Numerical prediction and characterization of shock-buffet in transport aircraft

Computational Fluid Dynamics (CFD) simulations are frequently used in the aerospace
industry to help reduce development times by cutting down on the need of extensive windtunnel campaigns. However, although design-point aerodynamics are well predicted, edge of
the envelope scenarios dominated by non-linear fluid phenomena can lead to uncertainties in
the accuracy of the results produced.
This work addresses the use of Reynolds-averaged Navier-Stokes (RANS) based simulations
in the prediction of unsteady shock-buffet phenomenon. Three studies are conducted: a 2D
validation study, a 3D validation study, and the pinnacle of this work which presents a novel
shock-buffet prediction on an industrially-relevant aircraft confguration.
Two dimensional shock-buffet predictions are presented as a confrmation of previous available knowledge. The dependency on turbulence modelling approaches is evident, with new
results showing that the full Reynolds Stress Model (RSM) is a more appropriate closure to
the RANS equations than other typically used eddy-viscosity-based models. However, this
implies additional computational costs (due to increased number of equations solved), and
inherited challenges associated with solver stability.
RANS-based simulations are then applied to a 3D confguration: the NASA Common
Research Model (CRM) wing-body test case. Complementary results to the AIAA CFD Drag
Prediction Workshop are produced. Novel results, outside the Drag Prediction Workshop
envelope, investigate the development and expansion of the shock-induced boundary layer
separation on the NASA CRM wing, however the steady RANS approach fails to accurately
predict this due to unsteady effects which are not accounted for.
Unsteady simulations in the shock-buffet regime of the wing-body NASA CRM are then
presented as the main novel contribution of this work. The complexity of the phenomenon is
revealed by unsteady shock oscillations coupled with shock-induced separation and vortex
shedding. The presence of shock-buffet cells is detected and helps understand shock dynamics.
A frequency analysis reveals the presence of multiple peak frequencies. A qualitative
comparison with experimental observation show similarity in the physics produced. Finally,
to further investigate the shock-buffet phenomenon, the effects of changing the Reynolds
number are presented.
Through industrial relevance, the current work can lead to decision making in the development
of the future generation of aircraft.