Panagiotis E. Mergos

Rocking isolation of a typical bridge pier on spread foundation

It has been observed that after some earthquakes a number of structures resting on spread footings responded to seismic excitation by rocking on their foundation and in some cases this enabled them to avoid failure. Through application to a standard bridge supported by direct foundations, this paper discusses the major differences in response when foundation uplift is taken into consideration. Special focus is given on the modifications of rocking response under biaxial and triaxial excitation with respect to uniaxial excitation. It is found that inelastic rocking has a significant isolation effect. It is also shown that this effect increases under biaxial excitation while it is less sensitive to the vertical component of the earthquake. Finally, parametric analyses show that the isolation effect of foundation rocking increases as the size of the footing and the yield strength of the underlying soil decreases.

Damage Analysis of Reinforced Concrete Structures with Substandard Detailing

The goal of this study is to investigate seismic behaviour of existing R/C buildings designed and constructed in accordance with standards that do not meet current seismic code requirements. In these structures, not only flexure, but also shear and bond-slip deformation mechanisms need to be considered, both separately and in combination. To serve this goal, a finite element model is developed for inelastic seismic analysis of complete planar R/C frames. The proposed finite element is able to capture gradual spread of inelastic flexural and shear deformations as well as their interaction in the end regions of R/C members. Additionally, it is capable of predicting shear failures caused by degradation of shear strength in the plastic hinges of R/C elements, as well as pullout failures caused by inadequate anchorage of the reinforcement in the joint regions. The finite element is fully implemented in the general inelastic finite element code IDARC2D and it is verified against experimental results involving individual column and plane frame specimens with non-ductile detailing. It is shown that, in all cases, satisfactory correlation is established between the model predictions and the experimental evidence. Finally, parametric studies are conducted to illustrate the significance of each deformation mechanism on the seismic response of the specimens under investigation. It is concluded, that all deformation mechanisms, as well as their interaction, should be taken into consideration in order to predict reliably seismic damage of R/C structures with substandard detailing.

A MODEL FOR NON-LINEAR DYNAMIC ANALYSIS OF SUB-STANDARD REINFORCED CONCRETE MEMBERS

Reinforced concrete (R/C) buildings designed according to older seismic codes represent a large part of the existing building stock; hence, it is important to be able to accurately assess their response to seismic actions. These substandard R/C buildings may contain structural elements that are prone to shear failure subsequent, or even prior, to yielding of their longitudinal reinforcement. This can potentially lead to loss of axial load resistance of vertical elements and initiate vertical progressive collapse of the building. In the present study, a model is put forward, which is able to capture the hysteretic behaviour of shear critical R/C elements up to the onset of axial failure. The model is able to account for gradual spread of inelasticity and shear-flexure interaction in the locations of the plastic hinges prior to the onset of shear failure as well as localisation of shear strains after the onset of shear failure; the latter constitutes a major advancement in the existing literature. The model is used to predict the behaviour of a cyclically loaded double-curvature specimen failing axially after shear failure and flexural yielding. The comparison of the analytical with the experimental results shows a very good agreement, both in terms of total response and of shear response. 1656 Available online at www.eccomasproceedia.org Eccomas Proceedia COMPDYN (2017) 1656-1666

Optimum design of reinforced concrete frames according to EC8 and MC2010 with Genetic Algorithms

