Yao Zheng

Zhejiang University | ZJU · School of Aeronautics and Astronautics

Parametric unsteady flow modeling plays a significant role in the fluid dynamics, since the unsteady flow problems are usually involved with complex physical phenomena. Currently, in the study of data-driven unsteady flow modeling, the convolutional neural network based autoencoder (CNN-AE) model has been widely used. While, its application in engi...

The ability to manipulate fluids has always been one of the focuses of scientific research and engineering application. The rapid development of machine learning technology provides a new perspective and method for active flow control. This review presents recent progress in combining reinforcement learning with high-dimensional, non-linear, and ti...

A deep learning framework is proposed for real-time transonic flow prediction. To capture the complex shock discontinuityof transonic flow, we introduce the residual network ResNet and deconvolutional neural networks to learn thenonlinear discontinuity phenomenon in transonic flow, which is affected by the Mach number, angle of attack, Reynoldsnumb...

An efficient meshing parallelism method based on a task graph model is presented for numerical simulation mesh generation. The proposed method can parallelize a meshing process without modifying the existing kernel code of meshing algorithms. Various novel schemes have been developed to improve the efficiency of the algorithm while not sacrificing...

Marching normal calculation is considered the focus of the commonly used advancing layer method for viscous mesh generation, and the local configuration of multiple normals for complex corners locally is a rather complicated problem that has not been completely solved yet. In some extremely complex corners, even the advanced commercial software can...

In order to further improve the trajectory tracking accuracy of quadrotor in unknown environment, a control method adding an iterative learning feedforward controller to the traditional feedback control architecture was proposed. Facing the difficulty of tuning learning parameters in the process of Iterative Learning Control (ILC), a method of tuni...

An automatic unstructured mesh generation approach is presented to discretize complex electronic packaging systems for finite element analysis. Various novel schemes are developed to resolve the common issues (models contain geometrical defects, models contain small but necessary features, simulation properties are predefined on models, etc.) to au...

Film cooling technology is significant for aero-turbines development. However, uncertainty on operating and manufacturing conditions would put the high-pressure turbines at the risk of unexpected problems. To extend the service life of the gas turbines, it is necessary to conduct an uncertainty analysis of the effects of the geometry parameters at...

An efficient advancing front algorithm is presented to create high-quality surface elements on CAD models. Various novel schemes are developed to improve the efficiency of the algorithm while not sacrificing its robustness and the resulting mesh quality. New features include a fast projection scheme that can perform robustly in case of badly define...

Rapid overall design method and related algorithms for morphing aircraft is presented. Considering the development trend of morphing aircraft in China and abroad, taking the overall design of Virgin Galactic SS2 and variable sweep angle aircraft as examples, this paper discusses the rapid overall parameter design method for morphing aircraft and it...

Deep reinforcement learning (RL) is capable of identifying and modifying strategies for active flow control. However, the classic active formulation of deep RL requires lengthy active exploration. This paper describes the introduction of expert demonstration into a classic off-policy RL algorithm, the soft actor-critic algorithm, for application to...

Due to the difficulty to establish an aerodynamic model and the complex disturbances, the flapping wing aerial vehicle (FWAV) cannot be accurately controlled at present. A framework of adaptive sliding mode controller based on an extended state observer and online learning is proposed in this paper. First, the extended state observer is established...

In the present study, a body-fitted and uniform initial particle arrangement method is proposed for mesh-free methods, such as SPH, MPS, and LSMPS. Single-layer boundary particles with/without normal direction vectors and multilayer boundary particles can be successfully generated. Typical tests show that the proposed method can successfully genera...

The design optimization taking into account the impact of uncertainties favors improving the robustness of the design. A Surrogate-Assisted Gradient-Based (SAGB) method for the robust aerodynamic design optimization of turbomachinery blades considering large-scale uncertainty is introduced, verified and validated in the study. The gradient-based me...

A comparative study of two micro-blowing temperature cases has been performed to investigate the characteristics of drag reduction in a subsonic flat-plate flow (where the freestream Mach number is 0.7) by means of Direct Numerical Simulation (DNS). With minute amount of blowing gas injected from a 32×32 array of micro-holes arranged in a staggered...

To take the potential aerodynamic advantages of ducted fan, a novel aircraft configuration utilizing tilt-duct eVTOL concept is developed in this study. The current paper focuses on the overall performance estimation of the tilt-duct aircraft via conceptual design and system level analysis for a typical eVTOL design mission with maximal take-off we...

Marching normal calculation is considered the focus of the commonly used advancing layer method for viscous mesh generation, and the local configuration of multiple normals for complex corners locally is a rather complicated problem that has not been completely solved yet. In some extremely complex corners, even the advanced commercial software can...

