l

THE I TNTVFR SIIY OF HONG KONG

B ACHEI .oR OF ENGD,JEERING· r EVEr. m EXAMINATION

DEPARTMENT OF CIVIL ENGINEERING

WIND ENGINEERING (CIVE3002)

May 16, 2002 Time : 9 : 3 0 -am - 12 : 3 0 pm 3 hours

Answer FOT ffi questions.

All questions cany equal marks.

I Ise of Electronic Calc1 dators·

Candidates taking examination that permits the use of calculators may use any calculator
which fulfils the following criteria:
(a) it should be self-contained, silent, battery-operated and pocket-sized;
(b) it should have numeral-display facilities only and should be used only for the purpose of
calculation~
(c) it should not have any printing device, alphanumeric keyboard, or graphic display; and
(d) it should not contain any recorded data or program.

It is the candidate's responsibility to ensure that the calculator operates satisfactorily and the
candidate must record the name and type of the calculator on the front page of the
examination scripts. Lists of permitted/prohibited calculators will not be made available to
candidates for reference, and the onus will be on the candidate to ensure that the calculator
used will not be in violation of the criteria listed above.

3
Density of air= 1.204 kglm
5 2
Kinematic viscosity of air= 1.5 X 10- m /s

l. To assess the wind loads and aeroelastic instability of a suspension bridge, a section
model is tested in the wind tunnel as shown in Fig. Q1a. The length of the section
model is 2 m and the geometric scale ofthe model is 1:150. The span ofthe prototype
bridge is 800 m and the full-scale natural frequencies of the bridge in the up-and-down
bending mode and the torsion mode are 0.14 and 0.19 Hz, respectively. By careful
adjustment of the suspension mechanis!l\ the natural frequency of the section model in
the up-and-down movement is tuned to 2.1 Hz.

(a) What should be the torsional natural frequency of the section model in order to
simulate the flutter behaviour? How is this frequency tuned in the suspendion
mechanism. Explain the mechanism of flutter in a suspension bridge deck.
(7 marks)

- 1 -
(b) During the aerodynamic instability tests, flutter instability is first observed to
occur at 10.5 m/s in the wind tunnel. Estimate the corresponding full-scale wind
speed for the onset of instability. Sketch the typical wind tunnel result of how
the vibration levels of the section model vary with wind speed and briefly explain
how the onset of instability can be identified from the result. (6 marks)

(c) Fig. Q 1b shows the non-dimensional wind turbulence spectrum. Sketch the
wind force spectrum (nSJJ(n)ld against n) on the suspension bridge deck at the
flutter wind speed. Take 200 m as the integral scale of turbulence. :Highlight the
natural frequencies of the bridge on the spectrum. If the suspension of the bridge
is made stiffer, what will be the effect on the flutter wind speed? (6 marks)

(d) With the suspension mechanism replaced by a force balance, static wind load
measurements are made on the section model. Under a wind tunnel speed at 7. 5
m/s, a drag force of3.5 Nand a torsional moment of0.14 Nm are measured on
the section model. Estimate the drag force and torsional moment on the entire
800 m span of the prototype bridge when it is under a full-scale wind speed of
80.0 m/s. (6 marks)

Fig Qla

tilitl
0.35
T
"b
(ij
c::
0.30
Il
-inSuu(n)/o}
I
~~
1C I

\~
2!-
·u; 0.25
I I

c::
Cl) I I
"0
e13 : I J I i '
0.20
Cl)
I IJ 1\
a.
rJl
-cac:: 0.15
7 I! I i\ ! ! I
'l I r\ I i !
0 ! !
(ii
c::
ID 0.10
I l 11 ! I I
~~
F

' i
E j I
~ I
z
c::
0 0.05 ......
/: I
'~ ....... ........
"""'
I i

0.00
I
10'2 2 3456710·1 2 34567100 2 34567101

Reduced frequency, St=nlu.x"V

Fig Q 1b Non-dimensional turbulence spectn 1m

- 2 -
2. Table Q2a shows the wind record of annual maximum hourly mean wind speeds in a city
over 20 years. The record is from anemometer station location on an open land terrain
and the measurement height is 10 m above ground. Turbulence intensity is measured to
have a value of 0. 15 at this height.

