Uso de machine learning para predição da taxa de sedimentação de fluidos magneto-reológicos

• Ano de defesa: 2025
• Autor(a) principal: Teixeira, Guilherme Silva
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Mato Grosso; Brasil; Instituto de Ciências Exatas e da Terra (ICET) – Araguaia; UFMT CUA - Araguaia; Programa de Pós-Graduação em Ciência de Materiais
• Palavras-chave em Português: Fluidos magneto-reológicos; Multiple linear regression; Decision tree; Gradient boosting; Random forest; CNPQ::CIENCIAS EXATAS E DA TERRA; Magnetorheological fluids
• Link de acesso: http://ri.ufmt.br/handle/1/7308

Magnetorheological fluids (MRFs) are smart materials whose rheological properties can be drastically altered by the application of an external magnetic field. These materials have a wide range of applications in industries such as construction, automotive, medicine, parts polishing, and others. Because of this, interest in understanding their properties has been growing in recent years. The material is composed of magnetic microparticles (e.g. ferrite, carbonyl iron, chromium dioxide, and iron oxide) dispersed in a carrier fluid (e.g. synthetic oil, mineral oil, water, and others) and additives. Due to the difference in density between the microparticles and the carrier fluid, in the absence of an external magnetic field, the particles settle by gravity. Several factors can improve the stability of these materials, such as size, shape, and coating of the microparticles, viscosity, density, and mixing of different carrier fluids, addition of nanoparticles of different compositions, and specific additives. However, testing the different combinations can be a challenge because it is an expensive material. At the same time, the application of machine learning (ML) models as an alternative to theoretical, experimental and computational methods has proven effective in solving fluid rheology problems. This tool emerges as an attractive approach, making it possible to investigate multiple variables with significantly fewer resources than other approaches. In this work, five machine learning models were used to predict the sedimentation rate of magnetorheological fluids based on several variables that are known to influence the stability of the material. The results showed that the Random Forest was the model with the best predictive capacity, based on the coefficient of determination and root mean square error.