Two-dimensional Bluff Bodies Research Articles

Active flow control for bluff body drag reduction using reinforcement learning with partial measurements

Active flow control for drag reduction with reinforcement learning (RL) is performed in the wake of a two-dimensional square bluff body at laminar regimes with vortex shedding. Controllers parametrised by neural networks are trained to drive two blowing and suction jets that manipulate the unsteady flow. The RL with full observability (sensors in the wake) discovers successfully a control policy that reduces the drag by suppressing the vortex shedding in the wake. However, a non-negligible performance degradation ( $\sim$ 50 % less drag reduction) is observed when the controller is trained with partial measurements (sensors on the body). To mitigate this effect, we propose an energy-efficient, dynamic, maximum entropy RL control scheme. First, an energy-efficiency-based reward function is proposed to optimise the energy consumption of the controller while maximising drag reduction. Second, the controller is trained with an augmented state consisting of both current and past measurements and actions, which can be formulated as a nonlinear autoregressive exogenous model, to alleviate the partial observability problem. Third, maximum entropy RL algorithms (soft actor critic and truncated quantile critics) that promote exploration and exploitation in a sample-efficient way are used, and discover near-optimal policies in the challenging case of partial measurements. Stabilisation of the vortex shedding is achieved in the near wake using only surface pressure measurements on the rear of the body, resulting in drag reduction similar to that in the case with wake sensors. The proposed approach opens new avenues for dynamic flow control using partial measurements for realistic configurations.

Investigation of the influence of the bluff-body temperature on a lean premixed DME/air flame approaching blowoff

In flames stabilized by a two-dimensional (2D) bluff body, the lean blowoff (LBO) limit can be extended by increasing the bluff-body temperature, which is verified experimentally in premixed dimethyl ether (DME)/air flames in this work. Two bluff-body temperatures of 573 K and 773 K are tested, respectively. High-speed OH* chemiluminescence (CL) imaging is applied to record the spatial distribution of the heat release. Particle image velocimetry (PIV) and OH planar laser-induced fluorescence (PLIF) are performed simultaneously to explore the flow field and the combustion characteristics. Based on the results of OH*-CL, the heat release of the area near the bluff body is strengthened in the case of higher bluff-body temperature, and the enhanced chemical reactivity will further influence the downstream area. The turbulent flame speed obtained by OH-PLIF is increased in the entire area of interest. Due to the higher velocity gradient adjacent to the bluff body in the upstream region (y < 35 mm), the flame surface density (FSD), brush thickness, and flame wrinkling ratio are not sensitive to the bluff-body temperature. However, in the downstream region (y > 35 mm), a noticeable change can be found in these three parameters due to a higher turbulent flame speed and a much lower velocity gradient. Based on the PIV results, a limited effect on the velocity field by increasing the bluff-body temperature by 200 K. Furthermore, the probability density functions (PDF) of the vorticity and strain rate on the flame front are computed. The results indicate that the enhanced flame stability leads to a more pronounced overlapping between the flame front and the high-vorticity region, further increasing the values of these two parameters.

Investigation of the applicability of conditioned particle image velocimetry under conditions close to lean blow-off

In recent years, conditioned particle image velocimetry (CPIV) has attracted much attention for flame front extraction. It is an economical and straightforward technique because the flame front can be obtained simply from Mie-scattering images. In the current work, Mie-scattering and hydroxyl planar laser-induced fluorescence (OH-PLIF) measurements were taken simultaneously to study the applicability of CPIV under conditions with varying equivalence ratios, and the reliable working range of the CPIV method and the source of bias were clarified quantitatively. The premixed dimethyl ether (DME)/air flames stabilized on a two-dimensional bluff body were tested. It is found that the accuracy of the CPIV method worsens as the equivalence ratio decreases. The bias of CPIV is supposed to be related to the flame structure and aerodynamics. The extraction deviation in the upstream region for the stable flames is more significant than that in the downstream area due to the intensified interaction between the shear layer and flame near the bluff body. However, for the flames approaching the lean blow-off (LBO), the bias in the upstream part is smaller than that in the downstream area, which is related to the “back-support” of the hot recirculation zone (RZ). In addition, the difference between the results obtained by CPIV and OH-PLIF is usually the preheat layer of flame and concave-wrinkled location of OH-PLIF filled with CH2O radicals, especially under conditions close to the LBO, which can be used to estimate the spatial distribution of CH2O.Graphical