(a) Carry out an extreme wind analysis to obtain the parameters of the extreme wind
distribution of annual maximum hourly mean wind speeds at the anemometer
station. Estimate the mean wind speed values with at return periods of 10 years,
l 00 years and 1000 years. ( 10 marks)

(b) A building, 150 m tall, is to be built on a site of the city terrain type in the same
city. Estimate the lOO-year return mean wind speed at the roof height of the
building. Use the wind characteristics given in Table Q2b. The power law is
used to describe both the variation of mean wind speed with height and the
variation of turbulence intensity with height. ( 5 marks)

(c) Estimate the turbulence intensity and the 100-year return 3-sec gust wind speed
at the roof height of the building in (b). (4 marks)

(d) If the expected life span of the building is 3 0 years, what is the probability that
the 100-year return wind speed values will be exceeded twice or more times in
the life span? (3 marks)

(e) If a risk of 0.05 is to be taken for the life span of 30 years, what return period
should be used? (3 marks)

Table Q2a Wind Record

20-year record of annual maximum hourly mean wind speed

(m/s) in ascending order
32.5 34.5 35.0 37.0 37.5
37.5 38.0 39.0 40.0 40.0
41.5 42.0 42.5 43.5 43.5
45.0 46.0 47.0 48.5 51.5

Table Q2b Wind characteristics

Terrain type: Open land City

Gradient height, m: 300 500
Zero-plane displacement, m 0 10
Power law exponent for mean wind speed profile: 0.15 0.36
Power law ex onent for. ur ~~ ~rit- ~r.t, .s-t. -0.15 -0.36
Gust wind speed (3 sec.):

- 3 -
3. A rectangular building, 180 m tall, 40 m by 30 m in plan, is to be built on a site of the
sub-urban terrain type. The variations of mean wind speed, 3-sec. gust wind speed, and
turbulence intensity with height are described by the power law as:
[J (z) ( z Jo.25
50m/s = 400m
U(z) ( z )0.14
70m/s = 400m
I u ( z) ( z ) -o .25
[Link] = 400m
When wind blows nonnally onto the wider face of the buildings, the following pressure
coefficients are to be used:

windward wall: Cp=O.S, qref= q(z),

leeward wall: Cp=-0.3, qref= qh,
side wall: Cp=-0.6, qref = qh

For wind pressure on the windward wall, the reference dynamic pressure varies with
height while for the other walls, the dynamic pressure at the roof height should always be
used. On the side walls, a local pressure factor of2.0 is to be applied to cladding located
within a distance of 5 m from the windward wall edge.

(a) Explain why it is needed to apply a local pressure factor on the 'Windward edge
of the side face. (2 marks)

(b) Calculate the mean base shear force and base overturning moment on the entire
building at this wind incidence. (6 marks)

(c) Calculate the design base overturning moment on the building using the quasi-
steady equivalent-gust method. Apply an area reduction factor of0.85.
(6 marks)

(d) Explain why an area reduction factor should be used in(~). (2 marks)

(e) Now the building is considered mildly dynamic with a damping ratio of 0.02.
Estimate the design base shear force and base overturning moment on the
building again using the dynamic gust factor method. The equations for the
method are given at the end of this question paper. (6 marks)

(f) The building is clad with curtain walls of unifonn size of 3 m by 3 m. Estimate
the largest design force on a wall cladding, allowing an internal pressure
coefficient of +0.2 or -0.2. (3 marks)

- 4 -
4. A tall building with a crucifonn-shaped plan is to be built in a suburban part of a city
(Fig. Q4). There are mostly low-rise buildings in the area. A wind tunnel test is carried
out to estimate the pedestrian-level wind environment of the proposed tall building. It is
found that point A has the windiest wind environment. The measured mean wind speed
ratios at point A are listed in Table Q4a. The parent wind data of the city are shown in
Table Q4b. It shows the distribution of hourly mean wind speeds at the gradient height
over 10 years. Very little directional variation is observed.

