. NASA Technical Memorandum 100048 Application of Empirical and Linear Methods to VSTOL Powered-Lift Aerodynamics Richard Margason and Richard Kuhn February 1988 National Aeronautics and Space Administration ~~ NASA Technical Memorandum 100048 Application of Empirical and Linear Methods to VSTOL Powered-Lift Aerodynamics Richard Margason, Ames Research Center, Moffett Field, California Richard Kuhn, V/STOL Consultant, Valencia, California February 1988 National Aeronautics and Space Administration Ames Research Center Moffett Field, California 94035 APPLICATION OF EMPIRICAL AND LINEAR METHODS TO VSTOL POWERED-LIFT AERODYNAMICS Richard Margason and Richard Kuhn Ames Research Center ABSTRACT borne flight. Proper design of VSTOL aircraft requires an accounting of these complex phenom- ena. Unfortunately, the flows involved are not This paper critically reviews available amenable to purely theoretical predictions at prediction methods and provides an assessment of this time. Improved prediction methods are their strengths and weaknesses. The methods are needed to guide experimental programs and to applied to selected problems which represent the reduce the testing requirements in the future, major aero/propulsion interactions for short as well as to provide an understanding of the takeoff and vertical landing (STOVL) aircraft of flow physics. One motivation of this paper is current interest. The first two problems deal to identify areas where modern computational with aerodynamic performance effects during fluid dynamics (CFD) methods can provide hover: a) out-of-ground effect, and improvements to the current computational capa- b) in-ground effect. The first problem can be bilities. The ultimate goal would be to develop evaluated for some multijet cases; however, the the ability to completely analyze the aerody- second problem is very difficult to evaluate for namics throughout the entire flight envelope of multijets. The ground-environment effects due an arbitrary VSTOL aircraft. to wall jets and fountain flows directly affect Traditionally, prediction techniques appli- hover performance. In a related problem c) hot- cable to VSTOL aircraft (1-8)* consist of gas ingestion affects the engine operation. potential-flow methods with empirically derived Both of these problems as well as jet noise corrections and/or adjustments to account for affect the ability of people to work near the viscous effects which significantly affect aircraft and the ability of the aircraft to aerodynamidpropulsion interactions. Some of operate near the ground. Additional problems these methods are simple to use but provide are d) the power-augmented lift due to jet-flap results with only limited applicability. Other effects (both in- and out-of-ground effects) and procedures build on potential-flow-panel e) the direct jet-lift effects during short methods. The input data tend to be complex and takeoff and landing (STOL) operations. The difficult to prepare, and often the results have final problem is f) the aerodynamic/propulsion limited usefulness. interactions in transition between hover and In recent years there has been rapid pro- wing-borne flight. Areas where modern computa- gress in the development of CFD. This progress tional fluid dynamics (CFD) methods can provide is due in part to improvements in digital com- improvements to the current computational cap- puters (9) which provide rapid computation abilities are identified. speeds (up to hundreds of millions of floating- point operations per second), and larger memo- FOR OVER THIRTY YEARS, vertical or short takeoff ries (up to hundreds of millions of words). and landing (VSTOL) aircraft have been developed This growth was clearly described in Jameson's primarily from experimental investigations survey (10) of the evolution of computational because there are many unique aerodynamic/ propulsion-induced effects which occur with these vehicles. These problems are especially *Numbers in parentheses designate significant at low speeds from hover to wing- references at end of paper. 1 methods for conventional aerodynamics. Numeri- VSTOL. One of the first stage I11 analyses was cal approximation methods were examined for a by Baker (12). This calculation of the flow hierarchy of models in ascending order of com- field caused by a jet in a crossflow used a plexity, ranging from the linearized potential- finite-element, pressure-interaction, parabolic flow equations to the Reynolds-averaged Navier- approximation to the Reynolds-averaged full Stokes equations. In a study (11) of the influ- Navier-Stokes equations. The other papers ence of CFD on experimental aerospace facili- (13-16) in the present session will present ties, four main stages (I-IV) of approximation current examples of several