Optimization of a Seamless Inlet Liner Using an Empirically Validated Prediction Method

An efficient multi-fidelity optimization methodology for turbofan intake acoustic liners is presented whereby the wall acoustic impedance is used as a design variable, in the frequency domain, to maximize forward fan noise suppression. Two methods are employed to predict the liner attenuation, a low-fidelity method based on solving the duct mode propagation for a treated circular duct with uniform mean flow using an eigenvalue code, and a high-fidelity method based on solving the duct mode propagation and radiation for a realistic inlet and flow using the finite element code ACTRAN. A validation of the prediction methodology is performed using a baseline seamless inlet liner, for which in situ impedance measurements, internal modal measurements, and far-field noise measurements are available. Both methods are then used to create response surfaces of the liner noise attenuation as a function of the liner impedance. The optimal wall impedance can then be extracted from these response surfaces. The speed of the low-fidelity method makes it possible to explore large sections of the design space at each of the target frequencies inexpensively, while allowing identification of wall impedance sub-domains where relative optima can be determined. The high fidelity method is employed on these sub-domains to provide more accurate and complete modeling. An empirically validated impedance model is then used to define a liner construction for a double-layer seamless liner that provides the best match to the optimum wall impedance.