Maj. Michael Iovnovich
DOI Number: N/A
Conference number: IFASD-2019-045
Flutter instability prediction of fighter aircrafts capable of carrying an extensive external stores inventory typically requires analytic assessment of thousands of configurations throughout the operational flight envelope. To enable realistic solution times, simplified unsteady aerodynamics models are commonly used, in which underwing stores aerodynamics is neglected. In the current study, the effects of underwing stores aerodynamics on flutter prediction capabilities are assessed using a heavyweight store configuration of the F-16 aircraft with multiple underwing stores and in correlation with flight test data. This examination is conducted at the high transonic flow regime using two unsteady aerodynamics solvers, namely, the linear, panel-based ZAERO solver and the nonlinear, Euler, ZEUS solver. This investigation showed that underwing stores modeling may in-fact considerably affect the predicted flutter onset speeds. Neglecting stores aerodynamics yielded both less realistic and nonconservative predictions by both ZAERO and ZEUS models. The comparison between the ZAERO and ZEUS models solutions revealed significant differences in steady and unsteady surface pressure distributions due to nonlinear flow phenomena. These differences also manifested in significant distinctions between the two models flutter predictions. To enable full aerodynamic modeling in industrial flutter survey applications, a superposition modeling approach is formulated and successfully validated using the studied test case. According to this method, introduction of a new store into the aircraft weapons inventory will only require a single aerodynamic model solution, then aerodynamic models for configurations including any permutation of the new store with other pre-existing stores in the database may be directly assembled with no additional computations required. This technique is found effective for superposition between linearized unsteady aerodynamics stores effects, even if the base flow of these effects is nonlinear. However, once interference between stores becomes dominant and nonlinear, this modeling approach is no longer valid.