Estudo de manifestaçoes patológicas em revestimento cerâmico de fachadas na cidade de Fortaleza-CE: análise, tratamento de dados e diagnóstico

Abstract

Sandy soils are commonly used in engineering works, whether in earth dams, containment works, roads, among others, for drainage and related works. Thus, it is of utmost importance to know the characteristics of the material to be used in the work, especially with regard to its hydraulic conductivity. In order to estimate the permeability coefficient of sandy soils, some researchers such as Hazen (1930), Kozeny-Carman (1938), Carrier (2003) and Chapuis (2004) determined mathematical models in which they correlated basic characterization parameters with hydraulic conductivity. Given this, this work has as main objective to determine a new more efficient mathematical formulation to estimate the permeability coefficient of sandy materials, through statistical analysis. For this research, seven sandy samples were selected, determining the permeability values under constant load for different voids indices. The study also included an adaptation of the necessary apparatus to the densification and permeability tests, both tests being performed in the same equipment. From the results of the obtained tests, several parameters that could influence the mathematical model were evaluated, and by an interactive process using the Solver® computational tool, the empirical equation for the proposed study was determined. In order to evaluate the model developed in this research, we compared the determination of the permeability coefficient by the equation of this research in relation to the different equations available in the literature, identifying that the proposed equation presents a smaller error than the other mathematical equations. To validate the proposed empirical model, the parameters of a distinct sample were used to verify the accuracy of the model under study and to compare it with other methodologies. The results obtained by the model were satisfactory, also presenting a considerably smaller error in relation to the other methods found in the literature.