Investigación sobre el uso de muros de fábrica para el reacondicionamiento sísmico masivo de estructuras portadas de hormigón armado de media alturaaplicación a construcciones existentes en países en vías de desarrollo dañadas por terremotos recientes

Résumé

Over the last few decades, the level of damage and the frequency of structural collapse incidents due to seismic action have been reduced in developed countries thanks to advance knowledge in earthquake engineering and the implementation of such in building codes. This subject has not emerged as a new, isolated science, but rather it is based on prior knowledge regarding the behavior of structures subjected to different types of loads. In general, the structural design for both gravitational seismic loads is regulated by national and/or community building regulations and codes such as EHE, NCSE-02 (Spanish building codes), and Eurocodes (European building codes). These codes, for their part, are regularly updated by committees of experts who are in charge of assessing and incorporating sufficiently detailed and well-founded knowledge. The obligation to implement or not legislation is at the behest of the government of each country. Generally speaking, not only do developed countries require specific codes of practice, but they also strive to cooperate internationally in improving these standards. The situation is quite different in developing countries. The limited economic resources of these countries means that they should be used to try to secure basic goods and services. Moreover, structural safety in these countries is not often addressed using scientific knowledge, but rather as a trade that transmits simple heuristic knowledge from master builder to labourer. Consequently, the damage caused by earthquakes in a developed country can be hugely different to that of a developing country. The ultimate goal of this Thesis is to bridge the gap between scientific and heuristic knowledge which is easily applicable and attainable in a context of limited economic and technical resources. The first aspect addressed in this Thesis is the influence of masonry infill walls on the seismic behavior of buildings. Masonry infill walls are considered non-structural elements in many building codes such as the Spanish building codes. There is, however, evidence that their presence can significantly increase the strength and lateral rigidity of the building. This Thesis begins from the quantification of the influence of these walls on the vulnerability of the buildings, investigating the real and recent scenario brought about by the Lorca earthquake in 2011. Chapter 3 of this Thesis presents the numerical research carried out in the City of Lorca after the earthquake in May 2011. The study began by the sizing and modelling of two real buildings in the city of Lorca. Each of the buildings was measured with and without masonry infill walls in order to be able to quantify their influence. The vulnerability study of each of the buildings was then carried out based on an energy-based damage index. Finally, the obtained results were compared with direct dynamic analysis in which the accelerogram of the 2011 Lorca earthquake was used obtaining results consistent with the damages observed through visual inspection. The conclusions of this study showed the great influence of masonry walls on the seismic behavior of the building and the need to (i) consider them as structural elements in the prediction of damage, and (ii) adequately design the distribution of these elements in order to benefit from their influence. After obtaining the results from the first part of the research, the possibility of using masonry infill walls as a reinforcement and seismic retrofitting solution in developing countries was considered. Masonry infill walls have several advantages that make them suitable to be used as seismic retrofitting elements in developing countries, including: (1) their low economic cost, (2) they are universal; they are made everywhere and have very similar properties, (3) technical or structural training for the worker who erects them is not necessary, (4) the walls themselves can be used as internal partitions or external enclosures, (5) they can be demolished and rebuilt easily in the case of earthquake damage. The research on the use of masonry infill walls as a seismic retrofitting and reinforcement solution was based on an extensive experimental study that included static and dynamic tests with a shake-table. Chapter 4 explains in detail the experimental campaign carried out in the Structures Laboratory of the University of Granada to evaluate the seismic behavior of masonry infill walls used as seismic retrofitting elements. In the dynamic tests, a reinforced concrete framed structure was used which had already been subjected to two dynamic test for the study of two types of hysteretic dampers. These previous test campaigns had caused a certain level of damage in the reinforced concrete frame structure, which appeared mainly in the formation of plastic hinges at the ends of the pillars. This reinforced concrete frame structure also had the advantage of being designed to only withstand gravitational loads, which is very common in existing structures in developing countries. Having previously been subjected to seismic simulations, this structure was also suitable to represent an existing building found in developing countries, which had been damaged by an earthquake and which was supposed to be retrofitted with the installation of masonry infill walls. For these reasons, the next step was to retrofit the frame by adding two masonry infill walls without openings on both frames in the direction of the earthquake. The dimensions of the bricks were adjusted to meet the scale factor for lengths of 2:5 with which the frame was planned and designed. Moreover, they were designed and made in accordance with the recommendations of the government of Haiti in terms