Improving the Manufacturing Process of a Constant Velocity Inner Race with the Application of Design of Experiments 2003-01-3578

One of the metal-mechanical industry characteristics is to present processes with many steps or operations. It is also known that the product and process quality is a result of the properly adjustment of the operations parameters during the whole production cycle. Nowadays, it is recognized that the quality improvement, when measured as variance of the nominal value (target), is reached more efficiently when the product and process is designed to reduce the variance, and not through usual final inspection. To reduce the variance it is needed to determine optimized operation parameters. When many factors are involved, the optimization becomes a complex task and needs the use of statistical tools. Among the available tools in literature, the Design of Experiments (DOE) is potentially useful in products development stages and in the improvement of existing processes and products. This paper presents a DOE application in a constant velocity inner race manufacturing process, with the aim of reducing the dimensional variance in its spline. It has been done a multi response optimization, considering the settings of five control factor simultaneously to optimize six response variables. The result obtained is an appreciable reduction of the spline dimensional variance, which represents a quality improvement and a cost reduction. Besides, there was a knowledge increase on the constant velocity inner race manufacturing process that can be extended to other components produced with similar process.