J. Wind Eng. Ind. Aerodyn.

159 (2016) 123–133

Contents lists available at ScienceDirect

Journal of Wind Engineering

and Industrial Aerodynamics
journal homepage: www.elsevier.com/locate/jweia

Interference effects between two high-rise buildings on wind-induced

torsion
crossmark
⁎
X.F. Yua, Z.N. Xiea, , X. Wangb, B. Caic
a
State Key Laboratory of Subtropical Building Science, South China University of Technology, WuShan Road 381, Guangzhou 510640, Guangdong, People's
Republic of China
b
State Key Laboratory Breeding Base of Mountain Bridge and Tunnel Engineering, Chongqing Jiaotong University, Chongqing 400074, People's Republic of
China
c
Department of Civil, Construction and Environmental Engineering, Iowa State University, Ames, IA 50011, USA

A R T I C L E I N F O A BS T RAC T

Keywords: The distribution variation and correlation of envelope interference factor (EIF) of the base torsion responses for
High-rise buildings principal building were studied by applying the rigid model wind tunnel test. The influence factor including
Base torsion response different breadth ratio (Br=0.4, 0.6, 0.8, 1.0, 1.2 and 1.4), height ratio (Hr=0.8, 1.0, 1.2 and 1.4), reduced
Interference effect velocity and approaching turbulence intensity were considered. Furthermore, the mechanism and occurrence
Reduced velocity
condition of wake vortex-excited resonance were also studied. Results showed that the maximum value of EIF
Vortex-excited resonance
was 1.9 when vortex-excited resonance failed to happen. The correlation coefficient of EIF between Br=0.8 and
Br=1.0 was 90.2%. Vortex-excited torsion resonance which induced from wake of interfering building occurred
when Br=0.4, and the maximum value of EIF reached up to 2.98. With increased Hr, the value of EIF increased.
Good correlation was shown among different height ratios. For breadth ratios range from 0.3 to 0.5, it was
suggested the reduced velocity should be out of the range of 3.33–5.56 to avoid occurring vortex-excited torsion
resonance.

1. Introduction
Wind-induced interference effect was always a hot spot in wind
engineering research since the collapse of England bridge thermal
power plant in 1965 (Armitt, 1980). However, most of current
researches on interference effect were focus on base pneumatic bending
moment, base response of bending moment and local pressure
distribution (Xie and Gu, 2004, 2007; Lam et al., 2008; Kim et al.,
2011; Hui et al., 2012, 2013; Yu et al., 2015), some of the research
results had been adopted by Chinese load code for the design of
building structures (GB50009-2012, Architectural Industry Press of
China, 2012).
Interference effects on torsional responses for high-rise buildings
were seldom considered because of cognitive and experimental limita-
tion, and previous researches in this respect only consider some simple
influence factors. However, it was shown that the interference effects
on torsional responses could not be ignored. Blessmann and Riera
(1985) had examined interference effect of torsion for two square cross
section tall buildings, it was discovered that the maximum torsional
moment coefficient was 3 times of isolated case. Zhang et al. (1994)
had studied interference effects on torsional responses for different
breadths and shapes of tall buildings by wind tunnel test on aeroelastic
model, it was found that torsional responses were significantly
enhanced by a factor of up to 2.2 times that of the isolated building
when vortex shedding from the interfering building was in resonance
with the natural frequency of the principal building. Zhang et al. (1995)
further experimentally investigated interference effects on torsional
response of a tall square cross-section building with structural asym-
metry. Results showed that the mean and standard deviation torsional
responses of the asymmetric principal building could be significantly
enhanced due to the presence of the interfering building.
Besides, few studies had examined the interference effect by
correlation method. Xie and Gu (2005) quantitatively analyzed the
correlation of base bending moment interference factors in different
breadths and heights of interfering building by the high frequency force
balance (HFFB) method. Measured results showed significant correla-
tions exist in the distributions of the interference factors estimated in
different configurations, and the correlation coefficient of mean inter-
ference factors was found in the range of 92~99%. Xie and Gu (2007)
further studied the correlation of mean and dynamics interference
effects of base-bending moment among 3 buildings with different
height ratios. Results showed the correlation coefficients of mean

⁎
Corresponding author.
E-mail address: znxie@scut.edu.cn (Z.N. Xie).

http://dx.doi.org/10.1016/j.jweia.2016.10.011
Received 18 February 2016; Received in revised form 11 October 2016; Accepted 20 October 2016
Available online 31 October 2016
0167-6105/ © 2016 Elsevier Ltd. All rights reserved.
X.F. Yu et al. J. Wind Eng. Ind. Aerodyn. 159 (2016) 123–133

100
interference factor reached up to 94%, while that of across-wind 600
dynamics interference factor also reached up to 83%. 570
G
In this paper, the time-varying external pressures on the surface of
principal building were obtained by applying the synchronous pressure
490
measurement technique in various interference cases, and the base F
torsion responses were calculated. Distributions and correlation of the
enveloped interference factor (EIF) of the peak base torsion responses 410
in different breadth ratios, height ratios, reduced velocities and E
approaching turbulence intensities were detailed analyzed. Finally,
the interference mechanism and occurrence condition of vortex- 330
D
induced resonance were further studied. 5 10 15 20 20 15 10 5

