AGB seminar room 3.01
An integrated computational environment for simulating structures in real fires
Liming Jiang, The University of Edinburgh
To protect the life and livelihoods of people in the built environment engineers are required to provide built infrastructure with a reasonably adequate level of resistance to the extremes of loads on structures resulting from natural and/or manmade hazards, such as earthquakes, tsunamis, hurricanes, fires etc. Given the complexity and uncertainty associated with these loads, traditional practice has often relied upon gross simplifications in representing them and enforced these representations through building codes and regulations, leading mostly to conservative designs. This however is not always the case as structural failures, where deficiencies of such prescriptive approaches can be identified as the cause, occur with an unnerving regularity. In recognition of this fact, there is an increasing consensus in the profession of structural engineering that the enforcement of sweeping simplified rules is perhaps not satisfactory in all cases. When the structure and the loadings expected to act upon is considered to be sufficiently complex and uncertain, engineers should be allowed, or even required, to consider alternative methods of engineering resistance delivering improved robustness (degree of safety against disproportionate failures) and resilience (the ability to recover rapidly from a disruptive event). This however is easier said than done, as alternative or the so-called performance-based engineering (PBE) approaches demand a much greater level of knowledge and understanding, supported by advanced computational tools for engineers to quantify as accurately as possible, the magnitudes of the expected loading and the response of the structure (ideally also quantifying the key uncertainties associated with the demand and resistance).
This presentation will provide details of an integrated computational environment for simulating structures subjected to real fires, being developed at Edinburgh (in collaboration with partners at Tongji University in China). This tool is based upon the OpenSees software framework developed at University of California, Berkeley, the reasons for this choice will be outlined. An introduction to the key issues associated with computational modelling of structures subject to fires will be discussed, followed by examples of previous modelling work undertaken on other platforms and discussion of the deficiencies of those approaches. The vision for the new tool and its conceptual structure will then be presented followed by the progress so far with examples and a discussion of the ongoing work.
Event Contact Name: