CFD INVESTIGATION OF UNSTEADY ROTORCRAFT AIRFOIL AERODYNAMICS
Open Access
- Author:
- Coder, James George
- Graduate Program:
- Aerospace Engineering
- Degree:
- Master of Science
- Document Type:
- Master Thesis
- Date of Defense:
- None
- Committee Members:
- Mark David Maughmer, Thesis Advisor/Co-Advisor
Mark David Maughmer, Thesis Advisor/Co-Advisor - Keywords:
- gurney flap
mite
airfoil
unsteady
aerodynamics
blade
rotor
helicopter
rotorcraft
CFD
aerospace engineering
grid
overset
RANS
DES - Abstract:
- As part of an effort to improve rotorcraft performance, the ability of the OVERFLOW computational fluid dynamics solver to predict the onset of dynamic stall and the influence of upstream miniature trailing-edge effectors (MiTEs) has been investigated. Static CFD validation cases and grid resolution studies were first performed for airfoils both clean and with Gurney flaps. It was determined that CFD generally performs well in predicting the zero-lift angle of attack and zero-lift pitching-moment coefficient for an airfoil, but the fully turbulent assumption causes the maximum lift coefficient to be generally overpredicted. Several trailing-edge stall suppression and double stall cases from the U.S. Army Dynamic Stall Data Package were used to provide experimental validation of the computational methods used in OVERFLOW for dynamic stall. The overprediction of the maximum lift coefficient resulted in the introduction of a quasi-steady scaling factor for these cases; the use of this scaling factor is shown to greatly improve the quality of the dynamic stall predictions. The aerodynamic response of a deploying and retracting upstream MiTE, which is a deployable Gurney flap, has also been investigated and validated for the purpose of providing “virtual wind tunnel” data for the development of an unsteady aerodynamic model. Moving the MiTE upstream of the trailing edge creates a lower-surface vortex which causes significant deviations from trailing-edge MiTE behavior that had previously been observed. This vortex was observed in both 2-D unsteady RANS and 3-D DES simulations, with the transient responses being qualitatively similar for both methods.