of volume of voids and compressive strength. The frame with masonry walls underwent several simulations with the shake-table which tried to reproduce two different scenarios: (1) that of a frequent earthquake followed by a replica with similar characteristics to the first earthquake; (2) that of a rare earthquake to study the behavior of the structure when subjected to large lateral displacements. From the experimental results, a numerical model was proposed to represent the behavior of the wall under random cyclic lateral loads. The model conceptualizes the wall as an equivalent diagonal bar that only works when subjected to compression. The model includes (i) the definition of lateral force displacement enveloping curve of the wall under load, and (ii) the proposal of a suitable combination of parameters to adjust the hysteretic behavior under cyclic loads of the wall using the Bouc-Wen algorithm. The following step was to carry out a numerical investigation in order to try and evaluate the effect of the construction of masonry infill walls as a measure of seismic retrofitting in Haiti. This study can be found in Chapter 5. We used field data collected by a team of researchers led by a professor (Santiago Pujol, Purdue University) who was part of the research project team within the framework in which this Thesis was carried out. This team had collected information from buildings in Haiti damaged by the 2010 earthquake. Buildings used for educational purposes were some of the buildings that suffered more damage from this earthquake. As a result of this, and due to the great importance of these types of buildings, the decision that this research would focus on buildings for educational use was made. From the field information obtained in Haiti, two configurations of building floor plans were proposed. The study of the actual data shows that buildings used for schools have always had between 2 and 3 stories, therefore two prototypes were established for each configuration; the first with two stories and the second with three stories. The four buildings were measured using Tricalc. In the absence of building codes in Haiti, Spanish building codes were used for gravitational loads only. The buildings were modeled on Idarc. In order to define the hysteretic behavior of the structural elements, the parameters calibrated in previous experimental campaigns were used for of beams and columns, and those obtained in the experimental campaign of this Thesis were used for the walls. One of the aspects which this Thesis has focused on has been to investigate the influence of the wall area by total floor area on the behavior of the building in the face of an earthquake. Therefore, the four buildings designed with different configurations and wall areas were studied. The number of walls in the study varied as did the layout of such. The methodology used to evaluate the influence of the number of walls was based on the study of fragility curves for each building, for different scenarios and damage states. For a given level of damage (Immediate Occupation, Limited Damage, Severe Damage, On the Verge of Collapse), a fragility curve provides the probability that this level of damage is exceeded for a given seismic intensity level characterized by the spectral shift (S_d ) ̅. The fragility curves follow a standard lognormal distribution defined by two parameters: (1) the average top spectral displacement of the building and (2) the standard deviation of the natural logarithm of the spectral shift. The parametric study of these values for all the buildings allowed us to propose fragility curves according to the geometric configuration of the building and the number of stories. Another damage index evaluated in this Chapter was the one defined by Hassan and Sozen (1997) based on the study of multiple post-earthquake scenarios and that allows us to determine whether or not a building is vulnerable in the face of an earthquake, on the understanding that it has a high probability of collapse or irreparable severe damage in the face of a seismic event. This index, which is presented in detail in Chapter 5, and which allows us to quickly and easily estimate the wall area needed to ensure that the building is not vulnerable to an earthquake. In summary, the aim of this study is to: (i) determine the wall area needed to ensure that the damage caused by an earthquake does not exceed a certain level; and (ii) propose fragility curves for the seismically retrofitted building with which studies of seismic risk can be carried out. Lastly, Chapter 6 is of great importance in that it offers guidelines of action as well as proposing simple seismic retrofitting strategies that can be implemented in the seismic codes of developing countries. Firstly, a guide is shown for the evaluation of the damage to a building after an earthquake. The first part describes how to carry out damage assessment in the building, that is the checks to be performed, the different types of damage that can be found in the structural elements and the determination of whether or not they are acceptable. Measurements are also proposed to repair beams and columns in case it is necessary to raise their earthquake-resistant capacity to a minimum level, before considering the strategy of retrofitting using masonry infill walls. The second part of this guide proposes the strategy of seismic retrofitting using masonry infill walls. To this end, and based on the formula proposed by Hassan and Sozen, the minimum area of brick walls necessary to ensure that the building survives a seismic event is determined. Following this, and based upon the results in Chapter 5, how to obtain approximate fragility curves for the retrofitted building which allow studies of seismic risk is shown. Finally, all of the above applies to the specific case of a school in Haiti that suffered severe damage and has been reinforced using the retrofitting strategy developed in this Thesis.