Aerodynamics and Wake Flow Characteristics of a Four-Cylinder Cluster

The aerodynamic behaviour and wake flow of a cluster of two-dimensional sharp-edged bluff bodies exhibits extremely complex unsteady phenomena in both near and far fields. Due to the high cost of wind-tunnel experiments and numerical simulations, a complete understanding of wake flows and a description of their characteristics are lacking. This paper presents large-eddy simulations (LES) in different flow/wind directions for a cluster of 2 times 2 aligned square cylinders, at a separation distance in streamwise and cross-wind directions equal to cylinder side length, and at Reynolds number {Re}=22,000 based on the single cylinder side length D. The case at 0^circ incidence shows an evident channel-type flow in the along-wind street/gap, and at its exit an irregularly pulsing jet with an intense shedding of large vortices. The wavelet analyses of the side force/lift coefficient and instantaneous velocities in the wake show that the characteristic length and time scales of the large vortical structures in the far-field wake are close to the cluster size 2D; this is the so called ‘cluster effect’. The cluster effect increases monotonically as the flow incidence angle increases. At a large incidence angle in the near-field wake, the cylinder-scale flow structures are much weaker compared to the cluster-scale structures. At the incidence angle of 45^circ, the overall wake flow and the aerodynamic characteristics are well scaled by the scale approximately equal to 2D. Nevertheless, the interaction between cylinders significantly affects the aerodynamics performance of the individual cylinders. The drag and lift coefficients of the individual cylinders differ substantially from each other in the cluster, and are significantly different from observations on a single isolated cylinder too.

MODULATION OF RECIRCULATION ZONE BEHIND A CUBICAL OBSTRUCTION BY THE VERTICALLY PLACED TURBULENT MULTIJETS IN THE FORM OF SHOWER

The present numerical work reports the application of jets for modulating and eradicating the recirculation zone in the wake region behind a fully submerged two-dimensional square cross-section bluff body for high Reynolds number flow. The wake region of the bluff body is characterized by a strong recirculation zone and vortex shedding, leading to unwanted vortex-induced vibrations that may instigate structural damage. Jets of varying Reynolds number and center-to-center distance are explored to look into the most effective jet Reynolds number and jet spread in completely modulating the recirculation region. The numerical simulation is based on a two-phase volume-of-fluid model with open channel boundary conditions. The standard k-&amp;omega; SST two-equation turbulence model is applied to close the time-averaged-momentum and continuity equations. The negative mean streamwise velocity signifies the recirculation region, and streamline contour plots are made to identify it. The variation in discharge velocity of the jet series is carried out for reallocation or complete disappearance of the recirculation region.

Harmonic forcing of a laminar bluff body wake with rear pitching flaps

A numerical study on the response of a two-dimensional bluff body wake subjected to harmonic forcing imposed by two rear pitching flaps is performed. The wake is generated by a rectangle at a height-based Reynolds number $Re=100$ , characterised by laminar vortex shedding. Two forcing strategies are examined corresponding to in-phase ‘snaking’ and out-of-phase ‘clapping.’ The effects of the bluff body aspect ratio ( $AR=1,2,4$ ), flapping frequency, flapping amplitude, flap length and Reynolds number are investigated. For the snaking motion, a strong fundamental resonance of the root mean square (r.m.s.) drag is observed when the wake is forced near the vortex shedding frequency. For the clapping motion, a weak subharmonic resonance is observed when the forcing is applied near twice the vortex shedding frequency resulting in an increase of the lift r.m.s. whereas the drag r.m.s. remains unaffected. Both resonances intensify the vortex shedding and a concomitant mean drag increase is observed for the snaking motion. Forcing away from the resonant regimes, both motions result in considerable drag reduction through wake symmetrisation and propulsion mechanisms. The formation of two vortex dipoles per oscillation period due to the flapping motion, which weaken the natural vortex shedding, has been identified as the main symmetrisation mechanism. A single scaling parameter is proposed to collapse the mean drag reduction of the forced flow for both motions over a wide range of flapping frequencies, amplitudes and flap lengths. Finally, the assessment of the performance of the forcing strategies has revealed that clapping is more effective than snaking.

