Offshore fire research is fundamental to the safety of installations and personnel in the offshore environment. In the UK and allied nations where oil and gas exploration and production activities upsurge, the likely release of hydrocarbon fuels and the resulting fire culminating in 'domino effect' has been a primary concern researchers try to handle. Most efforts to address this issue tend to provide solutions by recommending measures such as active and passive fire protection systems. The provision of these protections to structural components means additional weight, increased dimension and further cost. However, a large scale fire testing programme recently conducted at the Health and Safety Laboratory, Buxton as part of the Joint Industry Project (JIP) to investigate fire spread hazards of hydrocarbon pool fires has shown that it is possible that structural integrity may not be achieved within the fire resistance period for an unprotected steel deck. It is on this premise that this research attempts to investigate plate behaviour in a stiffened steel-plated deck under fire conditions. In an attempt to address this problem, a scoping study was conducted with some structural steel elements to determine the prediction level of the proposed numerical tool in calculating thermal and mechanical responses. The results gave an indication of the effective use of ABAQUS analysis code in modelling steel structures in fire conditions.To study this phenomenon, localised and uniform fire load models for spilled hydrocarbon fuels were developed to characterise the heat loads applied onto the deck structure. The aim of the research was mainly in three parts. Firstly to assess the behaviour of pool fires and then develop a method for characterising thermal radiation models of open, hydrocarbon pool fires. Secondly to apply this method to develop fire load models that include both the heating and the cooling regimes and consequently use these fire models as thermal loadings in modelling the response of steel-plated decks under fire conditions. Thirdly to conduct an extensive test to characterise the material property model of the butt-welded joints made with high strength steels at elevated temperatures in order to evaluate the performance of the abovementioned joint in this class of structure. The research findings include the following observations. From the thermal radiation model it is noted that experimental results from pool fire tests conducted at different times and locations can be very different. It is therefore suggested that when developing the thermal radiation models all the sub-models that are used to describe the geometric characteristics of the fire must be clearly defined. An experimental programme aimed at characterising the stress-strain model for butt-welded joints made with high strength steels shows that the model described in Eurocode 3 may not be suitable for this class of steel. Hence, it is recommended that further investigation should be conducted to review the existing model. An extensive numerical study was carried out on a simulated plated deck under running pool fires. It was discovered that the deformation capacity of the plated deck may be greatly dependent on the type of fire and its heating and cooling rates. It is also suggested that if hydrogen cracking is present in the weld, it is most likely that when the deck is stressed, like in the case of the HSL fire test, the weld would crack. To avoid the problem of hydrogen cracking, it is suggested that the appropriate welding consumables with less hydrogen content are used. The post-weld treatment method can also be applied to limit the formation of cold cracking within the weld. The above suggestions and recommendations are drawn from the results of the analyses to enable possible guidance for fire design of structure involving high strength steels.
|Date of Award||1 Aug 2014|
- The University of Manchester
|Supervisor||Colin Bailey (Supervisor)|