes [61], typical towers in the order of 60 m or less that
have a fundamental frequency in the order of 1 Hz or more have
a minor resonant component. That is because most of the wind
fluctuating energy is associated to small frequencies less than
1 Hz. Based on this, design codes normally neglect the resonant
component of the towers.
27
FC
Conductor
Conductor Tower
Fig. 2. Layout of the conductor and the applied force Fc (modified from the IEC
[69]).
2.3. Quantitative comparison between codes
This comparison focuses on assessing how different codes esti-
mate the main parameters affecting the design of TLs under synop-
tic wind loads. These parameters include: design wind velocity and
pressure, conductor gust response factor, forces transferred from
conductors to the tower and forces acting directly on towers. For
each quantity, a code ratio (CR) is obtained by normalizing the val-
ues predicted from the three codes [13,20,69] by the value
obtained from ASCE 74 [12]. The details of this comparison are
given below.
2.3.1. Design wind velocity and pressure
In general, the four codes use the wind velocity at 10 m eleva-
tion above the ground for open terrain exposure as the reference
velocity. The ASCE 74 [12] and the AS/NZS [13] use a 3 s averaged
wind velocity for the reference wind speed, V3sOpZ10. The BS [20]
and the IEC [69] use a 10 min averaged wind velocity, V10mOpZ10.
The codes allow for converting the velocity of the open terrain
exposure which is measured at 10 m height to obtain the velocities
for other exposures and heights, V(z). A comparison between the
expressions employed by various codes for the velocity V(z) is pro-
vided in Table 2. In order to allow for a direct comparison, velocity
expressions shown in the table are described as functions of the 3 s
gust wind speed for the open terrain exposure at 10 m height
V3sOPZ10. This is straight forward for the ASCE 74 [12] and the AS/
NZS [13] as their reference velocities are the 3 s averaged. For
the BS(2001) and the IEC [69], the conversion from the 10 min
averaged to the 3 s averaged is conducted using the formulas
provided in the two codes, where a gust velocity factor Gv is intro-
duced, as indicated in Table 2.
Design wind velocity, Vd(z) or V3s(z), is defined in the ASCE 74
[12] using the power law and is defined in the BS [20] and the
IEC [69] using the logarithmic law. The AS/NZS [13] provides a
table to describe the velocity variation, which is fitted in the cur-
rent study using the power law as indicated in Table 2. Design
pressure corresponding to the peak wind speed, qp, described by
the four codes is also compared in the table.
Fig. 3 illustrates the design pressure qpn(z) at different eleva-
tions for the sea side, open and sub urban terrain exposures nor-
malized by the design pressure qpOpZ10 at 10 m for open terrain
exposure as expressed by Eq. (4).
qp ðzÞ
qpn ðzÞ ¼ ð4Þ
qPoPZ10
As shown in the figure and indicated in Table A.1, at a height of
30 m for instance, design pressure obtained by the four different
codes are different. It is noticed from the results that the code ratio
CRp ranges between 0.91 and 1.13. The minimum value of CRp is
noticed in case of using AS/NZS [13] and IEC [69] approaches for
suburban and sea side exposures, respectively. The maximum
value of CRp is reached by using BS [19] approach for Suburban
exposure. These discrepancies are due to the slight differences in
the expressions and parameters used to describe the velocity pro-
files and consequently the pressure in each code.
28 H. Aboshosha et al. / Engineering Structures 112 (2016) 23–46

Sea side Open Sub urban

100 100 100
ASCE
AS/NZS
90 90 90
BS
IEC
80 80 80

70 70 70

60 60 60
Z (m)

Z (m)

Z (m)
50 50 50

40 40 40

30 30 30

20 20 20

10 10 10

0 0 0
0.5 1 1.5 2 0.5 1 1.5 2 0.5 1 1.5 2
P (z) P (z) P (z)
pn pn pn

Fig. 3. Normalized peak wind pressure Ppn(z) as defined by various codes.

2.3.2. Conductor forces and the gust factor GC
Expressions used in the codes for the force FC transmitted from
the conductor to the towers are summarized in Table 1. As indi-
cated from the table, the expressions are similar in the considered
four codes. The force depends on the angle h, which is the angle
between the incoming wind and the perpendicular to the line as
indicated in Fig. 2. The term GC is called the conductor gust
response factor, which is sometimes called the Span Reduction Fac-
tor (SRF) when used as a multiplier for the peak design pressure qp.
The Gc or the SRF accounts for the lack in the correlation due to
the increase of the conductor span S. As indicated from Table 2,
the AS/NZS, the BS and the IEC use a SRF that only depends on the
span length. According to Davenport [35], this SRF can be expressed
by Eq. (5), where the term JLs is called the joint acceptance function
which accounts for the lack of the correlation along the span S and
is expressed by Eq. (6).
pffiffiffiffiffi
1 þ 2IðzÞg f J Ls
SRF ¼ Im
1 þ 2IðzÞg f
Z 1 Z 1
Lx jn1n2j
J Ls ¼ u2v ðn1 Þu2v ðn2 ÞiFc ðn1 ÞiFc ðn2 Þeð
1 1
where I(z) is the turbulent intensity at the conductor elevation z; gf:
velocity peak factor; n1 and n2 are local axis which equals to 1, 0, 1
at the left tower, the tower of interest, the right tower as indicated
in Fig. 4; iFc: Influence line for the overall conductor force transmit-
ted to the tower iFc = 1 + n, where n = 1:0 and iFc = 1  n, where
n = 0:1; Ls: transverse length scale of the turbulence; Im : is an inte-
gral that depends on the distribution of the mean forces along the
R1
spans and the influence line Im ¼ 1 iFc ðn1 Þu2v ðn1 Þdn1 ; uv is the
velocity distribution function which equals to V(n)/V0, where V0 is
the velocity at any reference point. For the case of horizontal con-
ductors uv (n) = 1.
As indicated from Eqs. (5) and (6), SRF depends on the turbulent
intensity I(z), which vary with height, and on the turbulent length
scale Ls, which vary with the terrain exposure. Consequently, the
SRF is expected to vary with the height and the exposure type. This
is not the case for the AS/NZS [13] the BS [20] and the IEC [69], as
the expressions used for the Gv or the SRF do not depend on the
height or the exposure. Only the ASCE 74 [12] allows for changing
the Gv or the SRF with varying the conductor height and the terrain
exposure as indicated in Table 2.
As mentioned previously, the ASCE 74 [12] code uses a 3-s gust
velocity as the reference wind speed, however, when the gust
response of the conductor is calculated, the code uses a formula-
tion that is based on 10 min-averaged speed. The code introduces
a conversion factor Kv based on Durst curve. Other codes such as
the BS [20] and the IEC [69] use different conversion approaches.
This step in particular leads to differences in the order of 10–15%
[17].
It should be also mentioned that out of the four studied codes,
AS/NZS [13] is the only code which proposed a different expression
for the conductor gust factor under downburst wind. Following the
work of Holmes [61], which recognized a significant difference
between synoptic and downburst wind loading, AS/NZS [13] pro-
posed a different expression GCDb for the gust factor as indicated
ð5Þ
in Table 2.
A comparison between the normalized force Fcn transmitted
from the conductors to a typical tower, defined by Equation, is
Ls
Þ
dn1 dn2 ð6Þ shown in Fig. 5. The comparison is conducted for two conductors
that are placed at 20 and 40 m elevations for a span S ranging from
200 to 800 m. As indicated from the figure, and summarized in
Tables A.2 and A.3 the normalized forces FCn(20) and FCn(40) calcu-
lated by the four codes generally increases with the conductor
height and decreases with the increase of the conductor span.
qp
Fc
F cn ðzÞ ¼ 2
¼ SRF ð7Þ
qpOpZ10 SD cos ðhÞ qpOpZ10
The tables also show that the discrepancy in the code ratio for
conductor forces CRFc(20) for a conductor located 20 m above the
ground has a minimum value of 1.02, 0.99, 0.83 and a maximum
value of 1.4, 1.3, 1.42 for the sea side, open and suburban expo-
sures, respectively. For a conductor located at 40 m above the
ground, it is noticed that CRFc(40) has a minimum value of 1.01,
1.04, 0.81 and a maximum value of 1.44, 1.29, 1.31 for the sea side,
open and suburban exposures, respectively. In addition, the com-
parison shows that the minimum values of the code ratio CRFc
Table 2
Comparison between various codes.

Term Def. ASCE [12] AS/NZS [13] BS [20] IEC [69]

Vref Reference velocity V3sOpZ10 V3sOpZ10 V10mOpZ10 V10mOpZ10
 a1 pffiffiffiffiffi  z a=2 G K G
Vd(z) or V3s(z) (m/s) Design velocity
V 3s ðzÞ ¼ 1:418V 3sOpZ10 Zh V 3s ðzÞ ¼ b V 3sOpZ10 10m V 3s ðzÞ ¼ kt lnðz=z0 Þ 1:435
v V
3sOpZ10
r v
V 3s ðzÞ ¼ 1:355 V 3sOpZ10
Zg
V 3s ðzÞ ¼ G
v kt lnðz=z0 ÞV 10mOpZ10 V 3s ðzÞ ¼ G logðz=z0 Þ

v logð10=z0 Þ V 10mOpZ10
V 3s ðzÞ ¼ Gv V 10m ðzÞ V 3s ðzÞ ¼ G
v V 10m ðzÞ
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
G
v ¼ 1 þ lnðz=z0 Þ
2:28
G
v ¼ a lnðzÞ þ b
qp(z) (N/m2) Design wind pressure  2a   z a    
qp ðzÞ ¼ 0:60V 23sOpZ10 b V 3sOpz10 2
qp ðzÞ ¼ 0:613 1:435 lnðz=z0 Þkt K 2g qp ðzÞ ¼ 0:613 1:355
V 3sOpZ10 2 2 
qp ðzÞ ¼ 1:232V 23sOpZ10 Zh
Zg
10m K r Gv
FC Conductor force qp ðzÞGc Sd cos ðhÞ2 qp ðzÞGc Sd cos ðhÞ2 qp ðzÞGc Sd cos ðhÞ2 qp ðzÞG
C Sd cos ðhÞ
2

H. Aboshosha et al. / Engineering Structures 112 (2016) 23–46

pffiffiffi a1 pffiffiffiffi GC ¼ 1:3  a lnðSÞ G
c ¼ 1:575  0:106 lnðSÞ
ðS=210mÞ
GC Conductor gust 1þ13:23 jð10=Z h Þ FM BC GCNW ¼ 0:59 þ 0:41e
GC ¼
response factor K2 GCDb ¼ 1:0  0:3125 S200
1000 6 1:0
v
1
BC ¼ ð1 þ 0:8S=Ls Þ
FT Tower force F ¼ qp ðzÞGT ðC Dx Ax cosðhÞ2 þ C Dy Ay sin ðhÞ2 Þ F ¼ qp ðzÞGT ðC Dx Ax cos ðhÞ2 kh Þ þ ðC Dy Ay sin ðhÞ2 kh Þ F ¼ qp ðzÞGT kh ðC Dx Ax cos ðhÞ2 þ C Dy Ay sin ðhÞ2 Þ F ¼ qp ðzÞG
T kh ðC Dx Ax cos ðhÞ
2

þC Dy Ay sin ðhÞ2 Þ
GT Tower gust response pffiffiffi a1 pffiffiffiffi GT ¼ 1:0 GT ¼ 1:05 2

GT ¼
1þ13:23 jð10=Z h Þ FM BT G
T ¼
dz þezþf
G2
factor K 2v v
1
BT ¼ ð1 þ 0:373h=Ls Þ
8 9
CDx,y Drag coefficient for the >
> 4:0 / < 0:025 >
> C Dx;y ¼ 4:1/2  6:0/ þ 4:0 C Dx;y ¼ 4:1/2  6:0/ þ 4:0 C Dx;y ¼ 4:1/2  6:0/ þ 4:0
< =
towers 4:1  5:2/ 0:025 6 / < 0:44
C Dx;y ¼ kh ¼ 1 þ 0:55 sin ð2hÞ2 k2 kh ¼ 1 þ 0:2 sin ð2hÞ2 kh ¼ 1 þ 0:2 sin ð2hÞ2
>
> 1:80 0:045 < / < 0:69 >
> 8 9
: ; > 0:2 / 6 0:2 >
1:3 þ 0:7/ 0:7 < / < 1:0 >
< >
=
/ 0:2 < / 6 0:5
k2 ¼
>
> 1  / 0:5 < / 6 0:8 >
>
: ;
0:2 0:8 < /
Par. Different parameters Zg = 213, 274, 366 m b = 1.115, 1.000, 0.856, 0.669 z0 = 0.01, 0.05, 0.30, 1.0 m Kr = 1.08, 1.00, 0.85, 0.67
used in the code a = 11.5, 9.5, 7.0 a = 0.064, 0.130, 0.090, 0.185 kt = 0.17, 0.19, 0.22, 0.24 a = 0.291, 0.373, 0.494, 0.615
aFM = 10.0, 7.0, 4.5 for sea side, open, sub urban and urban terrains a = 0.073, 0.082, 0.098, 0.110 b = 1.05, 0.976, 0.912, 0.814
j = 103  3, 5, 10 for sea side, open, sub urban and urban d = 104  2.0, 2.0, 2.0, 2.0,
Ls = 51.8, 67.0, 76.2 m terrains, respectively e = 102  2.32, 2.74, 2.98,
for sea side, open and urban terrains, 3.84
respectively f = 1.47, 1.68, 2.27, 2.92
Kv = 1.43, and is used to z0 = 0.01, 0.05, 0.3, 1.0 m
for sea side, open and urban
terrains, respectively

V3sOpZ10: 3 Second gust velocity measured in open terrain at elevation z of 10 m; V10mOpZ10: 10 min averaged velocity measured in open terrain at elevation z of 10 m; subscript Op refers to an open terrain; qr: reference pressure; qp:
peak pressure; qm: mean pressure; FC: conductor force perpendicular to the line; FT: Tower force in the wind direction; GC: Gust factor for the conductor; GC3sEq: equivalent gust factor for the conductor in the IEC [69] code if the
peak pressure is used as the reference pressure; GT: Gust factor for the tower; GT3sEQ: Equivalent gust factor for the tower if the peak 3 s gust pressure is used; The superscript ⁄ indicates that the function is given in the standard in a
graph and is fitted here using the presented expression; The super script ⁄⁄ indicates the name of the these factors are given different name in the codes and the name assigned in this study is used for comparison purpose.
Subscripts NW refers to normal winds and Db refers to downburst wind.

