Proposta de uma metodologia baseada em análise modal operacional para monitoramento de fim de vida de ferramentas de corte

• Ano de defesa: 2009
• Autor(a) principal: Meola, Tatiana
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Uberlândia; BR; Programa de Pós-graduação em Engenharia Mecânica; Engenharias; UFU
• Palavras-chave em Português: Análise modal operacional; ERA; Identificação de forças; Monitoramento; Processo de torneamento; Ferramentas para cortar metais; Usinagem; Análise modal; Operational modal analysis; Forces identification; Monitoring; Turning process; CNPQ::ENGENHARIAS::ENGENHARIA MECANICA
• Link de acesso: https://repositorio.ufu.br/handle/123456789/14688

The indirect monitoring of cutting process by vibration signals measurements has comproved effectiveness, however it is very dependent on machine operating conditions and the vibration levels are not correlated directly with cutting process, like bearing vibrations for example. An alternative methodology is the use of dynamometers to measure cutting forces directly. However, due to transducers dimension and assembly difficulties, force measurements are not always feasible in a productive process. Indirect estimation of cutting forces by machine vibration measurements and classical techniques of dynamic forces identification have as drawback the difficult to get a reliable dynamic model. The dynamic system characteristics modify with tools changes and configuration of assemblage. To minimize these problems, this work presentes a methodology based on operational modal analysis (OMA) technique and Eigensystem Realisation Algorithm (ERA) for tools condition monitoring in turning process. OMA technique was used to estimate, in operating conditions, a modal metamodel to be used at forces identification by means ERA. To study several parameters influence at quality results of fitting model and at forces identification a modal model of turning machine tool port was developed from a finite elements model. The methodology was experimentally validated with success in a turning process, since estimated efforts for six different tools presented the same tendency with wear tool evolution.