seismic action Report of WG2: Earthquake resistance Dan Lungu*, - PowerPoint PPT Presentation

COST C26: Urban Habitat Constructions under Catastrophic Events Final Conference, 16-18 September 2010, Naples, Italy Characterization and modelling of seismic action Report of WG2: Earthquake resistance Dan Lungu*, Aurel Stratan** & Radu Vacareanu* * Technical University of Civil Engineering of Bucharest, Romania ** "Politehnica" University of Timisoara, Romania

State of the art  Ground motion produced at a site is characterised by the two orthogonal horizontal components and the vertical component  Ground motion representation for structural design may be in the form of: – Time histories – Power spectral density – Response spectra (elastic or inelastic)

Ground motion modelling  Main characteristics of the ground motion time history: – Peak ground acceleration (PGA), peak ground velocity (PGV) and peak ground displacement (PGD); – Motion duration; – Frequency content.  PGA and PGV – Very simple measures of severity of the ground shaking – Random variables – Prediction of the peak ground parameters at a site is the target of the probabilistic seismic hazard analysis (PSHA)  Motion duration t d – Time interval between two specified fractions of 2 E [a(t)] dt tot the total cumulative energy of accelerogram 0 – It represents the time interval over which the motion power is almost constant and near its maximum

Ground motion modelling  Frequency content - concept crucial for understanding the structural damage potential of ground motion.  Frequency content can be described: – Directly, by the power spectral density function (PSD), obtained from stochastic modelling of the acceleration process; – Indirectly, by the response spectra.  Stochastic measures of frequency content: – The dimensionless indicators (Cartwright & Longuet - Higgins) and q (Vanmarcke); – The f 10 , f 50 and f 90 fractile frequencies of the total cumulative power of PSD and the frequencies f 1 , f 2 and f 3 corresponding to the highest 1, 2, 3 peaks of the PSD.  Deterministic measures of frequency content are the control frequencies and corresponding control periods T B , T C , T D .

Modelling of response spectra EPA and EPV. Elastic response spectrum.  Effective peak acceleration (EPA) mean of SA 0.1 0.5 s EPA and effective peak velocity (EPV) 2.5 characterize the intensity of a mean of SV ground motion by averaging the 0.8 1.2 s EPV 2.5 effects of shaking on the structures most exposed to that spectral content  The elastic response spectrum is given as smoothed acceleration spectrum (for 5% damping) having a specified probability to be exceeded; 0.5 median, 0.1, etc.  Seismic input motions must be compatible with local soil condition, intensity of shaking, seismic source mechanism and hypocentral distance

Design spectrum Acceleration-displacement response spectra  Design spectra: – Behaviour factor q used to reduce spectral accelerations, accounting for inelastic structural response (ductility, redundancy and overstrength) – Lower bound a g  The acceleration- displacement response 300 Bucharest'86 - 16 free-field motions spectra (ADRS) 250 – An alternative representation of 200 SA , cm/s 2 response spectra 150 – Periods are represented by a 100 series of radial lines extending 50 from the origin of the plot. 0 0 1 2 3 4 5 6 7 8 9 SD , cm

Probabilistic seismic hazard assessment (PSHA)  A cornerstone position for the prediction of the strong ground motion likely to occur at a particular site  The general PSHA is based on the following methodology: – Identification of independent sources of seismic activity and determination of recurrence relationships; – Fitting the attenuation relationship on a ground motion parameter; – Calculating the peak ground motion 1.E-01 Vrancea seismic PGA subcrustal source parameter at the site with a specified Hazard curve for Iasi City probability of non-exceedance during 1.E-02 structure lifetime; – Delineation of isoseismal maps; 1.E-03 – Construction of uniform hazard PGA , cm/s 2 response spectra for design. 1.E-04 50 150 250 350 450

Contribution to the research development Seismic motion leading to exceptional actions on structures  Seismic action is characterised by high uncertainty and can be specified in probabilistic terms only  Mistakidis, E., Apostolska, R., Dubina, D., Graf, W., Necevska-Cvetanovska, G., Nogueiro, P., Pannier, S., Sickert, J.-U., Simões da Silva, L., Stratan, A., Terzic, U. outlined several phenomena that can lead to exceptional seismic action, related to near-fault effects and local site conditions  Near-fault effects: – Long-period, high-amplitude pulse in the forward-directivity region – Vertical component of the seismic action important

Seismic motion leading to exceptional actions on structures: local site conditions  Geotechnical conditions - soft soil layers: – Amplification of PGA – Amplification of spectral accelerations in the long-period range  "Trapping" of seismic waves inside basins: – amplification and – increase of duration of the seismic motion  Surface topography - amplification for irregular topographies, such as crest, canyon, and slope

Seismic action in urban habitats  Gioncu and Mazzolani: city-site interaction  City-site interaction influence of densely urbanized cities on the ground motion – Superposition of vibrations produced by buildings over soil vibrations coming from the source gives rise to a modification of the free-field motion – The largest effect is produced in the case of a dense constructed area situated in a soft soil – Each point on the surface can have different movements, explaining the strange and highly variable damage within identical building sets

Seismicity of Vrancea seismic source and soil conditions in Bucharest  Lungu, D., Arion, C., Calarasu, E. and Lungu, D., Vacareanu, R., Aldea, A.  Informations related to: – Seismicity of Romania – Seismic instrumentation – Available strong motion records – New seismic zonation map from the Romanian seismic design code  Seismic microzonation of Bucharest - a tool for urban planning and earthquake risk reduction

Selection of time-history records for dynamic analysis of structures  Stratan and Dubina - overviewed code requirements and selected references related to selection of time-history records  Code provisions: – how are the records obtained (through artificial generation, from existing recordings of past earthquakes, or through simulation); – compatibility between earthquake records and the seismic source, travel path and site characteristics; – matching between the target response spectrum and the ones of earthquake records and – number of records used and implications on result interpretation

Selection of time-history records for dynamic analysis of structures  Artificial accelerograms are generated using stochastic algorithms. Can be improved by accounting for some seismogenetic features or through semi-artificial accelerograms. Simple pulses can be used.  Simulated records are obtained through physical simulation of source and travel path mechanisms, and may account for site effects.  Recorded accelerograms are obtained from real seismic events in the past with similar source, travel path and local site conditions. Scaling usually necessary.

Simulation of accelerograms for fuzzy analysis of structures  Fuzzy stochastic tools for structural analysis and reliability assessment were investigated by Sickert, J.-U., Kaliske, M., Graf, W.  The uncertain character of both earthquake records and structural system behaviour was considered within a response history procedure.  Using the model fuzzy randomness, earthquake excitations are described as fuzzy random processes which represent a fuzzy set of real valued random processes.

Scenarios based earthquake hazard assessment  Romanelli, F., Peresan, A., Vaccari F. & Panza, G.F. investigated scenario earthquakes, also named neodeterministic seismic hazard assessment (NDSHA)  NDSHA - a hybrid method consisting of modal summation and finite difference methods  Artificial seismograms of the vertical, transverse and radial components of ground motion are computed at a predefined set of points at the surface

Recommendations for further development  There seems to be a gap between the existing knowledge on characterization of seismic action and the code provisions: – Lack of code provisions for near-fault effects – Behaviour factor q independent of response spectrum characteristics  Time-history records for nonlinear analysis difficult to obtain: – Time-histories compatible with the characteristics of the seismic source, travel path and site effects require expertise in seismology, that few structural engineers would have. – Code requirement of matching to code spectra requires scaling of records which alter the "seismological" compatibility. – A close collaboration between seismologists and structural engineers is needed to advance the current state of practice in structural analysis under seismic action.