FRACTIONAL FACTORIAL DOE PLANNING TO REDUCE INTERNAL REJECTIONS ON LEAK TEST OF A COMBUSTION ENGINE PRODUCTION LINE

Abstract

This work was applied in automotive industry in order to improve the efficiency of the combustion engines production leak test by using a fraction factorial DOE (Design of Experiments). The application of the anaerobic seal is done by a robot dispenser, and the main goal of this work is to demonstrate the application of fraction factorial DOE to improve the inputs of the application and eliminate leak fails on production leak test on the combustion engines at “T-joint” from the intersection between cylinder head, valve cover gasket and front cap. The method applied is experiment and the software used to plan the DOE and analyze the collected data is the Minitab. The inputs considered by the team application were four as: hose temperature, panel temperature, robot tip distance, and robot speed, as a result the input with the most significant effect identified on the DOE analysis is the robot tip distance. The outcome is zero leak engine on the “T-joint” at leak test, and then it was decided to evaluate two outputs: 1 - leak rate on leak test – this output showed no significance when analyzed the results from the DOE because it does not consider noise factors as environment temperature and humidity; 2 – seal bead diameter – this output showed to be significance on DOE, and is directly related to the root cause, because the engines failed with leak on production leak test are missing or with lack of anaerobic seal on the “T-joint”. The main result of this fractional factorial DOE is the best set of input parameters which exceeded the target of 80% improvement and achieved zero leak on the “T-joint” between cylinder head, valve cover gasket and front cap. This shows that fractional factorial DOE is an efficient method to be applied on the industry to improve process.