Active flow control using deep reinforcement learning with time delays in Markov decision process and autoregressive policy

Classical active flow control (AFC) methods based on solving the Navier–Stokes equations are laborious and computationally intensive even with the use of reduced-order models. Data-driven methods offer a promising alternative for AFC, and they have been applied successfully to reduce the drag of two-dimensional bluff bodies, such as a circular cylinder, using deep reinforcement-learning (DRL) paradigms. However, due to the onset of weak turbulence in the wake, the standard DRL method tends to result in large fluctuations in the unsteady forces acting on the cylinder as the Reynolds number increases. In this study, a Markov decision process (MDP) with time delays is introduced to model and quantify the action delays in the environment in a DRL process due to the time difference between control actuation and flow response along with the use of a first-order autoregressive policy (ARP). This hybrid DRL method is applied to control the vortex-shedding process from a two-dimensional circular cylinder using four synthetic jet actuators at a freestream Reynolds number of 400. This method has yielded a stable and coherent control, which results in a steadier and more elongated vortex formation zone behind the cylinder, hence, a much weaker vortex-shedding process and less fluctuating lift and drag forces. Compared to the standard DRL method, this method utilizes the historical samples without additional sampling in training, and it is capable of reducing the magnitude of drag and lift fluctuations by approximately 90% while achieving a similar level of drag reduction in the deterministic control at the same actuation frequency. This study demonstrates the necessity of including a physics-informed delay and regressive nature in the MDP and the benefits of introducing ARPs to achieve a robust and temporal-coherent control of unsteady forces in active flow control.

A new scaling for the flow instability past symmetric bluff bodies

A scaling law to predict the onset of the primary Hopf instability in the steady flow past two-dimensional symmetric bluff bodies is proposed. It uses a measure of the spatial extent of the separation bubble as the length scale and the largest reverse-flow speed within it as the velocity scale. The ensuing Reynolds number, evaluated at the onset of the primary Hopf bifurcation, collapses quite nicely for bodies of different shape and aspect ratio even when a small angle of attack perturbs the symmetry; its relative variation is one order of magnitude smaller than that of the usual Reynolds number defined with the free-stream velocity and the cross-stream body size. With the new scaling, it can be roughly assessed whether the steady flow past a two-dimensional bluff body is absolutely and globally unstable to two-dimensional perturbations without a computationally expensive stability analysis: only the inspection of the base flow is required. More importantly, the scaling provides an insight into the flow mechanism that produces the instability.

Effect of pulsatile flow to drag of a two-dimensional bluff body

Effect of pulsatile flow to drag of a two-dimensional bluff body

Effect of pulsatile flow to drag of a two-dimensional bluff body

Effect of pulsatile flow to drag of a two-dimensional bluff body

The influence of spanwise nonuniformity on lean blowoff in bluff body stabilized turbulent premixed flames