29
30 H. Aboshosha et al. / Engineering Structures 112 (2016) 23–46

due to the lack of correlation while increasing the tower height, as

n1, n2
n1, n2=-1 n1, n2=0 n1, n2=1
Fig. 4. Parameters required by the Joint Acceptance Function (JAF).
are noticed in case of AS/NZS [13] and the maximum values are
obtained from BS [20] and IEC [69]. This is related to the discrep-
ancy between the peak pressure qp(z) and between the expressions
used for the GC (i.e. SRF) recommended by the four codes. It is
noticed that the least force prediction is found by the ASCE 74
[12] code for the sea side and the open terrain exposures and by
the AS/NZS [13] for the suburban exposure. It is also noticed that
the maximum force for the sea side and open exposures is esti-
mated by the BS [20] and for the suburban exposure by IEC [69].
2.3.3. Tower forces
Expressions for the force due to the design pressure qp(z)
applied to a tower panel, Ft, are summarized in Table 2. This force
acts in the direction of the wind load. As indicated from the expres-
sions presented in the table, design codes use the gust response
factor for the tower GT to scale the design pressure qp(z). The AS/
NZS [13] uses a tower gust factor, GT, equal to 1.0 and the BS
[20] uses a factor of 1.05, which means that the two codes do
not in general allow for reducing the loads resulting from the lack
of correlation while increasing the tower height. On the contrary,
the ASCE 74 [12] allows for reducing the load on the tower panels
shown in Fig. 6. The code increases the gust factor GT with the
increase of the terrain roughness, which is the same trend for the
IEC [69] code. It should be mentioned that the tower gust factor
GT can be calculated using similar expressions to those in Eqs. (5)
and (6) while using the influence line of the base tower moment,
iBM, or the base tower shear iBS. According to these equations, rough
terrain exposures having large length scales Ls, would result in
higher gust factor GT than the smooth terrains, which agrees with
the trend predicted by the ASCE 74 [12] and the IEC [69] codes.
Similar to the conductor gust factor, the ASCE 74 [12] intro-
duces a conversion factor Kv to convert the gust factor formulation,
which is based on 10-min averaged velocity, to be used with a 3-s
reference velocity. This conversion leads to an average error of 10–
15%.
By comparing the expressions used by various codes to obtain
the tower forces FT, it appears that the ASCE 74 [12] is missing a
factor called Kh used to account for the increase of the projected
area subjected to the wind with the increase of the pitch angle h
up to 45°. This agrees with the findings of the wind tunnel testing
conducted by Mara et al. [93]. Normalized tower force, FTn, defined
by Equation is calculated for two different angles of attack and
compared as shown in Fig. 7 and summarized in Tables A.4 and
A.5. The comparison is conducted considering a square tower with
a height of 25 m that has equal projected area A and solidity ratio /
in both x and y directions. Two wind angle of attacks, 0° and 45°,
are considered in the comparison. Tables A.4 and A.5 summarize
the normalized tower forces FTn(0°) and FTn(45°) estimated using
the four codes. For the forces acting on tower panels in case of 0°
angle of attack, the discrepancy in the code ratio for tower forces
CRFt(0°) shows a minimum value of 1.08, 1.05, 0.68 and a maxi-
mum value of 1.53, 1.29, 1.27 for sea side, open and suburban ter-
rain exposures, respectively. These are compared to CRFt(45°), in
case of 45° angle of attack, which has a minimum value of 1.30,

Sea side z=20m Sea side z=40m

1.5 1.5

1 1
Cn

Cn
F

0.5 0.5

0 0
200 300 400 500 600 700 800 200 300 400 500 600 700 800
S (m) S (m)

Open z=20m Open z=40m

1.5 1.5

1 1
Cn

Cn
F

0.5 0.5

0 0
200 300 400 500 600 700 800 200 300 400 500 600 700 800
S (m) S (m)

Suburban z=20m Suburban z=40m

1.5 1.5

1 1
Cn

Cn
F

0.5 0.5

0 0
200 300 400 500 600 700 800 200 300 400 500 600 700 800
S (m) S (m)
ASCE AS/NZS BS IEC

Fig. 5. Conductor normalized force FCn.

 H. Aboshosha et al. / Engineering Structures 112 (2016) 23–46 31

Sea side Open Sub urban

60 60 60

55 55 55

50 50 50

45 45 45

Z (m)
Z (m)

Z (m)
40 40 40

35 35
35

30 30
30

25
25 25

20
20 20 0.8 0.9 1 1.1 1.2 1.3
0.7 0.8 0.9 1 1.1 0.7 0.8 0.9 1 1.1
GT GT
GT

ASCE AS/NZS BS IEC

Fig. 6. Gust tower response factors GT used in various codes.

Sea side θ=0o Sea side θ=45o

6 6

4 4
tn

tn
F

2 2

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

Open θ=0o Open θ=45o

6 6

4 4
tn

tn
F

2 2

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

Suburban θ=0o Suburban θ=45o

6 6

4 4
tn

tn
F

2 2

0 0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
φ φ
ASCE AS/NZS BS IEC

Fig. 7. Normalized tower force Ftn.

1.16, 0.82 and a maximum value of 1.85, 1.56, 1.53 for the same results is due to the difference in the design pressure in addition
type of exposure. As indicated from the reported values, it appears to the difference between the expressions used for the Kh and for
clearly that angle of attack 45° lead to higher values of CRFt. This is the drag coefficient Cd.
because ASCE 74 [12], used for normalization, does not take into
account the effect of change in the angle of attack (i.e. Kh = 1.0). FT
F Tn ¼ ð8Þ
Fig. 8 shows the variation of the forces with the change of the wind qpOpZ10 A
angle of attack, h. These figures are plotted considering a typical
solidity ratio of 20%. As inferred from the Kh equation, presented
in Table 2, tower forces increase with the angle h until a value of 3. Response of TLs under downburst wind loading
45° then tend to decrease. It is found that the IEC [69] code predicts
higher forces for the open and suburban exposures than those pre- A downburst is an intensive downdraft air that induces very
dicted by other codes. Maximum force for sea side terrain exposure strong wind in all directions when striking the ground. Fujita
is predicted by the BS [20] code. The discrepancy between the [48] defined a downburst as a mass of cold and moist air that drops
32 H. Aboshosha et al. / Engineering Structures 112 (2016) 23–46

Sea side φ =0.20

6

4
tn
F

0
0 10 20 30 40 50 60 70 80 90
θ

Open φ =0.20
6

4
tn
F

0
0 10 20 30 40 50 60 70 80 90
θ

Suburban φ =0.20
6

4
tn
F

0
0 10 20 30 40 50 60 70 80 90
θ
ASCE AS/NZS BS IEC

Fig. 8. Variation of the FTn with the wind angle of attack h.

