Rotating Blade Vibration Detection And Parameters Identification Technique Using Blade Tip-Timing

Large-sized rotating machinery such as aircraft engine, gas turbine, steam turbine, air blower etc. They are important equipments which are widely applied in aviation, ships, electric power, petrifaction, metallurgy and so on. Blades are the key component of large-sized rotating machinery and ensure the equipment working safely and high efficiency. Vibration is one of the main reasons which cause blades out of work. Vibration parameters identification is hard because the signal of BTT (Blade Tip-Timing) is badly sub-Nyquist-sampled. In this dissertation, the blade vibration parameters identification technique was discussed mainly base on the achievements of passed years. The model of the BTT system has been set up, and the feature of blade vibration signal has been analyzed in theory and simulation. New techniques for vibration parameters identification have been put forward. Many experiments to verify the technique have been done, and finally satisfied results were obtained. The main works are shown as follows:1. After force analysis of the rotating blade had been done, model of the whole BTT system which consists of blades model, exciting force model, and BTT sensor model has been set up. The models are the basic of BTT arithmetic analyzing in theory and simulation. The characters of synchronous and asynchronous vibration of blade under constant speed or varied speed have been contrasted by simulation.2. The difficulty of BTT signal processing under different condition has been discussed. Some typical BTT analyzing arithmetic has been deduced theoretically, such as speed-vector-end-tracking method (single parameter technique), two-parameter plot method, and autoregressive method. The merits and demerits of these methods have been analyzed, and the direction to search new method under sub-Nyquist-sampling is clear.3. The new technique called multi-sensors arbitrarily distributed for identifying blade vibration parameters has been introduced. The technique can be divided into three methods under different situations which are synchronous vibration under vary speed, synchronous vibration under constant speed and asynchronous vibration under constant speed, and two patents of invention are obtained. These methods have been discussed in detail by theory and simulation analysis. Much arithmetic methods are used in the technique, such as least squares, curve fitting, all phases FFT, engine-order traverse and so on. The Amplitude, frequency, and engine-order of blade vibration can be identified correctly by the new technique. The problem which caused by sub-Nyquist-sampled is can not identify all vibration parameters precisely. Now it has been solved by the new technique.4. Different distribution of sensors will affect the identified result of each method. This situation is discussed in the paper. The function of sampling point distributed range (DR) is build to evaluate the character of multi BTT sensors installation. And obvious level (ΔS %) is used for judging the result of engine-order traversing. The way to choose applicable distribution of sensors is studied out when using multi-sensors arbitrarily technique to identify vibration parameters.5. Many blade vibration online measuring experiments have been completed on different rotating machineries. The blade vibration analyzing arithmetic has been proved by great deal of experiment data. The result shows feasibility and validity of the arithmetic.