Brian E. Wake

Rotorcraft retreating blade stall control

Flow control to avoid or delay rotorcraft retreating blade stall can be an enabling technology for future high performance rotorcraft. Aerodynamic experiments and computations have indicated that appropriate unsteady excitation can delay boundary layer separation and stall on airfoils. Work is in progress to determine the control requirements for helicopter rotor blades at full scale Mach numbers, Reynolds numbers, and with unsteady pitching motions. Compact, powerful, and efficient flow actuation and control systems will be needed. Three actuation concepts were favorably evaluated during initial studies: electromechanical directed synthetic jets (DSJ), periodic flow modulation, and plasma actuation. Electromechanical DSJ and plasma actuators are being developed further and will be evaluated in full scale pitching blade section experiments. These experiments will determine the required control authority, validate the actuator concepts, and study open and closed loop control approaches. Computational studies are being performed of the combined external and actuator flow fields to determine preferred actuation geometries and operating points. System analyses are being used to quantify the benefits for representative aircraft configurations and missions. Copyright ÿ2000 by United Technologies Corporation. Published by the American Institute of Aeronautics and Astronautics, Inc., with Permission. Nomenclature A airfoil pitch rate, U c 2 / α&

Investigation of high-order upwinded differencing for vortex convection

Two model problems were formulated to investigate the benefits of using higher order upwinded differencing schemes to convect a vortex. In these model problems, one or two ideal vortices were specified at the upstream boundary and convected through a simple fluid domain in the direction of the vorticity vector. Numerical tests were performed using an implicit upwinded finite volume method with orders of accuracy varying from first to fifth order. The effect of grid refinement was examined to determine the relative advantage to higher order accuracy vs additional grid points. The calculations demonstrated that the fifth-order scheme with 14 points across the vortex diameter convected a well-aligned vortex almost perfectly. The fifth-order scheme provided a factor of 10 reduction in the number of points over the third-order scheme in each crossflow plane. The corotating double-vortex model problem was used to study the effect of vortex skewness relative to the grid. A vortex-convection skewness parameter was defined that indicates the number of grid points required in the axial direction for adequate preservation of a vortex skewed relative to the grid.

Investigation of Instabilities in a Lean, Premixed Step Combustor

A combined experimental and computational characterization of combustion instabilities in a lean, premixed backward-facing step combustor was performed. Specifically, the instabilities of interest were those encountered as the equivalence ratio was reduced to levels approaching the combustors lean extinction limit. A quasi-one-dimensional unsteady analysis with loss and heat-addition models was developed to simulate combustion instabilities for a premixed step combustor. Experimental results indicated that the magnitude of a longitudinal acoustic disturbance grew with decreasing equivalence ratio, until it eventually triggered a lower-frequency, high-amplitude instability. Numerical results compared with experimental data demonstrated that the analysis can capture the critical frequencies observed in the combustor model. Coupling the heat release with the step flow velocities in the analysis produced an instability at the dominant resonant frequency of the combustor. Experimentally, low-frequency instabilities were visible as a flapping of the flame and increased in severity with decreasing equivalence ratio until they caused combustor blowout.

Implementation of a rotary-wing Navier-Stokes solver on a massively parallel computer

An unsteady, compressible, three-dimensional, implicit Navier-Stokes solver (NSR3D) for helicopter and propeller applications has implemented using FORTRAN with 8X array extensions on the massively parallel connection machine (CM-2). The modifications to the original algorithm necessary to overcome communication bottlenecks and achieve reasonable computational efficiency on the CM-2 are described.

Combustion-Powered Actuation for Dynamic-Stall Suppression: High-Mach Simulations and Low-Mach Experiments

An investigation on dynamic-stall suppression capabilities of combustion-powered actuation (COMPACT) applied to a tabbed VR-12 airfoil is presented. In the first section, results from computational fluid dynamics (CFD) simulations carried out at Mach numbers from 0.3 to 0.5 are presented. Several geometric parameters are varied including the slot chordwise location and angle. Actuation pulse amplitude, frequency, and timing are also varied. The simulations suggest that cycle-averaged lift increases of approximately 4% and 8% with respect to the baseline airfoil are possible at Mach numbers of 0.4 and 0.3 for deep and near-deep dynamic-stall conditions. In the second section, static-stall results from low-speed wind-tunnel experiments are presented. Low-speed experiments and high-speed CFD suggest that slots oriented tangential to the airfoil surface produce stronger benefits than slots oriented normal to the chordline. Low-speed experiments confirm that chordwise slot locations suitable for Mach 0.3-0.4 stall suppression (based on CFD) will also be effective at lower Mach numbers.