Fig. 9. Downburst event. Source: http://lacasanaranja.wordpress.co.
suddenly from the thunderstorm cloud base, impinges on the
ground surface, and then horizontally diverges from the center of
impact as shown in Fig. 9. In basic terms, downbursts are down-
drafts with sufficient energy to reach ground level.
3.1. Field measurements and numerical modeling of downburst
3.1.1. Field measurements during downbursts
As mentioned in Section 1, wind field prediction for downbursts
is a major challenge. A limited number of field measurements for
downbursts are available in the literature. These include the North-
ern Illinois Meteorological Research (NIMROD) and the Joint Air-
port Weather Studies (JAWS) reported by Fujita [49], and the
FAA/Lincoln Laboratory Operational Weather Studies (FLOWS)
reported by Wolfson et al. [140]. The initial diameter (DJ) of the
downdraft reportedly ranges between 600 and 1700 m as provided
by Hjelmfelt [60]. Wilson et al. [138] used the Doppler weather
radar data from the JAWS project and reported the horizontal
and vertical profiles of the microburst. Choi [28] reported measure-
ments from wind and rain stations at Singapore using 20 m mast.
The study reported that gust speeds mostly take place during thun-
derstorms. The author compared the characteristics of the thun-
derstorm and non-thunderstorm winds and reported that the
turbulent effect and gust factors are higher in the first case. Later
on, Choi [28] carried out field measurements and experimental
thunderstorm simulations to investigate the wind profile with
height during thunderstorms at Singapore. The field measurements
were conducted considering five levels of measurements along a
150 m tower height while the experimental work adopted the
impinging jet methodology. The study emphasized the effect of
the distance from the thunderstorm center, intensity of the storm,
and the ground roughness. Orwig and Schroeder [105] compared
the metrological data of a downburst thunderstorm to those of
synoptic winds in a research project called 2002 Thunderstorm
Outflow Experiment. In their experiment, a linear array seven tow-
ers were utilized to record two high intensity wind events, a rear-
flank downdraft of a supercell, and a derecho. The analysis showed
significant differences with those events of non-thunderstorm
winds. Gunter and Schroeder [53] utilized two high-resolution
mobile Doppler radars to measure full-scale data of three thunder-
storm events in a project named SCOUT. The unique technology of
their experiments allowed analyzing the vertical profile of the
flow. The measurements showed different behavior between each
of the measured events in addition to the instantaneous dual Dop-
pler wind speed and direction of each of individual event. The anal-
ysis showed that the location of the maximum speed of each event
was higher in elevation than those of numerical simulations and
other metrological measurements. Lombardo et al. [87] analyzed
the archived data obtained Automated Surface Observing System
ASOS to identify a number of downburst thunderstorms and
compared them to ABL events. The results showed that a shorter
averaging times (15–60 s) can be used for downbursts compared
to 1–5 min for ABL winds. In addition, the study revealed that gust
factors correspond to those thunderstorm events differ between
thunderstorm types. Based on the date recorded at different ports
 H. Aboshosha et al. / Engineering Structures 112 (2016) 23–46
(a) Vortex ring model
(c) Cooling source model
Fig. 10. (a) Vortex ring model [111], (b) impinging jet model [111,73] and (c) cooling source model [129].
in Europe such as Genova, Savona, La Spezia, Livorno, and Bastia,
De Gaetano et al. [39] analyzed the set of data recorded using
through a semi- automated procedure to separate different events.
The authors stated that separating the synoptic and non-synoptic
events is a hard task due to the presence of third-class events that
has intermediate properties between the main two classifications
of winds. Solari et al. [123] reported the thunderstorm records that
have been recorded through ‘‘Wind and Ports” project. Using an in-
situ wind monitoring network, they analyzed the main properties
of a number of thunderstorm records detected in the Ports of
Genoa, La Spezia and Livorno during the period of 2011 to 2012.
Solari et al. [123] reported the mean values and the coefficient of
variation of three wind velocity ratios that are believed to have a
significant effect on structures.
The variable nature of downbursts, both spatially and in time,
makes full-scale study a difficult task.
3.1.2. Downburst numerical, analytical and empirical modeling
Due to the difficulty of field measurements for HIW events such
as downbursts, most of the research has relied on numerical mod-
eling or reduced-scale physical modeling (i.e. experiments) to pre-
dict the downburst wind field as well as the response of structures
to such events. Three numerical approaches in the literature are as
follows: (a) Ring Vortex Model, (b) Impinging Jet (Impulsive Jet)
Model, and (c) Cooling Source (Buoyancy-Driven) Model, Fig. 10.
The Ring Vortex Model [147,71,132,111] simulates the vortex ring
that is formed during the descent of the downdraft air column. As
reported by Savory et al. [111], the Ring Vortex Model is not accu-
rate in simulating the downburst field near the ground after the air
column touches the ground.
The impinging jet model is based on the analogy between an
impulsive jet impinging upon a flat surface and a downburst Fujita
[48]. The cooling source method is based on introducing a cooling
source inside the computational domain simulating the cooling
process in the cloud base. This cooling process increases the weight
of the cloud base and formulates the downdraft. Selvam and
Holmes [112] developed a two-dimensional, steady numerical
model for an impinging jet. Hangan et al. [57] improved this basic
33
(b) Impinging jet model
Ellipsoidal Cooling
Function
Computational Fluid Dynamics (CFD) model and validated their
CFD assumptions by carrying out pressure and velocity measure-
ments with an impinging jet. Kim and Hangan [73] compared their
simulated mean wind profile with full-scale data [141], experi-
mental results [41,42], as well as a generic 1/7 power law bound-
ary layer profile. Kim et al. [74] reported that downbursts can
produce different loading for tall structures and, consequently, dif-
ferent collapse modes. These CFD models of an impinging jet
yielded time series for a vertical (axial) component (VVR) of the
velocity field, as well as a radial (horizontal) component (VRD). At
a fixed point in space, these two velocity components are functions
of horizontal location relative to the center of the jet and height
relative to the ground.
Sengupta and Sarkar [113] simulated downbursts using the
impinging jet method employing K-epsilon, K-omega, Shear Stress
Transport (SST) and Large Eddy Simulation (LES) turbulence mod-
els and compared the resulting profiles with those from an exper-
iment. Their results showed a reasonable agreement between the
profiles obtained from the LES and from the experiment. The appli-
cability of using LES to simulate downbursts is also indicated from
the results of Hadžiabdić [55], Chay et al. [22], Gant [51], and
Aboshosha et al. [8]. Later, Zhang et al. [145] studied the down-
burst simulation using, experimentally, the steady impinging jet
model and, numerically, the cooling source model. The PIV mea-
surement of the impinging jet model showed a good match with
the average wind velocity obtained previously by field measure-
ments. However, the study showed inconsistencies in the transient
features of each model. The study provided a set of advantages and
disadvantages for each model.
Mason et al. [95] implemented the cooling source method based
on a dry, non-hydrostatic, sub-cloud and axisymmetric model. One
year later, Mason et al. [96] extended this work to a three-
dimensional model. In both papers, the Scale Adaptive Simulation
(SAS) method was used, which is an improvement for the unsteady
Reynolds-Averaged Navier–Stokes (URANS) method used to pre-
dict unsteady turbulent flow. However, Gant [51] reported that
the SAS method appears to be over-predicting the turbulent viscos-
ity of jet-type flows. Vermeire et al. [129] simulated downbursts
34 H. Aboshosha et al. / Engineering Structures 112 (2016) 23–46
Table 3
Aerodynamic roughness for different terrain exposures [137].
Terrain exposure
Smooth
Open
Roughly open (suburban)
Very rough (forest or urban)
using the cooling source approach with Large Eddy Simulation
(LES) and the results showed good agreement with Mason et al.
[95] and disagreement with impinging jet models.
Most studies have considered a single downburst column
whereas in a study that modeled two downbursts, representing
part of a downburst line, which commonly occur in nature. Ver-
meire [130] showed that two adjacent downbursts may result in
peak near-ground wind speeds that are 55% greater than those
associated with one of those downbursts on its own, together with
a 70% greater damage footprint. However, even cooling source
models neglect some of the other key physics of real events, nota-
bly the forcing associated with precipitation, and the fact that such
events are not axi-symmetric. Recently, LES simulation of complete
downburst-producing thunderstorms [103] have revealed the
complexity of downburst outflows, although circumferential aver-
aging of the wind field suggests that it may be possible to estimate
the highest wind speeds at any time within the event by using the
averaged wind field (essentially an axi-symmetric metric) multi-
plied by a consistent peak factor [104].
Kim and Hangan [73] studied different field and numerical
measurements of downbursts and were able to conclude a con-
stant estimate of 6 m/s for the translation velocity of the event
which was then included in their impinging jet simulation. Lin
et al. [82] conducted a physical and numerical comparisons
between the translating (slot jet) and non-translating (steady,
round impinging jet) simulations of downbursts. The study
revealed that the non-translating downburst underestimates
velocity profiles, on the other hand, the slot jet velocity profile
tends to exceed the observations. Although both simulations ade-
quately represents profile shape and height, one of the differences
between them is that the translating cooling source have shown
peak velocities at noticeably lower heights than the other profiles.
Design codes define the wind speed at any point as being the
sum of a mean wind component (non-turbulent) and a turbulent
component. Most studies have considered only the non-turbulent
component of downburst wind velocity. The turbulence character-
istics are essential for quantifying the peak loads and responses
experienced by surface structures as indicated by Kwon and Kar-
eem [77], Chen and Letchford [25,26], Chay et al. [22,23] and
Holmes et al. [63]. An innovative approach is introduced by Solari
et al. [124] which treated a SDOF structure subjected to thunder-
storm similar to response the spectrum approach usually used
for structural seismic design. In their approach, a thunderstorm
response spectrum is defined to allow the calculation of the max-
imum structural response. Similar to the previously denoted stud-
ies, the study emphasized the importance of considering the peak
records of thunderstorms.
Recent downburst simulations have considered the effects of
terrain roughness [95,96,129] and topography [97] on downburst
and synoptic winds. Terrain roughness in these studies was mod-
eled using neutral wall functions. Wall functions are typically used
to economically simulate the wall bounded flow using few number
of grid layers. Without using a wall function, simulation of wall-
bounded flows, such as downbursts, will be computationally very
expensive due to the required large number of grid layers near
the ground to model the rapidly varying velocity gradient. Terrain
roughness is typically characterized by the aerodynamic roughness
z0, summarized in Table 3 for different terrain exposures. Accord-
ing to Blocken et al. [18], the maximum roughness z0 that can be
z0 (m) accurately modeled by a wall function is governed by Dz > 60 z0,
0.005 where Dz is the height of the first grid layer. Mason et al. [95]
0.03 and Vermeire et al. [129] used a first grid layer with a height Dz
0.25 equal to 1.0 m, which shades doubts on their results for the terrain
1.00 roughness greater than 0.016 m.
The aforementioned structural studies of downburst loading
provided a good base of knowledge about mean flow characteris-
tics. However, the characteristics of the turbulence (e.g. turbulence
intensities or length scales) associated with downbursts are yet to
be fully clarified. Recently, Aboshosha et al. [7] characterized the
turbulent component of downburst wind field for 4 terrain expo-
sure conditions employing LES. Those terrain exposure were simu-
lated using fractal surfaces employing the modified roughness
gradient drag-based model [7], which does not have a limitation
on the maximum roughness z0 that can be used.
Many research attempts were done to simulate the downburst
wind field analytically. Oseguera and Bowles [106] and Vicroy
[131] developed an analytical model that counted for the mean
wind speed of stationary downbursts. The main drawback of their
models was that it eliminates the translation velocity. The model
also ignored the nonlinear variation of the downburst boundary
layer thickness. Later, Chay et al. [22] improved the model pro-
vided by Vicroy [131] by modifying several factors such as the
equation of the intensity-scaling factor so that the model could
predict the maximum out flow speed. They also considered the
effect of the distance from the downburst center on the maximum
wind speed. The model also took into account the effect of the
translation speed. However, Chay et al. [22] they did not consider
the continuity equation which controls the relation between the
vertical and the horizontal speed. In addition, their study did not
provide an analytical solution for the downburst turbulent compo-
nent. Abd-Elaal et al. [1] continued working on the previous mod-
els by incorporating varying characteristic lengths into the velocity
shaping functions and added external shaping functions for inten-
sity scalars. In Their model, Abd-Elaal et al. [1] were able to account
for both the radial and the vertical profiles of the downbursts tak-
ing into consideration the nonlinear effects of the downburst
boundary layer nonlinear development.
Different empirical simulations of the downburst wind filed are
found in the literature. Holmes and Oliver [62] developed an
empirical model of the horizontal wind component of the down-
burst. The model was developed using the basic knowledge of
the impinging jet model. The study provided two main equations
for the radial velocity and the corresponding storm translation
speed. The model was then validated using field measurement of
the downburst recorded at Andrews Air Force Base in 1983. Wood
et al. [141] proposed an empirical equation to predict the normal-
ized mean velocity profile using wind tunnel experimental data of
a downburst. Li et al. [80] developed an empirical 3D axisymmetric
steady state wind filed of downburst using the impinging jet basis.
The model considered the nonlinear effects of the downburst pro-
file by incorporating varying characteristic lengths into the velocity
shaping functions. In their study, a vertical shaping function for the
radial component of the downburst was developed and showed
good agreement with CFD results. However, Li et al. [80] reported
that the nonlinear effects could not be incorporated into the verti-
cal velocity model. Both studies accounted only for the steady state
flows of the downbursts. Abd-Elaal et al. [2] developed an empiri-
cal model that is able to predict the downburst turbulent compo-
nent using several recorded field measurements of this event.
The frequency features of the downburst is the main factor that
affects the structural response. Wang et al. [134] utilized a hybrid
time–frequency domain method using the stationary wavelet and
Hilbert transforms to model the instantaneous features of the
 H. Aboshosha et al. / Engineering Structures 112 (2016) 23–46 35

Velocity (m/s)

Time (sec)

Fig. 11. Contribution of the turbulent component to the overall wind velocity.

Tower one is the outflow of a rear-flank downdraft and another one is

Transmission Line
R The study recommended an empirical equation to calculate the
Downburst
DJ
Fig. 12. Downburst parameters with respect to the tower of interest.
downburst accurately. Orf et al. [102] conducted high-resolution
simulation of a downburst using three-dimensional cloud model.
The near ground outflow was analyzed using the circumferential
spatial averaging of the wind speed components. The analysis
showed that the general location of the maximum wind speed
was preserved; however, the averaging did not maintain the out-
flow structure, specifically the main horizontal vortex. Su et al.
[128] utilized the two full-scale downburst records, which
recorded near Reese Technology Center, Lubbock, Texas, where
Derecho as reported by Gast and Schroeder [50] to evaluate differ-
ent downburst simulation techniques such as the moving average,
discrete wavelet transform and empirical mode decomposition.
appropriate averaging window size. The study recommended a
carful choice for the window size it may differ for different down-
bursts. The analysis also showed that the discrete wavelet trans-
form as well as the empirical method had a good performance in
estimating the varying mean wind speed.
Fig. 11 shows a sample mean and turbulent wind velocity
record obtained from CFD simulation conducted by Aboshosha
et al. [7]. The figure indicates the contribution of the turbulent
component to the overall downburst wind velocity. An envelope
of the peak response can be approximated by scaling up the mean
component alone. Most quasi-static analyses of TLs have relied on
this scaled mean approach.
Turbulence intensity quantifies the magnitude of wind velocity
fluctuations relative to a mean wind velocity. Turbulence length
scales give indications of the size of the flow structures that affect
the span reduction factor and the gust factor [63]. The turbulence
associated with downbursts plays an important role when quanti-
fying loading for the design of transmission lines. For the case of
synoptic wind loading, design codes [12,13] apply a gust load fac-
tor to account for the reduction of the turbulence correlation with
space, the terrain type and the dynamic response of the structure.

250 250

200 200

R/Dj=1.00
Height (m)

150
Height (m)

150
R/Dj=1.00 R/Dj=1.10
100 R/Dj=1.10 100 R/Dj=1.20
R/Dj=1.20 R/Dj=1.30
50 R/Dj=1.30 50

0 0
0 0.2 0.4 0.6 0.8 1 1.2 0 0.2 0.4 0.6 0.8 1 1.2
Normalized Vertical Velocity V VRD/Vj Normalized Vertical Velocity V VRD/Vj

Fig. 13. Vertical profile of the horizontal/radial outflow wind speed [120,73]. Fig. 14. Vertical profile of the vertical/axial outflow wind speed [120,73].
36 H. Aboshosha et al. / Engineering Structures 112 (2016) 23–46

1.2 to the gust load factor for synoptic wind, whereby a factor was pro-
DJ=500
posed for scaling up conventional wind loads so as to match the
DJ=1000
1 loads resulting from gust-front winds. This approach contains var-
DJ=1500
ious components affecting the loading due to gust-front winds,
DJ=2000
such as a kinematic effects factor for the variation in the vertical
0.8
profile of wind speed, a pulse dynamic factor for the dynamic
effects introduced by the sudden rise in wind speed, a structural
J
V /V

0.6 dynamics factor for the non-stationarity of turbulence in gust-

10

front winds and a potential load modification factor to represent

0.4 the transient aerodynamics.

3.1.3. Downburst wind field characteristics

0.2
Downburst is a localized event which makes the wind field very
dependent on the spatial location relative to the downburst center.
0 Fig. 12 illustrates the spatial variables of the downburst relative to
0 0.5 1 1.5 2 2.5 3 3.5 4
R/DJ a transmission tower. These variables represent the jet diameter
(DJ), and the location of the downburst center relative to the tower
Fig. 15. Downburst reference horizontal velocity (V10) at 10 m height [120,73].
(represented by the polar coordinates R and h).
Shehata et al. [120] conducted an extensive parametric study to
investigate the downburst wind profiles. Fig. 13 illustrates the vari-
1.4 ation of the radial velocity, normalized with respect to the jet
velocity, along the height. The maximum velocity profile was
1.2 Maximum peak found to occur at an R/DJ value of 1.2. The absolute maximum
velocity is approximately equal to 1.1 VJ. El Damatty et al. [47]
1 reported that the value 1.1 VJ can be considered as a design refer-
ence velocity at 10 m height (V10). Vertical component is typically
0.8 Constant value less than the horizontal component close to the ground (i.e. 100 m
VRD/VJ

or less) as shown in Fig. 14.

