Experimental Investigation of the Grouped Blade Vibration for Steam Turbine by Noncontact Sensors

Abstract

All turbine blades have mistuned structures caused by manufacturing variations within the manufacturing tolerance, such as the geometrical deviations and variance of material properties. The mistuning effect has a known tendency to increase the dynamic stress, but it is also known to be difficult to predict the maximum vibration response before the operation. This paper studies the blade vibration of grouped blades in a low-pressure steam turbine. The study objectives are to characterize the vibration behavior of the grouped blade structure and to evaluate the maximum response of all blades in a stage experimentally. An experimental investigation is carried out in a vacuum chamber, and blades are excited by an air jet during start-up and shut-down. The circumferential blade amplitude distribution is measured by noncontact sensors (NCSs) and strain gauges (SGs). The circumferential blade amplitude distribution is found to differ depending on vibration modes and nodal diameters (NDs), but the relative tendency is almost the same for all types of operation at each mode and all NDs. Therefore, the median of all experimental results obtained with the NCSs is used in a comparison with calculation results and results of two theoretical curves obtained using equations from the literature. In comparing the measurement results and the calculation results, the circumferential blade amplitude distribution is not the same with all modes and NDs. However, the maximum amplitude magnification is about 1.5–1.8, and all measurement results are lower than the results for the two theoretical equations. This means the maximum response comparison to the tuned blade gives an evaluation on the safe side by the two theoretical equations.