250

15 10 5
2. Wind tunnel experiment C

2.1. Experimental method

20
170
B

20
Theoretically, three kinds of experimental methods including high

5 10 15
frequency force balance (HFFB), aeroelastic model (AEM) and high 90
frequency pressure integration (HFPI) can be used to conduct research A
100
on interference effect of torsional response. However, the measure Tap
range of a common high frequency force balance is always too large to
meet the precision of torsional component for its low signal-to-noise
Fig. 1. Principal building model and tapping locations on each tap floor (unit: mm).
ratio. Although AEM method can theoretically obtain the torsional
response, accelerometer can only get the resonance component but
cannot get the dominant average and background components. Thus it Table 1
Experimental models.
is usually to measurement displacement when adopted AEM method,
but the torsional displacement is angular rotation which is difficult to Dimensions (mm)
measure. Actually, AEM method can simulate the interaction between Experimental (Bp×Dp×Hp)a Breadth Height Locations
wind and building, but it is necessary to conduct wind tunnel models ratios ratios
(Bi×Di×Hi)b (Br=Bi/Bp) (Hr=Hi/
experiments in different velocities to obtain structural response in
Hp)
different velocities, which is a disadvantage for carrying massive wind
tunnel tests on interference effect. Thus, AEM cannot be as a main Principal building 100×100×600 – – 1
method to study interference effects. High frequency pressure integra- Interfering 40×40×600 0.4 1.0 64
building
tion (HFPI) method which based on synchronous pressure measure-
60×60×600 0.6 1.0 64
ment technique can be used to obtained wind-induced responses for 80×80×600 0.8 1.0 64
high-rise building. Actually, the precision of aerodynamic torsional 100×100×600 1.0 1.0 64
moment and base torsion responses obtained from HFPI method is 120×120×600 1.2 1.0 64
higher than that from HFFB method (Aly, 2013). Thus, HFPI method is 140×140×600 1.4 1.0 64
100×100×480 1.0 0.8 64
selected to investigate the interference effects between two high-rise
100×100×600 1.0 1.0 64
buildings on wind-induced torsion. 100×100×720 1.0 1.2 64
100×100×840 1.0 1.4 64
2.2. Experimental setup
a
Bp×Dp×Hp: dimension of principal building.
b
Bi×Di×Hi: dimension of interfering building.
The experimental models were consist of two rigid model: the
pressure model, called the principal building, and the other model,
building.
called the interfering building. Both the principal and interfering
Wind tunnel experiments were conducted in a Boundary Layer
buildings are square in section. The principal building model was
Wind Tunnel Laboratory. The test section of the wind tunnel was 3.0 m
100 mm×100 mm in plan and 600 mm in height. The length scale was
wide and 2.0 m high. The maximum block ratio is 2.4%. Exposure
set as 1:400, such that the model represented a full-scale tall building
category B with a power law exponent of 0.15 which represented a
with height of 240 m. Seven tap floors were arranged along the height,
suburban flat terrain, was simulated according to the Chinese Load
in which each tap floor had 28 pressure taps. The principal building
code (GB50009-2012, Architectural Industry Press of China, 2012).
model and tapping location on each tap floor was shown in Fig. 1. In
The simulated mean wind profile, turbulence intensity distribution,
wind tunnel test, the wind direction was kept constant, and DSM3200
and power spectrum at the height of the rooftop were shown in Fig. 3,
made by Scanivalve LTD was used to synchronously measure wind
in which standard velocity and turbulence profiles were simulated
pressure on the principal building. The shape of the principal building
according to Chinese Load code (GB50009-2012, Architectural
remained unchanged, and the interference effect of six kinds of breadth
Industry Press of China, 2012), UH was the mean wind speed at the
ratios (Br=Bi/Bp=0.4, 0.6, 0.8, 1.0, 1.2 and 1.4) and four height ratios
height of rooftop. To further investigate the influence of turbulence
(Hr=Hi/Hp=0.8, 1.0, 1.2 and 1.4) were considered, where Bi and Hi
intensity on the torsional interference effect, the normalized mean
were the breadth and height of the interfering building, Bp and Hp were
speed profile was kept unchanged, while the high and low turbulence
the breadth and height of the principal building, respectively. Table 1
profiles were simulated, as also shown in Fig. 3. The reference wind
showed the cases of the experimental model used in this study. The
velocity (standard velocity) was 11.4 m/s at the height of 0.6 m. The
position grid of the interfering building and image of wind tunnel test
test sampling time period was 65.536 s, and sampling frequency was
was shown in Fig. 2, in which “A” stood for the stationary principal
312.5 Hz.
building; “B” stood for the moving interfering building, x and y were
the distances between the two models in the longitudinal and trans-
verse directions respectively, and b was the width of the principal

124
X.F. Yu et al. J. Wind Eng. Ind. Aerodyn. 159 (2016) 123–133

Fig. 2. Coordinate grid and photo of wind tunnel test.

