Estudo da resistência ao embutimento de pinos de aço em peças estruturais de madeiras de reflorestamento
Carregando...
Data
2017-08-03
Autores
Pallarolas, Ernesto Abel Fernando Friedmann [UNESP]
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Estadual Paulista (Unesp)
Resumo
Este trabalho teve a finalidade de estudar a resistência ao embutimento de ligações efetuadas por pinos metálicos (pregos) em peças estruturais de madeira. Para tanto foram consideradas três principais abordagens: analítica, numérica e experimental. Inicialmente foi efetuado um levantamento teórico sobre o tema, com base na literatura disponível, para avaliação da normalização nacional e internacional pertinente ao assunto. Na sequência foram realizados ensaios experimentais em corpos de prova compostos por madeiras de reflorestamento de pinus e eucalipto com ligações metálicas efetuadas por pinos metálicos com diâmetro de 6,5 mm. Os ensaios foram realizados de acordo com as recomendações das normas: brasileira ABNT NBR 7190:1997, europeia EUROCODE 5:2004 e americana ASTM D5764:2007 com posterior comparação entre os resultados obtidos para a resistência ao embutimento. Foi realizada uma análise estatística dos resultados obtidos, incluindo um estudo da resistência ao embutimento corrigida pela umidade e pela densidade. Foram calculados os valores analíticos teóricos de resistência ao embutimento mediante fórmulas propostos por cada norma e comparadas às médias dos valores experimentais obtidos pelas mesmas. Foi proposta também a análise do valor ideal de resistência por meio de métodos de análise de regressão segmentada sobre as curvas de comportamento de resposta do material. Por fim foram realizadas simulações numéricas no software ABAQUS, que tem como base o Método dos Elementos Finitos (MEF) para avaliação das distribuições de tensões nas regiões da ligação. Os resultados obtidos mostram que existem divergências significativas entre os valores experimentais de resistência ao embutimento obtidos através das normas estudadas quando comparados entre si. Também mostram que os valores analíticos calculados superestimam a resistência ao embutimento experimental para todas as normas. E ainda, que a resistência ao embutimento quando corrigida pela densidade apresenta melhor qualidade em termos de dispersão para o conjunto de dados estudado. Conclui-se que os métodos estatísticos de regressão segmentada estudados, podem ser uma boa alternativa para a modelagem do comportamento mecânico da madeira na obtenção de valores característicos mais invariantes e padronizados para os parâmetros de resistência.
This work had the purpose of studying the embedment resistance of connections made by metallic pins (nails) in structural pieces of wood. For that, three main approaches were considered: analytical, numerical and experimental. Initially, a theoretical survey was carried out on the subject, based on available literature, for the evaluation of national and international standardization related to the subject. Experimental tests were carried out on test specimens composed of pine and eucalyptus reforestation woods species with metal connections made by pins with a diameter of 6.5 mm. The tests were carried out according to the recommendations of Brazilian ABNT NBR 7190:1997, European EUROCODE 5:2004 and American ASTM D5764:2007 standards for further comparison between the obtained results for the embedment resistance. A statistical analysis of the obtained results was carried out, including a study of the resistance corrected by the humidity and the density. The theoretical analytical values of resistance were calculated using formulas proposed by each standard and compared to the averages of the experimental values. It was also proposed the analysis of the ideal resistance value by segmented regression analysis methods on the curves of the embedment behavior data set generated by the standards. Finally, numerical simulations were performed in ABAQUS software, which is based on the Finite Element Method (MEF) to evaluate the stress distributions in the bond areas. The obtained results showed that there are significant divergences between the experimental values of resistance obtained through the studied standards when compared to each other. They also show that the analytical calculated values overestimate the experimental resistance for all studied standards. Also, that the strength corrected by the density presents better dispersion perform quality for the studied data set. It is concluded that the studied statistical methods of segmented regression can be a good alternative for modeling the mechanical behavior of the wood to obtain characteristic values more invariant and standardized for the resistance parameters.
This work had the purpose of studying the embedment resistance of connections made by metallic pins (nails) in structural pieces of wood. For that, three main approaches were considered: analytical, numerical and experimental. Initially, a theoretical survey was carried out on the subject, based on available literature, for the evaluation of national and international standardization related to the subject. Experimental tests were carried out on test specimens composed of pine and eucalyptus reforestation woods species with metal connections made by pins with a diameter of 6.5 mm. The tests were carried out according to the recommendations of Brazilian ABNT NBR 7190:1997, European EUROCODE 5:2004 and American ASTM D5764:2007 standards for further comparison between the obtained results for the embedment resistance. A statistical analysis of the obtained results was carried out, including a study of the resistance corrected by the humidity and the density. The theoretical analytical values of resistance were calculated using formulas proposed by each standard and compared to the averages of the experimental values. It was also proposed the analysis of the ideal resistance value by segmented regression analysis methods on the curves of the embedment behavior data set generated by the standards. Finally, numerical simulations were performed in ABAQUS software, which is based on the Finite Element Method (MEF) to evaluate the stress distributions in the bond areas. The obtained results showed that there are significant divergences between the experimental values of resistance obtained through the studied standards when compared to each other. They also show that the analytical calculated values overestimate the experimental resistance for all studied standards. Also, that the strength corrected by the density presents better dispersion perform quality for the studied data set. It is concluded that the studied statistical methods of segmented regression can be a good alternative for modeling the mechanical behavior of the wood to obtain characteristic values more invariant and standardized for the resistance parameters.
Descrição
Palavras-chave
Resistência, Embutimento, Normalização, Análise estatística de regressão segmentada, Resistance, Embedment, Standardization, Statistical analysis of segmented regression