03 Lect 17 Wind TunWindTunnenel

5/21/2018 03 Lect 17 Wind TunWindTunnenel

1/22

Wind-tunnel techniques

Wind loading and structural response

Lecture 17 Dr. J.D. Holmes


2/22


Original wind tunnel of W.C. Kernot - 1893

Contraction

Propeller

Gas engine

Fly wheelAir jet

Model mounted on 3-wheel

carriage


3/22


Two types : open circuit and closed circuit

Open circuit type (fan downstream of test section) :

Blowing type - fan upstream of test section :

test section nearly at atmospheric pressure

Fan

Test Section

DiffuserContraction

FLOW

FlowStraightener

Screen


4/22


Simulation of atmospheric boundary layer :

Natural growth method :

Boundary layer is grown naturally over surface roughness elements

Boundary layer thickness is usually too small to model complete

atmospheric boundary layer - use auxiliary tripping devices

10-15 m


5/22


Simulation of atmospheric boundary layer :

Methods for short test sections :

Other devices : triangular spires , graded grids

Counihan method

Fins

Castellated

barrierhT

Roughness

~4hT


6/22


Simulation of surface layer (30hB

Barrier

Roughness

Useful for model scales of 1/50 to 1/200 (e.g. low-rise buildings)


7/22


Simulation of hurricane boundary layers :

Near eye wall : steep profile up to about 100 metres - then nearly constant

Can use non-hurricane boundary layer for rougher terrain in wind tunnel

simulations

Turbulence is higher in hurricanes (but patchy)


8/22


Simulation of thunderstorm downburst by impinging jet :

Jet Working

section

Contraction Diffusing

section

Vertical board

Blower

Stationary downbursts only are modelled - continuous not

transient


9/22


Modelling rules - dimensional analysis :

s are non-dimensional groups associated with flow and structure

Non-dimensional response/pressure coefficients =

f(1, 2, 3etc)

s should be matched in full scale and model scale

Examples of s :

Iu, Iv, Iw - turbulence intensities

Uh/ - Reynolds Number (is kinematic viscosity)

E/aU2 - Cauchy Number (elastic forces in structure/inertial forces in flow)

s/a - density ratio (density of structure / air density)


10/22



For example, reduced frequency :

Non-dimensional numbers may not be independent

i.e. proportional to square root of the Cauchy Number divided by the density ratio

s

a

2

a

2

s

s

.

U

EK

U

L.

L

EK

U

Ln


11/22



Scaling requirements might be relaxed

Not possible to obtain equality of all non-dimensional groups

Judgement based on experience and understanding of mechanics of the

phenomena

Quality assurance manuals and standards for wind-tunnel testing arenow available - e.g. A.W.E.S. , A.S.C.E.


12/22


Measurement of local pressures :

Local pressures - measurement done with single measurement tap

Fluctuating and short duration peak pressures must be measured

Area-averaged pressures with multiple pressure taps manifolded together

Multiple input tubes -

single output tube toelectronic pressure sensor


13/22


Area-averaged pressures :

Discrete averaging overestimates continuous average fluctuating loads

Rddiscrete

averaging

Rccontinuous

averaging

RdRc

Assumed correlationfunction = exp (-Cr)

B

B

2

1.8

1.6

1.4

1.0

0.8

0.6

0.4

0

0 2 4 6 8 10

CB

Variance of averaged

panel force to variance

of point pressure

Overestimation depends on

correlation between point

pressures on the area


14/22


Frequency response of measurement system :

Require amplitude response ratio equal to 1.0 ( +/- small error) over a

defined frequency range

0

0.5

1

1.5

0 100 200 300 400

Frequency (Hertz)

Am

plituderatio System within +/- 15%

limits to 150 Hertz


15/22


Frequency response of measurement system :

Require phase response to vary linearly over a defined frequency range

0

50

100

150

200

250

300

350

0 100 200 300 400

Frequency (Hertz)

Phase

lag

(degrees

)

Time delay =(1/n) (phase angle / 2)

For constant time delay, phase

angle should be proportional tofrequency, n


16/22


Types of tubing systems :

Short tube : high resonant frequency

but amplitude response rises fast

Restricted tube :

restrictor tube damps resonant peak

Leaked tube :high pass filter, mean response is

also reduced

Transducer

volume

(a) Short tube

Restrictor

(b) Restricted

tube

Controlledleak

(c) Leaked tube


17/22


Overall loads on tall buildings :

Two techniques : aeroelastic models - resonant structural response is scaled

Base-pivotted aeroelastic model :

Gimbals

Springs

Strain

gauges

ElectromagnetAluminium

disc

h

motion of building in sway modes of vibration are reproduced - hence aeroelastic

(e.g. aerodynamic damping) forces are included

Uses equivalence of rigid body

rotation and movement of tall

building in first mode with linear

mode shape

Model should be scaled to have the

same density


18/22


Overall loads on tall buildings :

model building supported on a stiff

base balance to measure

aerodynamic applied forces

high-frequency base balance :

spectral densities of applied base

bending moments are measured andused to compute resonant

components in sway modes

mean and background aerodynamic forces only are measured

h

Six component strain gauge

balance

requires mode shape corrections,

special processing for coupled

modes, linked buildings


19/22


High-frequency base balance :

Support system should be made very stiff, and building

model light to keep frequency above measurement range

Frequency relationships :

U1(>U2)

U2

Simulated

building

frequency

Modelfrequency in

wind tunnel

Spectral

density

Usable frequency

range formeasurements


20/22


Full aeroelastic models :

Elastic properties are concentrated in a spine to which

non-structural segments are attached to give correct

aerodynamic shape and mass

Slender structures such as bridges and towers

Length scale ratio and velocity scale ratio chosen to suit

size and speed range of wind tunnel

Frequency then obtained by equality of reduced velocity :

p

s

m

s

U

Ln

U

Ln

Stiffness of spine obtained by requirement to keep

frequency of structure equal in model and full scale


21/22


Full aeroelastic models :

Segmented tower legs and deck


22/22

End of Lecture 17

John Holmes225-405-3789 [email protected]

Documents

03 Lect 17 Wind TunWindTunnenel