Key Technology Research On Non-contact Online Monitoring For Fume Turbine Blade Vibration

Large-sized rotating machinery which holds very important status is important equipment that is widely applied in national economy and national defence projects. Blade vibration super-standard is one of the main reasons that cause large-sized rotating machinery to failure. Especially for the key equipment——fume turbine which is working during rigorous environment of petrochemical and oil refining trades, its operating accidents emergence in endlessly and online monitoring system is urgent to develop. In order to ensure fume turbine set works safely, high-efficiently and chronically, to guarantee the safe production in enterprises and to reduce the huge economic loss caused by the accidents, key technology research on non-contact online monitoring for fume turbine blade vibration which is sponsored by NSFC project"The Key Technique Research of Tip-timing Vibration Parameter Measurement for Rotating Blade", NECT and"Research on Real-time Online Monitoring System for Fume Turbine Blade"project from SINOPEC, is discussed in this dissertation.Profound research on relative technology fitting for fume turbine and analyzing arithmetic for vibration parameters is mainly done in this dissertation, and it provides theoretical foundation for integral parameters measurement of asynchronous and synchronous vibration during variable operating cases of constant rotation speed monitoring and variable rotation speed testing. Main works are shown as follows:1. Mathematic model and circuit model of capacitive sensor is set up, and the effect to sensor's timing precision by system bandwidth is analyzed and simulated on computer. Thus system bandwidth of the sensor is designed with optimization, and finally satisfying result is obtained.2. Performance parameters of timing system are chosen properly, and then simple intellectual disposal to blade tip-timing signals is done on FPGA and true-false decision and circle-change disposal are done to vibration data on DSP. Finally the real-time pre-procession flow of vibration data is completed.3. Based on"5+2"measurement scheme,"5+2"frequency identification arithmetic is reinforced and perfected, and it is applied to identify frequency of synchronous vibration, and then theoretical deduce for synchronous vibration amplitude measurement by Fourier coefficient is done compendiously using new thoughtway, and application research is done at the same time.4. As to measuring asynchronous vibration's frequency, theoretical deduce and application of frequency measurement method by small-angle delay-sampling are researched, thus measurable frequency range is largely extended and the number of used sensors is rapidly reduced to two. As to measuring synchronous vibration's engine-order and amplitude, theoretical deduce and application of two-equal-angle method are researched, thus measurable vibration engine-order is extended and at the same time the number of used sensors is reduced to three.5. In order to measure synchronous vibration's engine-order and amplitude by a single sensor, blade vibration model and measurement model for speed-vector-end-tracking method are set up and its theoretical formulas are deduced and simulated by computer; Levenberg-Marquardt method is employed to fit vibration displacement curve by nonlinear least square fitting, so the center frequency of rotating speed during resonance can be obtained and this provides objective and exact theoretical base for resonance frequency measurement.6. The influence of rotating speed undulation discovered in experiment is eliminated by axes-set vibration suppression technology. Based on this technology and a great deal of experiment data, experimental verification is done to most of the arithmetic above. The result shows feasibility and validity of the arithmetic.