Search

A one-equation Reynolds-averaged Navier-Stokes (RANS) closure model has been established for bypass transition and laminar separation bubble (LSB)–induced transition. A new local indicator is proposed to describe the influence of turbulence intensities and pressure gradients, which makes the model Galilean invariant. Based on this new indicator, a novel and efficient transition criterion is formulated. For LSB-induced transition, an empirical correlation is developed to modify the intermittency factor and control the size of separation bubbles. Several flow cases, including flat plates with various pressure gradients, flat plates with LSBs, semicircular leading-edge flat plates, National Advisory Committee for Aeronautics (NACA) 0018 airfoil, and SD7003 airfoil, are employed for the model verification. The predictions show good agreement with the experimental data and large-eddy simulation data for different inlet conditions, which indicates the increment of free-stream turbulence intensity leads to the reduction of the size of laminar separation bubbles. [ABSTRACT FROM AUTHOR]

The characteristics of unsteady boundary-layer transition on a small-amplitude pitching NACA0012 airfoil are investigated using the γ−Reθt¯−Ar transition model, which consists of the γ−Reθt¯ transition model and the roughness amplification factor (Ar) transport equation. The present transition model is validated by three roughness surface cases, including a zero pressure gradient flat plate, the NACA0012 airfoil, and NREL-S814 airfoil with different roughness surface. The numerical results for these cases are in agreement with experimental data. Three pitching NACA0012 airfoils with smooth surface, fully distributed roughness surface, and roughness leading edge are simulated by the present model. The impact of reduced frequencies ranging from 0.019 to 0.301 is also taken into consideration. The results show that as the reduced frequency increases, the transition location moves downstream in the upstroke and upstream in the downstroke. With the influence of distributed roughness, the transition moves upstream and the time delay of the transition location is decreased. Moreover, the roughness leading edge enhances the asymmetry of the propagation speed and makes it more sensitive to the reduced frequency. [ABSTRACT FROM AUTHOR]

In traditional ways, the aerodynamic property of the aircraft is obtained by solving Navier-Stokes equations or performing tunnel experiments. However, these methods are time consuming for the aircraft design and optimization. In comparison, the deep learning technique is capable of handling high dimensional parameters and can describe compressible flow structures clearly and efficiently. For these, an efficient and accurate prediction approach based on the deep neural network is proposed for the compressible flows over the transonic airfoils in this study. By investigating the effects of the input coordinate features of the deep learning method on the prediction accuracy and robustness, the aerodynamic characteristics, such as lift, drag, and pitch coefficients, are obtained from the predicted flow fields. Results indicate that the proposed deep learning prediction method is with a high resolution and efficiency. It is promising to be extended to the optimization and design process of the supercritical airfoil. [ABSTRACT FROM AUTHOR]

Aerodynamic design optimization is a key aspect in aircraft design. The further evolution of advanced aircraft derivatives requires a powerful optimization toolbox. Reinforcement learning (RL) is a powerful optimization tool but has rarely been utilized in the aerodynamic design. It can potentially obtain results similar to those of a human designer, by accumulating experience from training. In this work, a popular RL method called proximal policy optimization (PPO) is proposed to investigate multi-object aerodynamic design optimization. By observing the aerodynamic performances of different airfoils, the PPO updates a reasonable policy to generate the optimal airfoils in a single step. In a Pareto optimization problem with constraints, the PPO requires only 15% of the computational time of the non-dominated sorted genetic algorithm (II) to achieve the same accuracy. The results from testing show that the agent learns a policy that can achieve ∼4.3%–10.1% improvements of the aerodynamic performance compared with the results of baseline. [ABSTRACT FROM AUTHOR]

In the preliminary aerodynamic design phase of transonic wing, inverse design of supercritical airfoils plays an important role in acquiring practical results. However, traditional inverse design approach in transonic regime highly dependents on the manual design experience and physical model simulation, which is time consuming and poses a challenge for its applications in engineering problems. In order to relieve these drawbacks, a novel method is proposed for inverse design of supercritical airfoils based on Wasserstein generative adversarial network (WGAN), genetic algorithm (GA) and deep convolutional neural network (DCNN). This method consists of three stages. First, design characteristics of transonic pressure coefficient (CP) distribution in physics are well captured and imitated by WGAN. Then, GA is used to select task-oriented CP via adjusting controllable latent variables in generator of WGAN. Next, physical-existing CP selected by GA is recovered to corresponding airfoil by DCNN with high efficiency and accuracy. Cases using the dataset sampled based on RAE2822 airfoil are carried out to show effectiveness of submodules involved. Finally, design performance of proposed multistage method is further verified by specific design goals. Test cases indicate that selected CP by GA according to the aerodynamic design criteria can be accurately reflected on inverse designed airfoils, which are confirmed by well agreements between high-fidelity simulation result and selected CP. The proposed method is a promising design approach that can quickly achieve optimal airfoils without significant loss of accuracy. [ABSTRACT FROM AUTHOR]