0.6 Mara et al. [93] found that the wind inclination can reach 20°
Minimum peak above horizontal in a critical case for a lattice tower. Fig. 15 shows
0.4 the variation of the reference velocity, V10, normalized with the jet
velocity, VJ, for different values of jet diameters. This figure indi-
0.2 cates that the absolute maximum velocity, which occurs at R/DJ
of 1.2, increases with decreased jet diameter. Furthermore, for
0 the time variation, the time history of the radial velocity compo-
0 100 200 300 400
nent (VRD), as well as the vertical velocity component (VVR), follows
Time (sec)
a trend with a maximum peak followed by a minimum peak as
Fig. 16. Typical time history of the horizontal/radial velocity [120,73]. shown in Figs. 16 and 17. The subsequent stabilization at a con-
stant value is an artifact of a simulation approach in which the
jet momentum source starts impulsively (correct) but then is left
0.6 on (non-physical). After the peak, wind speed converges to the
Maximum peak
steady state jet which is not realistic. It should be mentioned that
CFD simulations employing impinging jet properly simulate the
0.4
magnitude and the time instance of the peak velocities which are
the key parameters of the peak loads induced by downburst.
0.2
In the parametric study conducted by El Damatty et al. [47], the
Constant value
time of occurrence of the maximum radial velocity was found to be
VVR/VJ

proportional to the factor (DJ/VJ).

0 Table 4 summarizes the scope of previous studies focusing on
the downburst wind field and their main findings.

-0.2 3.2. Structural modeling of a transmission line under downburst

Minimum peak
loading

-0.4
0 100 200 300 400
Time (sec)
Fig. 17. Typical time history of the vertical/axial velocity [120,73].
In the case of typical transmission lines with support tower heights
of 50 m or less, dynamic response is usually neglected due to the
high aerodynamic damping of the conductors and the high natural
frequency of the tower. Treating a downburst outflow as a gust
front, Kwon and Kareem [77] presented an alternative analysis
framework called the gust-front factor approach, which is similar
3.2.1. Lattice tower and guy wires
Shehata et al. [120] developed a finite element model that sim-
ulates the tower members and the guys as a two-node, linear,
three-dimensional frame element with three translational and
three rotational degrees of freedom per node. Each tower member
was simulated by one element while each guy was modeled by five
elements. Rigid connections were assumed between the tower
members as these are physically connected using multi-bolted
connections that can transfer moments. The conductors were stud-
ied separately and then their reactions were reversed and applied
at the tower connection points. For the downburst loading, a pro-
cedure was developed to scale up a small impinging jet wind field
 H. Aboshosha et al. / Engineering Structures 112 (2016) 23–46 37

Table 4
Previous studies on downburst wind field.

Study Scope of the study Comments

Fujita [48,49] Categorized downbursts events based on field Events are categorized to downburst, microburst, and macroburst
Wolfson et al. [140] observations of actual events Gust factors correspond to downburst events differ between
Wilson et al. [138] thunderstorm types
Choi [28]
Orwig and Schroeder [105]
Gunter and Schroeder [53]
Lombardo et al. [87]
De Gaetano et al. [39]
Solari et al. [123]
Didden and Ho [41] Simulated downburst flow experimentally Simulations were based on the analogy between downburst and
Donaldson and Snedeker [42] impinging jets
Oseguera and Bowles [106] Modeled of downbursts analytically Proposed series of models that account for velocity changes in space
Vicroy [131] and time
Chay et al. [22]
Holmes and Oliver [62]
Wood et al. [141]
Li et al. [80]
Su et al. [128]
Abd-Elaal et al. [1,2]
Kim et al. [74] Conducted impinging jet simulation of downburst events Validated steady and unsteady simulation of downburst using
Kim and Hangan [73] experimental impinging jet test and field measurements
Hangan et al. [57]
Selvam and Holmes [112]
Sengupta and Sarkar [113]
Mason et al. [95–97] Simulated downbursts using the cooling source model and Elevation of the peak velocity increases with the increase of the
Vermeire et al. [129] studied terrain exposure effect and topography roughness. Hill tends to increase the magnitude of the peak velocity
Hadziabdic [55] LES simulations of downburst using impinging jet. Various Turbulence wind field was characterized and it is found that
Chay et al. [22] terrain exposures are modeled using fractal surfaces downburst turbulence is very well correlated in circumferential
Gant [51] direction
Aboshosha et al. [8]
Vermeire et al. [129]
Orf et al. [103,104]
Lin et al. [82] Assessed the effect of translating velocity of downburst Provided an estimate for the translating velocity of downburst and
Kim and Hangan [73] assessed its contribution to the peak profile
Solari et al. [124] Proposed a thunderstorm response spectrum The proposed criteria allows the calculation of the maximum
structural response
Wieringa [137] Terrain roughness Provided an estimate for the aerodynamic roughness
Blocken et al. [18]
Kwon and Kareem [77] Assessed the turbulent component of downbursts Turbulence component of downburst is essential to quantify peak
Chen and Letchford [25,26] responses of structures
Chay et al. [22,23] An expression for downburst span reduction factor
Holmes et al. [63]
Aboshosha et al. [7]
Holmes et al. [61]
Shehata et al. [120] Lattice transmission towers subjected to downbursts Downburst size, location, and angle of attack significantly affect the
Mara et al. [93] structure response

[57,120]. Shehata and El Damatty [117,118], and Shehata et al.
[119] studied a guyed tower while Darwish and El Damatty [29]
studied the behavior and the failure modes of a self-supported
tower. El Damatty and Aboshosha [45] studied both guyed and
self-supported towers to assess the behavior and the failure modes.
Ladubec et al. [78] improved upon the linear analysis by Shehata
and El Damatty [118] and studied the P–D effect in tower response
to a downburst wind field by using nonlinear space frame elements
to simulate the tower members. The study showed an increase of
20% in the peak axial forces in the chord members of the main legs,
as compared to the results from a linear analysis. Mara and Hong
[94] studied the inelastic response of TL tower subjected to down-
burst and synoptic wind fields. The study showed a dependence of
the tower capacity on the wind direction for both wind fields.
The inclusion of the turbulent component in the structural anal-
ysis may magnify the response due to the combined effects of the
fluctuating (background) component and the resonant component,
which means a lower failure capacity. Wang et al. [135] studied the
dynamic effect of a downburst on tall transmission towers. Wind
tunnel tests were conducted to determine the wind load coeffi-
cients of the transmission towers and then the towers were ana-
lyzed under downburst wind loading. Dynamic effects were
minor due to the relatively high natural frequency of the towers
compared to the event natural frequency.
Lin et al. [81] developed an aeroelastic model for a single span
of a transmission line. The guyed lattice tower was simplified to
an equivalent mast at a length scale of 1:100 while synoptic or
downburst wind loading were applied with a time scale of 1:10.
In either, case of atmospheric boundary layer or downburst wind
loadings, the structural response was generally quasi-static. Reso-
nant dynamic response was less evident with the downburst wind
than with the synoptic wind. Mara and Hong [94] studied the
inelastic response of TL tower subjected to downburst wind and
found that tower’s capacity depends on the wind direction.
3.2.2. Conductor lines
Finite Element Analysis (FEA) with the 2-D non-linear consis-
tent beam element [76] has been modified by Gerges and El
Damatty [52] to include the geometrical non-linear effects, and
was utilized by Shehata et al. [120] to accurately simulate conduc-
tor line properties and predict reactions for downburst loading.
Fig. 18 shows an element and each cable span is modeled as ten
elements. They considered the geometric non-linearity due to large
line span and relatively small line cross-section, which causes large
38 H. Aboshosha et al. / Engineering Structures 112 (2016) 23–46
Fig. 18. Consistent beam element coordinate systems and degrees of freedom (reproduced from Greges and El Damatty [52]).
Clamped End
Y
X
Springs simulating insulator flexibility
Fig. 19. Modeling of the transmission line under study (reproduced from Shehata et al. [120]).
displacements from wind loading, as was modeled by Desai et al.
[38] with a three-node iso-parametric finite element. The analysis
was performed in the horizontal and the vertical directions sepa-
rately to obtain the response under the radial and vertical down-
burst velocities, respectively. The 2-D element was acceptable for
downbursts as their associated velocities in the horizontal direc-
tion are much higher than those in the vertical direction and, thus,
decoupling between the two directions can be justified. Shehata
et al. [120] reported that modeling six conductor spans, as shown
in Fig. 19, is enough to obtain accurate results of the transmitted
forces from the conductor to an intermediate tower.
Savory et al. [111] studied the failure of transmission towers in
the cases of tornado and downburst wind loading. Conductor
forces were neglected and, as a result, failures were only associated
with tornado loading and no failure was observed with downburst
loading. On the contrary, Shehata et al. [120] predicted three differ-
ent failure modes for transmission towers with downbursts. This is
mainly due to the strong effect of a downburst upon a relatively
localized region of a transmission line. The most critical failure
mode was found to be due to the significant variation in the longi-
tudinal tensile forces from the lines upon the support towers. She-
hata et al. [120] revealed that this is the most critical failure mode
as the resultant, large, longitudinal force transmitted to the tower
cross arms leads to an out-of-plane bending moment in this region.
El Damatty and Aboshosha [45] discussed a similar mode of failure.
Aboshosha and El Damatty [4] conducted a parametric study to
check the expected values of the conductor longitudinal forces.
With a jet velocity of 40 m/s, the longitudinal force transmitted
to the tower due to the unbalanced tension was up to 60% of the
transverse force.
El Damatty et al. [47] showed that since the conductor struc-
tural response is highly nonlinear, different conductor types would
experience different responses based on the conductor characteris-
tics. They concluded that the main parameters that affect the lon-
gitudinal and transverse forces on a conductor are its material,
(Displacement and Rotation)
(Displacement Only)
(Rotation Only)
Clamped End
cross-sectional area, projected area, span, insulator length and
sag. The inclusion of the flexibility of the insulators at tower/con-
ductor connections, rather than assuming fully-hinged boundary
conditions, also has a significant effect on the natural frequencies
and mode shapes. Line sag is inversely proportional to the pre-
tensioning force. The level of the pre-tensioning force is found to
have a major influence upon the natural periods and mode shapes
of line vibrations due to the effect of sag on the conductor’s stiff-
ness [30]. In addition, environmental factors such as temperature,
highly affect the conductor sag where the actual sag at the time of
downburst occurrence may vary from reported values. High flexi-
bility and the expected nonlinear behavior of the conductors result
in a time consuming FEA procedure, since many iterations are
required to investigate different downburst configurations that
lead to the critical case of loading [4]. In view of this fact, there
is a need for a computationally efficient technique that can analyze
multi-span conductors under both vertical and horizontal HIW
loading and can take into consideration the conductor properties.
Irvine [70] derived a closed-form solution for the reactions of a
single-spanned conductor, where the loading can be fitted with a
third-degree polynomial. In addition, Yu et al. [144] considered
highly concentrated loads to derive an exact solution to calculate
the reactions for a single conductor span. Both solutions neglected
the flexibility of the insulators, which Darwish et al. [30] concluded
was important in quantifying the forces carried by the towers.
Although Winkelman’s earlier solution [139] accounted for the
insulator flexibility, the differences between tensile forces on con-
ductors in adjacent spans was neglected and, as such, the longitu-
dinal reactions that are transmitted from the conductors to the
supporting towers would be ignored. Based on the analytical solu-
tion for an elastic catenary, Ahmadi-Kashani and Bell [11] and Wei
et al. [136] developed cable elements to simulate a whole span.
Although this solution is more efficient, due to the reduction in
degrees of freedom, these elements were only developed for uni-
form wind load and are not suitable for HIW. Subsequently,
 H. Aboshosha et al. / Engineering Structures 112 (2016) 23–46 39

Table 5
Previous studies on the structural behavior of TL under downbursts.

