Resumen de Escala de gravedad de daños en edificios: de la asignación directa a la contrastación estadística
Félix Ruiz Gorrindo
To make the inspection and diagnosis of buildings is useful be able to quantify how far the existing deficiencies are serious or not, in order to facilitate the decision making and prioritize therapeutic interventions. In this context, the main objective of the thesis is to propose a gravity scale of damages in buildings and the subsequent validation. To this scale is required that there is low variability among technicians who use it, in order that assessments, conclusions and prioritizations are homogeneous regardless of the technician involved in the inspection of the building. To develop the proposed goal in the first place we study existing scales of general scope (Boufort, Douglas, Richter, Mohs, EVA, etc.) and in the field of construction. Subsequently we define the characteristics of the proposed scale and the way of application based on the direct assignment method, for assessing the gravity of constructive elements in buildings. Then is proposed a weighted calculation method based on the statistical distribution of the severity of existing damage, to determine the total degree of severity of the building. Finally, the scale and the proposed methodology have been applied to a real case of building, verifying that the used methods give consistent results. To validate the proposed scale and analyze its applicability a series of essays in which 33 images of building a population of 374 technical elements is performed, to allocate them a degree of gravity for each image according to generic definitions of the proposed scale, in order to analyze, among other things, if there is high or low variability among technicians. With the data collected is made statistical analysis, where different aspects are analyzed and relate, and mathematical techniques are used, such as descriptive statistics (arithmetic means, standard deviations, densities, frequencies, histograms, etc..), binary logistic regression and clustering. In particular, the techniques of clustering allow to determine what is the appropriate number of degrees to should have the scale to the operation thereof is optimal and the degree of variability among technicians is acceptable. In order to reduce the degree of variability between technicians produced by the direct assignment method is further proposed an alternative methodology for the specific case of façades that we call System of Evaluation of Façades (SEF). In the numerical part of this method, once obtained the different data of the façade from its graphical representation, the value of two proposed indicators that are measurable through variables is determined. On these indicators a series of calculations are made based on some mathematical expressions and methods that are proposed and justified, for determining the degree of severity of each zone of the façade with. To obtain some of the mathematical expressions some essays are made, as the survey to 5 experts in the issue of diagnosis of pathology of construction, in order to assess the indicator of probability of detachment in 197 images of specific zones of façades. The conclusions and proposals of this thesis have implications on the whole built park, since the scope of it is all kinds of building, regardless of aspects such as ownership (private or public), use (housing, sanitary, school, sports , etc.), or geographic location.
