Estudos para avaliar erros de sistemas de medição de processos de modelagem por deposição fundida

Abstract

Manufacturing accurate products with reduced defects through reduced variability is a challenge for additive manufacturing with regard to standardization and implementation in mass production. The variation of any manufacturing process can be related to the process itself, or to the measurement system. If a part is characterized as non-conforming and the variability found is attributed to the production process, actions are implemented in order to improve the capability. However, the implemented actions may not necessarily result in an improvement, if the process is already capable enough and the measurement error is unacceptable when comparing the process variations. So, conforming parts can be classified as bad, on account of deviations from the true value due to error in measurements. Therefore, it is necessary to investigate both the variability of the process and the measurement system established for assertive decision making. Thus, this dissertation presents a set of Measurement System Analysis (MSA) studies of a Fused Deposition Modeling process. The main objective of this research is the validation of the established measurement system, by carrying out Type I, Bias and Linearity, Repeatability and Reproducibility studies, and also the development of a method to plan, execute, analyze and simultaneously mitigate the systematic and random error of measurement systems, integrating the analysis of the Cgk capability index to the design of experiments. The methodological procedure, with a mixed approach, allowed the development of a Systematic Literature Review (RSL) study based on a sample of scientific articles extracted from the Web of Science (WoS) scientific knowledge base. Then the experimental design was performed to carry out the aforementioned studies, where a reduced set of bending specimens manufactured according to ISO 178 were evaluated repeatedly, by different operators, to determine the sources of variation. Contributions comprise clarification of the MSA's intellectual structure helping readers to plan and execute similar work, as well as a new method for analyzing measurements and measurement system capability. The statistical studies and the proposed method were implemented using data obtained from the 3d printing process of a Da Vinci 1.0 Pro printer.