2.3. Data analysis ∞

σξ = ∫0 H (ω) 2 SFT (ω) dω
(4)
To conveniently obtain the value of interference factor for engineers
1
in wind-resistant design of building structures, the peak value of base where H (ω) 2 = is the mechanical admittance
(ω T2 − ω2)2 + (2ζω T ω)2
torsion interference factor (IF) is defined as: function and SFT (ω) is the power spectrum density (PSD) of generalized
force FT . Then the dynamic component of angular displacement
The peak base torsion response with intefering building ∼ 0.577
IF = {θ } = [φ] gσξ is obtained. g = 2 loge (νT ) + is the peak factor
The peak base torsion response without intefering building (1) 2 loge (νT )
(Davenport, 1964), in which ν is the effective frequency for the
where, the peak value of base torsion response around z axis M̂z is response, T is the time interval over which the maximum response is
calculated by the following calculation procedure. required.
∼
According to classical differential equation of motion in the After the dynamic component of angular displacement {θ } is
∼
torsional direction, the structural response (angular displacement θ ) obtained, the dynamic torsion response component ({T }) for the
∼ ∼
for a building with n floors can be expressed as building is {T } = [K ]{θ }, and then the maximum and minimum values
⌢ ∼ ∼
of total torsion response are { T } = {T } + {T } and { T } = {T }−{T }, in
⏝
[I ]{θ ̈ (t )} + [C ]{θ ̇ (t )} + [K ]{θ (t )} = {P (t )} (2) which {T } is the mean torsion response component. Thus, the peak
⌢
torque for each floor of principal building is {Tˆ } = max{ Ti , T i }
⏝
where [I ], [C ] and [K ] are the mass moment of inertia, damping, and (i=1, 2, 3…, n).
stiffness matrixes, respectively. {P (t )} is the torque column vector Finally, the peak value of base torsion response M̂z can be obtained
which can be obtained by numerically integrating the pressures by summing all the peak torque as
recorded on each face of the principal building. When the first-order
n
torsional modal shape {ϕ} is adopted ({ϕ} is a column vector and
Mˆ z = ∑ Tˆi
assumed as linear distribution for convenience in this study), the Eq. i =1 (5)
(2) can be simplified as
where n is the number of floor for principal building.
{φ}T {P (t )} According to engineering experiences and previous research (Lam
ξ ̈ + 2ζω T ξ ̇ + ω T2 ξ = = FT
IT (3) et al., 2011), the fundamental frequency ω T and modal damping ratio ζ
of principal building in the torsional direction are set as 0.35 Hz and
where IT ={ϕ}T [I ]{ϕ}, ζ is the modal damping ratio, ω T and FT are
3.5%, respectively.
fundamental natural frequency and generalized force, respectively. ξ is
the generalized angular displacement which can be calculated by
solving the Eq. (3). Its standard deviation can be obtained as

Fig. 3. Simulated wind parameters, (a) mean speed profile, (b) turbulence intensity profile, and (c) spectra at the height of roof in standard case.

125
X.F. Yu et al. J. Wind Eng. Ind. Aerodyn. 159 (2016) 123–133

Table 2
Conversion relationship among basic wind pressure, wind speed at the height of rooftop, and reduced velocity.

w0 (kPa) 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90

VH (m/s) 40.75 43.22 45.56 47.78 49.91 51.95 53.91 55.80 57.63 59.40 61.12
Vr 2.91 3.09 3.25 3.41 3.56 3.71 3.85 3.99 4.12 4.24 4.37

y/b y/b
5 5
1.1
1.2 1.1 1.1
1.2 1.2
4 4
1.2 1.2
1.3
1.4 3 1.3 1 3
1.4 1.1
1.5 1
0.9 0.8
2 0.7 2
1.6 0.9 1.4
1.5 1.6
1.3 1.7 1 1.7 1
1.2
x/b 1.2 1.1 x/b 1.1
11 10 9 8 7 6 5
1
4 3 2 1 A00 10 9 8 7 6
1
5
0.9
4 3 2 1 A00
(a) Vr=4.37 (b) Vr=2.91
Fig. 4. Interference factor IF distributions of peak torsion responses at different reduced velocities, (a) Vr=4.37 (b) Vr=2.91.

4.37
3. Results and discussions EIF = max (IF (Vr ))
Vr =2.91 (7)
For each interference case, the base torsion response of principal
building are calculated for different basic wind pressures w0, and the 3.1. Effect of breadth ratio
corresponding reduced velocities Vr is shown in Table 2. The reduced
velocity is defined as: The cross-sectional width of interfering building has a significant
effect on the torsional interference effect. Fig. 6 shows envelope
VH
Vr = interference factors EIF in different breadth ratios Br. In this figure,
f0 b (6) the maximum envelope interference factor when Br equals 0.4 is
significantly higher than that of other breadth ratios, which is due to
where f0 is the fundamental frequency of principal building in the
vortex-induced resonance.
torsional direction, and b is the width of cross section of the principal
The envelope torsional interference factor EIF distribution of the
building, VH is design wind speed at the rooftop.
principal building are established for six kinds of different breadth
Given the basic configuration (both the interfering and principal
ratios (see Fig. 7). The main interference law is as follows:
buildings have equal width and height), IF distributions are plotted in
Fig. 4 when Vr equals 4.37 and 2.91. It shows that IF distributions have
(1) When Br equals 0.4, the maximum envelope interference factor
high similarity and have less noticeable changes among different
appears at the tendem position (x/b=3, y/b=0) and its value
reduced velocities. Fig. 5 shows the correlation between IF when Vr
reaches 2.98, while the minimum envelope interference factor
equals 4.37 and IFs in the other reduce velocities. It is also known that
appears at the position where x/b=1, y/b=1.8, and its value is 0.96.
the interference effects decreases slightly with decreasing reduced
The magnitude of the maximum envelope interference factor is
velocity. For engineering applications, a more representative envelope
much higher than that in other breadth ratios, which is caused by
value of peak torsional interference factor (EIF), as defined in formula
the weak-vortex-induced resonance. More details are presented in
(7), is adopted to analyze the base torsion responses between two high-
the 4th section. In the region {0.5≤x/b≤1.5, 0.9≤y/b≤1.8}, the
rise buildings.
envelope interference factor is less than 1, while it exceeds 1 in the
remaining regions. The amplified interference effect is particularly
remarkable in the region {2≤x/b≤5, y/b≤1.5}, in which the peak
1.8 Vr=4.12 torque at the base of the principal building increases more than
Vr=3.85 300% because of interference effect.
Vr=3.56 (2) When Br is between 0.6 and 1.0, the envelope interference factor is
1.6
Vr=3.25 greater than 1 only when x/b≥2, particularly, the peak torque
Vr=2.91 increases more than 50% in the region {3≤x/b≤5, 0.9≤y/b≤1.8};
1.4
when x/b < 2, the envelope interference factor is substantially less
than or equal to 1, that is because the wake flow of interfering
IF

