Seismic response of the built environment in densely populated urban areas

Earthquakes are a perennial hazard to engineered structures worldwide. This is particularly true of densely populated urban areas (cities) where recent large magnitude earthquakes have demonstrated that the density of structures has a significant multiplicative effect on the cost of seismic damage. It is common practice at present to determine seismic response of structures in isolation from the urban environment which surrounds them. In reality, buildings in densely packed urban areas are often surrounded by other structures which may either be similar in design or very different (e.g. a 40-storey building next to a 5 storey building). An urban area with a block of adjacent buildings may be considered as a complex system of structures which may interact with each other through the ground which connects them. The actual seismic response of a structure within a dense urban environment may be significantly different (and larger) to that predicted using current methods based on isolated structural behaviour. If the adjacent structures are closely enough spaced and vibrate out-of-phase with each other, damage can subsequently occur due to 'pounding' - transient collision between adjacent structures. This will be worsened if structure-to-structure interaction increases the building response compared to an isolated structure. Furthermore, the close packing/spacing of structures also makes the consequences of a catastrophic failure more critical, as significant collateral damage to the adjacent structures may occur. Although previous research into this area has provided good initial insights into structure-soil-structure interaction using numerical simulation, there remain a number of issues to be addressed before this knowledge and understanding can be practically applied to engineering design in densely-packed urban areas. Notably, this includes whether current elastic idealisations of soil behaviour are appropriate during large earthquakes when strains may be in the inelastic range, and whether the appropriate controlling parameters have been selected. The work described in this proposal will consist primarily of a programme of physical model tests to be conducted on small scale models of adjacent structures founded on real soil within a geotechnical centrifuge to address these needs. This work will represent the first systematic experimental study of the fundamental phenomena involved for structures founded on real soil. The results will subsequently be used to assess the validity of using current models for isolated structures to assess buildings in densely packed urban areas and assess the range of earthquake conditions for which current structure-soil-structure interaction models based on linear elastic soil behaviour are applicable. The database of physical test data will additionally provide a useful database of observed physical behaviour which may be used to calibrate future more advanced numerical modelling work.

The results of the study will be of relevance to town planners, government and insurers who are faced with seismic hazards and who are responsible for the huge value of the building stock of population centres located in seismic regions, especially in earthquake prone cities. Improved estimates of the risk posed to both the individual structures within urban areas and to the areas themselves as a whole will reduce the costs and impact associated with insuring building stock against this type of natural disaster and assessing seismic risk. By providing improved estimates of seismic response, remedial action can be more effectively targeted. Although the impacts detailed above will only be fully realised by completing an extended body of research in this area over a longer period, the work proposed represents an important step towards achieving these goals and will underpin the development of such a body of research. Within the UK, consulting engineers and academics will be the main beneficiaries. The impacts of the proposed work for these beneficiaries will mainly consist of the improvements to knowledge and enhanced ability of engineers to design a seismically robust built environment. UK based engineering consultancies now undertake a significant proportion of their work (> 50% in some cases) overseas which is estimated to be worth around 2.8 billion. Much of this is located in regions of moderate to high seismic risk, such that the ability to design successfully against the damaging effects of earthquakes is vitally important. It is proposed to hold an industry workshop, in which representatives of UK engineering firms undertaking work in seismic areas will be invited to Dundee for a one day workshop. This will include the opportunity to witness one of the centrifuge tests live. The data will be discussed in a workshop-type forum in which the delegates can fully engage with the discovery of new knowledge in collaboration with the academic staff. This will help stimulate interest in the industrial community and encourage future industrial collaboration with Dundee and provide the opportunity to 'sand-pit' to develop future directions for extending the research programme that meet the needs of the engineering profession. Advertisement and organisational support will be made possible through the PI's personal contacts, through the Scottish Universities' Geotechnical Network, the Innovation Portal in Dundee and the Scottish Construction Centre which is based at the University. Opportunities to present the work to UK specialist bodies at evening meetings (with web casting) will also be sought (e.g. through ICE's Society for Earthquake and Civil Engineering Dynamics, SECED). It is also proposed to incorporate the research findings within the applicant's Earthquake Engineering module which is taught as part of the new MSc Earthquake and Offshore Geotechnical Engineering at Dundee. To the (international) academic community, impact will be more direct and in the shorter term, consisting of the addition of both new/improved knowledge and the availability of a database of high quality physical model test data for validation studies. This will be disseminated through presentation at international conferences and symposia, and publication in internationally leading journals. The development of earthquake engineering research at Dundee which will be further advanced by this project will enable the University to play a key role in the UK National Earthquake Engineering Simulation (UK-NEES) network, funded by EPSRC (Grants EP/D079691/1, EP/D079101/1, EP/D080088/1) and which is connected internationally to the wider NEES network.