|
|||||
New Zealand Engineering 1999 March Construction You can Teach a Dog New Tricks Behaviour of multi-storey steel frames
buildings in severe fires The title also reflects the fact that while steel has a long and well established history in structural and civil engineering, major advances in research, design guidance and fabrication capability are making it one of the most dynamic and versatile structural materials going into the new millennium. (This needs to be stated up front, for the benefit of those uncharitable readers who may be thinking that the "old dog" refers to the author!) Behaviour
of multi-storey steel frames buildings Design of multi-storey steel framed buildings for fire resistance has traditionally been undertaken on the premise that the building will collapse unless the beams and columns are insulated from temperature rise under fully developed fire conditions. Since 1990, a growing body of evidence from both actual fires and realistic fire tests has indicated that this is not likely to be the case, at least for buildings with uninsulated floor support beams supporting a composite concrete slab and insulated (passive fire protected) columns. Our understanding of the behaviour of such buildings in severe fires and the inelastic reserve of strength available from them has increased dramatically during the 1990s. This understanding has come from actual fires, realistic fire tests, analytical modelling of buildings and has led to the formulation of a draft design procedure for high temperature response. The actual fire case histories and realistic fire tests are from overseas, especially the United Kingdom, however, HERA is actively involved in advanced analytical modelling work in conjunction with the University of Canterbury and we have formulated the draft design procedure mentioned above. An illustration of the inelastic reserve of strength available from a complete steel-framed building is shown in Figures 1 and 2. The test building is a full-sized replica of a typical modern eight-storey office building and was tested in a joint British Steel, Building Research Establishment project at the BRE Cardington Large Building Test Facility in Bedford, England. A series of large-scale, realistic, fully developed fire tests were undertaken on the second level of this test building to determine the inelastic nature of building response and the inelastic reserve of strength available. Fuel used in the example shown was actual office furniture, with 20 percent plastics content, plus wood cribs, to develop a fire load energy density equivalent to 45.6 kg of wood/m2 of floor area, which is near the maximum credible limit for an office. The columns were protected, the beams and floor slab unprotected and all floors above the fire floor were loaded to slightly in excess of the New Zealand fire emergency design dead and live load. The aim was to generate maximum structural fire severity and hence the greatest inelastic response. This was achieved; Figure 1 shows the distorted floor system following the fire and Figure 2 the very high temperatures reached in one of the primary beams and the fire (gas) temperatures reached at that location. The peak steel temperature recorded was almost 400oC in excess of the limiting temperature, for that member and load condition, given by NZS 3404 (the Steel Structures Standard), however, there was no local or overall collapse and the floor continued to function as an effective fire separation. Following the fire, the floor above the fire floor was load tested and was still able to resist the full factored limit state design dead and live load without failure. We have since been able to produce a pattern of primary beam behaviour through advanced analysis similar to that observed experimentally, using a 17-storey Auckland office building as the model. These studies have included the influence of lateral loading (eg. from wind) in conjunction with fire, variations in the steel strength, the fire load, the fire spread and the available ventilation. We have also developed a draft design procedure using assumptions relating to member temperature distribution and behaviour. These assumptions are currently being compared against the UK Cardington LBTF test results and modifications to the procedure being made as required. The ultimate aim of this research is to publish general design and detailing requirements to allow the routine use of unprotected steel floor support beams in multi-storey office and general commercial buildings. While this is still some time away, tangible benefits are already flowing through to the design community from this fire engineering research, such as:
In addition, the HERA Steel Structures Analysis Service has undertaken multi-variable regression analysis on the corrosion rate test data previously obtained by BRANZ at 156 sites around New Zealand, to determine the relationship between the first year corrosion rate and readily available site location and meteorological data. This allows the first year corrosion rate for structural steel at any site to be determined, including allowance for microclimatic effects. Rapid connection design HERA Report R4-96, Structural Steelwork Estimating Guide, provides a comprehensive method and the necessary data for accurate pricing of steelwork. Connections form a significant component of steelwork cost and so R4-96 contains connection component sizes and details, for costing purposes. The HERA Steel Structures Analysis Service has now gone to the next stage and published HERA Report R4-100, Structural Steelwork Connections Guide. This guide presents standard details and design capacities for the range of commonly used connections in HERA Report R4-96. Use of the guide in selecting, documenting and detailing steelwork connections will save considerable money and design time. Between 70 and 90 percent of the connections in a typical multi-storey steel building are covered by Report R4-100. CAD details of the connections are included in this report. G Charles
Clifton, HERA structural
engineer |
|||||
 |
|
|
|