New observations are reported of the processes leading up to lean blowoff (LBO) in a two-dimensional confined bluff body flame by visualizing LBO from a spanwise perspective. Time-resolved OH* chemiluminescence imaging simultaneous with particle image velocimetry are used to characterize the spanwise structure and dynamics. For reacting conditions, newly observed large-scale corner vortex structures (CVS) are present downstream of the bluff body inside the recirculation zone near the span walls. The CVS are orthogonal to the conventional recirculation zone observed in a bluff body flow, and the CVS are absent under non-reacting conditions. As the equivalence ratio is decreased, the vorticity and circulation of the CVS increase, contributing to significant nonuniformity across the spanwise direction both inside the recirculation zone and on the surrounding combustion. Approaching LBO, increased quantities of vanishing combustion near the span walls implies increased quantities of reactants entering and diluting the recirculation zone. Since cold reactant entrainment into the recirculation zone has been suggested as a possibly important LBO contributing factor, the current study suggests that this wake disruption mechanism begins much earlier than previously thought. The CVS are also the final regions of the flow to extinguish before LBO is complete, emphasizing that LBO is a highly three-dimensional process. A bluff body with an aspect ratio three times larger was also investigated, and the LBO process is observed to remain qualitatively similar with significant spanwise nonuniformity and containing the CVS. This paper intends to improve the three-dimensional understanding of two-dimensional bluff body stabilized premixed flames approaching LBO.

Numerical investigation of splitter plate effect on bluff body using finite volume method

In aerodynamics study, whenever there is a flat/blunt face, it will result in the maximum drag effect on the flat body. In these types of problems, the splitter plates have found a significant role in reducing the drag force. The splitter plate will separate/bifurcate the flow at the front face of the body. The present study aims to find the fluid flows over a body by varying the length of the splitter plate. In the study, the pressure and Mach number effects near the surface of the body. Also, it is crucial to study the fluid flows for higher velocity. A two-dimensional rectangular bluff body of a 50x60 mm dimension with a splitter plate controller is analyzed using a rectangular fluid domain—the splitter plate located at the front face of the body as a passive control method. Further study of this article, the parametric effect of the splitter plate with respect to the height of the body is considered. The ANSYS Fluent is used to simulate the results using a pressure-based solver because the flow is incompressible. The k-e turbulent model is used to simulate the outcomes and validated them with the wind tunnel experimental results. Based on the present results, it has been realized that the existing model can be utilized for the study of fluid flow over a bluff body. The simulations result in an essential effect of the upstream splitter plate on the separating the flows with a turbulent flow. Results indicate that the splitter plate is useful in separating the flow, which results in reducing the drag.

Aerodynamics of a two-dimensional bluff body with the cross-section of a train

Experiments on the aerodynamics of a two-dimensional bluff body simplified from a China high-speed train in crosswinds were carried out in a wind tunnel. Effects of wind angle of attack α varying in [−20°, 20°] were investigated at a moderate Reynolds number Re = 9.35 × 104 (based on the height of the model). Four typical behaviors of aerodynamics were identified. These behaviors are attributed to the flow structure around the upper and lower halves of the model changing from full to intermittent reattachment, and to full separation with a variation in α. An alternate transition phenomenon, characterized by an alteration between large- and small-amplitude aerodynamic fluctuations, was detected. The frequency of this alteration is about 1/10 of the predominant vortex shedding. In the intervals of the large-amplitude behavior, aerodynamic forces fluctuate periodically with a strong span-wise coherence, which are caused by the anti-symmetric vortex shedding along the stream-wise direction. On the contrary, the aerodynamic forces fluctuating at small amplitudes correspond to a weak span-wise coherence, which are ascribed to the symmetric vortex shedding from the upper and lower halves of the model. Generally, the mean amplitude of the large-amplitude mode is 3 times larger than that of the small one. Finally, the effects of Reynolds number were examined within Re = [9.35 × 104, 2.49 × 105]. Strong Reynolds number dependence was observed on the model with two rounded upper corners.

