Zhihua Chen

Verified

Zhihua verified their affiliation via an institutional email.

Learn more

Verified

Zhihua verified their affiliation via an institutional email.

Learn more

• PhD
• Professor at Southern University of Science and Technology

The propagation of a flame is investigated experimentally and theoretically for a large, horizontal combustion tube containing a mixture of air and aluminum powder with pre-existing turbulence. One end of the tube is closed and the other is connected to a large dump-tank. Twenty dispersion systems are used on the tube to produce a uniform suspensio...

Numerical simulations are carried out to investigate the influence of geometric parameters on the flow and heat transfer characteristics of rectangular, trapezoidal and triangular shaped microchannel heat sinks. A good agreement is found between numerical results and experimental data when the variation of water viscosity with temperature is taken...

The three-dimensional flow characteristics of the compressible vortex ring generated by under-expanded circular jets with two typical pressure ratios, i.e. $n=1.4$ (moderate) and 4.0 (high), are investigated numerically with the use of large-eddy simulations. Our results illustrate that these two pressure ratios correspond to different shock struct...

The flow dynamics around tandem cylinders are complex, with significant engineering implications, especially in applications like high-rise buildings. This study presents a jet flow control framework for two tandem cylinders with a Reynolds number (Re) of 100, based on deep reinforcement learning. We compare two control strategies: (1) a single-age...

Conventional missile guidance and control systems are difficult to adapt to complex environments, and new missile guidance systems with fast and accurate decision-making capabilities and high robustness are urgently needed to meet the needs of current and future military and aerospace missions. This study presents a deep reinforcement learning (DRL...

The effects of droplet evaporation on turbulence characteristics are systematically investigated in a three-dimensional spatially developing supersonic mixing layer at the convective Mach number of 1.2 via direct numerical simulation. With the point-particle approach, the fluid–droplet interactions are achieved through the two-way coupling in the E...

To swiftly and precisely assess the IGBT’s thermal safety during the circuit design phase, this paper proposes a method of establishing an IGBT thermal network model by electrothermal coupling simulation and parameter identification. This model integrates electrothermal coupling simulation techniques and parameter identification methods. Firstly, t...

To investigate the inception of cavitation, the morphology of supercavity, and the detailed flow field influenced by the surface morphology of underwater high-speed projectiles, a high-precision large eddy simulation method was developed. The method accurately simulates cavitation inception and its fine structure under the influence of small-scale...

Computational efficiency and precision pose a classic contradiction in aerodynamic shape optimization. To address this challenge, this study introduces an effective optimization framework based on multi-fidelity fully connected neural network (MFFCN). The framework utilizes transfer learning (TL) to train a multi-fidelity surrogate model that estab...

This work proposes a novel surrogate model (noted as HCP-PIGN) combining two groups of neural networks: i.e., the physics-informed and the graph convolutional neural networks (noted as PINN and GCN). It aims to tackle the existing challenges: pixelated pre-processing of data and large amounts of training data. For predicting 2D steady-state heat co...

The oblique water entry of a hollow cylinder at various entry angles is numerically studied. The formation characteristics of the internal and external cavities, the curling splash, and the underwater rotation of the cylinder are revealed and analyzed. Our results show that asymmetric left- and right-attached cavities form near both the inner and o...

Early studies on Rayleigh-Taylor instability (RTI) primarily relied on the Navier-Stokes (NS) model. As research progresses, it becomes increasingly evident that the kinetic information that the NS model failed to capture is of great value for identifying and even controlling the RTI process; simultaneously, the lack of analysis techniques for comp...

The jet can be applied to the yaw control of a projectile, however, the complex interaction of the jet with the supersonic mainstream makes the flow field complex and yaw force unpredictable. To reveal the evolution of flow structures under different pressure ratios (PRs), or momentum flux ratios, a transverse sonic jet injected into a supersonic l...

n high latitudes during winter, the icing period is inevitable. Expanding submarine-launched marine equipment in a low-temperature ice environment holds significant engineering value. However, the existence of floating ice will unavoidably increase the nonlinearity of the submarine-launched vehicle during the water-exit process. The dynamic fluid i...

Deep reinforcement learning (DRL) has gradually emerged as a novel and effective method for intelligent control of conjugate heat transfer. Through proper training, DRL agent usually can find a better control strategy than the one optimized manually. For numerous numerical simulation-based investigations, the promising results have only been obtain...