Study Scope of the study Main finding/conclusion

Shehata et al. [119,120]
Shehata and El Damatty
[117,118]
Darwish and El Damatty [29]
Ladubec et al. [78]
Kwon and Kareem [77]
Darwish et al. [30]
El Damatty et al. [47]
El Damatty and Elawady [46]
Aboshosha and El Damatty [5,9]
Mara and Hong [94]
Aboshosha and El Damatty [6]
Investigated the behavior of TLs (tower and conductor) under
downbursts loading
Proposed a new framework for the gust factor called the Gust
Front Factor (GFF) approach
Studied the effect of turbulent component of the downburst on
the TL conductor response
Studying the critical load patterns of a downburst acting on a
TL conductor
Developed tools in the form of numerical technique and a
closed form solution for the conductor reaction due to
downburst
Studied the inelastic response of TL tower subjected to the
wind coming from various directions
Used statistical approach [35] to evaluate the span reduction
factor for the conductor reactions using downburst parameters
obtained from LES [7]
Internal forces in the tower members depends on the relative
location of the downburst to the tower
The study resulted in a new gust factor approach that works for
synoptic and non-synoptic wind
This new GFF converges to the Davenport’s GF when used for
synoptic wind
A 20% increase in the maximum wire forces is due to the
turbulence inclusion and negligible dynamic effect is noticed are
noticed
Three load cases were proposed representing the worst case
scenario of a downburst load acting on a TL
The technique is found to be 180 times faster than FEM, while the
closed form solution allows for a direct prediction of the
maximum conductor reactions
Yield and maximum tower’s capacities vary with the wind
direction
Span reduction factor was found higher than the factor for normal
wind. The obtained factors well-agrees with the field
measurements Holmes et al. [61]

Aboshosha and El Damatty [5] developed a numerical technique to
analyze multi-spanned conductors under HIW. This solution is the
first semi-closed form solution for a multi-spanned conductor sys-
tem under non-uniform loading in both the vertical and horizontal
directions, where insulator rigidity is considered. The technique
was approximately 185 times faster than FEA. Later, Aboshosha
and El Damatty [9] proposed an engineering method suitable for
practitioner engineers to evaluate the maximum conductor reac-
tions due to downburst.
Darwish et al. [30] modified the two-dimensional nonlinear
finite element model of the transmission lines developed by She-
hata et al. [120] to study the dynamic characteristics of the conduc-
tors under turbulent downburst loading. The turbulence
component was extracted from full-scale data and then added to
the mean component of the downburst wind field developed by
Kim and Hangan [73]. Large deformations and the pre-tension
loading were modeled. The aerodynamic damping was determined
with an expression given by Davenport [32], which was derived for
synoptic winds and assumed an average velocity in the calculation
of the aerodynamic damping in order to overcome the localized
nature of downbursts as the wind velocity varies with time and
also spatially along the conductor length. The study concluded that
the resonant component due to the turbulence is negligible due to
the large aerodynamic damping, the dynamic response is mainly
due to the background component of wind velocity fluctuations
and turbulence accounted for almost 20% of the conductor deflec-
tion and reactions. Aboshosha et al. [10] studied the dynamic and
quasi-static response of single span and multi spanned lines under
both the synoptic and downburst loads. For multi-spanned lines,
the study showed that the resonant component contribution com-
pared to the peak reactions is in order of 6% assuming different
wind intensities. However, for single span conductor, the resonant
component contribution to the peak response was higher (in order
of 16%) at low speeds.
Holmes et al. [63] isolated the turbulent component of the
downburst velocity and produced a peak load reduction factor
for the spatial variation along the longitudinal direction from a
set of impinging jet CFD data [74]. The span reduction factor was
found to range between 1 and 0.8 for a separation of 720 m with
downburst wind loading. A similar result was found by Aboshosha
and El Damatty [6] using downburst wind field obtained from LES.
The effect of terrain roughness on SRF remains uncertain.
A summary of previous studies on the behavior of TLs (includ-
ing either tower, conductor or both) is provided in Table 5. The
table shows the scope of the studies and their main findings.
3.3. Behavior of transmission lines subjected to downbursts and
corresponding critical load cases
A large research has been conducting in the University of Wes-
tern Ontario regarding the behavior of the transmission line struc-
tures subjected to downburst wind events. Shehata et al. [120],
Shehata and El Damatty [117], Shehata et al. [119], and Darwish
and El Damatty [29] adopted quasi-static analysis procedures (rel-
evant for a mean component that varies slowly with time) with the
downburst wind field from Hangan and Kim [56] as an input. She-
hata et al. [120] reported that the effective period for downburst
wind speed variation ranged between 20 and 22 s while the vibra-
tion frequencies for the transmission tower and conductor were
0.58 s and 8.25 s, respectively. As such, no strong dynamic effect
was evident.
Using the structural analysis model developed by Shehata et al.
[120], Shehata and El Damatty [117] conducted a parametric study
to investigate the critical downburst configuration by varying the
jet diameter (DJ) and the location of the downburst center relative
to the tower (R). A guyed transmission tower, which collapsed in
Manitoba in 1996 due to a downburst event, was used to perform
this parametric study. The critical downburst parameters, in terms
of the size of the event and its location relative to the tower, lead-
ing to maximum forces in the tower members, were identified. The
study revealed that the critical downburst parameters vary based
on the type and location of the members. Unsurprisingly, the chord
members, diagonal members and cross arm members had different
critical downburst configurations. Shehata and El Damatty [118]
extended their numerical scheme by including a failure model
for the tower members, which was used to study the progressive
collapse of the tower failure. An optimization routine was imple-
mented by Shehata et al. [119] to predict the critical downburst
parameters and the corresponding forces on a transmission line
by an automated procedure. This finite element-optimization tech-
nique was validated by comparing the maximum forces and critical
downburst parameters in a number of tower members to the cor-
responding values obtained from an extensive parametric study.
40 H. Aboshosha et al. / Engineering Structures 112 (2016) 23–46
Fig. 20. Field evidence of lattice tower failure at cross arms (modified from
Manitoba Hydro [92]).
Shehata and El Damatty [118] conduct a failure analysis for a
guyed tower subjected to a microburst where they compared the
internal forces developing in each of the tower member to its
capacity calculated using the ASCE (1992) standards. In their study,
they considered a microburst jet velocity of VJ = 70 m/s. The analy-
sis showed that a failure region was generated at the cross arm and
the guys zones. This is justified since the downburst longitudinal
reaction affected the cross arm zone; this load configuration is
not considered in the load cases recommended by the standards.
The authors also indicated a failure mechanism when the virtual
Fig. 21. (a) VRD distribution along tower height with h = 90° and 0°, R/DJ = 1.3 and
DJ = 500 m and (b) VRD distribution over six line spans with h = 0°, R/DJ = 1.3 and
DJ = 500 m [46].
line connecting the center of the downburst and the tower of inter-
est is perpendicular on the line. Under this load case, the conduc-
tors will experience significant transverse reaction. This force
will cause a large shear force in the guys area. This load case is sim-
ilar to the normal wind load case required by the codes. Thus, the
main difference between the downburst under this load configura-
tion and the synoptic wind load would be the difference in both
wind speeds. Shehata and El Damatty [118] reported that one of
the major failure mechanisms for this type of structures occurs
when the virtual line connecting the downburst center to the cen-
ter of the transmission line is parallel to the line. This causes a zero
reaction contribution from the conductors. This load case caused
failures initiation in the chord members as well as the guys zone.
This was then confirmed by the late study conducted by Darwish
and El Damatty [29] who studied the behavior of self-supported
tower, Aboshosha and El Damatty [3] who studied one guyed
tower and another self-supported tower, and El Damatty and Ela-
wady [46] who studied the behavior of three guyed and three
self-supported towers. Fig. 20 shows a failure of cross arms docu-
mented by an electric energy provider.
Darwish et al. [30] investigated the importance of the transla-
tion velocity of the downburst, using three critical downburst con-
figurations. It was found that the tower failed at the same radial
velocity regardless of the contribution of the translational compo-
nent to this velocity. Therefore, Darwish et al. [30] reported that
there is no need to consider the translation velocity of the down-
burst and it is sufficient to vary the location of the downburst in
space for a large number of separate stationary events. In addition,
Darwish et al. [30] found almost no variation in the dynamic char-
acteristics of the conductors under the different loading configura-
tions. Darwish and El Damatty [29], El Damatty and Aboshosha
Fig. 22. (a) VRD distribution along tower height with h = 30°, R/DJ = 1.6 and
DJ = 500 m and (b) VRD distribution over six line spans with h = 30°, R/DJ = 1.6 and
DJ = 500 m (modified from El Damatty and Elawady [46]).
 H. Aboshosha et al. / Engineering Structures 112 (2016) 23–46 41

Table A.1
Pressures at 30 m elevation qpn(30) normalized by the peak pressure at 10 m for open terrain exposure.

Col. No. 1 2 3 4 5 6 7
Codenexposure qpn(30)ASCE qpn(30) AS/NZS CRp(30) (AS/NZS) qpn(30) BS CRp(30) (BS) qpn(30)IEC CRp(30) (IEC)
Sea side 1.43 1.43 1 1.47 1.03 1.30 0.91
Open 1.26 1.33 1.05 1.31 1.04 1.22 0.97
Suburban 0.98 0.89 0.91 1.11 1.13 1.02 1.04

qpn: normalized pressure calculated using Eq. (4).

Code ratio for pressure CRp using ASCE 74 [12] as a datum for normalization.
CRp(30) (AS/NZS) = Column (2)/Column (1).
CRp(30) (BS) = Column (4)/Column (1).
CRp(30) (IEC) = Column (6)/Column (1).

Table A.2
Normalized forces transferred from a 20 m elevated conductor, Fcn (20).

Col. no. 1 2 3 4 5 6 7
Coden L Fcn (20) ASCE Fcn (20) AS/NZS CRFc(20) (AS/NZS) Fcn (20) BS CRFc(20) (BS) Fcn (20) IEC CRFc(20) (IEC)
Sea side exposure
200 0.90 1.02 1.12 1.24 1.37 1.12 1.23
300 0.87 0.94 1.08 1.20 1.39 1.07 1.23
400 0.84 0.88 1.05 1.17 1.39 1.04 1.23
500 0.82 0.85 1.03 1.15 1.39 1.01 1.22
600 0.81 0.83 1.02 1.13 1.4 0.99 1.21
700 0.80 0.82 1.02 1.12 1.4 0.97 1.21
800 0.79 0.81 1.02 1.11 1.4 0.95 1.20
Open exposure
200 0.84 0.90 1.07 1.03 1.23 1.08 1.29
300 0.80 0.82 1.03 0.99 1.24 1.03 1.29
400 0.77 0.78 1.01 0.96 1.24 1.00 1.29
500 0.75 0.75 0.99 0.94 1.24 0.98 1.29
600 0.74 0.73 0.99 0.92 1.24 0.95 1.29
700 0.73 0.72 0.99 0.91 1.24 0.94 1.28
800 0.72 0.72 0.99 0.90 1.24 0.92 1.28
Suburban exposure
200 0.70 0.62 0.88 0.77 1.08 0.98 1.39
300 0.67 0.57 0.85 0.73 1.09 0.94 1.40
400 0.64 0.54 0.84 0.70 1.09 0.91 1.41
500 0.62 0.52 0.83 0.68 1.09 0.89 1.42
600 0.61 0.51 0.83 0.66 1.08 0.87 1.42
700 0.60 0.50 0.84 0.64 1.08 0.85 1.42
800 0.59 0.50 0.84 0.63 1.07 0.84 1.42

Fcn: normalized conductor force calculated using Eq. (7).

Code ratio for conductor forces CRFc using ASCE 74 [12] as a datum for normalization.
CRFc(20) (AS/NZS) = Column (2)/Column (1).
CRFc(20) (BS) = Column (4)/Column (1).
CRFc(20) (IEC) = Column (6)/Column (1).

[45], Aboshosha and El Damatty [4] and El Damatty et al. [47] per- (2) At h = 90°, R/DJ = 1.30 and DJ = 500 m, the same wind load
formed additional parametric studies to investigate the critical profile is applied on the tower face parallel to the major axis
downburst configurations. The studies agreed that changing the of the line, as in the previous case, while no wind load is
location of the downburst (R/DJ and h) has a stronger effect on applied on the conductors.
the value of the axial force in all tower members when compared (3) At h = 30°, R/DJ = 1.60 and DJ = 500 m, the vertical profiles of
to changing the downburst size (DJ) which has a minor effect. In velocity in the direction perpendicular and parallel to the
addition, the ratio of the span (L) to the jet diameter, DJ, plays a line can be approximated as having uniform values of 0.8
vital role in the existence of these critical cases. El Damatty et al. VJ and 0.47 VJ, respectively, to stipulate the tower wind load.
[47] proposed three critical load cases to calculate static forces act- A difference in downburst radial velocity distribution at the
ing upon towers and conductors under downburst loading, where two adjacent conductors results in a nonlinear longitudinal
the loads should be applied on both the tower of interest as well force acting on the tower, Fig. 22. The transverse reaction
as adjacent conductor spans from each side. The three cases are of the wires can be estimated assuming a uniform velocity
shown in Figs. 13 and 14 and explained below: of 0.65 VJ.
Comparison between load cases described above and load cases
(1) At h = 0°, R/DJ = 1.30 and DJ = 500 m, the downburst outflow
for synoptic wind shows two main differences (1) Downburst has
is modeled as impacting perpendicular to the major axis of
an oblique case of loading (h = 30°) causing differential conductor
the transmission line. The radial velocity applied on the sup-
tension on the sides of the tower. Synoptic wind does not generate
port tower can be considered to be equal to 1.1 VJ. To
differential conductor. (2) Downburst load cases in the transverse
approximate the conductor’s transverse reaction, an equiva-
and longitudinal directions (h = 0° and 90°) are close to synoptic
lent uniform velocity distribution with a magnitude of 1.06
wind load cases, but the load magnitude is mostly higher. Load
VJ can be used, as shown in Fig. 21.
magnitude for downburst requires further research.
42 H. Aboshosha et al. / Engineering Structures 112 (2016) 23–46

Table A.3
Normalized forces transferred from a 40 m elevated conductor, Fcn (40).

