Theoretical Predictions of F-16 Fighter Limit Cycle Oscillations for Flight Flutter Testing

Abstract

A computational investigation of the flutter onset and limit cycle oscillation behavior of various F-16 fighter weapons and stores configurations is presented. A nonlinear harmonic balance compressible Reynolds-averaged Navier–Stokescomputational fluiddynamic flowsolverisusedtomodeltheunsteadyaerodynamicsoftheF-16wing. Slender body/wing theory is used as an approximate method for accounting for the unsteady aerodynamic effects of wing-tip launchers and missiles. Details of the computational model are presented along with an examination of the sensitivity of computed aeroelastic behavior to characteristics and parameters of the structural and fluid dynamic model. Comparisons with flight-test data are also shown. I. Introduction T HE SEEK EAGLE Office at Eglin Air Force Base performs an essential task in clearing new aircraft/stores configurations through flight tests for safe and effective operation. Many of these flighttestsarefortheF-16aircraftwhichcontinuestobeaworkhorse for the U.S. Air Forcewith continually new stores (missiles, bombs, and fuel tanks) being considered for aircraft operations. Similar aeroelastic flight tests are expected for future fighter aircraft as they go into service in the coming years. The number of needed flight tests is projected to be well beyond the financial and staff resources available. Hence there is a pressing need to identify the most critical aircraft/store configurations for the limited flight-test resources available and also insofar as possibly reduce the number of flight tests needed. Virtual flight testing may be the answer. Using new improved computational capability that provides much more rapid solutions, computational simulation can help identify the most critical aircraft/ store configuration and also hasthe potential of reducingthe number ofneeded flighttestsifconfidencecanbeestablishedinthecapability of simulations to correlate with flight-test data. A new methodology has been developed to produce these computer simulations based upon the notion that because the response is periodic in time, the solution need only be obtained over a single period of oscillation in time. By avoiding the traditional time marching solution which computes the long transient before a steady-state periodic oscillation is reached, computational times are reduced by a factor of 10–100. This enables a sufficiently rapid solutiontomakesuchsimulationsapracticalrealityforthe flight-test engineer and support team. Future developments of this methodology hold the promise of further substantial reductions in computational cost and are being vigorously pursued. Also further refinements in the physical fidelity of the simulation models are being considered.