Deflexiones de la Guadua angustifolia Kunth para culmos sometidos a esfuerzos por flexión con cargas permanentes en el ambiente de Bogotá. Modelo de Burger y modelo de Findley.
Cargando...
Autores
Chavarro Morales, Sergio Andrés
Director
Tipo de contenido
Document language:
Español
Fecha
Título de la revista
ISSN de la revista
Título del volumen
Documentos PDF
Resumen
En el presente trabajo de investigación se modelaron las deflexiones iniciales y diferidas de elementos a flexión de Guadua angustifolia Kunth (GAK) bajo cargas permanentes para diferentes niveles de esfuerzo en el ambiente de Bogotá. Se realizaron ensayos a flexión de culmos de guadua en un montaje con cargas puntuales localizadas lo más cerca posible a los tercios de su longitud y se clasificaron en dos grupos: ensayos de corta duración y de larga duración. En las probetas de corta duración se determinaron conicidades, longitud de entrenudos, asimetría en la aplicación de la carga, distribución de esfuerzos a flexión y módulo de elasticidad longitudinal teniendo en cuenta los anteriores aspectos, el cual fue comparado con el que se determina utilizando el Reglamento Colombiano de Construcción Sismo Resistente (NSR-10) y las Normas Técnicas Colombianas (NTC5525). En los ensayos de larga duración se estableció y se aplicó una carga permanente para diferentes niveles de esfuerzo que varían entre el 95% y 30% del esfuerzo de falla promedio obtenido en los ensayos estándar de corta duración y se registraron las deflexiones diferidas y tiempos de falla de las probetas. Finalmente se aplicaron para cada una de las curvas de deflexión en el tiempo los modelos de Burger y Findley determinando la relación de sus parámetros para su obtención. Con los registros de deflexiones en el total del tiempo de la duración de los ensayos, se obtuvo como resultado que los modelos de Burger y de Findley tienen al menos un coeficiente de determinación de 0,7689 y 0.9837, respectivamente. Igualmente se analizaron los modelos de Burger y Findley con los registros de las deflexiones en un rango de tiempo menor hasta en un 12.5% del total del tiempo de la duración del ensayo obteniendo un coeficiente de determinación de al menos 0.9019 y 0.5932, respectivamente. De los resultados se puede concluir que el modelo de Findley es más efectivo describiendo la totalidad de la curva de deflexiones basado en todos los datos del ensayo. En cambio el modelo de Burger predice con mayor exactitud la curva de deflexiones diferidas a partir de un rango de datos iniciales.
Abstract. In this research, initials and deferred deflections of the Bamboo Guadua angustifolia Kunth under permanent loads for different stress levels on Bogotá's environmental conditions were modeled. Bending tests were performed with loaded points, located at thirds of the samples length. Samples were classified into two groups once the test were performed: short term and long term test. In short term specimens, properties such conicity, internode length, asymmetry in the applied load, bending stress distribution and modulus of elasticity were determined and compared with the methodology of the Colombian Regulations of Earthquake Resistant Constructions (NSR-10) and the Colombian Technical Standards (NTC5525). On the other hand, long term tests were performed on samples with different stress levels ranging between 95% and 30% of the average failure stress obtained according to the standard short term test. Deferred deflections and failure times of the long term samples were also registered. Burger and Findley models were applied to each one of the time deflection curves establishing the relation of their obtained parameters. Findley and Burger models have at least a coefficient of determination of 0.7689 and 0.9837, respectively, when models were estimated with deflections records between the total duration of the tests. In contrast, Burger and Findley models determined in a range of shorter time were analyzed (12.5% of the total duration) obtaining a coefficient of determination of at least 0.9019 and 0.5932, respectively. The quantitative and qualitative analysis of the data showed that the Findley model is more accurate describing the entire deflection curve when total data test results are available, whereas the Burger model predicts more accurately the deferred curve deflection when a initial data range is used.
Abstract. In this research, initials and deferred deflections of the Bamboo Guadua angustifolia Kunth under permanent loads for different stress levels on Bogotá's environmental conditions were modeled. Bending tests were performed with loaded points, located at thirds of the samples length. Samples were classified into two groups once the test were performed: short term and long term test. In short term specimens, properties such conicity, internode length, asymmetry in the applied load, bending stress distribution and modulus of elasticity were determined and compared with the methodology of the Colombian Regulations of Earthquake Resistant Constructions (NSR-10) and the Colombian Technical Standards (NTC5525). On the other hand, long term tests were performed on samples with different stress levels ranging between 95% and 30% of the average failure stress obtained according to the standard short term test. Deferred deflections and failure times of the long term samples were also registered. Burger and Findley models were applied to each one of the time deflection curves establishing the relation of their obtained parameters. Findley and Burger models have at least a coefficient of determination of 0.7689 and 0.9837, respectively, when models were estimated with deflections records between the total duration of the tests. In contrast, Burger and Findley models determined in a range of shorter time were analyzed (12.5% of the total duration) obtaining a coefficient of determination of at least 0.9019 and 0.5932, respectively. The quantitative and qualitative analysis of the data showed that the Findley model is more accurate describing the entire deflection curve when total data test results are available, whereas the Burger model predicts more accurately the deferred curve deflection when a initial data range is used.