Numerical investigation of pre-mixed step-combustor instabilities

A quasi-one-dimensional unsteady analysis with loss and heat addition models was developed to simulate combustion instabilities for a premixed step combustor. This analysis utilizes a third-order upwindbiased finite-volume scheme to solve the Euler equations. Time accuracy was achieved by using Newton sub-iterations within each time step. The ability of the solver to capture the resonant acoustic frequencies was examined by modeling a speaker radiating broad-band noise in a simple duct. The additional complexities of the experimental combustor models, such as skin-friction losses, losses due to sudden area change, and the addition of fuel, air, water and heat were all modeled and included in the governing equations as source terms. Preliminary results compared with experimental data for several combustion rigs indicate that the analysis can capture the critical frequencies of the combustor model. Coupling the heat release with the step flow velocities, produced an instability at the dominant resonant frequency of the combustor.

AN EFFICIENT PROCEDURE FOR THE NUMERICAL SOLUTION OF THREE-DIMENSIONAL VISCOUS FLOWS

A solution procedure is described for the numerical solution of steady and unsteady compressible v iscous flow past wing-a1 one, and rotor configurations. T his procedure solves the three-dimensiona l unsteady Navier-Stokes equations in a body-fitted coordinate system. A finite difference procedure of second order spatial accuracy and first order temporal accuracy is used to discretize the governing e quations, and a hybrid time marching scheme is used to advance the solution from one time level to the next. This orocedure lends itself to efficient solution on In this work, a hybrid time-marching tech- nique for the numerical solution of the three- dimensional unsteady Navi er-Stokes e quations is proposed, as an a1 ternative to existing explicit (MacCormack or Runge-Kutta) and implicit (ADI, LU) time marching techniques. This technique has in the past been applied to the numerical solution of the unsteady ree-dimensional Euler equations with success. 3y Although in this work it has been used with standard spatial d iscretization techniques, it may be used with improved spatial discretization schemes such as high order upwind schemes. the c urrent generation vector machines. In un- steady applications involving oscillating wing In the n ext several sections the solution surfaces or rotating rotor blades, the surface algorithm is described in detail. The algorithm is motion is treated exactly, by allowing the body- then tested f or solution accuracy and stability by fitted grid to rotate or deform. Sample steady and considering a number of cases, such as flow over a helicopter rotor blade in hover, and transonic unsteady calculations are presented for fixed and transport wing appl cati ens. Detailed CPU time rotary wing configurations. Detailed surface pressure and integrated load comparisons with estimates and memory requirements are given to experiments are given. illustrate the suitability of these t echniques for engineering use.

High-Speed Experiments on Combustion-Powered Actuation for Dynamic Stall Suppression

This work documents high-speed wind-tunnel experiments conducted on a pitching airfoil equipped with an array of combustion-powered actuators. The main objective of these experiments was to demonstrate the stall-suppression capability of combustion-powered actuators on a high-lift rotorcraft airfoil (the VR-12) at relevant Mach numbers. Through unsteady pressure measurements at the airfoil surface, it was shown that combustion-powered actuators could positively affect the stall behavior of the VR-12 at Mach numbers up to 0.4. Static airfoil results demonstrated 25 and 50% increases in poststall lift at Mach numbers of 0.4 and 0.3, respectively. Deep dynamic stall results showed cycle-averaged lift coefficient increases up to 11% at Mach 0.4. Furthermore, it was shown that these benefits could be achieved with relatively few pulses during the downstroke and with no need to preanticipate the stall event. The flow mechanisms responsible for stall suppression were investigated using particle image velocimetry...

Modified Hybrid Navier–Stokes/Free-Wake Method for Hovering Rotor Analysis

The analysis of rotors in hover has always been a significant challenge. The presence of strong vortices underneath the rotor requires a large amount of grid points to properly capture the vorticit...

Numerical and Experimental Study of Centrifugally Driven Flow Inside a Rotating Duct

Rotor blades experience very high centrifugal forces that can be used to pump air to the outboard region of the blade through an internal duct, which can be used for flow control. Analysis or design of such systems requires accurate prediction capability. To validate current Reynolds-averaged Navier–Stokes simulation methodologies, an experiment was performed using a rotating pipe, and simulation results were compared to the measured data. A quasi-one-dimensional code was also compared to experiment as a lower-order simulation tool for faster solutions. The test and simulations include several combinations of steady inlet and exit conditions as well as an unsteady inlet valve operation at several rotational speeds. The quasi-one-dimensional code showed good correlation for steady inlet and exit conditions with boundary conditions obtained from experiment. Navier–Stokes methods also showed good agreement with measured data for pressure and mass flow rate at most conditions, while properly capturing complex...