Miano A., Ebrahimian H., Jalayer F., Vamvatsikos D., Prota A. (2020). Impact of sampling techniques on uncertainty propagation with non-linear dynamic analysis. Proceedings of the 17th World Conference on Earthquake Engineering (17WCEE), Sendai, Japan
Abstract | The Method: Quantifying the impact of modelling uncertainty on the seismic performance assessment is a crucial issue for existing buildings, considering the partial information available related to material properties, construction details and the uncertainty in the capacity models. It has been proved that the effect of structural modelling uncertainties on the seismic performance of existing buildings can be comparable to that of uncertainty in ground motion representation. In this work, the impact of different sampling techniques such as Standard Monte Carlo simulation and Latin Hypercube sampling with Simulated Annealing on the uncertainty propagation with non-linear dynamic analysis has been investigated. Two alternative non-linear dynamic analysis procedures, namely, Incremental Dynamic Analysis and Cloud Analysis are explored. The types of uncertainty encompass record-to-record variability, structural modelling parameters and the fragility model parameters. A one-to-one sampling approach has been adopted in which each of the ground motion records is paired up with a different realization of the structural model. The Application: The case-study structure consists of three stories with a semi-embedded story. The structure lies on soil type B (according to national Italian code NTC 2018 site classification). The building is constructed in the 1960s and is designed for gravity loads only. The structure is composed of bi-dimensional parallel frames, without transversal beams. The main central frame in the structure is used herein as structural model. The finite element model of the frame is constructed, using OpenSees, assuming that the non-linear behaviour in the structure is modelled as distributed plasticity. The Beam-with-hinges element from the library of OpenSees is used to model the distributed plasticity. As the uniaxial material from OpenSees library, Pinching4 Material is used. The points on the backbone curve are defined as cracking, yielding, spalling and the ultimate, respectively. These points are obtained based on moment-curvature analysis of beam-column elements subjected to flexure and axial force. The lateral force-deformation response of the element is obtained by considering as a spring the flexural-compression response of the element (section analysis for normal stresses) which is acting in series with a shear spring and a spring representing the fixed-end rotations. The total lateral force deformation response of the element considers the interaction between the shear, bar-slip and the axial-flexural response. A large ground motion set of 160 records from NGA West2 Database, ITACA (Italian Accelerometric Archive), and recent Iranian recordings (International Institute of Earthquake Engineering, IIEES, personal communication) has been employed. Cloud Analysis and IDA have been implemented with the complete set of 160 un-scaled records. Moreover, Cloud Analysis and IDA have been carried out with subsets of respectively 50 and 30 ground motion records.