Analysis of a mass production balancing method for turbochargers

  • David Quashie

Student thesis: Master's Thesis

Abstract

Purpose of research
The research considers an existing application of high-speed balancing of turbochargers in mass production at Cummins Turbo Technologies. The method used is for each part to be run on a high-speed balance station which measures vibration and speed. The machine attempts to minimise the vibration response by removing material from a single correction plane, the compressor locknut. As a mass production process, cycle time, efficiency and yield are significant considerations. The purpose of the research is to assess and analyse the process used: to evaluate the equipment and method in light of the available literature and propose and assess any potential options to improve efficiency.

Key findings
The machinery and standards used in the existing high-speed balance process are examined and assessed. The balancing equipment used is found not to apply any numerical method to determine where to make a correction, instead removing material at a fixed offset from the vibration phase angle at a fixed speed. This leads to repeated attempts to correct on some units and even scrap or rework of units where multiple attempts have been made. This has a negative impact on the overall output of the process. A numerical method based on influence coefficients is proposed and assessed in MATLAB using test data from real parts. This method, the least squares method as detailed in Goodman (1964) is shown to offer a potential improvement in the effectiveness of reducing residual vibration of the test units.

Significance of the research
The challenges of balancing turbochargers are quite specific. Rotating speeds are very high (100000-250000rpm is typical) so any residual vibration from unbalance will consequently be higher than a lower speed rotor with the same level of unbalance. This could be taken to mean that special care must be taken over each unit to minimise the vibration and risk of damage or noise. However, the volume of units that must be made mean that this is not possible, and a method must be found that can minimise vibration effectively whilst meeting required production rates of 1-2 minutes per unit. The market for turbocharger balancing equipment is limited with few suppliers and high costs. A turbocharger high-speed balance machine costs from $250,000-750,000 depending on capabilities and production rate. Cummins Turbo Technologies has approximately 15 production lines worldwide using the equipment detailed in this research, so to upgrade all these lines to a newer and more capable machine would cost an estimated $7,500,000. If a method can be found to make the existing machines more efficient at minimal cost, this would be a significant saving to the company.

Major conclusions
The major conclusions are that a least-squares influence co-efficient method could offer a significant improvement in the efficiency of the current balancing process. It is also found that the reporting of vibration phase angle in the existing equipment is incorrect, which is further complicating the task of making correction cuts on the turbocharger assemblies.
Date of Award22 Apr 2022
Original languageEnglish
SupervisorJohn Allport (Main Supervisor) & Simon Barrans (Co-Supervisor)

Cite this

'