1.2 building offsets parts of flow fluctuations, reducing torsional

response at the base of the interfesred building.
1.0 (3) When Br is between 1.2 and 1.4, the maximum envelope inter-
ference factor is as high as 1.9 in the region {1.5≤x/b≤3, y/b=0.9};
0.8 while the envelope interference factor is less than or equal to 1 and
as minimum as 0.7 in the region {1≤x/b≤3, 1.8≤y/b≤2.7}.
0.8 1.0 1.2 1.4 1.6 1.8
IF, Vr=4.37 Besides, as seen from Fig. 8, the amplification interference effect is
more obvious with breadth ratio in parallel configuration because of
Fig. 5. Regression analyses of interference factor IF at different reduced velocities. channeling effect. In the critical position, peak torque at the base of the

126
X.F. Yu et al. J. Wind Eng. Ind. Aerodyn. 159 (2016) 123–133

3.5
Br=0.4 Br=0.6
3.0 Br=0.8 Br=1.0
2.5 Br=1.2 Br=1.4

2.0

EIF
1.5

1.0

0.5
0 4 8 12 16 20 24 28 32 36 40 44 48 52 56 60 64
Work cases
Fig. 6. Envelope interference factor EIF in different work cases.

principal building increases up to 42%. ⎧ 0.490EIF + 0.572 ; Br = 0.6, ρ = 0.751, ε = 0.084

⎪
⎪ 0.831EIF + 0.147
For the non-resonance case, using the working case that Br equals ; Br = 0.8, ρ = 0.902, ε = 0.078
RIF = ⎨
1.0 as a reference, the linear regression analyses were completed ⎪1.009EIF − 0.047 ; Br = 1.2, ρ = 0.886, ε = 0.103
between the envelope interference factor EIF and reference values for ⎪ 0.957EIF + 0.012
⎩ ; Br = 1.4, ρ = 0.778, ε = 0.151
the other four different breadth ratios, as showed in Figs. 9 and 10. The
(8)
regression relationship of the EIFs is summarized as:
where RIF is the linear regression inteference factor, ρ is the correala-

y/b y/b
5 1.1 5
1.1

4 1.1 4
1.1
1.2
3 1.2 3
1.1

1.2
1.2 1 1.3 1
2 2
1.4
1.3 1.5 1.6 1.7 0.96 1.5
1.9
1.4 2.2 1 1
1.2
x/b 2.5
x/b 1.1
11 10 9 8 7 6 5 4 3
3
2 1 A00 11 10 9 8 7 6 5 4 3 2 1 A00
(a) Br=0.4 (b) Br=0.6

y/b y/b
5 5
1.1
1.1 1.2
4 1.2 1.1 4
1.2
1.1 1.3 1.3
1.2
1.2

3 1.4
1.3 3
1.3 1.1
1 1.4 1
1.4 1.5
1.6 0.9 2 0.9 2
0.8
1.5
1.2 1 1.3
1.6 1.7 1
x/b 1.1
x/b 1.2 1.1
11 10 9 8 7 6 5 4
1
3 2 1 A00 11 10 9 8 7 6 5
1
4
0.9
3 2 1 A00
(c) Br=0.8 (d) Br=1.0

y/b y/b

1.1
5 1.1
5
1.1 1.1
1.2 1.2
1.2 4 1.2 4
1.3

1.3 1.3
1 3 1
3
1.3

0.9 0.8
1.4
2 1.4 0.7 2
1.4

0.8
1.5 1.5
1.3 1.6 1.7 1.8 1 1.3 1.6 1.7 1.9 1
1.2
x/b 1.2 1.1 1 x/b 1.1 1 0.9 0.8
11 10 9 8 7 6 5
0.9
4 3 2 1 A00 11 10 9 8 7 6 5 4 3 2 1 A00
(e) Br=1.2 (f) Br=1.4
Fig. 7. Envelope interference factor EIF distributions of the base torsion responses in different breadth ratios, (a) Br=0.4 (b) Br=0.6 (c) Br=0.8 (d) Br=1.0 (e) Br=1.2 (f) Br=1.4.

