turbines, and propellers are low to moderate. Based on wing/blade chord, they are typically less than 2 10 and often are
4
only on the order of 10 to 10 . By comparison, civilian airplanes are characterized by Reynolds numbers ranging from a
6
few millions to 80 10 for the Boeing 747 at cruising velocity. Recent experimental investigations of low Reynolds
number aerodynamics [4, 3, 9] reveal several features of such ows that complicate their prediction compared with high
Reynolds number ows. Low Reynolds number flows are often dominated by the effects of flow separation. Separation
greatly influences lift and drag, and thus the flight stability of UAV's as well as the efficiency of wind turbines. It also
causes unsteadiness in turbine flows which is a determining factor in high cycle fatigue (HCF) of turbomachinery
components.
The physical origin of flow separation is qualitatively well understood: the attached laminar boundary layer developing on
a wing is subjected to an adverse pressure gradient due to the airfoils curvature which causes it to separate. Immediately
behind the separation point there is an eectively stagnant ow region, the so-called dead air region, followed by a
reverse ow vortex. The interface between the separated ow moving away from the wing and the recirculating ow in the
vicinity of the wing results in a shear layer with an inectional mean velocity prole. This shear layer experiences
Kelvin-Helmholtz instabilities that develop into turbulence after generating the rst characteristic spanwise vortices.
Further downstream, the separated turbulent ow reattaches and gradually evolves into the classical turbulent boundary
layer. The above picture emerges from numerous experimental investigations, e.g. [4, 3, 9], as well as from direct
numerical simulations (DNS) results [7, 12, 8, 1, 5, 6].
experimental agreement was high numerical resolution (472x72x64 mesh points) and a high order numerical method,
requirements dicult to satisfy in simulations of practical ows often performed with low order nite dierence or nite
volume methods (e.g. commercial codes). Similarly, Eisenbach and Friedrich [2] performed LES of ow separation on an
5
airfoil at high angle of attack at Re = 10 using cartesian grids. This case also required very high resolutions between 50
and 100 million mesh points. Therefore, the question remains: can LES produce suciently accurate results for laminar
separation bubble ow with drastically reduced resolution, say around 1% of DNS resolution, commonly achievable for
fully turbulent ows?
NACA-0012 airfoil at Rec =5 10 at 5 degrees of incidence with resolution reduced drastically from that used in DNS.
Ideally, a curvilinear coordinates code would be used for this problem (if available). Another Ph.D. student, Giacomo
Castiglioni, has started simulations on this problem using the immersed boundary (IB) method code INCA developed in
Adams group at Munich. While we have made progress using INCA there are fundamental issues with the accuracy of the
IB method at the rigid boundaries which impacts the separation predictions. A curvilinear-coordinates-based simulation
would be useful in disentangling the numerical issues of the IB method from the modeling issues of LES. We could also
consider using Stanford IB codes if issues of numerical accuracy at the rigid boundaries have been addressed and are not
considered serious. Basically any validated numerical code would be of interest if the numerical errors are expected to be
much less than the eects of a SGS model.
References
[1] M. Alam and N.D. Sandham. Direct numerical simulation of short laminar separation bubbles with turbulent
reattachment. J. Fluid Mech., 410:128, 2000.
[2] S. Eisenbach and R. Friedrich. Large-eddy simulation of ow separation on an airfoil at a high angle of attack and re
5
[14] X. Wu and P. Moin. Transitional and turbulent boundary layer with heat transfer. Phys. Fluids, 22:085105, 2010.
[15] Z. Yang and P.R. Voke. Large-eddy simulation of boundary-layer separation and transition at a change of surface
curvature. J. Fluid Mech., 439:305333, 2001.