EFFECT OF INLET TURBULENCE INTENSITY ON TRANSPORT PHENOMENA OVER BLUFF BODIES

The effect of turbulence intensity on the transport phenomena over two-dimensional bluff bodies is studied in this work. Simulation is performed for an infinite circular cylinder and a square cylinder with the same hydraulic diameter D, which is also the nondimensional length scale. The computation is performed to cover the cross flow enclosing the laminar and turbulent regime for Reynolds numbers ranging up to 200,000 and inlet turbulence intensities from 5% to 40%. The Reynolds-averaged Navier-Stokes and energy equations are solved along with the transition shear stress transport (SST) model for the closure of turbulence. The results are initially validated with prior works for drag and thermal performance and experimental correlations. The analysis quantifies the influence of inlet turbulence intensity on enhancing heat transfer from the bluff bodies. However, the drag and pressure coefficients are not affected by the variation of inlet turbulence intensity.

Spanwise resolution requirements for the simulation of high-Reynolds-number flows past a square cylinder

In the past, the coarse spanwise resolution was widely used for high-Reynolds-number flow simulations of a two-dimensional bluff body with the purpose of reducing the computational resources. The effects and requirements of spanwise resolution (dz/D, where D is the width of the cylinder) remain unclear for direct numerical simulation (DNS) or large-eddy simulation (LES). This numerical study, without any explicit subgrid-scale model, focuses on the typical flow around a square cylinder at the Reynolds number of 22,000. dz/D varies between 0.08 and 0.013. The requirement is recommended based on the systematical examination of aerodynamic pressure forces, vortex shedding in the wake, and flow structures in the separating shear layers. dz/D ≈ 0.04 is suggested sufficient for the prediction of mean and r.m.s aerodynamic pressures acting on the cylinder, mean and r.m.s velocity fields and small-scale turbulent motions in the wake. The coarser dz/D causes longer vortex formation length behind the cylinder and smaller fluctuating velocities around the cylinder (except the spanwise component). However, a stricter dz/D ≤ 0.02 is suggested necessary for accurately predicting the shear layer behaviors. Only if this requirement is satisfied, the subharmonic and harmonic of the Kelvin-Helmholtz instability can be captured, together with the process of laminar-turbulent transition in the shear layer. The shear layer when dz/D ≤ 0.02 is composed of interconnected small-scale spanwise and streamwise vortices. The shear layer has greater integrity along the spanwise direction and flaps as a whole at the same pace with vortex shedding. However, the shear-layer spanwise vortices are discrete when dz/D>0.02 as the streamwise vortices are not sufficiently reproduced.

Effect of the spanwise computational domain size on the flow over a two-dimensional bluff body with spanwise periodic perturbations at low Reynolds number

To effectively control the flow over a two-dimensional bluff body, three-dimensional perturbations including spanwise-varying shape modifications and surface actuations have been suggested by previous studies. One of the computational issues is how big the spanwise computational domain size (Lz) should be as compared to the wavelength of these spanwise-varying perturbations (λ), to accurately estimate the effect of these perturbations. To investigate this issue, we consider three different two-dimensional bluff bodies with spanwise periodic perturbations: wavy square cylinder, wavy circular cylinder, and spanwise-varying blowing and suction on a circular cylinder at Re = 100. The spanwise length scales of largest vortical structures existing in the wake significantly depend on Lz, and they are larger than λ, reaching 2λ, 3λ, and even bigger than or equal to 12λ for the cases considered. These largest structures are composed of multiple of elementary vortical structures whose length scales are smaller than or equal to Lz. The amounts of drag reduction by periodic perturbations at Lz > λ are smaller than those at Lz = λ, indicating that the computational domain size equal λ overestimates their performances.

