Modelling and experimental validation of active and passive eddy current sensors for blade tip timing

Abstract

To monitor the vibration of blades in rotating machinery, the contactless method called Blade Tip Timing (BTT) is widely used. blade vibration and clearance are important diagnostic features for condition monitoring, including the detection of blade cracks. To perform the BTT technique, optical sensors were widely used by industry due to their high accuracy, but the main drawback of these systems is their low tolerance to the presence of contaminants. To overcome this downside, eddy current sensors are a good alternative for health monitoring applications in gas turbines due to their insensitivity to contaminants and debris. This type of sensor has been used by many researchers, predominantly on the experimental side to investigate BTT systems and there is a lack of modelling to support the measurement system design. This paper fills the gap between experiments and modelling for the particular case of a blade rotating past eddy current sensors. Hence the novelty of this paper is the simulation of the BTT application using detailed quasi-static finite element models of the electro-magnetic field to estimate the outputs from active and passive eddy current sensors. A test rig composed of a bladed disk with 12 blades clamped to a rotating shaft was designed and manufactured in order to validate the proposed models with experimental measurements. Finally, a parametric study is presented to show the effect of the blade tip clearance and the rotational speed on the accuracy of the BTT measurement. This leads to better understanding of the sources of error in the time of arrival of the blades passing the sensor and hence insight into the blade vibration measurement accuracy.