Dimensional and Geometrical Errors in Vacuum Thermoforming Products: An Approach to Modeling and Optimization by Multiple Response Optimization

In the vacuum thermoforming process, the product deviations depend on several parameters of the system, which make the analysis, the computational modeling, and the optimization of errors a multi-variable process with conflicting objectives. In this sense, the aim of this work was to study the dimensional and geometrical errors as well as the optimization (minimization) of these errors in one typical vacuum thermoforming product made of polystyrene (PS). In particular, it was intended to predict and minimize errors in a range of ideal tolerances using Multiple Response Optimization (MRO) Models. Thus, through the fractional factorial design (2(kP)), initial experimental tests were performed using proposed measurement procedures, and Analysis of Variance being the data analysis is discussed. Following that, the MRO models were implemented which were also validated to represent the sample data. Through this analysis of the results, it can be concluded that the regression models of errors are not linear functions, hence, the developed models are valid for the studied process, and finally that the validation results proved the efficiency of MOR models developed, but these models will not be able to generalize to new situations in a range far from the values studied.