Numerical investigation of Type II non-Newtonian de/anti-icing fluid effects on take-off performance for general aviation aircraft

Abstract

<p>Ground icing, while preventable with glycol based freezing point depressant fluids, accounts for nearly 40-50 civil accidents a year according to one account. More viscous than Type I fluids, Type II fluids are inherently non-linear in their shear stress-rate of strain relationship having smaller relative viscosity at higher shear rates. The non-linearity makes properly scaled wind-tunnel testing difficult and computational methods are employed in this study to look at the aerodynamic effects of the deicing fluid on global performance during typical take-off maneuvers for general aviation. The method is tested on a two-dimensional NACA 0012 airfoil under typical take-off simulation parameters;A modified PANEL method arrives at potential flow solutions which account for the accelerating freestream, rotation maneuver, and shed vorticity while time-dependent Boundary Layer equations are solved using an implicit finite difference scheme. Viscid-inviscid interaction is accomplished in an inverse method through the specification of normal velocities induced on each panel during potential flow calculations to account for displacement thickness effects. Deicing fluid motion is driven by shear stresses at the interface of the fluid and gas-dynamic boundary layer and pressure gradients based on the outer flow solution. Slip velocities and shear stresses are then matched at the interface to insure kinematic and dynamic continuity. The displacement thickness effect of the deicing fluid is accounted for in the viscid-inviscid interaction;The deicing fluid is assumed Newtonian in this study and exhibits a fluid bucking effect which may point to reasons for reported losses in lift. The large shear stresses toward the leading edge drag the fluid to the center of the airfoil while large pressure gradients in the back push the fluid to the center. The buckling phenomena is shown to be brought on by (1) increased fluid viscosity, (2) deeper initial depths of deicing fluid and (3) higher rotation speeds where shear stresses and pressure gradients are larger. In simulations which did not exhibit fluid buckling, the effect on maximum lift coefficient was found to be minimal. The current programming is not equipped to handle this aspect of fluid stability and remains an issue for further investigation.</p>