Dynamic Stall

Fortunately, engineers and technologists do not wait until everything is completely understood before building and trying new devices. Even so, an improved understanding of fundamental unsteady fluid flow processes can serve to stimulate new innovations, as well as improvements in the performance, reliability, and costs of many existing machines. Therefore, research in unsteady fluid dynamics seems assured a lively future in modem industrial societies.

William J. McCroskey (1975)

9.1 Introduction

The phenomenon of dynamic stall has long been known to be a factor that limits helicopter performance. The problem of dynamic stall usually occurs on the rotor at high forward flight speeds or during maneuvers with high load factors and is accompanied by the onset of large torsional airloads and vibrations on the rotor blades – see, for example, Tarzanin (1972), McCroskey & Fisher (1972), McHugh (1978), Bousman (1998), and Isaacs & Harrison (1989). Whereas for a fixed-wing aircraft, stall occurs at low flight speeds, stall on a helicopter rotor will occur at relatively high airspeeds as the advancing and retreating blades begin to operate close to the limits where the flow can feasibly remain attached to the airfoil surfaces. As shown previously by Fig. 7.1, the advancing blade operates at low values of AoA but close to its shock induced flow separation boundary. The retreating blade operates at much lower Mach numbers but encounters very high values of AoA close to stall. Because of the time-varying blade element AoA resulting from blade flapping, cyclic pitch inputs, and wake inflow, the flow separation and stall ultimately occurs on a rotor in a very much more dynamic or time-dependent manner. This stall phenomenon is, therefore, is referred to as “dynamic stall.” Despite the fact that the static stall characteristics of airfoils have been discussed extensively in Chapter 7, the problem of flow separation and airfoil stall must now be reassessed from a nonsteady perspective.

Following the general definition given by McCroskey and colleagues (1976, 1982), dynamic stall will occur on any airfoil or other lifting surface when it is subjected to time – dependent pitching, plunging or vertical translation, or other type of nonsteady motion, that takes the effective AoA above its normal static stall angle. Under these circumstances, the physics of flow separation and the development of stall have been shown to be fundamentally different from the stall mechanism exhibited by the same airfoil under static (quasi-steady) conditions (i. e., where k = 0). Dynamic stall is, in part, distinguished by a delay in the onset of flow separation to a higher AoA than would occur statically. This initial delay in stall onset is obviously advantageous as far as the performance and operational flight envelope of a helicopter rotor is concerned. However, when dynamic flow separation does occur, it is found to be characterized by the shedding of a concentrated vortical disturbance from the leading edge region of the airfoil. As long as this vortex disturbance stays over the airfoil upper surface, it acts to enhance the lift being produced. Yet, the vortex flow pattern is not stable, and the vortex is quickly swept over the chord of the blade by the oncoming flow. This produces a rapid aft movement of the center of pressure, which results in large nose-down pitching moments on the blade section and an increase in torsional loads on the blades. This is the main adverse characteristic of dynamic stall that concerns the rotor analyst, for which the effects have proved difficult to predict.

As discussed by Beddoes (1979, 1983) and Wilby (1984, 1996, 1998), the consideration of dynamic stall in the rotor design process will more accurately define the operational and overall performance boundaries of the helicopter. Generally the rotor will be first designed so that the onset of high hlade loads, aeroelastie problems or limits in fUrrk+ performance are not limiting factors on the basis of linear and nonlinear quasi-steady aero­dynamic assumptions (i. e., using the blade element representations described in Chapter 7). Nonlinearities in the airloads associated with dynamic stall can introduce further effects that give rise to dangerously high blade stresses, vibrations, and control loads. One such nonlinear phenomena is called stall flutter.’ Because of the significant hysteresis in the airloads as functions of AoA that takes place during dynamic stall, and also because of the possibilities of lower aerodynamic damping, an otherwise stable elastic blade mode can be­come unstable if flow separation is present. Therefore, the onset of dynamic stall generally defines the overall lifting, propulsive, and aeroelastie performance limits of a helicopter rotor.

The accurate prediction of the combination of AoA and Mach number on the blade section that will produce dynamic stall onset, as well as the prediction of the subsequent effects of dynamic stall on rotor loads and performance, is not an easy task. The phenomenon of dynamic stall is not fully understood and is still the subject of much research on both experimental and numerical fronts — see Carr (1988). The very large number of publications on the phenomenon (see bibliography for this chapter) illustrates the importance of dynamic stall in the more complete aerodynamic and aeroelastie analysis of the helicopter rotor and the difficulties in both measuring and predicting the phenomenon. The complicated nonlinear physics of dynamic stall means that the behavior can only be completely modeled by means of numerical solutions to the Navier-Stokes equations (using computational fluid dynamics or CFD). This, like other CFD problems that involve unsteady, compressible, separated flows, the solution to dynamic stall problems is a formidable task that is not yet practical, but see the detailed discussion in Section 14.10.1. However, since the early 1990s, the rapid increase in computer resources has enabled considerable progress to be made in modeling dynamic stall by means of CFD approaches – see, for example, Srinivasan et al. (1993), Ekaterinaris et al. (1994), Landgrebe (1994), Barakos et al. (1998), and Spentzos et al. (2004). CFD methods will eventually prevail, but for the most part these methods are currently impractical for routine use in helicopter rotor analyses or design studies. For engineering analyses, the modeling of dynamic stall also remains a particularly challenging problem. To this end, a large number of semi-empirical models have been developed for use in helicopter rotor analysis and design codes. A brief discussion of some of these methods will be described in this chapter, along with a demonstration of their general capabilities in predicting dynamic stall induced airloads. While giving good results, these models are not strictly predictive tools and can really only be used confidently for conditions that are bounded by validation with experimental data. Such data, unfortunately, are not easy to obtain, requiring extensive wind tunnel testing on airfoils and wings, but remains fundamental to the success of such engineering models.

1 This is different from classical flutter, which involves fully attached flow.