127
X.F. Yu et al.
Fig. 8. Envelope interference factor EIF as a function of building spacing in parallel
arrangement.
Fig. 9. Linear regression analysis of envelope interference factors EIF between Br=0.8
and Br=1.0.
Fig. 10. Linear regression analyses of envelope interference factors EIF in different
breadth ratios.
tion coefficient denoting the corrlation degree of EIFs between Br≠1.0
and Br=1.0, ε is the residual denoting the accuracy of regression
analysis. It is shown from the formula (8) that the correlation between
EIF (Br=0.8) and EIF (Br=1.0) is the best (correlation coefficient
ρ=90.2%), followed by the correlation between EIF (Br=1.2) and EIF
(Br=1.0), while the correlation between EIF (Br=0.6) and EIF (Br=1.0)
is the worst. Besides, the EIFs of Br≠1.0 can be predicted from the EIF
distribuion of two identical buildings as shown in Fig. 6(d).
128
J. Wind Eng. Ind. Aerodyn. 159 (2016) 123–133
Fig. 10 shows the variations of the RIF with various Br of the
interfereing building. It is noted that EIF and the reference value
becomes more relevant with increased Br when Br≤1. However, it
becomes less relevant with increased Br when Br≥1. It means that the
amplification effect is enhanced with the increase of the breadth of the
upstream building when Br≤1 in non-resonace cases, while shielding
effect becomes more remakable with increased breadth of the upsteam
building when Br≥1.
3.2. Effect of height ratio
Fig. 11 shows the envelope interference factors EIF at different
height ratios Hr. It can be seen that when height ratio Hr equals 0.8,
EIF is significantly lower compared to EIFs at the other height ratios.
However, the change of height ratio has insignificant effect on the
envelope interference factors EIF when Hr≥1.0.
Fig. 12 plotted the distribution of the envelope interference factor at
different height ratios Hr , showing that the distributions remain
similar at different height ratios Hr. The main interference law is as
follows:
(1) With increased height ratio Hr, the envelope interference factor
EIF tends to increase. The maximum EIF when Hr equals 0.8 is at
least 30% smaller than that when Hr≥1.0. Therefore, the amplifi-
cation interference effect should be focused on the situation when
interfering building is higher or equal to the principal building.
(2) In tandem arrangement, with increased height ratio, the shielding
region shifts away from the principal building, and the EIF in the
shielding region approaches to 1.
Using working case when Hr equals 1.0 as a reference, similar linear
regression analyses are completed between the envelope interference
factors EIF and the reference value at three other height ratios Hr, as
shown in Figs. 13 and 14. The regression relationship of the EIFs is
summarized as:
⎧ 0.682EIF + 0.290 ; Hr = 0.8, ρ = 0.883, ε = 0.071
⎪
RIF = ⎨1.125EIF − 0.132 ; Hr = 1.2, ρ = 0.976, ε = 0.049
⎪1.173EIF − 0.157 ;
⎩ Hr = 1.4, ρ = 0.968, ε = 0.060
(9)
where RIF is the linear regression inteference factor. It is shown from
the formula (9) and Fig. 14 that the correlation between EIF (Hr=1.2)
and EIF (Hr=1.0) is the best (correlation coefficient ρ=97.6%), followed
by the correlation between EIF (Hr=1.4) and EIF (Hr=1.0), and that the
correlation EIF (Hr=0.8) and EIF (Hr=1.0) is the worst. The correlation
coefficient between EIF and the reference value has an insignificantly
growth with increased Hr, while the correlation is essentially the same
when the interfering building reaches a certain height, indicating that
the distribution of the envelope interference factor EIF is essentially
the same for this case.
3.3. Effect of reduced velocity
In the basic configuration described above (Br=1, Hr=1), correlation
analysis has been completed to the interference factor IF at different
reduced velocities. It is found that the change of reduced velocity has
little influence on the torsional interference effect. However, wake-
vortex-induced resonance occures when Br=0.4, the effect of reduced
velocity is more obvious on the interference factor distribuiton. Thus,
the interference effect at different reduced velocities when Br=0.4
should be closely analyzed because of the risk of vortex-induced
resonance.
Fig. 15 shows the distribution of interference factor IF at different
reduced velocities Vr when Br=0.4. It can be clearly found that the
amplification effect is gradually increasing with the increased Vr. When
X.F. Yu et al. J. Wind Eng. Ind. Aerodyn. 159 (2016) 123–133

2.5
Hr=0.8 Hr=1.0
Hr=1.2 Hr=1.4
2.0

EIF
1.5

1.0

0.5
0 4 8 12 16 20 24 28 32 36 40 44 48 52 56 60 64
Work Cases
Fig. 11. Envelope interference factors EIF in different work cases.

Fig. 12. Envelop Interference factor EIF distributions of the base torsion responses at different height ratios, (a) Hr=0.8, (b) Hr=1.0, (c) Hr=1.2, and (d) Hr=1.4.

Fig. 13. Linear regression analysis of the envelop interference factors EIF between Fig. 14. Linear regression analyses of the envelop interference factors EIF in different
Hr=1.2 and Hr=1. height ratios.

129
X.F. Yu et al. J. Wind Eng. Ind. Aerodyn. 159 (2016) 123–133

y/b y/b
1 1 5 5
1 4 4
1
1.1
1.1 3 3

1.1
1
1.1 1
1.2
2 1.2
1.3
2
0.9 1.6 1.8 0.9 1.1
1.3 1.5
1.4
1.5 1 1.4 2 1
x/b x/b 2.1

11 10 9 8 7 6 5 4 3 A00 11 10 9 8 7 6 5 4 3 2 1 A00
(a) Vr=2.91 (b) Vr=3.56

y/b y/b
1 5 1 1
5
1 1
4 4
1

1
3 3
1.1 1.1 1 1.1
1
0.9 2 2
1.3 0.9 1.1
1.3 1.4 1.6 1.2
1.2 1.1 1.4
1.9 1 1.6
1.9
1
1.1 2.2
x/b 2.6 x/b 1.1
2.3
2.6
11 10 9 8 7 6 5 4
2.8
3 2 1 A00 11 10 9 8 7 6 5 4 3
3
2 1 A00
(c) Vr=4.12 (d) Vr=4.37
Fig. 15. Interference factor IF distributions of the base torsion responses at different reduced velocities (Br=0.4), (a) Vr=2.91, (b) Vr=3.56, (c) Vr=4.12, and (d) Vr=4.37.

