Democritus University of Thrace (DUTH)
Empirical and analytical assessment of kinematic soil-structure interaction for low to mid-rise buildings in Greece
Abstract
dc:descriptionDynamic soil-structure interaction (SSI) refers to the way that soil deformations, occurring upon arrival of seismic waves, affect the seismic response of an overlying structure and subsequently, how the response of the structure affects the subsoil through additional deformations coming from inertial loads transferred from the structure to its foundation. Relative to the fixed-base approach, consideration of SSI leads to a more realistic estimate of the soil-structure system response, as well as increased fundamental period and damping characteristics. Increased damping may arise due to soil hysteretic response and radiation of waves emanating from foundation movement and travel through the soil half-space during earthquake shaking. Moreover, the presence of foundation, being stiffer than the surrounding soil and not able to follow its movement, causes variations between foundation and free-field motions, which are more pronounced for soft soil sites. Significant theoretical and analytical research has been published over the last decades on SSI, either inertial or kinematic. Moreover, improvement of recording instruments over the decades has led to an increase inground motion recording stations and enhancement of the records’ quality. Pairs of foundation and free-field ground motion records have come up in ground motion networks worldwide which have been utilized either to empirically assess the transfer functions between foundation and free-field motions, or to correlate ground motion parameters. The effect of SSI on the recorded motions within buildings has been recognized by many researchers working on the development of Ground Motion Prediction Equations (GMPEs). Many of them have tackled this problem by ignoring records from buildings through implementation of empirical rules based on macroscopic features, such as number of storeysand existence of a basement floor. Others have tried to include correction factors within the GMPEs to account for SSI based on the same features. Based on the above, the objective of the present PhD dissertation is to correlate ground motion properties using, as a case study test-bed, the recordings of the Hellenic National Accelerometric Network (HNAN) in Greece, run by the Institute of Engineering Seismology and Earthquake Engineering (ITSAK-EPPO), the National Observatory of Athens (NoA) and University of Patras (UPatras) obtained at the basement of carefully instrumented and documented single or multi-storey buildings. A wider aim, that is not analytically analyzed inthis work, but highlights its importance, is the possibility to use the outcomes of the SSI impacton the earthquake records obtained within instrumented buildings to draw corrective procedures that can predict the “equivalent building-free” ground motions and assess the error induced by SSI on GMPEs and seismic hazard that have been produced utilizing suites of ground motions recorded within buildings. Initially, the effect of kinematic soil-structure interaction is highlighted through comparison between foundation and adjacent free-field recorded motion data. The recordings reported come from stations belonging to three different strong motion networks, namely the Hellenic National Accelerographic Network (HNAN) in Greece, the Center for Engineering Strong Motion Data (CESMD) in California, USA and the Building Research Strong Motion Network (BRI) in Japan. Since Greece is the region of main focus in this thesis, the selection of sites from CESMD and BRI networks is affected by the site characteristics of the HNAN. The sites and recordings of each network are presented. Recording-based transfer functions are estimated for all the sites considered, to highlight the filtering of high-frequency components due to the presence of the building. Moreover, ground motion intensity measures and frequency content parameters are calculated, and comparison between foundation and free field motions is carried out. Subsequently, a sub-structure analysis method for SSI systems is described. The method adopted, regarding kinematic interaction, is based on existing analytical expressions, calibrated to recorded data and incorporates both, foundation embedment and base-slabaveraging effects, to estimate the foundation input motion (FIM). The foundation impedance functions are calculated according to existing, well-known analytical expressions, and the analysis procedure is concluded by performing a dynamic analysis in frequency domain of SDOF model of the super structure, supported on the foundation impedance functions and excited by the FIM. The analysis procedure is implemented on well documented sites belonging to the HNAN. It is initially validated through comparison between analytical and empirical transfer functions between foundation and free-field. Parametric SSI analyses are performed for each of the sites to establish correlations between foundation and free-field motion intensity and frequency content. The analysis procedure includes equivalent linear ground response analyses of the site soil profiles under rock outcropexcitations, in order to produce multiple free-field ground motions. These motions along with the effective properties of the soil layers are later utilized as input data to perform the SSI substructure analyses