Optimization of LM2500 Gas Generator and Power Turbine Trim-Balance Techniques

Abstract

A procedure has been developed by the U.S. Navy to trim-balance, in-place, the gas generator and power turbine rotor of the LM2500 Marine Gas Turbine Engine. This paper presents the theoretical background and the techniques necessary to optimize the procedure to balance the gas generator rotor. Additionally, a method was developed to trim balance LM2500 power turbines. To expand the implementation of both gas generator and power turbine trim-balancing, a capability has to be developed to minimize the effort required (trial weight runs, etc.). The objective was to be able to perform consistently what are called “First-Shot” trim balances. First-Shot trim balances require only one weight placement to bring the engine vibration levels to within the specified goals (less than 0.002 of an in. maximum amplitude) and that being the final trim weight. It was realized that the Least-Squares Influence-Coefficient Method, even with a good set of averaged influence coefficients, can lead to a number of trial weight experiments before the final trim weights can be placed. The method used to maximize the possibility of obtaining a First-Shot trim balance was to use modal information to tailor the influence coefficient sets to correct the most predominant and correctable imbalance problem. Since the influence coefficients were tailored, it became necessary to be able to identify, in the initial vibration survey, the type of response a particular LM2500 has. Using modal information obtained from a LM2500 rotor dynamics model and from the early trim-balance efforts, it was possible to identify the modal response of a given LM2500 and optimize the trim balance of that engine. With these improved techniques a 70 percent success rate for First-Shot trim balance has been achieved and the success rate of the trim balance procedure, as a whole, has been near 100 percent.