Estimation of turbine blade natural frequencies from casing pressure and vibration measurements

Abstract

Non-contact measurement of gas turbine rotor blade vibration is a non-trivial task, with no method available which achieves this aim without some significant draw-backs . This paper presents a truly non-contact method to estimate rotor blade natural frequencies from casing vibration measurements at a single engine operating speed. An analytical model is derived to simulate the internal casing pressure in a turbine engine including the effects of blade vibration on this pressure signal. It is shown that the internal pressure inside a turbine contains measureable information about the rotor blade natural frequencies and in-turn the casing vibration response also contains this information. The results presented herein show the residual, pressure and casing vibration, spectrum can be used to determine the rotor blade natural frequencies with validation provided for the analytical model by experimental measurements on a simplified test rig. A simulated blade fault in one of the rotor blades is introduced with successful estimation of the simulated faulty blade natural frequency.