A comparative study about riblets-controlled turbulent boundary layers has been performed to investigate the turbulence characteristics associated with drag reduction in a compressive flat-plate flow (where the free-stream Mach number is 0.7) by means of direct numerical simulations (DNSs). With a setting of the triangular riblets (s+≈30.82, h+≈15....

Based on an integrated annular combustor-turbine experimental platform, the effects of wall temperature on thermoacoustic instability were experimentally studied in a fixed case with an equivalence ratio Φ=0.82 and a combustion power P=15.5kW. Compared with the typical independent annular combustor test, the outlet with the component of the turbine...

Electric propulsion aircraft tends to be a promising way to reach the aviation carbon reduction goals in addition to evolutionary/revolutionary aircraft technologies and operational strategies. In this paper, we perform analysis and optimization studies for hybrid electric propulsion aircraft. To enhance the reliability of the study, verified aircr...

In the present study, the SPH method is implemented to simulate the sloshing in the propellant tank of a reusable rocket vehicle. The SPH method with the surface tension model and LES model is used. The complex three-dimensional geometry is discretized usingSPHparticles using aCAD-compatible body-fitted particle generator.The narrowband surface rec...

Critical slowing down phenomena occur during the transition process of various dynamical states, such as bifurcations. The eigenvalues of dynamical systems can be regarded as an indicator of critical slowing down of impending bifurcation. Adaptive, locally linear models can extract local eigenvalues of the nonlinear dynamical system by segmenting a...

A novel data-driven nonlinear reduced-order modeling framework is proposed for unsteady fluid-structure interactions (FSIs). In the proposed framework, a convolutional variational autoencoder model is developed to determine the coordinate transformation from a high-dimensional physical field into a reduced space. This enables the efficient extracti...

Inspired by the albatross, a dynamic soaring simulation system with distributed pressure sensors was constructed in this study. The advantage of the system lies in harvesting the energy from wind shear layer and estimating the wind information using a pressure-based sensor system. Specifically, the dynamic soaring simulation system contains an offl...

In the present work, we propose an optimization framework based on the active learning method, which aims to quickly determine the conditions of tandem flapping wings for optimal performance in terms of thrust or efficiency. Especially, multi-fidelity Gaussian process regression is used to establish the surrogate model correlating the kinematic par...

An indoor positioning method of unmanned aerial vehicle (UAV) based on improved MSCKF was proposed aiming at the problem that the indoor positioning of UAV is prone to drift. A high robustness and low delay detection method was proposed under the framework of multi-state constraint Kalman filter (MSCKF). The pose of UAV was calculated with the help...

Impinging atomization of two jets is numerically simulated using the GPU (Graphics Processing Unit) accelerated moving particle semi-implicit (MPS) method. The acceleration ratio of the GPU-accelerated code is 11.75, which proves the validity of GPU acceleration. Three jet impact modes under different impinging velocities are successfully simulated...

For the poor maneuverability of flapping wing micro aerial vehicles (FWMAVs), a deep reinforcement learning (DRL) based local path planning method (IL-PPO2) is proposed with the assistant of demonstration learning in an unknown environment. Firstly, due to the limited visual angle of a stereo camera on a FWMAV, a “Heart” algorithm is proposed to re...

Prediction of complex fluid flows from sparse and noisy sensor measurements is widely applied to many engineering fields. In the present study, a novel compressed sensing reduced-order modeling framework has been proposed to predict unsteady flow fields from sparse and noisy sensor observations, which includes offline learning and online forecastin...

The transition from the low-amplitude, aperiodic fluctuation to thermoacoustic instability is usually treated as a Hopf bifurcation, but the transition incorporates complex dynamical behaviors. Intermittent oscillations commonly occur in the transition process wherein chaotic states alternate with limit cycle oscillations and both the chaotic state...

This paper presents a novel pipeline that improves the efficiency of narrowband surface reconstruction for particle-based fluids. Different from previous narrow band methods, we only construct the scalar field in the outer surface region. A method with robust criteria for detecting the outer surface vertices of the scalar field is then proposed to...

The interaction between an incident shockwave and awavy end-wall in a three-dimensional geometry is numerically simulated by using a high-order finite-difference solver with a ghost-cell immersed boundary method. The aim is to discover the differences of the unsteady propagation characteristics of triple bifurcation points and transverse waves at d...

Compared with the quasi-steady combustion process, the large pressure oscillation in the combustion unstable process contains rich physical characteristics and more severe structural damage phenomenon. The active control method is more flexible and efficient by controlling some parameters of the combustion system through the actuator, so as to dest...

