Abstract:
Flutter is an undesirable phenomenon which occurs in elastic structures due to interaction
of aerodynamic, inertial, and elastic forces on a body. Studies on such problems
are multi-disciplinary, requiring including physics of structures and physics of fluid flow
into account. This study focused on comparison of flutter boundary prediction using
a panel methods-based linear solver (NASTRAN) and RANS-based non-linear solver.
Although Nastran (SOL 145) can solve flutter problems with the help of linear aerodynamic
models (doublet lattice method based aerodynamics) and works well for simple
aerodynamic models in the subsonic regime, it fails in the transonic regime (due to nonlinearities
in flow). The Navier Stokes (RANS) based CFD model is used to capture
the transonic flow non-linearities. For this, separate solvers for structure and CFD are
used, and the two are connected with a third module to efficiently transfer the data
in real-time. A validation case of AGARD 445.6 wing successfully implemented with
the coupled solver. The AGARD 445.6 wing is a 45-degree sweptback wing with a thin
airfoil cross-section. The nonlinearities related to shock waves are minimum on such
profiles. The linear aeroelastic solver yielded closely followed results to the coupled
solver in predicting flutter boundary. Literature suggested that the angle of sweep is
a vital function of shock-wave strength. The low sweep model produces strong shock
waves, which are often termed a singularity in flow physics, thus significantly reducing
the accuracy of linear solvers. These observations attained our attention to quantify the
effect of angle of sweep on aeroelastic response estimation of the two solvers in transonic
and supersonic flow regimes. The results compiled showed that about 22% difference is
induced in flutter estimation of the two solvers for an angle of sweep-back of 30 degrees
compared to 4.6% with the angle of sweep-back of 45 degrees in transonic flow. While
for the supersonic flow regime, having a maximum of 8.3% for 45 and 11.9% with 30 degrees
models. These studies are further extended by performing sensitivity analysis for
iv
the Agard 445.6 flutter boundary concerning its material properties, semi-span length,
and angle of sweep-back. Also, a case of existing aircraft wings has been analyzed for
an aeroelastic response via the coupled framework.
