SIMULATION OF FLOATING RING BEARING ON REDUCER GEARS OF ROLLER MILL

• Main Author: Siena Pari, Hagi
• Format: Final Project
• Language: Indonesia
• Subjects: Teknik (Rekayasa, enjinering dan kegiatan berkaitan)
• Online Access: https://digilib.itb.ac.id/gdl/view/70119
• Institution: Institut Teknologi Bandung

Floating ring bearings generally have a simple construction, especially when compared to other bearings such as ball bearings and journal bearings that consist of many components that need to be precisely arranged. These bearings are used in reducer gears on roller mills in cement plants. In this reduser, there is a floating ring bearing system that is used to facilitate the rotation of the planetary gear system around a shaft. The reducer system is designed to run for 170,000 hours, but failed after only 17,400 hours on the planetary gear. This study used a suitable simulation application to analyze the cause of roller mill reducer failure based on the design of the bearing and found whether the bearing would work properly or fail before the designed service life. Simulations were carried out with Simscale, Ansys CFX, and Ansys Fluent applications on 3 floating ring bearing models. The parameters used to test the success of the simulation include: stability of the solver graph for the outlet domain at the end of the simulation, fluid area velocity, velocity magnitude, velocity distribution, pressure gradation and visualization. The results obtained were that the Ansys Fluent application is more suitable for simulating floating ring bearings because it can produce more realistic simulation results than the Simscale application with the same model and parameters. There were obstacles in the process of making the bearing model mesh caused by the thin volume of fluid between the bearing components, mesh conflicts at the meeting area with different cell volumes and dimensions, and the complex shape of the bearing model. It is recommended in future research to make the pieces of the model separately and then put them together to form the desired model.