Film cooling has shown great potential in protecting hot section of high-pressure turbine from melting down. A counter-rotating vortex pair (CVP) is produced downstream of the cooling hole and reduces the lateral diffusion of the jet. To enhance the film cooling effectiveness, a semi-sphere vortex generator (SVG) is proposed to be installed downstr...

In present paper, mesh adaptation is applied for curing the pathological behaviors of the enhanced time-accurate upwind scheme (Loh & Jorgenson, AIAAJ 2016). In the original ETAU (enhanced time-accurate upwind) scheme, a multi-dimensional dissipation model is required to cure the pathological behaviors. The multi-dimensional dissipation model will...

In order to further improve the trajectory tracking accuracy of quadrotor drone in unknown environment, a control method that added an iterative learning feedforward controller to the traditional feedback control architecture was proposed. Facing the difficulty of tuning learning parameters in the process of iterative learning control, a method of...

In order to study the characteristics of flame flickering in premixed conical flame, the premixed propane-air flame was used as the experimental object and the velocity fields of flame flickering were experimentally studied by particle image velocimetry. Dynamic mode decomposition was used to distinguish and reconstruct the main features of the vel...

Purpose Sizing functions are crucial inputs for unstructured mesh generation since they determine the element distributions of resulting meshes to a large extent. Meanwhile, automating the procedure of creating a sizing function is a prerequisite to set up a fully automatic mesh generation pipeline. In this paper, an automatic algorithm is proposed...

Topology optimization as a computational approach is used to find an optimal structure with specific objectives such as an ultimately lightweight design. The optimization is normally performed under a series of constraint functions to ensure expected structural performance for safety and resilience. To promote the use of topology optimization in st...

Plenary Lecture, Presented at the 21st IACM Computational Fluids Conference (CFC 2021), Hangzhou, China, October 17-21, 2021.

The recently proposed anisotropic mesh adaptation based on a Riemannian metric field is regarded as a promising approach to control the computational cost of computational fluid dynamics simulations. To ensure that computations of the aeroengine combustor have high fidelity and are cheap, this study extends anisotropic mesh adaptation to simulate u...

In this paper, local flame and flow properties of propagating premixed turbulent propane/air flames were derived by two-dimensional time-resolved Mie scattering and PIV methodology in a MICCA-type annular combustor. The curvature of flame front, local flame speed and the evolution of leading points during the light-round process were analyzed in de...

View Video Presentation: https://doi.org/10.2514/6.2021-3321.vid Despite the low noise and zero carbon emissions of electric propulsion aircraft, the low specific energy of batteries has limited the application of electric propulsion only to light aircraft. Via retrofit and new designs, the current study aims at studying and evaluating the overall...

For annular combustors, the interactions between different modal structures (longitudinal, azimuthal spinning, standing, and slanted) remain a challenge in ongoing studies of thermoacoustic instabilities. The present work develops an annular combustor named TurboCombo, similar to the multiple-injector annular combustor named MICCA, but with a conve...

In this paper, an entirely parallel pipeline for creating realistic visualization of particle-based fluid is presented, which significantly improves surface reconstruction and foam generation performance. Firstly, we develop a spatial hashing grid-based strategy. Basis on this strategy, a novel narrow band surface reconstruction method is presented...

For film cooling in high pressure turbines, it is vital to predict temperature distribution on the blade surface downstream of the cooling hole. This temperature distribution depends on the interaction between the hot mainstream and the coolant jet. Deep learning techniques have been widely applied in predicting physical problems such as complex fl...

In this article, an entirely parallel pipeline for creating realistic visualization of particle‐based fluid is presented, which significantly improves surface reconstruction and foam generation performance. Firstly, we develop a spatial hashing grid‐based strategy. Based on this strategy, a novel narrow band surface reconstruction method is present...

Direct numerical simulation (DNS) is used to investigate the turbulent flat-plate boundary layer with localized micro-blowing. The 32 × 32 array of micro-holes is arranged in a staggered pattern on the solid wall, located in the developed turbulent region. The porosity of the porous wall is 23%, and the blowing fraction is 0.0015. The Reynolds numb...

In the present work, an efficient active flow control strategy in eliminating vortex-induced vibration of a cylinder at Re = 100 has been explored by two machine learning frameworks, from active learning to reinforcement learning. Specifically, an adaptive control scheme by a pair of jets placed on the poles of the cylinder as actuators has been di...

Correlating combustion performance parameters to the main operating variables of combustors with mathematical expressions contributes to reducing the number of experiments and simplifying the design procedure of gas turbines. The application of empirical formulations meets the requirement with finite precision. The present study aims at adopting sy...

Using axisymmetric nozzle flow field as the reference flow field， and nonlinear truncation and offset was carried out on the nozzle， a new three-dimensional streamline tracking truncated offset hypersonic nozzle was designed by means of streamline tracking and multi-objective surrogated-based optimization method．Numerical simulations were carried o...