Vr=4.37, the interference effect due to vortex-induced resonance is very position.

significant, and the maximum interference factor is 1.87 times as that (2) In tandem arrangement: the envelope interference factor EIF is
when reduced velocity Vr=2.91. In addition, the amplification inter- constantly larger than 1 and there is no sheilding effect in low
ference effects are the most significant when the interfering building is turbulence intensity. In standard turbulence intensity, the envel-
located exactly upstream 3 times as the width of principal building. ope interference factor EIF is less than 1 in the range 3.0≤x/b≤5.0.
In high turbulence intensity, the envelope interference factor EIF is
3.4. Effect of turbulence intensity less than 1 in the range 1.0≤x/b≤8.0. It can be concluded that the
higher turbulence intensity, the more turbulence of approaching
To understand the influnce of turbulence intensity on the torsional flow weaken by vortex shedding of interfering building. Thus the
interference effect, the same rigid model is tested again in the wind sheilding effect is more obvious, and the shielding area is greater.
tunnel with increased or decreased turbulence intensity and the same (3) In oblique arrangement: In the oblique upstream region {x/b≤2.5,
normalized wind speed profile (see Fig. 3). Taking into account the 1≤y/b≤3}, the envelope interference factor EIF is relatively small,
interference effects in vortex-induced resonance and non-resonance but it varies significantly with turbulence intensity. The minimum
cases are significantly different, a non-resonant case (Br=1, Hr=1 as the EIF is 0.68 in low turbulence case, while it is 1.0 in high turbulence
representative) and the resonant case (Br=0.4, Hr=1) are discussed, case.
respectively.
Linear regression analysis between EIF in high or low turbulence
3.4.1. Non-resonance case intensities and EIF in standard turbulence (EIF (S)) is done, respec-
In the basic configuration (Br=1, Hr=1), Fig. 16 shows the tively. The results, as shown in Fig. 17 and formula (10), indicate the
distributions of envelope interference factors EIF in different turbu- envelope interference factor EIF in low and standard turbulence
lence intensities. It can be seen that the turbulence intensity has intensities has a good correlation and the correlation coefficient is up
significant effect on the torsional interference effect. to 95.2%. With increased turbulence intensity, the correlation becomes
weakened, and the correlation coefficient of EIF in high turbulence and
(1) Overall, the amplification interference effect is more significant standard turbulence is only 72.2%.
with decreased turbulence intensity. The maximum envelope
⎧ 2.327EIF (S) − 1.328 ; Low turbulence, ρ = 0.952,
interference factor in the low turbulence intensity is 1.81 times ⎪
as much as that of high turbulence intensity, which is because the ⎪ ε = 0.137
RIF = ⎨
effect of vortex shedding from interfering building on the down- ⎪ 0.555EIF (S) + 0.456 ; High turbulence, ρ = 0.722,
stream principal building is larger when the turbulence intensity is ⎪
⎩ ε = 0.104
lower, whereas the vortex is offset by the approaching flow
fluctuation when the turbulence intensity is higher, and so at this (10)
time the interference effect was insignificant. Table 3 listed the
maximum and minimum envelope interference factor EIF and its
corresponding position in different turbulence intensities. It can be
seen that the most remarkable shielding region gradually transfers 3.4.2. Resonance case
from oblique upstream of principal building to a tandem upstream For the resonance case (Br=0.4, Hr=1.0), Fig. 18 shows the
distribution of envelope interference factor EIF in high turbulence

130
X.F. Yu et al. J. Wind Eng. Ind. Aerodyn. 159 (2016) 123–133

Fig. 16. Envelope interference factor EIF distributions of the peak base torsion responses in different turbulence intensities, (a) High turbulence intensity, (b) Standard turbulence
intensity, and (c) Low turbulence intensity.

Table 3 interference effect is more significant. d) The most remarkable region

The maximum and minimum EIF and corresponding location in different turbulence of the interference effects is at unstream tandem position about 2 times
intensities.
to 3 times of the width of principal building.
Turbulenceintensity The maximum value of EIF The minimum value of EIF Formula (11) gives linear regression result of the envelope inter-
and corresponding location and corresponding location ference factor EIF between the high turbulence and the standard
turbulence. It can be seen that under the working condition of vortex-
High 1.64 (2, 0.9) 0.76 (3, 0)
induced resonance, EIF between high turbulence and standard turbu-
Standard 1.74 (3, 0.9) 0.82 (1, 1.8)
Low 2.97 (3, 0.9) 0.68 (1, 1.8)
lence has a good correlation and the correlation coefficient reaches
92.8%, which is different from that of non-resonant cases (correlation
coefficient is only 72.2%).
2.7
Tested RIF = 0.434EIF (S) + 0.549 ; High turbulence, ρ = 0.928,
2.4 Fitting curve
ε = 0.067 (11)
EIF (low turbulence)

2.1

1.8
3.4.3. Comparison of two cases
1.5 Fig. 19 shows the comparison of linear regression results in non-
1.2 resonance and resonance cases. It indicates that the regression analysis
results in both cases are close in the same turbulence intensity, and the
0.9 resonance results in high turbulence intensity are smaller than the non-
0.6 resonance result in low turbulence intensity. In engineering applica-
tion, the simplified formula (12) can be used to analyze the torsional
0.3 interference effects between different turbulence intensities.
0.6 0.8 1.0 1.2 1.4 1.6 1.8
⎧ 2.327EIF (S) − 1.328; Low turbulence, non−resonance
EIF (standard turbulence) RIF = ⎨
⎩ 0.555EIF (S ) + 0.456 ; High turbulence
Fig. 17. Regression analyses of envelope interference factors EIF in between low and
(12)
standard turbulence intensities.

