No_1_January_1969 — Page 34

Direct open space velocity recorders. (Velocity Meters)

Open space force

recorders. (Anemometers)

WIND

Applied force

recorders. (Pressure Gauges)

Structural response

recorders. (Sway Gauges ) Strain Gauges

Statistical Theory

Statistical Theory

Statistical Theory

Statistical Theory

Velocity Spectra.

Theory

Open Space Pressure

Spectra.

Applied Pressure

Spectra.

Theory

Structural Response

Spectra.

Aerodynamic Theory

Study of:

Statistical Theory

(1) Aero-admittance of point

Computed open space pressure spectra.

(2) Dimensional aero-admittance

of test building.

Study of:

(1) Mechanical Admittance of test

building,

(2) Structural Dynamics,

(3) Damping characteristics.

METEOROLOGY

Study of:

(1) Aerodynamic Theory

(2) Aero-admittance of velocity

meters.

(3) Aero-admittance of Anemometers.

Fig. 2. Idealised flow diagram for study of wind effects on structures

plicated conditions prevailing in a wind

storm.

Such random functions are only known in the past, that is, up to the present instant in time, but never in the future. Nobody can predict the wind velocity five seconds hence, for it is simply beyond our capability to express it as an explicit function of time in the future. But we can analyse its past record and on a statistical basis we can make predictions into the future.

The limited number of observations made to date indicate that wind load- ing is random and stationary and this has led to the adaptation of the method of spectrum analysis for stationary random time series for solution of problems associated with wind load- ing. By comparing wind spectra with pressure spectra, and pressure spectra with response spectra, the aerodyna- mic admittance and the mechanical admittance of the structure may be obtained. Then, by using the extreme value theory gust factors may be ob- tained.

The spectral relation between load- ing and response calls for a knowledge of:

(i) the natural frequencies of the building

sure on the surface of the building, and cross-correlation of pressures at different points.

(iv) the behaviour of building ma- terials under dynamic loads.

Although application of the theory of random functions to the problem. of wind loading seems logical, its vali- dity still requires checking. The as- sumptions are made that natural wind is a stationary, ergodic, Gaussian ran- dom process. Stationary and ergodic mean that it is independent of the time origin and that any element of the ensemble is representative of the ensemble as a whole.

Acceptance of these assumptions means that analysis of a short-term wind record, taken at any time, gives characteristics which are representative of the atmospheric turbulence at the recording point and this seems reason- able. But the assumption of a Gaussian distribution requires elaboration in ob- servation in order to verify the validity of the assumption, and the limited observations made to date in Hong Kong do not appear to justify this assumption.

Full-scale research

It is because of these considerations that a full-scale experimental research (ii) damping characteristics of the is being currently undertaken in Hong building

Kong. The scope of this work is il- (iii) the distribution of wind pres- lustrated diagrammatically in Fig. 2.

38

Conclusions

STRUCTURAL DYNAMICS

An experimental building is now under construction at Cape D'Aguilar. It is rectangular in shape, measuring 60 ft. x 30 ft. on plan and is ten storeys or 100 ft. tall. The framework is of exposed structural steel construc- tion with beam and column connec- tions fully welded. The floors are part- ly of reinforced concrete, solid slab construction and partly of composite construction and the roof consists of a solid reinforced concrete slab.

Glass curtain wall cladding has been adopted for the four vertical faces except for a 4 ft. depth next to the foundation raft where light- gauge metal panels have been sub- stituted to avoid damage due to flying debris. A light-gauge metal parapet is fitted on to the roof with detach- able metal panels which may be varied in height from zero to 4 ft. in order to measure the effects of parapet height on the negative pressures de- veloped on the roof.

The structural framing and floor and roof slabs are so constructed that the building can be divided over its full vertical height into four sections, as shown in Fig. 4, each capable of acting completely independently of the others. Under high-velocity winds the separate sections will be coupled to- gether with shear connectors at every floor-level so that the whole building can act as a single monolithic unit.

Far East BUILDER, January 1969.