The moving particle semi-implicit (MPS) method and GPU parallel acceleration technology were carried out to simulate the two jets impinging atomization．The influence of the change of jet diameter ratio and momentum ratio on the atomization characteristic parameters was analyzed．The simulations successfully capture the primary atomization process of...

In the present study, an effective optimization framework of aerodynamic shape design is established based on the multi-fidelity deep neural network (MFDNN) model. The objective of the current work is to construct a high-accuracy multi-fidelity surrogate model correlating the configuration parameters of an aircraft and its aerodynamic performance b...

In this work, a new concept of annular combustor with 16 oblique-injecting swirling burners is innovatively proposed to induce a gyratory flow motion for better circumferential mixing and ignition performance. The cold flow fields were obtained by the PIV method, which clearly verified the existence of the gyratory flow motion. The light-round sequ...

In conventional linear parameter-varying (LPV) modelling and gain scheduling control design for turbojet engines, the linearisation is performed at a set of equilibrium points, and the validity of such LPV models is ensured near the equilibria. However, the linear model can only provide an approximate description of the engine’s state when the syst...

In this paper, numerical improvements are implemented for solving for the pressure in the moving particle semi-implicit (MPS) method for free-surface flow simulations. The tensile instability problem is solved using a dynamic stabilization (DS) algorithm. The low numerical diffusion of this algorithm is shown through numerical tests. A free-surface...

Routing design of aircraft Electrical Wiring Interconnection System (EWIS) is time-consuming and error-prone. A solution, which automatically routes the EWIS inside the aircraft Digital MockUp (DMU), has been proposed and presented in the previous publications. The DMU, however, includes over-detailed features, which hardly influence the routing re...

Since the small initial variation amplitude of soft fault leads to the low detection efficiency of fault detection algorithm based on traditional neural network observer，a soft fault detection algorithm for Flapping Wing Micro Aerial Vehicle(FWMAV)based on multistep neural network observer and adaptive threshold was proposed. Firstly，a multistep pr...

As an indirect noise source generated in the combustion chamber, entropy waves are widely prevalent in modern gas turbines and aero-engines. In the present work, the influence of entropy waves on the downstream flow field of a turbine guide vane is investigated. The work is mainly based on a well-known experimental configuration called LS89. Two di...

In this contribution, we described a Dynamic Mode Decomposition (DMD) analysis of a turbulent channel flow database at a moderate friction Reynolds number Reτ ≈ 950. More specifically, a composite-based DMD analysis was conducted, employing hybrid snapshots assembled by skin friction Cf (tk ) and either instantaneous Reynolds stress u'v'(x; t_k) or...

Vortex identification is important for understanding the physical mechanism of turbulent flow. The common vortex identification techniques based on velocity gradient tensor such as [Formula: see text] criterion will consume a lot of computing resources for processing great quantity of experimental data. To improve the vortex identification efficien...

Synthetic Jet Actuator (SJA) works cyclically with the directionally transportation of fluid near the exit, and the paper presents the performance of a loudspeaker driven SJA in static flow field. Time-Resolved Particle Image Velocimetry (TR-PIV) system is used to measure the flow field characteristics near the SJA slot exit with input voltage chan...

Heat transfer is an important phenomenon that exists in many industrial applications, especially for gas turbines, aeronautical engines. In this work, two different turbulence models ([Formula: see text] and SAS model) are used to investigate the effects of inlet turbulence on wall heat transfer and the characteristics of flow field in a well-known...

In this article, the Stackelberg game strategy is coupled with the adjoint method and applied in single-objective and two-objective aerodynamic shape optimizations. In the proposed method, two types of player (leader and follower) are involved, and each of these players is responsible for the optimization of one objective function by adjusting a su...

This paper constructs an optimization framework based on the data-informed self-adaptive quasi-steady model. The framework aims at achieving a specific aerodynamic force coefficient by optimizing the kinematic parameters of the flapping motion of an ellipsoid wing. All the model coefficients of this quasi-steady model are calibrated empirically by...

Dynamic Mode Decomposition (DMD) techniques have risen as prominent feature identification methods in the field of fluid dynamics. Any of the multiple variables of the DMD method allows to identify meaningful features from either experimental or numerical flow data on a data-driven manner. Performing a DMD analysis requires handling matrices V∈Rnp×...

The ignition process of an annular combustor with multiple injectors was investigated by experiments and simulations. In the numerical simulation, the unsteady Reynolds average Navier-Stokes (RANS) method, adaptive mesh refinement (AMR) and University of California, San Diego (UCSD) detailed chemistry were applied. The simulation predicted light-ro...