Earthquake-induced structural pounding between insufficiently separated buildings has been repeatedly observed during ground motions. The reports after earthquakes indicate that it may lead to some local damage in the case of moderate seismic events or may result in considerable destruction or even collapse of colliding structures during severe earthquakes.
The aim of the present work is to conduct a detailed investigation on pounding-involved response of two equal height buildings with different dynamic properties, paying a special attention to modelling the non-linear effects taking place during impact as well as observed in the structural behaviour as the result of ground motion excitation.
First, a verification of the effectiveness of different models of structural pounding is conducted by comparing the results of numerical analysis with the experimental results. The study indicates that, among the models compared, the proposed non-linear viscoelastic model can be considered as the most precise one in simulating the pounding force during single impact as well as in modelling the pounding-involved response of structures under earthquake excitation.
Then, theoretical and experimental approaches are employed in order to determine the parameters of the non-linear viscoelastic model. The analytical study leads to derivation of a formula describing quite precisely the relation between the impact damping ratio and the coefficient of restitution, which is a parameter widely used and studied in the literature. The results of the experimental study show a wide range of coefficients of restitution and impact stiffness parameters determined for different building materials for various prior-impact velocity and mass values of the colliding elements.
The idea of impact force response spectrum for earthquake-induced structural pounding between two insufficiently separated buildings is considered in the next part of the work. The presented examples of spectra show that a selection of different structural parameters may have a substantial influence on the pounding force value. The results indicate that impact force response spectra provide valuable information on the peak pounding force, which can be expected during the earthquake, and thus they might serve as a very useful tool for the design purposes of closely-spaced structures.
In the next part of the work, the non-linear response analysis as well as the parametric study are conducted for earthquake-induced pounding of two adjacent buildings modelled as elastoplastic multi-degree-of-freedom lumped mass systems. The results of the study show that collisions have a significant influence on the lighter building causing substantial amplification of the response due to yielding, whereas the behaviour of the heavier building has been found to be influenced only negligibly.
Finally, a detailed three-dimensional pounding-involved response analysis is conducted under uniform and non-uniform (stochastically generated) earthquake excitation using the FEM with the non-linear model of material behaviour, including stiffness degradation model of concrete under cyclic loading. Also the results of the FEM analysis indicate that collisions may lead to a significant increase of the response of the lighter structure as well as may result in a substantial increase of damage induced. On the other hand, the analysis show that the response of the heavier structure is only slightly influenced by the structural interactions. A further study indicates that the non-uniform ground motion excitation, due to spatial seismic effects connected with the propagation of seismic wave, may lead to a considerable change of pounding-involved behaviour of both structures.
Spis treści:
CONTENTS
LIST OF SYMBOLS
1. INTRODUCTION
1.1. Damage due to earthquake-induced structural pounding
1.2. Research on structural pounding during earthquakes
1.3. Aim and scope of study
2. MODELS OF EARTHQUAKE-INDUCED STRUCTURAL POUNDING
2.1. Modelling approaches
2.1.1. Classical theory of impact
2.1.2. Modelling the pounding force during impact
2.1.2.1. Linear viscoelastic model
2.1.2.2. Non-linear elastic model
2.1.2.3. Non-linear viscoelastic model
2.2. Experimental verification of pounding force models
2.2.1. Impact between a ball and a hemisphere
2.2.2. Impact between a pendulum striker and a pile
2.2.3. Impact between a ball and a rigid surface
2.2.3.1. Concrete-to-concrete impact
2.2.3.2. Steel-to-steel impact
2.2.3.3. Timber-to-timber impact
2.2.4. Pounding between a bridge girder model and an abutment
2.2.5. Pounding of steel tower models
2.2.6. Pounding of tower models with different impacting materials
2.2.6.1. Concrete-to-concrete pounding
2.2.6.2. Steel-to-steel pounding
2.2.6.3. Timber-to-timber pounding
2.2.7. Results of experimental verification of pounding force models
2.3. Summary and conclusions
3. THEORETICAL AND EXPERIMENTAL DETERMINATION OF PARAMETERS FOR THE NON-LINEAR VISCOELASTIC MODEL OF EARTHQUAKE-INDUCED STRUCTURAL POUNDING
3.1. Analytical relation between the impact damping ratio and the coefficient of restitution
3.1.1. Energy loss during impact
3.1.2. Maximum deformation and relative velocity during restitution period
3.1.3. Approximate function 1
3.1.4. Approximate function 2
3.1.5. Verification of accuracy of analytical formulations
3.2. Experimental determination of the coefficients of restitution and the impact stiffness parameters for different building materials
3.3. Summary and conclusions
4. POUNDING FORCE RESPONSE SPECTRUM UNDER EARTHQUAKE EXCITATION
4.1. Pounding force response spectrum for elastic buildings
4.1.1. Numerical model
4.1.2. Effect of damping ratio
4.1.3. Effect of gap size between buildings
4.1.4. Effect of mass
4.1.5. Effect of time lag
4.1.6. Pounding force spectra for different earthquakes
4.2. Pounding force response spectrum for inelastic buildings
4.2.1. Numerical model
4.2.2. Pounding force spectra for one elastic and one elastoplastic building
4.2.3. Pounding force spectra for two elastoplastic buildings
4.3. Summary and conclusions
5. ANALYSIS OF POUNDING-INVOLVED RESPONSE OF BUILDINGS MODELLED AS MULTI-DEGREE-OF-FREEDOM LUMPED MASS SYSTEMS
5.1. Numerical model
5.2. Response analysis
5.2.1. Response in the longitudinal direction
5.2.2. Response in the transverse direction
5.2.3. Response in the vertical direction
5.3. Parametric study
5.3.1. Effect of gap size between buildings
5.3.2. Effect of storey mass
5.3.3. Effect of structural stiffness
5.3.4. Effect of structural damping
5.3.5. Effect of yield strength
5.4. Summary and conclusions
6. POUNDING-INVOLVED RESPONSE ANALYSIS USING FINITE ELEMENT METHOD
6.1. Pounding between the hospital main building and the stairway tower in the Olive View Medical Center during the San Fernando earthquake
6.1.1. Description of the main building and its damage
6.1.2. Description of the stairway tower C and its damage
6.2. Non-linear model of colliding structures
6.2.1. Model of reinforced concrete
6.2.1.1. Concrete
6.2.1.2. Reinforcing steel
6.2.2. FEM models of the main building and the stairway tower
6.2.3. Ground motion excitation
6.3. Response analysis under uniform earthquake excitation
6.3.1. Response in the longitudinal direction
6.3.2. Response in the transverse direction
6.3.3. Response in the vertical direction
6.3.4. Structural damage due to pounding-involved response
6.4. Response analysis under non-uniform earthquake excitation
6.4.1. Stochastic determination of ground motion records for different locations
6.4.1.1. Correlation function of ground motion field
6.4.1.2. Conditional probability density function
6.4.1.3. Numerical simulation procedure
6.4.1.4. Numerical generation of input ground motion records for different support locations of the main building
6.4.2. Pounding-involved response analysis
6.4.2.1. Response in the longitudinal direction
6.4.2.2. Response in the transverse direction
6.4.2.3. Response in the vertical direction
6.5. Summary and conclusions
7. FINAL REMARKS
7.1. Conclusions
7.2. General remarks
7.3. Original elements of the study
ACKNOWLEDGEMENTS
REFERENCES
SUMMARY IN ENGLISH
SUMMARY IN POLISH
APPENDIX A
