An extended version of the previously presented linearized one-dimensional turbomachinery performance model is described. The current version of the model is capable of performing forced response and flutter simulations on several stationary and/or rotating bladerows. The amplitude of the perturbation is assumed small thus impact of the perturbations of several sources may be superimposed. The distortion propagation analysis may be performed in the early stages of the design process, or whenever a quick solution is desirable, having only minimal information about the studied geometry.
The approach has a block structure, where each block represents a bladed passage or the empty space between. The blocks contain linearized gas relations that relate the gas state to known changes of enthalpy, entropy and momentum. The blade blocks are represented using the extended semi-actuator disk theory, where the flow inside the passage assumed to be one-dimensional [9]. The model considers frequency scattering for the rotating bladerows and is also using a complete package of linearized loss- and deviation correlations, providing more realistic results. The approach extends the previously presented methods by Amiet [2], [3] and Kaji&Okazaki [4], [5], being capable to handle harmonic distortions of various wavelength-to-chord ratios. Minimal assumptions are made about the studied geometry and nature of the gas, allowing to perform unsteady flow analysis not only on the idealized cases, but on more complex, realistic geometry.
Appropriate non-reflecting boundary conditions are applied at the boundaries of each block, using the hyperbolic characteristic theory, thus facilitating multi-blaredow domains setup and allowing running more complex cases, involving both forced response and flutter.
A number of idealized cases presented by Amiet and Kaji & Okazaki are reproduced to validate the model against the reference data, where a good comparison is achieved. The approach is also tested on the HP compressor of modern design for forced response simulations. Several multi-bladerow cases are also studied.
