CYCLONE ‘LARRY’ March 2006
Dr. John Holmes of JDH Consulting spent three days in the Innisfail district of North Queensland in March 2006 investigating the effects of Cyclone ‘Larry’ to buildings and other structures, shortly after the event. He accompanied researchers from the Bureau of Meteorology, Geoscience Australia and the James Cook Cyclone Testing Station.
Although initially classified by the Bureau of Meteorology as a Category 5 cyclone at landfall, subsequent analysis of recorded wind speeds and the failure of simple structures such as road signs indicated that, in fact, it was a low Category 4 event. The original assessment as a Category 5 was of considerable concern to Dr. Holmes, in his role of Chair of the wind actions sub-committee of Standards Australia, as the current design wind speed for ultimate limit states design on the Queensland coast (Region C) in the Australian Wind Actions Standard is at the Category 4 level. A Category 5 cyclone has never previously been recorded as crossing the Queensland coast.
For the general maximum gust wind speeds of 50-60 m/s (10 metres in open terrain) seen at locations near the eye wall of the cyclone, the performance of newer structures (built since Cyclones ‘Althea’ and ‘Tracy’ in the 1970s, and ‘Winifred’ in 1986 (also affecting the Innisfail area) was generally satisfactory. However, some newer houses on a ridge in East Innisfail had failed – probably due to the lack of account of the topographic speed-up effects, as required by AS/NZS1170.2.
Many older buildings had failed, however; the failures were very often attributed to the development of high internal pressures, resulting from debris damage to window glass, or the failure of roller doors due to direct wind pressure. The latter has been of particular concern to the Standards Committee for some time, and it has introduced an amendment to AS/NZS1170.2 requiring roller doors to be treated as dominant openings for internal pressure assessment. This would be applicable to all types of extreme winds, not just cyclones. Unfortunately roller doors have a very poor track record in extreme wind events; John Holmes believes that it should be possible for manufacturers to supply cyclonic resistant locking/retaining systems for roller doors.
Prevention of debris generation due to roof failure of older buildings, and of small structures such as garden sheds was important, as large debris items, flying at speeds approaching cyclonic wind speeds, had the potential to cause failure of newer buildings, that had themselves been correctly designed for wind loads exceeding those experienced.
The performance of engineered non-building structures was generally good
in ‘Larry’, with the notable exception of a chimney stack
at the Mourilyan Sugar Mill. The latter was apparently suffering from
corrosion. Interestingly the chimney was equipped with helical strakes.
The latter are effective at mitigating cross-wind vibrations at low-medium
wind speeds, but would have significantly increased the wind loads during
the Cyclone, because of the increased drag coefficient. Failure of some
small communications towers, including a mobile telephone tower, were
also observed during ‘Larry’.