Design and Analysis of Swirl Recovery Vanes for an Isolated and a Wing Mounted Tractor Propeller

Abstract

In light of the energy crisis of the early 1970's, NASA and industry gained a renewed interest in high-speed propellers for improved propulsive efficiency and explored the idea of swirl recovery vanes (SRV) to generate a net thrust from the residual swirl in the propeller slipstream. After this first effort on the aerial application of SRV, only recently research is resumed. When a wing is introduced in the slipstream of a propeller, for instance for a wing-mounted tractor-propeller, conclusions drawn on SRV in isolated condition may not hold. The objective of this research is to gain an improved understanding of the aerodynamic interaction between the propeller and swirl recovery vanes in an isolated configuration and wing-mounted tractor arrangement in the cruise condition and in a high-thrust condition. This study is realized by performing a series of transient Reynolds-averaged Navier-Stokes CFD simulations of a propeller with and without SRV in an isolated and installed configuration. Throughout this research the 6-bladed propeller of the European APIAN project is used. Available experimental propeller performance, blade pressure and slipstream measurements are used to validate the isolated propeller CFD model. Within the limitations of fully turbulent modelling of the boundary layer by means of automatic wall functions, good agreement is found with the experimental data, including the existence of a conical separation vortex at low advance ratios. Simulated performance and slipstream results are presented of the APIAN propeller with SRV designed for the APIAN-INF test program in the DNW-LLF. PIV measurements in a plane spanned by the radial and rotation axis provide a comparison of the slipstream velocity components and vorticity. This simulation combined with the PIV measurements enables an extensive description of the structure of root and tip vortices induced by the propeller blades and swirl recovery vanes. It is found that the propulsive efficiency increase by the addition of SRV is only 0.57% which is much lower than the design prediction of 1.8%. Therefore this design is not used in the remainder of the research and new SRV designs are proposed. An SRV analysis tool based on lifting-line theory modified for non-uniform inflow is presented. In combination with an optimisation routine, this tool allows for the design of SRV for an isolated propeller. From a simplified analysis of an elliptical vane in a uniform swirl flow, it is concluded that optimisation for maximum SRV thrust is preferred over complete swirl recovery to reach the highest gain in propulsive efficiency. Four designs are presented: Design 1 is optimised for the cruise condition with a constraint on stall for the high-thrust condition. Design 2 is optimised for the high-thrust condition with a constraint on the cruise condition for zero or positive efficiency benefit. These are designs where the SRV have a fixed pitch in flight. Also two variable pitch designs are proposed. The effect of cropping and the number of vanes on the propulsive efficiency is investigated as well for the objective of design 1. Design 1 and 2 are used in CFD simulations behind the isolated propeller to validate the predictions from the SRV analysis tool. In general the simulation results show that SRV lead to an increase in propulsive efficiency by increasing the system thrust over a wide range of advance ratios, with minor effect on the system power. Gains in propulsive efficiency of 0.39% and 0.20% are found in the cruise condition and 2.62% and 3.07% in the high-thrust condition for design 1 and 2 respectively. For high advance ratios the prediction is very accurate, while towards lower advance ratios the tool overpredicts the propulsive efficiency gain. The difference is within the limits that can be explained by the set assumptions. Design 1 proves that it is possible to increase the propulsive efficiency of an operating point close to the point of maximum propeller propulsive efficiency. Design 2 shows that if a larger increase in propulsive efficiency at low advance ratios is desired, the design can be changed at the cost of propulsive efficiency benefit at higher advance ratios, for a fixed SRV pitch design. Downstream of the SRV, somewhat less than half of the swirl is recovered on average. An expansion of the slipstream boundary is present, which is the result of the interaction of propeller blade and vane tip vortices. In the last part the wing of a Fokker 50 is introduced behind the propeller and SRV design 1. The loading on the wing induces an upwash upstream of the wing, resulting in a deviation from the SRV design inflow that is different for each vane by such a degree that flow separation degrades the SRV performance to a large extent. Therefore a change in the SRV design is made by turning each vane over an angle to obtain the time- and radial-average design inflow in the cruise condition. For future research it is recommended to find a different design for each vane. Since the effect of the wing upwash on the SRV inflow field varies with advance ratio and with wing loading and thus varies in flight, a variable pitch SRV design is recommended where the pitch of each vane is adjusted individually. For the cruise condition the increase in propulsive efficiency by the addition of SRV without considering differences in wing drag is found to be 0.93%, which is considerably higher than without wing, mainly due to the increased propeller propulsive efficiency, but partly by increased SRV thrust as well. 2.14% for a medium-thrust condition, which is very similar to the value without wing. For a wing-mounted tractor-propeller conclusions on SRV performance can only be drawn from the complete force balance of thrust and lift of the propeller, SRV, wing and nacelle. Considering the drag of all components, the net increase in propulsive efficiency by the addition of SRV is found to be -0.14% for the cruise and 1.00% for the medium-thrust condition with a net increase in lift of 0.35% and net decrease in lift of 0.55% respectively. Careful optimisation of SRV taking the wing into account as well as the lift as a constraint will most likely result in a performance benefit, since already with this non-optimised design an increase in thrust or lift can be found depending on the advance ratio. The propeller slipstream greatly affects the wing lift and drag distribution by its increased axial velocity and introduced swirl. It is concluded that SRV reduce some of the effects of the propeller on the wing lift and drag distribution by a reduction of the swirl, resulting in a smaller deviation from the wing loading without propeller. A design procedure for SRV should include the wing for instance by an additional lifting line and optimise for combined SRV and wing maximum thrust with a constraint on the net lift. This may lead to SRV designs more focussed on providing the optimal inflow for the wing in order to reduce the wing drag.