and standards turbulence. It can be seen that: a) The distribution law of

envelope interference effect in different turbulence intensities is almost
invariant, but the envelope interference factor EIF significantly change.
b) The amplification effect in standard turbulence is significantly
greater than that in high turbulence, and the former maximum
envelope interference factor (2.98) is nearly 1.5 times as that of the
latter. c) In the range of {2≤x/b≤7, y/b≤2.7}, the amplification
4. Wake vortex-induced resonance mechanism
The distributions of envelope interference factor EIF in different
breadth ratios Br have been analyzed in Section 3.1. It is found that the
interference law when Br=0.4 is significantly different from that in
other breadth ratios, and the amplification effect is more obvious at the

131
X.F. Yu et al. J. Wind Eng. Ind. Aerodyn. 159 (2016) 123–133

y/b y/b
1
5 5
1 4 4
1 3 3
1.1
2 2
1.1 1.2
1 1.3 1
1.4 1 1
1.6
x/b 1 2 1.8 x/b
11 10 9 8 7 6 5 4 3 2 1 A
0
0 A
0
0

(a) High turbulence intensity (b) Standard tu rbu lence intensity

Fig. 18. Envelope interference factor EIF distributions of the peak base torsion responses in different turbulence intensities (Br=0.4), (a) High turbulence intensity, (b) Standard
turbulence intensity.

Low turbulence (non-resonance case)

3.2
High turbulence (non-resonance case)
2.8 High turbulence (resonance case)

2.4

2.0
RIF

1.6

1.2

0.8

0.4

0.0
0.0 0.4 0.8 1.2 1.6 2.0 2.4
EIF (Standard turbulence)
Fig. 19. Regression analyses of envelope interference factors EIF in different turbu-
lences intensities.
most remakable interference position. In which, the envelope inter-
ference factor EIF is nearly doubled than that in non-resonance case,
which is caused by wake vortex-induced resonance.
Fig. 20 shows the variation of IF at the most significant interference
position (x/b=3, y/b=0) with reduced veloctiy Vr when Br=0.4. It can
be known that the interference factor IF increases significantly with the
reduced velocity, which is significantly different from the variation in
non-resonant case, and when reduced velocity reaches 4.3, i.e. nearly
reduced velocity in vortex-induced resonance case, the interference
factor IF reaches the maximum value. Fig. 21 indicates the correspond-
ing power spectral density, and the power spectra when breadth ratio
2.8 3.2
Fig. 21. PSD of generalized force in the most remarkable interference location.
Br=0.6, 0.8, 1.0 and isolated case are also plotted in the figure for
comparison. It can be seen that the reduced frequency corresponding
to basic wind pressure (0.9 kPa) coincides with the reduced frequency
corresponding to peak power spectrum of generalized force when
breadth ratio Br=0.4, which means the wake vortex-induced resonance
phenomenon occurs.
To further explore the occurring condition of vortex-induced
resonance phenomenon in different cross-sectional width of interfering
building, the following formula (13) can be obtained from formula (6):
VH
f0 =
Vr b (13)

While the vortex shedding frequency ft of the upstream interfering

3.0 building is:

2.8 St VH
ft =
Br b (14)
2.6
where, St is the strouhal number of the interfering building.
2.4
Thus, the ratio of vortex shedding frequency ft of interfering
2.2 building to natural frequency f0 of principal building can be obtained
IF

from formula (13) and (14):

2.0
ft SV
1.8 = t r
f0 Br (15)
1.6
When the frequency ratio shown by formula (15) equals to 1,
1.4 namely the natural frequency of principal building is consistant with
2.8 3.0 3.2 3.4 3.6 3.8 4.0 4.2 4.4 the vortex shedding frequency of interfering building, the vortex-
induced resonance phenomenon occurs. At this time, principal building
Vr
shows a greater torsional resonance response, and the interference
Fig. 20. Interference factors IF vary with the reduced velocity in the most remarkable factor is much higher than that of non-resonance cases.
interference location. For square section building in exposure category B, the observed

