TY  - UNPB
AV  - public
EP  - 361
A1  - Littler, JD
UR  - https://discovery.ucl.ac.uk/id/eprint/10123414/
N2  - Although previous attempts have been made to make full-scale measurements of the response of tall buildings to wind loading, none of them have obtained data of sufficient accuracy to enable the validity of prediction methods to be assessed. This thesis gives details of the major full-scale tests that have been carried out and the reasons why the data obtained was not of high accuracy.
The dynamic characteristics of Hume Point, the building selected for the long-term testing, were found by forced vibration testing at a number of amplitude levels. The results from this testing were used to calculate suitable recording parameters for the acquisition and analysis of data. Hume Point is 66.9m tall and over 300m away from other buildings of comparable height.
Anemometers mounted on a 20m mast on the roof of the building and up to fourteen accelerometers were used to measure the wind impinging on the
building and the building's dynamic response respectively.	Two tracking 
lasers were used to observe the motion of a plumb line suspended in the building and so determine the building's quasi-static response. Thermistors were used to measure the temperature of all four faces of the building.
Over 2,250 hours of data were obtained in almost 8,000 records each 1024 seconds long. Response spectra calculated from records obtained under
similar wind conditions were averaged together.	Using this selective 
ensemble averaging technique, the rms acceleration of Hume Point was obtained to an accuracy of ±10% for a range of wind speeds and directions. The response in all three fundamental modes was found to be proportional to the wind speed raised to between the power of 2.7 and the power of 3.3. Allowing for possible variance error, the response of Hume Point is symmetrical about both the west and north faces except for winds blowing directly onto the east or west faces. The response in all three fundamental modes is greatest for winds blowing onto the west face.
Better correlation was obtained for mean temperature difference between opposite faces and mean quasi-static displacement than for mean wind speed
and mean quasi-static displacement.	In general the correlation between 
displacement and wind speed within an individual record was better than that for the mean values of records collected under similar wind conditions.
Both wind tunnel tests and a recent calculation method predicted the along-wind response of Hume Point to within ±50% when the correct natural
frequency and other building parameters were used in the predictions.	The 
wind tunnel tests predicted the across-wind response to within a factor of 2, whilst the calculation method overestimated the actually measured
across-wind response by between 1.6 and 5 times.	However, using the 
actually measured building parameters can give a false idea of the accuracy
of prediction methods.	For example, using the suggested estimates of the 
building parameters rather than the actually measured values increased the overestimate of across-wind response predicted by the calculation method to between 5.5 and 12 times.
M1  - Doctoral
PB  - UCL (University College London)
ID  - discovery10123414
Y1  - 1991///
N1  - Unpublished
TI  - The response of a tall building to wind loading
ER  -