RANS and DDES simulations of a horizontal-axis wind turbine under stalled flow condition using OpenFOAM

This paper presents a numerical study of a horizontal axis wind turbine under stalled flow condition with the tip speed ratio λ=4.17. Both RANS and DDES simulations are performed using OpenFOAM in order to predict the aerodynamic characteristics of the wind turbine rotor. In addition, numerical results are compared with experimental data where a good agreement is achieved. The complex and unsteady root flow characteristics are comprehensively investigated. Numerical result shows that in terms of aerodynamic blade loading, DDES performs much better than RANS in aerodynamic prediction, where the error has been reduced from 20% to 5%. The Strouhal numbers are in the range 0.16≤St≤0.20 for blade sections r/R=0.25, 0.60 and 0.82, indicating two-dimensional bluff body vortex shedding occurs in stall or deep stall conditions. A very low Strouhal number is observed at r/R=0.92 with a value of 0.018. The low value of St corresponds the flow at r/R=0.92 section is in the near post-stall regime, with three-dimensional surface flow typologies, with periodic switching between stalled and unstalled conditions.

Separation delay through contoured transverse grooves on a 2D boat-tailed bluff body: Effects on drag reduction and wake flow features

Abstract The effectiveness of properly contoured transverse grooves in delaying the flow separation occurring on a two-dimensional boat-tailed bluff body is assessed through numerical simulations. The body has a cross-section with a 3:1 elliptical forebody and a rectangular main part followed by a circular-arc boat tail. Three-dimensional Variational Multiscale Large Eddy Simulations are carried out at R e = D u ∞ ∕ ν = 9 . 6 × 1 0 4 , using a mixed finite-volume/finite-element method. The introduction of one contoured groove on each of the boat-tail lateral surfaces produces a significant delay of flow separation and a consequent increase of the base pressure, with a global drag reduction of the order of 9 . 7 % . The wake dynamical structure remains qualitatively similar to the one typical of blunt-based two-dimensional bodies, with quantitative variations that are consistent with the reduction in wake width caused by boat tailing and by the grooves. The introduction of the grooves leads also to a regularization of the vortex shedding downstream of the body, which is more correlated in the spanwise direction. Finally, a few supplementary simulations show that the effect of the grooves is also robust to the variation of the geometrical parameters defining their location and shape.

Periodic vortex shedding phenomenon for various separation distances between two plane turbulent parallel jets

In the present study, two plane parallel jets with various separation distances between the two jets (d) are numerically simulated using two-dimensional unsteady RANS equations. The Reynolds number based on the jet width w is Re=10,000. The results reveal that when the separation distance between the two jets lies in the range 0.6⩽d/w⩽1.4, the flow field demonstrates a periodic vortex shedding phenomenon close to the nozzle plate similar to what would be expected in the near wake region for flow over a two-dimensional bluff body. On the contrary, for d/w=0.5 and 1.5, the near flow field region remains to be steady with two counter rotating stable vortices in the recirculation zone. Within the range 0.6⩽d/w⩽1.4, the time series of streamwise and transverse velocity signals display a pure sinusoidal oscillation. In this range, for a given d/w, the fast Fourier transform of both the velocity signals exhibits a single dominant frequency peak at a same value of Strouhal number (non-dimensional vortex shedding frequency) that gradually decreases with the progressive increase in the separation distance between the two jets.

Transition of a Steady to a Periodically Unsteady Flow for Various Jet Widths of a Combined Wall Jet and Offset Jet

In the present paper, a dual jet consisting of a wall jet and an offset jet has been numerically simulated using two-dimensional unsteady Reynolds-Averaged Navier–Stokes (RANS) equations to examine the effects of jet width (w) variation on the near flow field region. The Reynolds number based on the separation distance between the two jets (d) has been considered to be Re = 10,000. According to the computational results, three distinct flow regimes have been identified as a function of w/d. For w/d ≤ 0.5, the flow field remains to be always steady with two counter-rotating stable vortices in between the two jets. On the contrary, within the range of 0.6 ≤ w/d &lt; 1.6, the flow field reveals a periodic vortex shedding phenomenon similar to what would be observed in the wake of a two-dimensional bluff body. In this flow regime, the Strouhal number of vortex shedding frequency decreases monotonically with the progressive increase in the jet width. For w/d ≥ 1.6, the periodic vortex shedding is still evident, but the Strouhal number becomes insensitive to the variation of jet width.