Abstract
A lack of reliable experimental data on transonic blade flutter in real turbomachines hampers the further improvement of computational design predictions for off-design operation regimes of newly-built machines. Acquiring unsteady pressure distribution on blades in real turbomachines already in operation is practically impossible. The goal of this work is to explore if an approximate unsteady pressure distribution can be created experimentally in a simple aerodynamic tunnel by composing a sequence of blade surface steady pressures acquired for gradually varying blade incidence angle offsets. An essential condition for such an approximation is the assumption that the dynamic pressure component induced by the blade motion is substantially smaller than the flow pattern changes caused by the variation of interblade channel geometry. The methodology is proposed for blade sections with prevailing two-dimensional flow. A dedicated test facility, called the blade flutter module (BFM), has been built and used for this purpose. The BFM is a linear cascade consisting of five transonic airfoils that can be operated either in a static or a dynamic regime. For the dynamic operation, any of the blades can be forcedly and independently oscillated at frequencies of up to 400 Hz with a maximum angular amplitude of 3 deg. The obtained results confirm that within the range of the test conditions, the proposed compounded quasi-dynamic approach exhibits similar characteristics to dynamically acquired unsteady blade pressures. This is true for a test range of a maximum inlet Mach number of 1.09, maximum blade oscillating frequency of 100 Hz, and measurement of unsteady pressure distribution on a blade suction surface. The corresponding blade chord based reduced frequency is 0.21.
