Quantification of Uncertainties in Experimental Modal Parameters Estimation: An Industrial Case Study
Diego Fernandes Rodrigues1; Omar Rachid Murad1; Marcelo Areias Trindade2; Leopoldo Pisanelli Rodrigues de Oliveira2
1 Whirlpool Latin America; 2 Escola de Engenharia de São Carlos - Universidade de São Paulo
doi:10.20906/CPS/USM-2016-0056
Resumo
In an industrial metrology environment, surrounded by a large amount of variation sources, where time to experiment execution must be minimized without jeopardizing results quality, knowledge about the measurement processes is a key aspect in guaranteeing confidence in the outcome results. A Measurement System Evaluation (MSE) applied to a modal testing rig allows qualifying and/or quantifying the measurement process according to its reproducibility, precision, discrimination and stability, obtaining the contribution of each variation source in the experimental modal analysis to a determined confidence interval. The methodology presented in this paper differs from the usual MSE due the use of process mapping to organize the steps of the measurement process, tracing the sources of variation, applying the rational subgrouping concept in order organize these sources in an experimental tree. Once executed and analyzed with control charts, this method allows the evaluation of the variation in each step of the measurement process, the analysis of the discrimination threshold, measurement precision, measurement reproducibility and stability of the measurement process. The selected case study presents a MSE of a modal testing rig and shows that it is possible to evaluate the expected variation in the first natural frequency measured on a washing machine component, in the context of a workshop metrology quality control test.
Palavras-chave: Measurement System Evaluation; Experimental Modal Analysis; Uncertainty Quantification