Experimental analysis of Reynolds effect on flow detachment and sudden flow release on a wall-mounted hump

Highlights

• The mean flow unsteady behavior clearly impacts the growth of the recirculated region.

• Flow acceleration imprints a momentum boost that abates the flow detachment.

• The momentum boost impact is still present after the mitigation of the acceleration.

• The dynamic evolution suggests that pulsating actuation is effective and efficient.

Abstract

The boundary layer detachment limits compact power plants' operation at moderate Reynolds number in adverse pressure conditions. Flow detachment is promoted by the lack of momentum in the near-wall region when exposed to adverse pressure gradients. Transient flows or periodic flow perturbations may delay or prevent the flow detachment. The present investigation experimentally analyzes the behavior of separated flows based on an ad-hoc wall-mounted hump. The near wall flow region detachment and recirculated flow evolution under sudden flow release were experimentally characterized. The extension of the separated region and its dynamic development were monitored through surface pressure, temperature measurements, and hot-wire traverses. Comparing the 3D hump performance during steady state experiments against sudden flow acceleration runs, we report for the first time the temporal response of a diffusive passage exposed to a sudden flow acceleration and the impact of the acceleration on the boundary layer detachment. Due to the sudden flow release, the near-wall region can overcome the adverse pressure gradient. However, as the flow acceleration dilutes, the boundary layer detaches, and the recirculated flow region develops. The comparison of experimental results against 3D transient Computational Fluid Dynamics simulations, using the standard gas turbine industry approach, demonstrates the ability of Unsteady Reynolds Averaged Navier-Stokes models to predict the dynamic performance of this phenomenon.

Introduction

Unmanned Air Vehicles propulsive solutions, serial and parallel hybrid engine concepts, and extended range propulsion concepts promote the design and evolution of efficient and lighter power plants with an extended range of operations [1], [2]. However, the operation at elevated altitudes is limited because of potential flow separation along diffusive passages, mitigating the overall power plant performance [3]. For this reason, over the last decades, both passive ([4], [5], [6]) and active ([7], [8], [9], [10]) flow control techniques were introduced to abate the boundary layer detachment at reduced Reynolds numbers. The introduction of passive flow control elements alters the aerothermal performance at the design point, locally boosting the losses and cutting the power plant efficiency. On the other hand, active flow control techniques only operative when needed while maintaining the optimum performance at design conditions. Hence, the operation at low Reynolds numbers can broaden the operational range while maintaining optimum efficiency around the design point. In this sense, Greenblatt et al. [11] experimentally analyzed the performance of steady flow suction preventing the generation of recirculated flow regions. The control of flow separation becomes pivotal in the low-pressure turbines, where their operation under unsteady free-stream condition modulate the boundary layer separation and reattachment, [12], [13], [14]. Similarly, the influence of unsteady flow conditions or unsteady operation strongly modulates the performance of wind turbine geometries as presented by Leu et al. [15]. Hence, if we want to propose a practical and effective flow control approach, we must first understand the dynamics of flow separation under unsteady flow conditions. Periodic wall humps were experimentally investigated by Rapp and Manhart [16] to disclose the influence of periodic perturbations over diffusive passages, identifying the reduction of the recirculated flow region as a consequence of the increased near wall flow momentum.

The study of the boundary layer detachment and reattachment process with or without actuation is usually done using wall-mounted humps, such as Pescini et al. [17] and Martinez et al. [18]. Their experiments focused on using dielectric barrier discharge actuators to mitigate the boundary layer detachment at reduced Reynolds numbers; plasma-induced momentum was proven to abate the boundary layer detachment. Plasma actuation based on Dielectric Barrier Discharge actuators was also proposed by Benard et al. [19] to reattach the flow over diffuser walls. Along these lines, Seifert and Pack [20] introduced periodic flow blowing and suction in a diffusive geometry operating at elevated Reynolds number to abate the boundary layer detachment.

Saavedra and Paniagua analyzed the dynamic performance of the near wall flow separation and reattachment phenomena while operating under unsteady free-stream conditions via Unsteady Reynolds Average Navier Stokes and Large Eddy Simulations [21]. The description of the dynamic evolution of the boundary layer detachment and reattachment could lead to further effective and efficient boundary layer regulation solutions that abate the flow separation with minimal energy requirements. The computational studies presented in [21] indicated the impact of free-stream bursts of speed and flow fluctuations over the near wall flow separation and recirculated flow extension. The test article was designed to suffer boundary layer detachment driven by adverse pressure gradients under low Reynolds number operation while maintaining the attached flow at higher Reynolds numbers. The numerical analysis under sudden flow discharge revealed the flow acceleration's influence boosting the near-wall region's flow momentum and preventing the separation.

This manuscript explores the flow separation evolution during rapid flow release experiments. The 2D geometry characterized in [21] was evolved into a 3D variant of the diffusive wall mounted test article. Investigations at uniform and stable mean flow conditions revealed the test article's performance over an extended range of Reynolds numbers, identifying the limits of operation for designed and recirculated flow conditions. Also, blowdown experiments were performed to characterize the influence of varying free-streams conditions over the flow separation and recirculated flow region progression. Abrupt flow discharge tests simulate the influence of sudden operational line alteration or the flow establishment in diffusive passages within the propulsion system, representative of the fluctuating conditions caused by upstream rotating wakes. Higher accuracy eddy-viscosity models are required to properly simulate the aerodynamics present in complex flow configurations with flow straining, streamline curvature, alternating acceleration-deceleration processes, or shear layer detachment. Scale resolving models must be applied to fully characterize these phenomena, such as DNS, LES, or wall-modelled LES,[22]. However, to test the accuracy and temporal dynamic performance of the commonly used approach in the gas turbine industry in linear eddy-viscosity models, we compare the empirical findings against 3D URANS using the transitional k-ω SST turbulence closure.