Hadrien Mamelle, Gabriel Broux, Nicolas Forestier, Zdenek Johan, Eric Garrigues
DOI Number: N/A
Conference number: IFASD-2019-148
Innovative aircraft configurations comprising intersecting surfaces like U-tails exhibit aerodynamic phenomena which are computationally challenging to predict in view of flutter behavior assessment. From the subsonic domain, where aerodynamic interactions occur between the tail surfaces, to the transonic domain, where shock waves and separated flow at the tails intersection increase significantly the flow field complexity, it is necessary to validate not only aerodynamic models but also the aeroelastic coupling strategy involved in flutter computations. To this aim, a comprehensive experimental database has been acquired for several U-tail configurations in subsonic and transonic domains on a heavily instrumented wind tunnel flutter mock-up. This article presents aerodynamic and flutter correlations between experimental and numerical results which are computed with an in-house linearized Navier-Stokes solver with linearized turbulence. The overall strategy set up to obtain these correlations is discussed in detail to highlight the rigorousness required at every stage of the process to be able to compare properly normalized experimental data with adequate numerical data. Starting from accurate steady pressure field predictions in subsonic and transonic domains, unsteady pressure sensitivities induced by a pitching motion are analyzed and point out that linearized Navier-Stokes results are in very good agreement with the measured data, including in the corner area at the tails intersection. A tuned FEM modal basis representative of the mock-up dynamic behavior is used to compute a linearized pressure database which is used to perform frequency-domain flutter computations. The computed flutter diagrams and the predicted flutter pressure variation with respect to the Mach number are in excellent agreement with the measured ones. Thanks to the availability of flutter onsets measurements, flutter mode shape and pressure have been extracted and offer an uncommon opportunity to perform correlations at the flutter point. The predicted bending/torsion flutter mechanism is identical to the measured one and good flutter mode pressure correlations are obtained, providing an additional way to validate aerodynamic predictions. The flutter mode pressure is further analyzed by separating the contributions from the real and imaginary parts of the flutter mode shape to get a deeper understanding of the measured data. Altogether, the results presented in this article constitute a milestone toward numerical tools validation and demonstrate the relevance of the computational strategy adopted at Dassault Aviation in performing accurate aerodynamic and intersecting surfaces configurations. flutter predictions for intersecting surfaces configurations.