Stage II and Stage to the full Navier-Stokes equations were identi- 111 approximations applied to either STOL- or fied. For completeness in relation to short STOVL-aircraft-related flow fields. takeoff and vertical landing (STOVL) aero/ This paper critically reviews available propulsion interactions, 'the present authors stage 0 and stage I prediction methods, and have added a Stage 0 to represent the empiri- attempts to provide an assessment of their cally based methods, and Stage Ib to represent strengths and weaknesses. Existing methods will linearized inviscid methods which include an be discussed based on six selected problems approximate model for jets: which represent the major aero/propulsion inter- actions for STOVL aircraft of current inter- Stage Approximation est. This paper presents examples using avail- able methods. The first two problems deal with 0 Empirical methods aerodynamic performance effects during hover: a) out-of-ground effect and b) in-ground I Linearized inviscid effect. Both problems are reasonably well la Linearized inviscid plus boundary understood for the case of a single jet. The layer (BL) first problem can be evaluated for some multijet Ib Linearized inviscid plus modeled jets cases; however, the second problem is very difficult to evaluate for multijets. The next I1 Nonlinear inviscid three problems are related to operations near IIa Nonlinear inviscid plus interacting BL the ground: c) the ground-environment effects due to wall jets and due to surface breakdown, I11 Reynolds-averaged Navier-Stokes d) hot-gas ingestion effects on the engine operation, and e) short takeoff and landing IV Full Navier-Stokes (large-eddy (STOL) ground effects. These problems affect simulations with small-scale both the ability of people to work near the turbulence modeling) aircraft and the ability of the aircraft to operate near the ground. The final problem Stage 0 has been the mainstay method for extra- f) is the aerodynamic/propulsion interactions in polating experimental data beyond the original transition between hover and wing-borne flight. wind tunnel test conditions up to the present time for VSTOL aero/propulsion interactions. NOMENCLATURE Stage I is currently used for engineering anayl- sis of conventional airplanes. The Navy V/STOL A jet area Aerodynamics, Stability and Control Manual (4,5) BL butt line utilizes Stages 0, I, and Ib methods, and repre- CD drag coefficient sents the current state of the art in V/STOL CL lift coefficient computational methods. For conventional air- CM pitching-moment coefficient planes without propulsion-induced effects, CT thrust coefficient limited to moderate use is also being made of CU jet momentum coefficient stage I1 solutions; Stage I11 is expected to be cL section lift coefficient widely used in the next 10 years. The principal C pressure coefficient I items pacing introduction of Stage I11 into the P C' section jet momentum coefficient aerodynamic design process are: a) development DM plate or jet diameter of improved turbulence models; b) geometry grid d jet diameter generation difficulties and limitations; EVD elementary vortex distribution c) development of more efficient and reliable FS fuselage station numerical algorithms; and d) development of more H,h height powerful scientific computers. upflou jet thickness At this time some initial efforts have been turbulent intensity made to use stage I1 and I11 approximations for 2 up from ground plane to undersurface Kh factor for effect of forward speed on UP suckdown V ground vortex L lift W wake e stagnation line X x direction LID lift improvement device m free stream M Mach number NPR nozzle pressure ratio Superscript P,P pressure I 9 dynamic pressure section R radius, radial distance *n Reynolds number MULTI-JET OUT-OF-GROUND EFFECT HOVER S wing area INTERACTIONS S' distance between jet impingement points on a ground plane STOL short takeoff and landing During hover, there is a base loss due to STOVL short takeoff and vertical landing interactions between the lifting jets and the T thrust, temperature lower surface of the aircraft which results in a V velocity - distribution of induced suction pressures which /q/qj loss. 'e 'e effective velocity ratio, produce a lift Additional performance VSTOL vertical or short takeoff and landing losses include inlet-flow distortion, hot-gas WL water line ingestion, hot-day conditions, control bleed, x,y,z Cartesian coordinates internal nozzle flow, thrust vectoring, and a angle-of-attack static ground effect. There are many items 6 sideslip angle related to the details of the aircraft design AL aero/propulsion lift-interference which determine the magnitude of the losses. increment Even though the sum of these losses may be only AM aero/propulsion lift-interference