Date of this Version
Grinding burn has been one of the re-occurring issues at a leading automotive manufacturer. Historically when grinding burn occurred, trial and error was used to delicately find a set up process settings that provided good-enough-to-get-by results, but not optimal. Some changes include changing the coolant manifold to improve the coolant flow and rate, changing the grinding wheel, changing the type of coolant. The experiments have proved to introduce some improvements, however the occurrence of grinding burns have not been reduced or quantified and the cause of the grinding burn has not been identified or quantified. Grinding is a complex phenomenon and there are many variables that play a role in grinding burn and have to be studied as a complete system. This thesis aims at studying the cause of grinding burn in crankshaft systematically using World Class Manufacturing’s Advanced Kaizen methodology.
While the Advanced Kaizen provides the framework to approach the grinding burn, Design of experiments is the tool used in the Kaizen to study the grinding burn. In production scenarios, running a design of experiment poses many challenges, which range from availability of resources, time, material batches, some factors being hard to change etc. These challenges and restrictions can affect the three main principles of DOE namely, Randomization, Replication, and blocking. Hence selecting the right type of DOE and performing the right kind of analysis becomes crucial. This thesis focuses on this objective and introduces Fractional Factorial Split plot Design to analyze the grinding burn issue.
Adviser: Kamlakar P. Rajurkar