132
X.F. Yu et al.
Table 4
Reduced velocities as vortex-induced resonance occurs in different breadth ratios.
Br 0.3 0.4 0.5 0.6 0.8 1.0 1.2
Vr 3.33 4.44 5.56 6.67 8.89 11.11 13.33
strouhal number (St) is around 0.09 in wind tunnel tests, so the
reduced velocity is shown in formula (16) when vortex-induced
resonance phenonenon occurs:
Br Br
Vr = = = 11.11Br
St 0.09
Corresponding to the six kinds breadth ratios in present paper, the
reduced velocity of vortex-induced resonance is shown in Table 4. It is
shown that the reduced velocity Vr=4.44 when vortex-induced reso-
nance occurs when Br=0.4, which is similar to the maximum value of
the range of reduced velocities 2.91–4.31 in present paper. For the
other breadth ratios, the reduced velocities corresponding to vortex-
induced resonance are much greater than the range of reduced
velocities 2.91–4.31. That is why the envelope interference factor is
the biggest only when breadth ratio Br=0.4. in six kinds of breadth
ratios configurations.
Taking into account the practical engineering application, the
resonance reduced velocities when the breadth ratios Br=0.3 and
Br=0.5 are also added in Table 4. The corresponding basic wind
pressures are 0.55 kPa and 1.45 kPa, respectively, which are likely to
occur in coastal areas. However, the resonance reduced velocity is 6.67
when Br=0.6, and its corresponding basic wind pressure is 2.1 kPa,
which is much larger than that in most areas; In addition, there is no
interfering building with too small cross-section width (Br < 0.3) in
practice. Thus, it should be avoided that the reduced velocity of
principal building is in the range of 3.33–5.56 for interfering building
with small cross-section width (Br=0.3–0.5) in wind-resistant design of
high-rise buildings.
5. Conclusions
(1) In non-resonance case, with increased breadth ratio Br, the area of
significant interference effect is gradually closed to principal
building, and the largest envelope interference factor EIF was
1.9. The amplification effect is more and more significant and
maximum peak torque response increases by 42% due to the
channeling effect. The correlation of EIFs between Br=0.8 and
Br=1.0 is the best, and the correlation coefficient ρ reaches up to
90.2%.
(2) When breadth ratio Br=0.4, the wake of vortex shedding of
interfering building causes vortex-induced resonance of principal
building. The envelope interference factor EIF reaches the max-
imum value 2.98 when the reduced velocity Vr is 4.37.
(3) With increased height ratio Hr, the envelope interference factor
EIF tendes to increase. The amplification effect when Hr≥1 needs
to be highly paid attention. The EIFs in different height ratios Hr
has good correlation.
J. Wind Eng. Ind. Aerodyn. 159 (2016) 123–133
(4) In non-resonant case, the effect of reduced velocity Vr on torsional
interference effects is small. However, the torsional amplification
effect gradually increases with the increase of reduced velocity Vr
1.4
15.56 when vortex-induced resonance occurs. The maximum envelope
interference factor EIF when Vr=4.37 is 1.87 times as that when
Vr=2.91.
(5) With decreased turbulence intensity, the torsional amplification
effect is more and more significant. In vortex-induced resonance
case, the correlation coefficient ρ of EIFs in high turbulence
intensity and standard turbulence intensity is 92.8%, which is
much higher than that in non-resonant case (ρ=72.2%).
(16)
(6) For the interfering building with small cross-section width
(Br=0.3–0.5), it should be avoided that the reduced veloctiy of
principal building falls within the range of 3.33–5.56 to occur
vortex-induced resonance.
Acknowledgements
We gratefully acknowledge the joint support by the National
Natural Science Foundation of China (51408227) and the
Fundamental Research Funds for the Central Universities
(2015ZM001).
References
Aly, A.M., 2013. Pressure integration technique for predicting wind-induced response in
high-rise buildings. Alex. Eng. J. 52, 717–731.
Architectural Industry Press of China, 2012. Load code for the design of building
structures, (GB50009-2012), Beijing.
Armitt, J., 1980. Wind loading on cooling towers. J. Struct. Div. 106, 623–641.
Blessmann, J., Riera, J.D., 1985. Wind excitation of neighbouring tall buildings. J. Wind
Eng. Ind. Aerodyn. 18, 91–103.
Davenport, A.G., 1964. Note on the distribution of the largest value of a random function
with application to gust loading. Proc. Inst. Civ. Eng. 28, 187–196.
Hui, Y., Tamura, Y., Yoshida, A., 2012. Mutual interference effects between two high-rise
building models with different shapes on local peak pressure coefficients. J. Wind
Eng. Ind. Aerodyn. 104–106, 98–108.
Hui, Y., Yoshida, A., Tamura, Y., 2013. Interference effects between two rectangular-
section high-rise buildings on local peak pressure coefficients. J. Fluids Struct. 37,
120–133.
Kim, W., Tamura, Y., Yoshida, A., 2011. Interference effects on local peak pressures
between two buildings. J. Wind Eng. Ind. Aerodyn. 99, 584–600.
Lam, K.M., Leung, M.Y.H., Zhao, J.G., 2008. Interference effects on wind loading of a
row of closely spaced tall buildings. J. Wind Eng. Ind. Aerodyn. 96, 562–583.
Lam, K.M., Zhao, J.G., Leung, M.Y.H., 2011. Wind-induced loading and dynamic
responses of a row of tall buildings under strong interference. J. Wind Eng. Ind.
Aerodyn. 99, 573–583.
Xie, Z.N., Gu, M., 2004. Mean interference effects among tall buildings. Eng. Struct. 26,
1173–1183.
Xie, Z.N., Gu, M., 2005. A correlation-based analysis on wind-induced interference
effects between two tall buildings. Wind Struct. 8, 163–178.
Xie, Z.N., Gu, M., 2007. Simplified formulas for evaluation of wind-induced interference
effects among three tall buildings. J. Wind Eng. Ind. Aerodyn. 95, 31–52.
Yu, X.F., Xie, Z.N., Zhu, J.B., Gu, M., 2015. Interference effects on wind pressure
distribution between two high-rise buildings. J. Wind Eng. Ind. Aerodyn. 142,
188–197.
Zhang, W.J., Kwok, K.C.S., Xu, Y.L., 1994. Aeroelastic torsional behaviour of tall
buildings in wakes. J. Wind Eng. Ind. Aerodyn. 51, 229–248.
Zhang, W.J., Xu, Y.L., Kwok, K.C.S., 1995. Interference effects on aeroelastic torsional
response of structurally asymmetric tall buildings. J. Wind Eng. Ind. Aerodyn. 57,
41–61.
133