The need for cost-efficient seismic design in limited time has led to the development of automated structural optimization methodologies. Genetic Algorithms (GA) belong to the class of stochastic, nature-inspired heuristic optimization algorithms. GA can be easily implemented and applied to advanced structural problems since they don’t require use of gradients of cost or constraints functions. Furthermore, they are able to identify global optima as opposed to local optimum solutions. Early efforts to optimise seismic design of concrete structures were based on traditional seismic design code approaches. The last two decades, performance- and deformation-based seismic design methodologies have emerged. These methodologies provide enhanced structural damage control for different earthquake intensities reducing both economic losses and human casualties. Lately, a fully-fledged performance- and deformation based design methodology has been incorporated in the fib Model Code for Concrete Structures MC2010, which is meant to serve as a guidance document for future design codes of concrete structures. To the best of author’s knowledge, there exists no study investigating optimum seismic design solutions in accordance with MC2010. The aim of this study is to develop a GAbased seismic design optimization framework for reinforced concrete frames in accordance with the provisions of both Eurocode 8 (EC8) and MC2010. Application of this framework to reinforced concrete frames is conducted and comparisons of the optimum solutions obtained by the two seismic design guidelines are made. The advantages and disadvantages of the two seismic design methodologies, in the context of structural optimization, are highlighted and discussed.

Displacement-Based Seismic Design of Symmetric Single-Storey Wood-Frame Buildings with the Aid of N2 Method

This paper presents a new methodology for the displacement-based seismic design of symmetric single-storey wood-frame buildings. Previous displacement-based design efforts were based on the direct displacement-based design approach, which uses a substitute linear system with an appropriate stiffness and viscous damping combination. Despite the fact that this method has shown to produce promising results for wood structures, it does not fit into the framework of the Eurocode 8 (EC8) provisions. The methodology presented herein is based on the N2 method, which is incorporated in EC8 and combines the non-linear pushover analysis with the response spectrum method. The N2 method has been mostly applied to reinforced concrete and steel structures. In order to properly implement the N2 method for the case of wood-frame buildings, new behavior factor–displacement ductility relationships are proposed. These relationships were derived from inelastic time history analyses of 35 SDOF systems subjected to 80 different ground motion records. Furthermore, the validity of the N2 method is examined for the case of a timber shear wall tested on a shake table and satisfactory predictions are obtained. Last, the proposed design methodology is applied to the displacement-based seismic design of a realistic symmetric single-storey wood-frame building in order to meet the performance objectives of EC8. It is concluded that the simplicity and computational efficiency of the adopted methodology make it a valuable tool for the seismic design of this category of wood-frame buildings, while the need for extending the method to more complex wood-frame buildings is also highlighted.

SHEAR HYSTERESIS MODEL FOR REINFORCED CONCRETE ELEMENTS INCLUDING THE POST-PEAK RANGE

Reinforced concrete (R/C) buildings designed according to older seismic codes represent a large part of the total building stock; hence, it is important to accurately and efficiently assess their response to actions induced by natural hazards, such as earthquake. Substandard R/C structural elements are prone to shear failure subsequent, or even prior, to yielding of their longitudinal reinforcement. This can potentially lead to loss of axial load bearing capacity of vertical elements and initiate progressive collapse of the building. So far, there have been efforts to model the full-range behaviour of such elements following a macro-modelling approach, usually based on quite a limited amount of experimental results, especially with respect to the post-peak part of their response, and adopting assumptions that are not entirely appropriate. In the present study, an extensive database of shear and flexure-shear critical rectangular R/C columns has been compiled, to the purpose of investigating R/C member post-peak response and calibrating the models mentioned below. It includes both monotonic and cyclic tests, the latter constituting the majority, it spans a broad range of design, material and loading parameters and the majority of the specimens have been tested up to axial failure. A shear macro-model is developed, which is able to capture the full hysteretic behaviour of R/C elements. In addition to the behaviour up to peak shear resistance, an effort is made to properly capture the post-peak response, calibrating an empirical model for the descending branch directly, instead of indirectly defining it through shear and axial failure that has traditionally been the case. The angle of the shear failure plane is an important parameter of this model, hence an empirical relationship has been developed for it, as well. The onset of axial failure constitutes a vital aspect of post-peak response, since it signals the initiation of a process of loss of an individual R/C element’s axial load-bearing capacity simultaneously with the redistribution of vertical loads to neighbouring ones; thus, it was also closely examined and empirical models were derived.