Previsão da contração e da variação dimensional em componentes injetados através de simulação computacional

• Ano de defesa: 2014
• Autor(a) principal: Kiam, Thyago Morette
• Orientador(a): Pessan, Luiz Antonio
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de São Carlos; Câmpus São Carlos
• Programa de Pós-Graduação: Programa de Pós-Graduação em Ciência e Engenharia de Materiais - PPGCEM
• Palavras-chave em Português: Moldagem por injeção; Simulação; Contração; Empenamento; DOE
• Área do conhecimento CNPq: CIENCIAS EXATAS E DA TERRA
• Link de acesso: https://repositorio.ufscar.br/handle/20.500.14289/7715

Both, average linear shrinkage and dimensional variation behavior for injection-molded components influence its final application performance, hence the project overall quality and profit [1]. The use of simulation software for estimating dimensional variation is still incipient due to some factors such as lack of proficient users, code simplifications and lack of raw material proper characterization. On the Brazilian context, this third item is particularly critical, once the coefficients for warpage results improvement for the main commercial code available on the market, Autodesk Moldflow Insight (AMI); requires a very specific characterization named ‘corrected residual in-mold stress’ (CRIMS), just generated by the software developers in a quite complex and low cost-benefit process for national resin suppliers. The present work used computational experimentation based on commercial code AMI 2013 aiming both: to generate CRIMS coefficients from correlation studies involving optimization algorithm, based on Cellere and Luchetta work [2]; and to perform a case study proposing a method that make use of optimization algorithm and Monte Carlo Simulation in order to drive both, design and process decisions; in the end providing a estimative of the dimensional variation behavior on early design and mold development phases. The results level of error obtained with the CRIMS coefficients generated on this work is close to the results generated from coefficients generated by the supplier for the same material (confidential data) that represents the state-of-the-art of correction coefficients generation; confirming the potential for the method showed in this work. The adaptable nature of the method described on Cellere and Lucchetta paper [2] was maintained.