Col. no. 1 2 3 4 5 6 7
Coden L Fcn (40) ASCE Fcn (40) (AS/NZS) CRFc(40) (AS/NZS) Fcn (40) BS CRFc(40) (BS) Fcn (40) IEC CRFc(40) (IEC)
Sea side exposure
200 1.00 1.11 1.11 1.42 1.42 1.17 1.17
300 0.96 1.02 1.06 1.38 1.43 1.12 1.17
400 0.93 0.97 1.03 1.34 1.43 1.09 1.16
500 0.92 0.93 1.01 1.32 1.44 1.06 1.15
600 0.90 0.91 1.01 1.30 1.44 1.04 1.15
700 0.89 0.90 1.01 1.28 1.44 1.02 1.14
800 0.88 0.89 1.01 1.27 1.43 1.00 1.13
Open exposure
200 0.95 1.07 1.13 1.21 1.28 1.14 1.21
300 0.90 0.99 1.09 1.16 1.28 1.10 1.21
400 0.88 0.93 1.06 1.13 1.29 1.06 1.20
500 0.86 0.90 1.04 1.11 1.29 1.03 1.20
600 0.84 0.88 1.04 1.09 1.29 1.01 1.20
700 0.83 0.87 1.04 1.07 1.28 0.99 1.19
800 0.82 0.86 1.04 1.05 1.28 0.98 1.19
Suburban exposure
200 0.81 0.70 0.86 0.94 1.15 1.05 1.29
300 0.77 0.65 0.84 0.89 1.15 1.00 1.30
400 0.74 0.61 0.82 0.86 1.14 0.97 1.30
500 0.73 0.59 0.81 0.83 1.14 0.95 1.30
600 0.71 0.58 0.81 0.81 1.14 0.93 1.30
700 0.70 0.57 0.81 0.79 1.13 0.91 1.30
800 0.69 0.56 0.82 0.77 1.12 0.90 1.30

Fcn: normalized conductor force calculated using Eq. (7).

Code ratio for conductor forces CRFc using ASCE 74 [12] as a datum for normalization.
CRFc(40) (AS/NZS) = Column (2)/Column (1).
CRFc(40) (BS) = Column (4)/Column (1).
CRFc(40) (IEC) = Column (6)/Column (1).

4. Conclusions employed in each code in addition to the various methods used

This section summarizes codes guidelines related to response of
TLs under synoptic and non-synoptic winds. Gaps exist in the
codes related to the response of TLs under synoptic wind and pos-
sible ways of covering these gaps are summarized below:
 The tower frequency and the conductors’ aerodynamic damp-
ing, below which the resonant component needs to be
accounted for, require further investigations.
 The assumption of the minor dynamic effect of the conductors
due to the high aerodynamic damping is true for high wind
speeds. This assumption requires further investigations under
low wind speeds. Vibration of the conductors under low wind
speed might lead to fatigue problems in the cross arms.
 Most of TL codes are valid for towers with an overall height less
than 60 m. For taller towers, the dynamic effect of the towers
may be accounted for using the gust factor approach while con-
sidering the resonant component. Accounting for the resonant
component by the gust factor approach can be done using a fre-
quency domain analysis, which is typical applicable for linear
structures. Research needs to be done to modify the frequency
domain analysis in order to account for the nonlinear behavior
expected for tall towers.
 Design pressure obtained by the AS/NZS [13], BS [20] and IEC [69]
codes and normalized by values obtained from ASCE 74 [12]
shows that the discrepancy in the code ratio CRp is ranging
between 0.91 and 1.13. The minimum value of CRp is noticed in
case of using AS/NZS [13] and IEC [69] approaches for suburban
and sea side exposures, respectively. The maximum value of
CRp is reached by using BS [20] approach for Suburban exposure.
This discrepancy is due to the different velocity profiles
in the codes to covert between the wind speeds averaged over
10 min and 3 s.
 The differences between the four codes in terms of design pres-
sure, gust response factors for the conductors, GC, and the tower,
GT, and expressions for the tower force FT lead to larger differ-
ences between the forces estimated by those codes. The dis-
crepancy of the code ratio for conductor forces CRFc is found
to be ranging between 0.81 and 1.44. For tower forces code ratio
CRFt, a discrepancy range of 0.68 and 1.85 is noticed. Assessment
of the accuracy of the estimated forces based on various codes
can be conducted based on wind tunnel testing or full scale
measurements of real towers.
 The ASCE 74 [12] code uses 3-s gust velocity as the reference
wind speed, however, when the gust response of the structure
is calculated, the code uses a formulation that is based on
10 min-averaged wind speed. The code introduces a conversion
factor Kv based on Durst curve, which is different than the
approaches adopted by the BS [20] and the IEC [69]. This con-
version leads to differences in the order of 10–15%.
 The ASCE 74 [12] does not account for the increase of the tower
force for skewed angle of attack up to 45°. This appeared to
underestimate the tower forces as found from a previous wind
tunnel study.
 Design codes separate between the gust factor for the conductor
and that for the tower. For the cases of no dynamic effect, this
may overestimate the forces transmitted at the tower base as
such an assumption means full correlation between the conduc-
tor and the tower forces. A reduction factor can be applied to the
conductor forces in this case. For the cases of potential dynamic
effects, the dynamic properties of the system, represented in
the modal frequencies, mode shapes and modal aerodynamic
 H. Aboshosha et al. / Engineering Structures 112 (2016) 23–46 43

Table A.4
Normalized forces Ftn acting on a tower panel for angle of attack 0°.

Col. no. 1 2 3 4 5 6 7
Code n u Ftn (0°) ASCE Ftn (0°) AS/NZS CRFt(0°) (AS/NZS) Ftn (0°) BS CRFt(0°) (BS) Ftn (0°) IEC CRFt(0°) (IEC)
Sea side exposure
0.0 3.46 4.39 1.27 4.64 1.34 3.84 1.11
0.1 3.10 3.84 1.24 4.05 1.31 3.36 1.08
0.2 2.65 3.30 1.25 3.49 1.32 2.89 1.09
0.3 2.20 2.86 1.30 3.02 1.37 2.51 1.14
0.4 1.75 2.52 1.44 2.65 1.52 2.20 1.26
0.5 1.56 2.26 1.45 2.38 1.53 1.98 1.27
0.6 1.56 2.09 1.34 2.21 1.42 1.83 1.17
0.7 1.55 2.02 1.35 2.13 1.37 1.76 1.14
0.8 1.61 2.03 1.26 2.15 1.33 1.78 1.11
0.9 1.67 2.14 1.28 2.26 1.35 1.87 1.12
1.0 1.73 2.34 1.35 2.47 1.43 2.05 1.18
Open exposure
0.0 3.67 3.94 1.07 3.94 1.07 4.16 1.13
0.1 3.29 3.44 1.05 3.44 1.05 3.63 1.11
0.2 2.81 2.96 1.06 2.96 1.06 3.13 1.11
0.3 2.33 2.57 1.10 2.57 1.11 2.71 1.16
0.4 1.85 2.26 1.22 2.26 1.22 2.38 1.28
0.5 1.65 2.03 1.23 2.03 1.23 2.14 1.29
0.6 1.65 1.88 1.14 1.88 1.14 1.98 1.20
0.7 1.64 1.81 1.10 1.81 1.11 1.91 1.16
0.8 1.71 1.82 1.07 1.82 1.07 1.92 1.13
0.9 1.77 1.92 1.08 1.92 1.08 2.03 1.14
1.0 1.84 2.10 1.14 2.10 1.14 2.22 1.21
Suburban exposure
0.0 4.40 3.37 0.77 3.08 0.70 4.91 1.12
0.1 3.94 2.95 0.75 2.69 0.68 4.28 1.09
0.2 3.37 2.54 0.75 2.32 0.69 3.69 1.10
0.3 2.79 2.20 0.79 2.01 0.72 3.20 1.15
0.4 2.22 1.93 0.87 1.76 0.79 2.81 1.26
0.5 1.98 1.73 0.88 1.58 0.80 2.52 1.27
0.6 1.98 1.61 0.81 1.47 0.74 2.34 1.18
0.7 1.97 1.55 0.79 1.41 0.72 2.25 1.14
0.8 2.05 1.56 0.76 1.43 0.70 2.27 1.11
0.9 2.12 1.64 0.77 1.50 0.71 2.39 1.13
1.0 2.20 1.80 0.82 1.64 0.75 2.61 1.19

Fct: normalized force acting on tower calculated using Eq. (8).

u: Solidity ratio.
Code ratio for tower forces CRFt using ASCE 74 [12] as a datum for normalization.
CRFt(0°) (AS/NZS) = Column (2)/Column (1).
CRFt(0°) (BS) = Column (4)/Column (1).
CRFt(0°) (IEC) = Column (6)/Column (1).

damping of the combined conductor-tower system have to be
obtained, while accurately modeling the connection between
the tower and the conductor. Then, these dynamic properties
can be used through the gust factor approach proposed by Daven-
port [32] which is adopted by design codes for dynamically sen-
sitive structures. An expression for the aerodynamic damping of
the combined system is also required.
Current guidelines related to response of TLs under non-synoptic
winds are summarized below:
 In most of the available codes, no clear definition is given for the
wind field associated with the downburst events.
o Only AS/NZS:7000 [13] has defined the downburst wind
field as a cold air column that fall vertically from a great
height and strike the ground causing the wind draft to
radiate from the impact site.
o ASCE 74-2010 considers the downburst wind speed simi-
lar to the F2 tornado scale.
 Regarding the velocity profile along the height, AS/NZS:7000
[13] provides the vertical distribution of the wind speed. How-
ever, due to the localized nature of downbursts, the velocity
profile will depend on the downburst size and its relative loca-
tion to the tower. This will lead to a number of critical load
cases, which should be incorporated in the design codes.
 Effect of terrain and topography on the downburst wind field
needs further research to be considered in the design codes.
 Along the line spans, and due to the localized nature of down-
bursts, a nonsymmetrical distribution of the wind pressure
along cables’ length can happen. This can lead to an unbalanced
longitudinal force causing an out-of-plane moment on the
tower cross arms. Such a load case needs to be considered in
the downburst analysis of the towers and thus in the codes of
practice.
 Span reduction factor is only given in the AS/NZS:7000 [13] for
the case of downburst event. However, the factor provided in
this code does not account for the terrain effect or the elevation
of the conductor above the ground.
 For flexible tall towers, the dynamic effect should be considered
through the gust factor approach proposed by Kwon and Kar-
eem [77]. However, this approach requires further research.
Acknowledgment
The authors would like to thank the CEATI International (www.
ceati.com) and its members from the Transmission Overhead Line
& Extreme Events Mitigation program – 28 leading transmission
electrical utilities form all over the globe for their financial support
44 H. Aboshosha et al. / Engineering Structures 112 (2016) 23–46

Table A.5
Normalized forces Ftn acting on a square tower panel for angle of attack 45°.

Col. no. 1 2 3 4 5 6 7
Code n u Ftn (45°) ASCE Ftn (45°) AS/NZS CRFt(45°) (AS/NZS) Ftn (45°) BS CRFt(45°) (BS) Ftn (45°) IEC CRFt(45°) (IEC)
Sea side exposure
0.0 3.46 4.88 1.41 5.56 1.61 4.61 1.33
0.1 3.10 4.26 1.37 4.86 1.57 4.03 1.30
0.2 2.65 3.67 1.38 4.19 1.58 3.47 1.31
0.3 2.20 3.34 1.52 3.63 1.65 3.01 1.37
0.4 1.75 3.07 1.75 3.19 1.82 2.64 1.51
0.5 1.56 2.88 1.85 2.86 1.84 2.37 1.52
0.6 1.56 2.55 1.64 2.65 1.70 2.20 1.41
0.7 1.55 2.35 1.52 2.55 1.65 2.12 1.37
0.8 1.61 2.26 1.40 2.58 1.60 2.14 1.33
0.9 1.67 2.38 1.42 2.71 1.62 2.25 1.35
1.0 1.73 2.60 1.50 2.97 1.71 2.46 1.42
Open exposure
0.0 3.67 4.37 1.19 4.73 1.29 4.99 1.36
0.1 3.29 3.82 1.16 4.13 1.26 4.36 1.33
0.2 2.81 3.29 1.17 3.56 1.27 3.75 1.34
0.3 2.33 2.99 1.28 3.08 1.32 3.25 1.40
0.4 1.85 2.75 1.48 2.71 1.46 2.86 1.54
0.5 1.65 2.58 1.56 2.43 1.47 2.56 1.55
0.6 1.65 2.29 1.39 2.25 1.36 2.38 1.44
0.7 1.64 2.11 1.28 2.17 1.32 2.29 1.39
0.8 1.71 2.02 1.19 2.19 1.28 2.31 1.35
0.9 1.77 2.13 1.20 2.31 1.30 2.43 1.37
1.0 1.84 2.33 1.27 2.52 1.37 2.66 1.45
Suburban exposure
0.0 4.40 3.74 0.85 3.70 0.84 5.89 1.34
0.1 3.94 3.27 0.83 3.23 0.82 5.14 1.31
0.2 3.37 2.82 0.84 2.78 0.83 4.43 1.32
0.3 2.79 2.56 0.92 2.41 0.86 3.84 1.37
0.4 2.22 2.36 1.06 2.12 0.95 3.37 1.52
0.5 1.98 2.21 1.12 1.90 0.96 3.03 1.53
0.6 1.98 1.96 0.99 1.76 0.89 2.80 1.42
0.7 1.97 1.80 0.92 1.70 0.86 2.70 1.37
0.8 2.05 1.73 0.85 1.71 0.84 2.73 1.33
0.9 2.12 1.83 0.86 1.80 0.85 2.87 1.35
1.0 2.20 2.00 0.91 1.97 0.90 3.14 1.43

