Combinação de técnicas de delineamento de experimentos e elementos finitos com a otimização via simulação Monte Carlo
Carregando...
Data
2019-07-23
Autores
Oliveira, José Benedito da Silva
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Estadual Paulista (Unesp)
Resumo
A Estampagem a Frio é um processo de conformação plástica de chapas metálicas, que possibilita, por meio de ferramentas específicas, obter componentes com boas propriedades mecânicas, geometrias e espessuras variadas, diferentes especificações de materiais e com boa vantagem econômica. A multiplicidade destas variáveis gera a necessidade de utilização de técnicas estatísticas e de simulação numérica, que suportem a sua análise e adequada tomada de decisão na elaboração do projeto das ferramentas de conformação. Este trabalho foi desenvolvido em uma empresa brasileira multinacional de grande porte que atua no setor de autopeças, em seu departamento de engenharia de projetos de ferramentas, com o propósito de reduzir o estiramento e a ocorrência de trincas em uma travessa de 6,8 [mm] de aço LNE 380. A metodologia proposta obtém os valores dos fatores de entrada e sua influência na variável resposta com o uso de técnicas de Delineamento de Experimentos (DOE) e simulação pelo método de Elementos Finitos (FE). Uma Função Empírica é desenvolvida a partir desses dados, com o uso da técnica de regressão, obtendo-se a variável resposta y (espessura na região crítica), em função dos fatores influentes xi do processo. Com a Otimização via Simulação Monte Carlo (OvSMC) insere-se a incerteza nos coeficientes desta Função Empírica, sendo esta a principal contribuição deste trabalho, pois é o que ocorre, por via de regra, na prática com problemas experimentais. Simulando-se por FE as ferramentas de estampar projetadas com os inputs da OvSMC, valida-se estatisticamente os resultados, constatando-se que o valor da variável resposta esta dentro do Intervalo de Certeza (IC) fornecido pela OvSMC. A validação prática dos resultados junto a um grupo de especialistas se deu pela constatação de que houve uma redução de cerca de 25% do estiramento inicial, que na visão destes eliminará a ocorrência de trincas na prática. Confirmou-se também que o conceito de variar a altura, portanto o ponto de reação do estampo inferior, de modo a retardar a ação do punção na região crítica sujeita à tração, colabora para a preservação da espessura inicial do produto estampado. Esse ganho conceitual significativo e aplicável a outros componentes estampados de característica similares demonstra, junto às outras vantagens já mencionadas, a contribuição da metodologia proposta na aplicação em sua totalidade na elaboração dos projetos de ferramentas.
Cold Stamping is a process of plastic forming of sheet metal, which enables specific tools to obtain components with good mechanical properties, varied geometries and thicknesses, different material specifications and with a good economic advantage. The multiplicity of these variables generates the need to use statistical techniques and numerical simulation, which support their analysis and adequate decision making in the design of the conformation tools. This work was developed in a large multinational Brazilian company that operates in the auto parts sector, in its engineering department of tools projects, with the purpose of reducing the stretch and the occurrence of cracks in a cross beam of 6.8 mm ] of LNE 380 steel. The proposed methodology obtains the values of the input factors and their influence on the variable response with the use of Experimental Delineation (DOE) techniques and simulation by the Finite Element (FE) method. An empirical function is developed from these data, using the regression technique, obtaining the response variable y (thickness in the critical region), as a function of the influencing factors xi of the process. Monte Carlo Simulation Optimization (OvSMC) inserts the uncertainty in the coefficients of this Empirical Function, which is the main contribution of this work, since this is usually the case in practice with experimental problems. By simulating the stamping tools designed with the OvSMC inputs, the results are statistically validated, showing that the value of the response variable is within the Certainty Interval (CI) provided by OvSMC. The practical validation of the results with a group of experts was verified by the fact that there was a reduction of about 25% of the initial stretch, which, in their view, will eliminate the occurrence of cracks in practice. It has also been confirmed that the concept of varying the height, hence the reaction point of the lower stamp, in order to retard the action of the punch in the critical region subject to traction, contributes to preserving the initial thickness of the stamped product. This significant conceptual gain, which is applicable to other similar printed components, characterizes together with the other advantages already mentioned, the contribution of the methodology proposed in the application in its entirety in the elaboration of tool designs.
Cold Stamping is a process of plastic forming of sheet metal, which enables specific tools to obtain components with good mechanical properties, varied geometries and thicknesses, different material specifications and with a good economic advantage. The multiplicity of these variables generates the need to use statistical techniques and numerical simulation, which support their analysis and adequate decision making in the design of the conformation tools. This work was developed in a large multinational Brazilian company that operates in the auto parts sector, in its engineering department of tools projects, with the purpose of reducing the stretch and the occurrence of cracks in a cross beam of 6.8 mm ] of LNE 380 steel. The proposed methodology obtains the values of the input factors and their influence on the variable response with the use of Experimental Delineation (DOE) techniques and simulation by the Finite Element (FE) method. An empirical function is developed from these data, using the regression technique, obtaining the response variable y (thickness in the critical region), as a function of the influencing factors xi of the process. Monte Carlo Simulation Optimization (OvSMC) inserts the uncertainty in the coefficients of this Empirical Function, which is the main contribution of this work, since this is usually the case in practice with experimental problems. By simulating the stamping tools designed with the OvSMC inputs, the results are statistically validated, showing that the value of the response variable is within the Certainty Interval (CI) provided by OvSMC. The practical validation of the results with a group of experts was verified by the fact that there was a reduction of about 25% of the initial stretch, which, in their view, will eliminate the occurrence of cracks in practice. It has also been confirmed that the concept of varying the height, hence the reaction point of the lower stamp, in order to retard the action of the punch in the critical region subject to traction, contributes to preserving the initial thickness of the stamped product. This significant conceptual gain, which is applicable to other similar printed components, characterizes together with the other advantages already mentioned, the contribution of the methodology proposed in the application in its entirety in the elaboration of tool designs.
Descrição
Palavras-chave
Processo de Estampagem., Elementos Finitos., Delineamento de Experimentos - DOE., Metodologia de Superfície de Resposta - MSR., Incerteza., Simulação Monte Carlo - SMC., Otimização via Simulação Monte Carlo - OvSMC., Stamping Process, Finite Element, Desing of Experiments - DOE, Response Surface Methodology - RSM, Uncertainty, Monte Carlo Simulation - SMC, Monte Carlo simulation Optimization - OvSMC., Estampagem (trabalhos em metal), Monte Carlo - Método de, Planejamento experimental