Este trabalho avalia a utilização de técnicas de planejamento estatístico de experimentos com misturas envolvendo variáveis de processo no estudo de formulações para a produção de revestimentos cerâmicos, buscando minimizar o caráter empírico encontrado industrialmente nesta etapa do processamento. Misturas de três matérias-primas (argila, feldspato e areia), utilizadas industrialmente para a produção de revestimentos, foram ensaiadas sob condições padronizadas visando à quantificação da influência de cada matéria-prima na resistência mecânica e na absorção de água das peças após a queima. As influências de duas variáveis de processo, pressão de compactação e temperatura de queima, também foram avaliadas. Os resultados obtidos demonstram o potencial de utilização das técnicas de planejamento estatístico de experimentos no estudo e desenvolvimento de formulações para revestimentos cerâmicos.
Maintenance actions on industrial equipment are essential to reduce expenses associated with equipment failures. Based on a well-fitted model, it is possible, through the estimated parameters, to predict several functions of interest, such as the cumulative average and reliability functions. In this paper, a new frailty model is proposed to analyze failure times of repairable systems subject to unobserved heterogeneity actions. The Non-Central Gamma distribution is assumed to the frailty random variable effect. The class of minimal repair models for repairable systems is explored considering an approach that includes the frailty term to estimate the unobserved heterogeneity over the systems’ failure process. Classical inferential methods were used to parameter estimation and define the reliability prediction functions. A simulation study was conducted to confirm the properties expected in the estimators. Two real-world data known in literature were used to illustrate the estimation procedures and validate the proposed model as a viable alternative to those already established in the literature. The results obtained highlight the potential of our proposed approach, particularly for industries dealing with such systems, where unquantifiable factors may impact equipment failure times.
In this article, we propose an approach for incorporating continuous and discrete original outcome distributions into the usual exponential family regression models. The new approach is an extension of the works of Suissa ( 1991 Suissa, S. (1991). Binary methods for continuous outcomes: A parametric alternative. J. Clin. Epidemiol. 44:241–248. [Crossref], [PubMed], [Web of Science ®] [Google Scholar] ) and Suissa and Blais ( 1995 Suissa, S. , Blais, L. (1995). Binary regression with continuous outcomes. Stat. Med. 14:247–255. [Crossref], [PubMed], [Web of Science ®] [Google Scholar] ), which present methods to estimate the risk of an event defined in a sample subspace of an original continuous outcome variable. Simulation studies are presented in order to illustrate the performance of the developed methodology. Real data sets are analyzed by using the proposed models.