Fct: normalized force acting on tower calculated using Eq. (8).

u: Solidity ratio.
Code ratio for tower forces CRFt using ASCE 74 [12] as a datum for normalization.
CRFt(45°) (AS/NZS) = Column (2)/Column (1).
CRFt(45°) (BS) = Column (4)/Column (1).
CRFt(45°) (IEC) = Column (6)/Column (1).

of this research. Also, the authors would like to thank the anon-
ymous reviewers for their insightful comments which enhanced
the quality of the paper.
Appendix A
Results for the synoptic winds obtained using various codes.
See Tables A.1–A.5.
References
[1] Abd-Elaal E, Mills JE, Ma X. An analytical model for simulating steady state
flows of downburst. J Wind Eng Ind Aerodyn 2013;115:53–64.
[2] Abd-Elaal E, Mills JE, Ma X. Empirical models for predicting unsteady-state
downburst wind speeds. J Wind Eng Ind Aerodyn 2014;129:49–63.
[3] Aboshosha H, El Damatty A. Capacity of electrical transmission towers under
downburst loading. In: ASEA-SEC-1, Perth, November 28–Decmber 2; 2012.
[4] Aboshosha H, El Damatty A. Downburst induced forces on the conductors of
electric transmission lines and the corresponding vulnerability of towers
failure. In: CSCE 2013 conference, 29 May–1 June, Montreal, Canada; 2013.
[5] Aboshosha H, El Damatty A. Effective technique for the reactions of
transmission line conductors under high intensity winds. Wind Struct
2014;18(3):235–52.
[6] Aboshosha H, El Damatty A. Span reduction factor of transmission line
conductors under downburst winds. J Wind Eng 2014;11(1):13–22.
[7] Aboshosha H, Bitsuamlak G, El Damatty A. Turbulence characterization of
downbursts using LES. J Wind Eng Ind Aerodyn 2015;136:44–61.
[8] Aboshosha H, Bitsumalak G, El Damatty A. LES of lower ABL flow in the built-
environment using roughness modeled by fractal surfaces. J Sust Cities Soc
2015;19:46–60.
[9] Aboshosha H, El Damatty A. Engineering method for estimating the reactions
of transmission line conductors under downburst winds. J Eng Struct
2015;142:198–216.
[10] Aboshosha H, El Damatty A. Dynamic effect for transmission line conductors
under downburst and synoptic winds. Wind Struct Int J 2015;21(2):241–72.
[11] Ahmadi-Kashani K, Bell AJ. The analysis of cables subject to uniformly
distributed loads. Eng Struct 1988;10(3):174–84.
[12] American Society of Civil Engineers (ASCE). Guidelines for electrical
transmission line structural loading. ASCE manuals and reports on
engineering practice, No. 74, New York, NY, USA; 2010.
[13] Australian Standard (AS) 1170.2. Australian standard for minimum design
loads on structures – Part 2: wind loads, Standard Australia, North Sydney,
Australia; 2002.
[14] Australian Standard/New Zealand Standard (AS/NZS) 7000. Overhead line
design detailed procedures. Standards Australia Limited/Standards New
Zealand, North Sydney, Australia; 2010.
[15] Bachmann et al. Vibration problems in structures. Basel: Birkhauser; 1995.
[16] Battista R, Rodrigues R, Pfeil M. Dynamic behavior and stability of
transmission line towers under wind forces. J Wind Eng Ind Aerodyn
2003;91:1051–67.
[17] Behncke RH, Ho TCE. Review of span and gust factors for transmission line
design. In: Proceedings of the 2009 electrical transmission and substation
structures conference, 8–12 November, Fort Worth, TX, USA, published in
‘‘Electrical transmission and substation structures 2009: Technology for the
next generation”, 2010, edited by M.W. Vogt, sponsored by Electrical
Transmission Structures Committee of the Structural Engineering Institute
of ASCE; 2009. p. 209–220.
[18] Blocken B, Stathopoulos T, Carmeliet J. CFD simulation of the atmospheric
boundary layer: wall function problems. Atmos Environ 2007;41(2):238–52.
 H. Aboshosha et al. / Engineering Structures 112 (2016) 23–46
[19] British Standard (BS) 8100. British standard for lattice towers and masts: Part
1 – code of practice for loading, London, UK; 1986.
[20] British Standard (BS) 50341. Overhead electrical lines exceeding AC
45 kV: Part 1 – General requirements – Common specifications, London,
UK; 2001.
[21] Boddy D, Rice J. Impact of alternative galloping criteria on transmission line
design. In: Proc ASCE electrical transmission and substation structures 2009,
ASCE, Reston, VA; 2009.
[22] Chay MT, Albermani F, Wilson R. Numerical and analytical simulation of
downburst wind loads. Eng Struct 2006;28(2):240–54.
[23] Chay MT, Wilson R, Albermani F. Gust occurrence in simulated non-
stationary winds. J Wind Eng Ind Aerodyn 2008;96(10):2161–72.
[24] Chabart O, Lilien JL. Galloping of electrical lines in wind tunnel facilities. J
Wind Eng Ind Aerodyn 1998;74–76:967–76.
[25] Chen L, Letchford CW. A deterministic–stochastic hybrid model of
downbursts and its impact on a cantilevered structure. Eng Struct 2004;26
(5):619–29.
[26] Chen L, Letchford CW. Parametric study on the along-wind response of the
CAARC building to downbursts in the time domain. J Wind Eng Ind Aerodyn
2004;92(9):703–24.
[27] Cook NJ. The designer’s guide to wind loading of building structures Part 1.
Building research establishment, London, UK; 1985.
[28] Choi ECC. Wind characteristics of tropical thunderstorms. J Wind Eng Ind
Aerodyn 2000;84:215–26.
[29] Darwish MM, El Damatty AA. Behavior of self supported transmission line
towers under stationary downburst loading. Wind Struct 2011;14(5):481–4.
[30] Darwish M, El Damatty A, Hangan H. Dynamic characteristics of transmission
line conductors and behaviour under turbulent downburst loading. Wind
Struct 2010;13(4):327–46.
[31] Davenport AG. Buffeting of a suspension bridge by storm winds. ASCE J Struct
Div 1962;88(3):233–64.
[32] Davenport AG. Gust loading factors. ASCE J Struct Div 1967;93(3):11–34.
[33] Davenport AG. Gust response factors for transmission line loading. In:
Proceedings. 5th international conference on wind engineering, Fort Collins,
Colorado. Pergamon Press; 1979. p. 899–909.
[34] Davenport AG. Gust response factors for transmission line loading. In:
Cermak JE, editor. Proc of the fifth international conference on wind
engineering, Fort Collins, CO, July 1979. New York (NY): Pergamon Press;
1980.
[35] Davenport AG. How can we simplify and generalize wind loads?. In:
Presented at the third asia-pacific symposium on wind engineering.
December 13–D15. Hong-Kong, Keynote Lecture; 1993.
[36] Davison et al. Alcoa overhead engineering data no. 4. Alcoa.
[37] Dempsey D, White H. Winds wreak havoc on lines. Transm Distrib World
1996;48(6):32–7.
[38] Desai YM, Yu P, Popplewell N, Shah AH. Finite element modelling of
transmission line galloping. Comput Struct 1995;57(3):407–20.
[39] De Gaetano P, Repetto MP, Repetto T, Solari G. Separation and classification of
extreme wind events from anemometric records. J Wind Eng Ind Aerodyn
2014;126:132–43.
[40] Den Hartog JP. Transmission line vibration due to sleet. Trans AIEE
1932;51:1074–86.
[41] Didden N, Ho CM. Unsteady separation in a boundary layer produced by an
impinging jet. J Fluid Mech 1985;160:235–56.
[42] Donaldson CD, Snedeker RS. A study of free jet impingement, Part 1. Mean
properties of free and impinging jets. J Fluid Mech 1971;45:281–319.
[43] Durst CS. Wind speeds over short periods of time. Meteorol Mag
1960;89:181–6.
[44] Dyke PV, Laneville A. Galloping of a single conductor covered with a D-section
on a high-voltage overhead test line. J Wind Eng Ind Aerodyn 2008;96:
1141–51.
[45] El Damatty A, Aboshosha H. Capacity of electrical transmission towers under
downburst loading. In: Proceedings of the 1st Australasia and South-East Asia
structural engineering and construction conference, 28 November–2
December, Perth, Australia; 2012. p. 317–22.
[46] El Damatty A, Elawady A. Critical load cases for lattice transmission line
structures under downbursts. In: The 14th international conference on wind
engineering, ICW14, June 21–25; 2015.
[47] El Damatty A, Hamada A, Elawady A. Development of critical load cases
simulating the effect of downbursts and tornadoes on transmission line
structures. In: Proceedings of the 8th Asia-Pacific conference on wind
engineering, 10–14 December, Chennai, India; 2013.
[48] Fujita TT. ‘‘The downburst: microburst and macroburst”, SMRP research
paper 210. USA: University of Chicago; 1985.
[49] Fujita TT. Downbursts: meteorological features and wind field characteristics.
J Wind Eng Ind Aerodyn 1990;36:75–86.
[50] Gast KD, Schroeder JL. Supercell rear-flank downdraft as sampled in the 2002
thunderstorm outflow experiment. In: Proceedings of the 11th international
conference on wind engineering, Lubbock, Tex; 2003. p. 2233–40.
[51] Gant SE. Reliability issues of LES-related approaches in an industrial context.
Flow Turbul Combust 2009;84:325–35.
[52] Gerges RR, El-Damatty AA. Large displacement analysis of curved beams. In:
Proceedings of the Canadian society of civil engineering conference, 3–5 June,
Montreal, Canada, ST 100; 2002.
[53] Gunter WS, Schroeder JL. ‘‘High-resolution full-scale measurements of
thunderstorm outflow” winds. J Wind Eng Ind Aerodyn 2015;138:13–26.
45
[54] Gurunga CB, Yamaguchia H, Yukinob T. Identification and characterization of
galloping of Tsuruga test line based on multi-channel modal analysis of field
data. J Wind Eng Ind Aerodyn 2003;91:903–24.
[55] Hadžiabdić M. LES, RANS and combined simulation of impinging flows and
heat transfer. Ph.D thesis. University of Sarajevo; 2005. 185p.
[56] Hangan H, Kim J. Numerical simulations of impinging jets with application to
downbursts. J Wind Eng Ind Aerodyn 2007;95(4):279–98.
[57] Hangan H, Roberts D, Xu Z, Kim J. Downburst simulation. Experimental and
numerical challenges. In: Proceedings of the 11th international conference on
wind engineering, 1–5 June, Lubbock, TX, USA; 2003.
[58] Harikrishna P, Shanmugasundaram J, Gomathinayagam S, Lakshmanan N.
Analytical and experimental studies on the gust response of a 52 m tall steel
lattice tower under wind loading. Comput Struct 1999;70:149–60.
[59] Havard DG, Pohlman JC. Control of galloping conductors by detuning Paper
22–05. Paris (France): CIGRE; 1980.
[60] Hjelmfelt MR. Structure and life cycle of microburst outflows observed in
Colorado. J Appl Meteorol 1988;27:900–27.
[61] Holmes JD. Recent developments in the specification of wind loads on
transmission lines. J Wind Eng 2008;5(1):8–18.
[62] Holmes JD, Oliver SE. An emperical model of a downburst. Eng Struct
2000;22:1167–72.
[63] Holmes JD, Hangan HM, Schroeder JL, Letchford CW, Orwig KD. A forensic
study of the Lubbock-Reese downdraft of 2002. Wind Struct 2008;11
(2):137–52.
[64] Holmes JD. Along-wind response of lattice towers: Part I – derivation of
expressions for gust response factors. Eng Struct 1994;16:287–92.
[65] Holmes JD. Along-wind response of lattice towers: II – aerodynamic damping
and deflections. Eng Struct 1996;18:483–8.
[66] Holmes JD. Along-wind response of lattice towers – III. Effective load
distributions. Eng Struct 1996;18:489–94.
[67] Houle S, Ghannoum E, Hardy C. Static and dynamic testing of transmission
lines subjected to real wind conditions. In: CIGRE symposium, Leningrad;
1991.
[68] Indian Standard (IS) 875. Indian standard code of practice for design loads
(other than earthquake) for buildings and structures, Part 3: Wind loads, New
Delhi, India; 1987.
[69] International Electrotechnical Commission (IEC) 60826. Design criteria of
overhead transmission lines. Geneva, Switzerland; 2003.
[70] Irvine HM. Cable structures. Cambridge (MA, USA): MIT Press; 1981.
[71] Ivan M. A ring-vortex downburst model for flight simulations. J Aircraft
1986;23:232–6.
[72] Jones KF. Coupled vertical and horizontal galloping. ASCE J Eng Mech
1992;118(1):92–107.
[73] Kim J, Hangan H. Numerical simulations of impinging jets with application to
downbursts. J Wind Eng Ind Aerodyn 2007;95(4):279–98.
[74] Kim JD, Hangan H, Ho TCE. Downburst versus boundary layer induced loads