In the aerodynamic design, optimization of the pressure distribution of airfoils is crucial for the aerodynamic components. Conventionally, the pressure distribution is solved by computational fluid dynamics, which is a time-consuming task. Surrogate modeling can leverage such expense to some extent, but it needs careful shape parameterization schemes for airfoils. As an alternative, deep learning approximates inputs-outputs mapping without solving the efficiency-expensive physical equations and avoids the limitations of particular parameterization methods. Therefore, this paper presents a data-driven approach for predicting the pressure distribution over airfoils based on Convolutional Neural Network (CNN). Given the airfoil geometry, a supervised learning problem is presented for predicting aerodynamic performance. Furthermore, we utilize a universal and flexible parametrization method called Signed Distance Function to improve the performances of CNN. Given the unseen airfoils from the validation dataset to the trained model, our model achieves predicting the pressure coefficient in seconds, with a less than 2% mean square error. [ABSTRACT FROM AUTHOR]

Abstract Since being proposed, the HLLEM-type schemes have been widely used because they are with high discontinuity resolutions and can be easily applied to the other system of hyperbolic conservation law. In this paper, we conduct theoretical analyses on the HLLE-type schemes' performances at low speeds. By realizing that the excessive numerical dissipations corresponding to the velocity-difference terms of the momentum equations make these schemes incapable of obtaining physical solutions at low speeds, we adopt the function g to control such dissipation. Also, we borrow the HLLEMS scheme's construction and damp the shear waves in the vicinity of the shock to avoid the shock anomaly's appearance. The moving contact discontinuity case and the Sod shock tube case show that the HLLEMS-AS scheme we propose in this paper can capture contact discontinuities and shocks as sharply as HLLEMS scheme. The Quirk's odd–even test case and the hypersonic inviscid flow over a cylinder case demonstrate that HLLEMS-AS is robust against the shock anomaly. The inviscid low-speed flow around the NACA0012 airfoil case indicates that HLLEMS-AS is with a high resolution at low speeds. The turbulent flow past a backward facing step case demonstrates the shear wave capturing ability of the HLLEMS-AS scheme. These properties suggest that HLLEMS-AS is promising to be widely used in both cases of low speed and high speed. [ABSTRACT FROM AUTHOR]

Abstract By extending Bradshaw's assumption from the free shear flows to the wall-bounded flows, the Kinetic energy Dependent Only turbulence model (KDO) established a new Reynolds stress constitution. The Bradshaw function (τ 12 / k) and the coefficient of the dissipation term are the only two empirical coefficients. They were both calibrated with the turbulent Reynolds number, Re k = ρ k 1 / 2 d / μ , which depends on the wall distance. Once the wall distance is replaced with "flow–structure–adaptive" parameters, such as the eddy viscosity ratio, r = μ t / μ , the model could naturally capture various transition phenomena. The improved KDO model is assessed by some test cases including the classic bypass transition of T3A and T3B boundary layers, the natural transition of the T3A boundary layer, and the separation bubble induced transition of Aero-A airfoil. The assessments show that the improved KDO model is not capable of capturing the precise process of the laminar-turbulent flow transition, but the model can accurately predict the transition onset locations. The model does not include specific transition mechanisms; however, for high Reynolds numbers and complex flows with different types of transition, the predictions are reliable. [ABSTRACT FROM AUTHOR]

Circulation Control (CC) has been attracting increasing attention because of its potential to replace traditional control surfaces. One of the bottlenecks that restrict the further application of the CC technology to the transonic flight control is the separation of the supersonic CC jet, which is usually accompanied by hysteresis. In this study, the hysteresis effect of the CC technology on the ONERA OAT15A supercritical airfoil is numerically studied with the transonic freestream conditions M a = 0.73 , α = 2.5 °, and R e c ≈ 3.0 × 10 6. Results indicate that there is a significant hysteresis between the jet separation and the reattachment. Furthermore, the mechanisms of the CC jet separation and reattachment are obtained by analyzing the development of the jet flow. Finally, four different CC devices are investigated to show that the hysteresis effect of the CC technology can be effectively reduced by increasing the radius-to-slot ratio, setting a step between the slot exit and the Coanda surface, and employing a convergent-divergent nozzle. [ABSTRACT FROM AUTHOR]