(a) Fit an appropriate statistical distribution to the parent wind data and estimate the
mean wind speed at the gradient height with a probability of exceedance of (i)
So/o and (ii) 10%. (7 marks)

(b) Estimate the probability that the mean wind speed at the pedestrian level at point
A wil1 be higher than 8 m/s. (7 marks)

(c) Explain why point A has the windiest pedestrian-level wind environment.
Furthennore, explain the observed variation of ground-level wind speed with
wind direction at point A (Table Q4a). (4 marks)

(d) Locate other regions where high pedestrian-level wind speeds are likely to occur
around the building. (3 marks)

(e) Suggest some architectural precautions and remedial measures to reduce the
level of high pedestrian-level winds. (4 marks)

A
•

Fig Q4 Test building, wind direction and measurement point A

- 5 -
Table Q4a Ratios ofgronnd-level wind speed at point A to the gradient wind speed
(meaSJJred in the wind tJmnel test)

Wind angle, B(deg.) UA/Ug

0 0.159
30 0.201
60 0.268
90 0.552
120 0.725
150 0.437
180 0.248
210 0.543
240 0.635
270 0.143
300 0.106
330 0.193

Table Q4b Parent wind data (1 0 years ofhm1rly mean wind speed at gradient height)

Mean wind speed, Ug, m/s Number of hours

V<3 4920
3 <V<6 11400
6 <V <10 18980
10<V<16 26340
16 <V<22 15390
22 <V< 28 7290
28 <V<34 2610
34 <V< 41 390
V>41 280
Total: 87600

- 6 -
5. (a) A wind tunnel for wind loading studies have a working section 15 m long, 3 m
wide and 2 m tall. Suggest a suitable geometric scale of the wind tunnel tests.
Do you think it is possible to simulate the full atmospheric boundary layer by the
natural growth of a turbulent boundary level in the wind tunnel? If no, what can
be done to achieve the simulation of the natural wind? (5 marks)

(b) The roof of a grandstand is cantilevered at the rear of the grandstand. What is
the most probable aeroelastic effect on the roof? Describe the mechanism.
(5 marks)

(c) A telephone pole is 10 m tall and 0.4 m in diameter. It has a first-mode

structural frequency at 10 Hz. At what wind speed will the pole undergo severe
vibration due to the lock-in mechanism? Along which direction will the wind-
induced vibration occur? If the structural frequency were one order of
magnitude higher or lower, what would be the significance of lock-in vibration?
Take a value of St = 0.2 for the Strouhal number of vortex shedding from a
circular cylinder. ( 5 marks)

(d) The dynamic gust response factor method due to Prof A. G. Davenport is often
used to estimate the dynamic wind load of a tall building. Describe the principle
behind the method by explaining the relations connecting the velocity spectrum,
force spectrum and response spectrum.

The expression for the dynamic gust response factor G is given at the end of this
examination paper. Explain the meanings of the various terms, i.e. g, lu, B, S, E,
c;, in the expression. How is G used in the estimation of dynamic wind load?
(5 marks)

(e) Give five examples to illustrate the effect of atmospheric stratification on the
dispersion behaviour of a plume. In relation to the situation of stack emission
from a chimney on the roof of a building, explain briefly the effect of
"aerodynamic downwash". How can it be avoided and is there an even worse
situation? (5 marks)

END OF PAPER

- 7 -
Equations and Infannation:

Dynamic gJtst factor:

(h is the height in metres of the building and U }t is the mean wind speed at that height. b
and windward breadth of the building. Tis normally taken as one hour or 3600 sec.)

G = 1+2glu
H
1
·
'

0.47 N

Lh = 1000 ( -
h J
0.25
·
10 '
46
n =-
o h

Fisher-Tippet Type 1 distribution·

P(U) = exp(-exp(-a(U -M)))

Weihr dl distrihr rtion ·

P(U) 1 exp( -(U I C)k)

Dynamic magnification factor:

- 8 -