and obtain the foundation motions. Correlation between intensity parameters and frequency content parameters of foundation and free-field motions are plotted, and regression expressions are derived. Analysis and recorded data regressions proveto satisfactorily fit. Further evidence of the analysis procedure suitability is acquired by presenting simulation results of the available recordings for each site and comparing the computed to the recorded foundation motion in terms of wave form, Fourier amplitude and response spectrum. Further work builds upon the confidence on the adopted analysis method which was obtained through earlier findings. At first, the sub-structuring analysis method, which was applied, isextrapolated to account various soil conditions, for structures having similar characteristics to those already addressed. A process to generate random soil profiles belonging to a specific category is followed, through Monte Carlo simulation and variation of soil properties such as: layer thickness, shear wave velocity and depth to bedrock. Subsequently, the same SSI analysis procedure is followed to derive direct relationships between intensity measures of foundationand free-field motions, as well as relationships expressing the relative variation of intensity measures of the two motions, with respect to frequency content, for various soil conditions. The same analysis procedure is extrapolated to structures with various heights for each of the ground types considered. This is made possible by inducing additional storeys to the mathematical 3D models of the buildings of the well documented sites. The original buildings possessed 2 to 3 overground storeys and 5, 8 and 10-storey versions of them are built. The intention of this was to investigate the effect of increased, with respect to the original state, inertia of buildings on the relationships produced in the previous section. The proposed analytical expressions are validated against the available recordings, per ground type. Given the complexity of the studied phenomena and the simplicity of the analysis method, it is deemed that the proposed expressions simulate to the recordings’ observation sadequately, with errors ranging from 10% to 17%. The role of shear wave velocity is further investigated by adding a VS parameter in the analytical expressions of each ground type and validation of the new expressions, through recordings, was performed again. Various VS parameters are investigated and evaluated. At last, the analysis data for all ground types as well as the recordings were classified together in order to construct global intensity variation expressions by considering the investigated VS parameters. The resulting analytical expressions, presented satisfactory validation and analytical errors. The analytical expressions developed, form a useful tool in estimating foundation intensity measures and frequency content, from free-field motions and vice versa. Depending on the knowledge level of soil profile of a site, the various forms of analytical expressions can be used. At last, the variation in peak response measures between foundation and free-field motions was studied through estimation of the high-frequency spectral decay parameter, “κ”. For this, all the recordings’ data considered in this dissertation, are utilized. The calculation of “κ” isbased on Anderson and Hough’s (1984) definition, where as an automated signal energy-based procedure is developed to calculate the S-wave windows of each waveform, accounting for any errors in defining it. Site-specific restrictions regarding the upper limit, f2, of the frequency range along which κfnd is computed, are imposed, based on the empirical transfer functions between foundation and free-field motions. Based on the computation results, non-linear regression analyses are performed to relate free-field motions’ κff, and foundation motions’ κfnd, to Mw, Repi and VS30. Simulation of ground motions though the stochastic method is performed, utilizing the regression expression to define “κ”, to assess the effectiveness of “κ”variation in reproducing the variation in intensity between free-field and foundation motions. The variation between the intensity measures of foundation and free-field motions is related to soil-structure interaction. However, this variation is adequately described by one, sole empirical parameter (“κ”). Therefore, the performance of the regressed expression, combined with stochastic simulations, is deemed adequate and promising for implementation to procedures which aim at correcting recorded foundation motions to obtain peak response parameters of corresponding free-field ground motions, under the circumstances imposed by the considered recordings.
Degree
thesis:*- Grantor dc:publisher
- Democritus University of Thrace (DUTH)
- Year dc:date
- 2020
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Sotiriadis, Dimitrios
- Σωτηριάδης, Δημήτριος
Subjects
dc:subject × 14- Αλληλεπίδραση εδάφους - κατασκευής
- Κινηματική αλληλεπίδραση
- Ανάλυση εδαφικής απόκρισης
- Καταγραφές ισχυρής κίνησης
- Αριθμητική αποσύζευξη
- Soil-structure interaction
- Kinematic interaction
- Site response analysis
- Strong motion recordings
- Numerical sub-structuring
- Επιστήμες Μηχανικού και Τεχνολογία
- Επιστήμη Πολιτικού Μηχανικού
- Engineering and Technology
- Civil Engineering
Rights
dc:rights- Statement dc:rights
-
- BY
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/47666