Search Results

The architecture and training of artificial neural networks are briefly described. Five applications of these networks to design and analysis problems are presented; three in aerodynamics and two in flight dynamics. The aerodynamics cases are those of a harmonically oscillating airfoil, a pitching delta wing, and airfoil design. The flight dynamic examples involve control of a super maneuver and a decoupled control case. It is demonstrated that highly nonlinear aerodynamic cases can be generalized with sufficient accuracy for design purposes. It is shown that although neural networks generalize well on the aerodynamic problems, they appear lacking comparable robustness in modeling dynamic systems. It is also shown that generalization appears to become weak outside of the training domain.

Flow field computations and, in particular, that of pressure, skin friction, and heat transfer (for high speed flights) are the primary parameters in the design of aerospace vehicles. Most computational schemes based on either the inviscid Euler equations or various forms of the Navier-Stokes equations are remarkably accurate in the predictions of pressure distributions. However, computations of skin friction and heat transfer particularly at high speeds have been a source of considerable difficulty. Problems arise not only due to the grid resolution but also due to the particular numerical scheme employed. To address the difficulty associated with accurate computations of the velocity and temperature gradients, a comparative investigation of several Navier- Stokes codes is undertaken. Previous studies with regard to the effect of grid resolution are incorporated into the current investigation.

Two-dimensional and three-dimensional leading edge ice shapes for a finite wing model computed with the NASA Glenn LEWICE 2.0 and LEWICE3D Version 2 ice accretion codes are compared with experimental data from icing tunnel tests. The wing model had 28° leading edge sweep angle, 1.52-m (60-in) semispan and an airfoil section representative of business jet wings. Experimental wing leading edge ice shapes were obtained at the NASA Glenn Icing Research Tunnel (IRT) for six icing conditions. Tests conditions included angles of attack of 4° and 6°, airspeeds ranging from 67.06 m/s (150 mph) to 111.76 m/s (250 mph), static air temperatures in the range of -11.28°C (11.7°F) to -2.78°C (27°F), liquid water contents of 0.46 g/m₃, 0.51 g/m₃, and 0.68 g/m₃, and median volumetric diameters of 14.5 μm and 20 μm.

This paper compares the environmental durability of currently used as well as some of the potential aircraft aluminum alloy skin materials. A simple test was developed to evaluate the environmental durability by simultaneous application of fatigue loads and aerated salt water attack. Alclad alloys showed excellent resistance to corrosion fatigue. On the other hand, 7475 alloy and new potential material 6013 alloy experienced inter-granular corrosion at the fatigue crack origin area. The test results also indicated the significance of corrosion preventive coatings to increase the age of aircraft.

A basic one-component Doppler Global Velocimetry (DGV) system has been developed at Wichita State University. This system was evaluated on a round axisymmetric jet. The results are compared with measurements made using traditional Constant Temperature Anemometry (CTA) and Preston tube measurements at 3.2, 9.6 and 16.0 jet exit diameters downstream and along the jet centerline. The DGV results show similar trends to these measurements. Software corrections for camera misalignment, optical distortions, and laser frequency variations were necessary to assure data quality. Results indicate good agreement between the DGV and CTA measurements exist.

Results from a two-dimensional computer model developed at Wichita State University (WSU) for bleed air system analysis are compared with experimental data from icing tunnel tests performed with a wing model equipped with a hot air ice protection system. The computer model combines a commercial Navier-Stokes flow solver with a steady-state thermodynamic analysis model that applies internal flow heat transfer correlations to compute wing leading edge skin temperatures and the location and extent of the runback ice. The icing tunnel data used in the validation of the computer model were obtained at the NASA Icing Research Tunnel using representative in-flight icing conditions and a range of bleed air system mass flows and hot air temperatures. Correlation between experiment and analysis was good for most of the test cases used to assess the performance of the simulation model.

A new class of aerodynamic control devices have recently been designed specifically for wind turbine applications. These new controls were tested to evaluate their effectiveness in modulating wind turbine power output and for slowing or stopping a wind turbine in high wind or loss of generator situations. While these control devices were developed specifically for wind turbine applications, there exists the possibility that alternate aviation uses exist. In particular, these trailing-edge control devices were evaluated for reducing aircraft landing distances, generating rapid rates of descent, deep stall or spin recovery and for high angle of attack control.

The flow about multi-element airfoil configurations is investigated using the unsteady Reynolds averaged Navier-Stokes equations. An explicit scheme is used to advance the solution in time while a finite difference scheme is applied to discretize the flux terms. An algebraic and two one-equation turbulence models are used to model turbulence. The domain about each multi-element airfoil is discretized with structured Chimera grids. The multi-element configurations presented in this paper include two airfoils with slotted flaps and an airfoil with a 50% chord vented aileron deflected at 90 degrees. Subsonic flow computations are performed for attached and separated flow conditions. The computational results obtained with the CRTVD code developed at Wichita State University are in good correlation with wind tunnel data and with computational results obtained with the INS2D computer code developed at NASA Ames research center.

Two new concept of flexible fixture subsystem (FFS) for aircraft wing spar assembly are introduced in this paper. The advantages and characteristics of FFS are discussed and compared with the current assembly method and fixtures. The objective of FFS is to replace the dedicated tooling and be able to quickly reconfigure itself for new types of spars. The fixture enables a family of spars to be mounted and assembled in the same tooling. Left- and right-hand side spars, varying lofts(spar cap angles), height, and depths are all accommodated on the same tool, within its envelop.

The effect of Gurney flaps on a NACA 0011 airfoil was investigated. Gurney flaps provide a substantial increase in lift while the penalty in drag is small. With the Gurney flap, the airfoil pressure distribution shows increased suction on the upper surface and higher pressure on the lower surface compared to the clean airfoil. This change in pressure is most profound on the lower surface just in front of the Gurney flap. Since separation occurs on the upper surface upon stall, this higher pressure condition on the lower surface continues into the post-stall regime. Thus, the NACA 0011 airfoil with Gurney flaps generates lift coefficients greater than one even under post-stall conditions.

An experimental methodology for investigating the effects of anti-icing fluids is presented in this paper. A wing model was designed, fabricated, and instrumented for testing anti-icing fluids in a wind tunnel facility. In addition, a video capturing method was developed and used to document fluid behavior during simulated takeoff tests. The experiments were performed at the Wichita State University 2.13-m by 3.05-m (7-ft by 10-ft) wind tunnel facility with two pseudoplastic fluids representative of Type IV anti-icing fluids. The experimental data obtained included fluid wave propagation speeds, chordwise fluid thickness distributions as a function of time, and boundary layer velocity profiles for the clean and fluid contaminated wing model at select chordwise stations. During simulated takeoffs with initial fluid depths of either 4 mm or 2 mm, the fluids were observed to thin in the forward (upstream) regions of the wing model and accumulate in the aft regions.