for tall buildings. Wind Struct 2007;10(5):481–94.
[75] Kishnasamy S. Wind and ice loading research in ontario hydro. In:
Proceedings of the fifth US national conference on wind engineering,
Lubbock, Texas; 1985.
[76] Koziey B, Mirza F. Consistent curved beam element. Comput Struct 1994;51
(6):643–54.
[77] Kwon DK, Kareem A. Gust-front factor: new framework for wind load effects
on structures. ASCE J Struct Eng 2009;135(6):717–32.
[78] Ladubec C, El Damatty AA, El Ansary AM. Effect of geometric nonlinear
behaviour of a guyed transmission tower under downburst loading. In:
Proceedings of the international conference on vibration, structural
engineering and measurement (ICVSEM 2012), 19–21 October 2012,
Shanghai, China. Trans Tech Publications; 2012. p. 1240–9.
[79] Li CQ. Stochastic model of severe thunderstorms for transmission line design.
Probab Eng Mech 2000;15(4):359–64.
[80] Li C, Li QS, Xiao YQ, Ou JP. A revised empirical model and CFD simulations for
3D axisymmetric steady-state flows of downbursts and impinging jets. J
Wind Eng Ind Aerodyn 2012;102:48–60.
[81] Lin WE, Savory E, McIntyre RP, Vandelaar CS, King JPC. The response of an
overhead electrical power transmission line to two types of wind forcing. J
Wind Eng Ind Aerodyn 2012;100(1):58–69.
[82] Lin WE, Orf LG, Savory E, Novacco C. Proposed large-scale modelling of the
transient features of a downburst outflow. Wind Struct 2007;10(4):315–46.
[83] Loredo-Souza AM, Davenport AG. The effects of high winds on transmission
lines. J Wind Eng Ind Aerodyn 1998;74–76:987–94.
[84] Loredo-Souza AM, Davenport AG. A novel approach for wind tunnel
modelling of transmission lines. J Wind Eng Ind Aerodyn 2001;89:1017–29.
[85] Loredo-Souza AM, Davenport AG. Wind tunnel aeroelastic studies on the
behaviour of two parallel cables. J Wind Eng Ind Aerodyn 2002;90:407–14.
[86] Loredo-Souza AM, Davenport AG. The influence of the design methodology in
the response of transmission towers to wind loading. J Wind Eng Ind Aerodyn
2003;91(8):995–1005.
[87] Lombardo FT, Smith DA, Schroeder JL, Mehta KC. Thunderstorm
characteristics of importance to wind engineering. J Wind Eng Ind Aerodyn
2014;125:121–32.
[88] Lou W, Sun B, Tang J. Aeroelastic model investigation and spectral analysis of
a tall lattice tower. Adv Struct Eng 2000;3(2):119–30.
[89] Lilien JL, Erpicum M, Wolfs M. Overhead line galloping. Field experience
during one event in Belgium on last February 13th, 1997. In: 8th int conf
atmospheric icing of structures, Reykjavik, Iceland; 1998.
46 H. Aboshosha et al. / Engineering Structures 112 (2016) 23–46
[90] Macdonald JHG, Griffiths PJ, Curry BP. Galloping analysis of stranded
electricity conductors in skew winds. Wind Struct 2008;11(4):303–21.
[91] Ma WY, Liu QK, Du XQ, Wei YY. Effect of the Reynolds number on the
aerodynamic forces and galloping instability of a cylinder with semi-elliptical
cross sections. J Wind Eng Ind Aerodyn 2015;146:71–80.
[92] Manitoba Hydro. Bipole 1 & 2 HVDC transmission line wind storm failure on
September 5, 1996 – review of emergency response, restoration and design of
these lines. Transmission and Civil Design Department, 98-L1/1-37010-
06000; 1999. 54 p.
[93] Mara TG, Galsworthy JK, Savory E. Assessment of vertical wind loads on
lattice framework with application to thunderstorm winds. Wind Struct
2010;13(5):413–31.
[94] Mara TG, Hong HP. Effect of wind direction on the response and capacity
surface of a transmission tower. Eng Struct 2013;57:493–501.
[95] Mason MS, Wood GS, Fletcher DF. Numerical simulation of downburst winds.
J Wind Eng Ind Aerodyn 2009;97(11–12):523–39.
[96] Mason MS, Fletcher DF, Wood GS. Numerical simulation of idealized three-
dimensional downburst wind fields. Eng Struct 2010;32(11):3558–70.
[97] Mason MS, Wood GS, Fletcher DF. Numerical investigation of the influence of
topography on simulated downburst wind fields. J Wind Eng Ind Aerodyn
2010;98(1):21–33.
[98] Mehta KC, Kadaba R, Kempner L. Conductor response to extreme winds. In:
Proceedings. Second international symposium on probability methods
applied to electric power systems, September 20–23, 1988, Oakland, CA;
1988. p. 9-1–9-11.
[99] Mehta KC, Kadaba R. Wind induced response of transmission line conductors.
In: Sun TF. Recent advances in wind engineering. Proceedings of the second
Asia-Pacific symposium on wind engineering, Beijing, China. vol. 2. Pergamon
Press; 1989. p. 748–55.
[100] Mehta KC, Kadaba R. Field data analysis of electrical conductor response to
winds. J Wind Eng Ind Aerodyn 1990;36:329–38.
[101] Momomura Y, Marukawa H, Okamura T, Hongo E, Ohkuma T. Full-scale
measurements of wind-induced vibration of a transmission line system in a
mountainous area. J Wind Eng Ind Aerodyn 1997;72:241–52.
[102] Orf L, Oreskovic C, Savory E, Kantor E. Circumferential analysis of a simulated
three-dimensional downburst producing thunderstorm out flow. J Wind Eng
Ind Aerodyn 2014;135(2014):182–90.
[103] Orf LG, Kantor E, Savory E. Simulation of a downburst-producing
thunderstorm using a very high-resolution three-dimensional cloud model.
J Wind Eng Ind Aerodyn 2012;104–106:547–57.
[104] Orf LG, Oreskovic C, Savory E. Circumferential analysis of simulated 3-D
downburst-producing thunderstorm outflows. In: Proceedings of the 6th
European and African conference on wind engineering, Cambridge, UK, July;
2013.
[105] Orwig KD, Schroeder JL. Near-surface wind characteristics of extreme
thunderstorm outflows. J Wind Eng Ind Aerodyn 2007;95:565–84.
[106] Oseguera RM, Bowles RL. ‘‘A simple, analytics 3-dimensional downburst
model based on boundary layer stagnation flow” NASATM- 100632. Hampton
(VA): NASA Langley Research Center; 1988.
[107] Ohkuma T, Marukawa H. Galloping of overhead transmission lines in gusty
wind. Wind Struct 2000;3(4):243–53.
[108] Peil U, Nölle H. Guyed masts under wind load. J Wind Eng Ind Aerodyn
1992;41–44:2129–40.
[109] Rawlins CB. Analysis of conductor galloping field observations single
conductors. IEEE Trans Power Apparatus Syst 1981;PAS-100(8):3744–51.
[110] Savory E, Parke GAR, Disney P, Toy N. Wind-induced transmission tower
foundation loads: a field study-design code comparison. J Wind Eng Ind
Aerodyn 2008;96:1103–10.
[111] Savory E, Parke GAR, Zeinoddini M, Toy N, Disney P. Modelling of tornado and
microburst-induced wind loading and failure of a lattice transmission tower.
Eng Struct 2001;23(4):365–75.
[112] Selvam RP, Holmes JD. Numerical simulation of thunderstorm downdrafts. J
Wind Eng Ind Aerodyn 1992;41–44:2817–25.
[113] Sengupta A, Sarkar PP. Experimental measurement and numerical simulation
of an impinging jet with application to thunderstorm microburst winds. J
Wind Eng Ind Aerodyn 2008;96:345–65.
[114] Shan L. Characteristics study of transmission line mechanical research center
(TLMRC) wind tower data. Interim report no. EPRI TR-100906. Project 2016-
03. Palo Alto (CA): Electric Power Research Institute; 1992.
[115] Shan L. Overview of EPRI conductor wind loading experiments. In:
Proceedings of the seventh US national conference on wind engineering.
Los Angeles, California; 1993.
[116] Shan L. Evaluation of the results of several full-scale conductor wind loading
experiments. EPRI TR-100479; 1994.
[117] Shehata AY, El Damatty AA. Behaviour of guyed transmission line structures
under downburst wind loading. Wind Struct 2007;10(3):249–68.
[118] Shehata AY, El Damatty AA. Failure analysis of a transmission tower during a
microburst. Wind Struct 2008;11(3):193–208.
[119] Shehata AY, Nassef AO, El Damatty AA. A coupled finite element-optimization
technique to determine critical microburst parameters for transmission
towers. Finite Elem Anal Des 2008;45(1):1–12.
[120] Shehata AY, El Damatty AA, Savory E. Finite element modeling of
transmission line under downburst wind loading. Finite Elem Anal Des
2005;42(1):71–89.
[121] Simiu E, Scanlan RH. Wind effects on structures. 3rd ed. New York: Wiley;
1996.
[122] Sissenwine N, Tattleman P, Granthan DD, Gringorten II. Extreme wind speeds,
gustiness and variations with height. Air Force Cambridge Research
Laboratories. Technical report 73-0560. Aeronomic Laboratory, Project
8624; 1972.
[123] Solari G, Burlando M, De Gaetano P, Repetto MP. Characteristics of
thunderstorms relevant to the wind loading of structures. Wind Struct
2015;20(6):763–91.
[124] Solari G, De Gaetano P, Repetto MP. Thunderstorm response spectrum:
fundamentals and case study. J Wind Eng Ind Aerodyn 2015;143:62–77.
[125] Solari G. Gust buffeting. I: peak wind velocity and equivalent pressure. J
Struct Eng 1993;119(2):365–82.
[126] Solari G. Gust buffeting. II: dynamic along-wind response. J Struct Eng
1993;119(2):383–97.
[127] Sparling B, Wegner L. Estimating peak wind load effects in guyed masts.
Wind Struct 2007;10(4):347–66.
[128] Su Y, Huang G, Xu Y. Derivation of time-varying mean for non-stationary
downburst winds. J Wind Eng Ind Aerodyn 2015;141(2015):39–48.
[129] Vermeire BC, Orf LG, Savory E. Improved modelling of downburst outflows for
wind engineering applications using a cooling source approach. J Wind Eng
Ind Aerodyn 2011;99(8):801–14.
[130] Vermeire BC, Orf LG, Savory E. A parametric study of downburst line near-
surface outflows. J Wind Eng Ind Aerodyn 2011;99(4):226–38.
[131] Vicroy DD. ‘‘A simple, analytical, axisymmetric microburst model for
downdraft estimation” NASA-TM-104053. Hampton (VA, USA): NASA; 1991.
[132] Vicroy DD. Assessment of microburst models for downdraft estimation. J
Aircraft 1992;29:1043–8.
[133] Volpe H. Wind effects on conductors for span factors. Final project report, line
design section, division of facilities engineering. Portland (OR): Bonneville
Power Administration; 1989.
[134] Wang L, McCullough M, Kareem I. A data-driven approach for simulation of
full-scale downburst wind speeds. J Wind Eng Ind Aerodyn 2013;123
(2013):171–90.
[135] Wang X, Lou W, Li H, Chen Y. Wind-induced dynamic response of high-rise
transmission tower under downburst wind load. J Zhejiang Univ 2009;43
(8):1520–5.
[136] Wei P, Bingnan S, Jinchun T. A catenary element for the analysis of cable
structures. Appl Math Mech 1999;20(5):0253–4827 [English Edition].
[137] Wieringa J. Updating the Davenport roughness classification. J Wind Eng Ind
Aerodyn 1992;41–44:357–68.
[138] Wilson JW, Roberts RD, Kessinger C, McCarthy J. Microburst wind structures
and evaluation of Doppler radar for airport wind shear detection. J Climate
Appl Meteorol 1984;23:898–915.
[139] Winkelman PF. Sag-tension computations and field measurements of
Bonneville Power Administration. AIEE paper 59-900; 1959.
[140] Wolfson MM, DiStefano JT, Fujita TT. Low-altitude wind shear characteristics
in the Memphis, TN area. In: Preprints of the 14th conference on severe local
storms, October 29–November 1, Indianapolis, IN, USA. American
Meteorological Society; 1985. p. 322–7.
[141] Wood GS, Kwok KCS, Motteram NA, Fletcher DF. Physical and numerical
modelling of thunderstorm downbursts. J Wind Eng Ind Aerodyn
2001;89:535–52.
[142] Yan Z, Li Z, Savory E, Lin W. Galloping of a single iced conductor based
on curved-beam theory. J Wind Eng Ind Aerodyn 2013;123(Part A):
77–87.
[143] Yasui H, Marukawa H, Momomura Y, Ohkuma T. Analytical study on wind-
induced vibration of power transmission towers. J Wind Eng Ind Aerodyn
1999;83:431–41.
[144] Yu P, Wong P, Kaempffer F. Tension of conductor under concentrated loads. J
Appl Mech – Trans Am Soc Mech Eng 1995;62(3):802–9.
[145] Zhang Y, Hu H, Sarkar PP. Modeling of microburst outflows using impinging
jet and cooling source approaches and their comparison. Eng Struct
2013;56:779–93.
[146] Zhou Y, Kareem A. Gust loading factor: new model. J Struct Eng 2001;127
(2):168–75.
[147] Zhu S, Etkin B. Model of the wind field in a downburst. J Aircraft
1985;22:595–601.