The present study introduces a novel optimization framework that combines a Graph Convolutional Neural Network surrogate model with Genetic Algorithms (GCN-GA). This framework can optimize the heat sources layout in horizontal annuli, achieving the highest heat transfer efficiency. The predictive GCN surrogate model utilizes graph data as training...

Based on ideal compressible magnetohydrodynamics (MHD) equations, the interface instability induced by the interaction between planar shock wave and the lightw gas (Helium) cylinder under the influence of the magnetic field with different directions was investigated numerically using the CTU(Corner Transport Upwind)+CT (Constrained Transport) algor...

This paper presents a novel deep learning-based surrogate model for steady-state natural convection problem with variable geometry. Traditional deep learning based surrogate models are more or less limited by the requirement for large amounts of training data, loss of accuracy due to pixelization of the original data, the low accurate prediction ne...

With the shrink of critical dimensions of semiconductor devices to a few nanometers, atomic layer etching (ALE) has become an important technique to achieve single-atom resolution. ALE decouples plasma etching into two self-limiting reaction processes:passivation and etching processes, allowing for sequential removal of material atomic layer by lay...

In recent years, the realm of deep learning has witnessed significant advancements, particularly in object detection algorithms. However, the unique challenges posed by remote sensing images, such as complex backgrounds, diverse target sizes, dense target distribution, and overlapping or obscuring targets, demand specialized solutions. Addressing t...

To alleviate the computational burden associated with the computational fluid dynamics (CFD) simulation stage and improve aerodynamic optimization efficiency, this work develops an innovative procedure for airfoil shape optimization, which is implemented through coupling the genetic algorithm (GA) optimizer with the aerodynamic coefficients predict...

A three-dimensional numerical model is established to simulate the tandem water entry at different time intervals between two cylinders. The evolution of cavity, pressure field, velocity field and multi-scale vortex structures during water entry are studied. The motion characteristics of the tandem cylinders are analyzed to investigate the coupling...

This paper introduces a novel surrogate model for two-dimensional adaptive steady-state thermal convection fields based on deep learning technology. The proposed model aims to overcome limitations in traditional frameworks caused by network types, such as the requirement for extensive training data, accuracy loss due to pixelated preprocessing of o...

Spherical robots have fully wrapped shells, which enables them to walk well on complex terrains, such as swamps, grasslands and deserts. At present, path planning algorithms for spherical robots mainly focus on finding the shortest path between the initial position and the target position. In this paper, an improved A* algorithm considering energy...

This paper investigates the performance of several most popular deep reinforcement learning (DRL) algorithms applied to fluid flow and convective heat transfer systems, providing credible guidance and evaluation on their characteristics and performance. The studied algorithms are selected by considering the popularity, category, and advancement for...

To study the flow characteristics of the rapid mixing process of fuel and air in supersonic combustion chamber, this paper is based on the compressible Navier-Stokes equation, a sonic jet injected into a Mach 2.68 crossflow over a flat plate with a jet-to-crossflow pressure ratio of 36 is investigated numerically with the use of large eddy simulati...

In this paper，the Richtmyer—Meshkov instability is studied numerically by using the high resolution Roe scheme based on the two dimensional unsteady Euler equation，which is caused by the interaction between shock wave and the helium circular light gas cylinder with different component distributions．The numerical results are used to further discuss...

The Kelvin-Helmholtz instability (KHI) stems from the velocity shear in a single continuous fluid or a velocity difference across the interface between two distinct fluids. The effect of the transverse magnetic field on the KHI of the supersonic mixing layer is investigated by numerical method. An algorithm with corner-transport-upwind and constrai...

The flow of a weakly conductive fluid (i.e. seawater) can be controlled by Lorentz force, which holds promising applications in drag reduction. In this paper, the drag reduction on a stationary sphere in the weakly conductive fluid with the Lorentz force on the sphere surface are numerically investigated at Re = 300. The relations among the vortex...

To evaluate the damage assessment of high-power microwaves on complex electronic systems, a damage assessment method combining top-down system analysis and bottom-up comprehensive evaluation is proposed. A UAV damage tree model was constructed, and the FAHP-CRITIC method was used to assign weights to system components. The pulse energy transfer and...

This work proposes an innovative approach for supersonic flow field modeling around airfoils based on sparse convolutional neural networks (SCNN) and Bézier generative adversarial network (GAN), where 1) the SCNN model is built to end-to-end predict supersonic compressible physical flow fields around airfoils from spatially-sparse geometries and 2)...

In the current paper, the zero-mass synthetic jet flow control combined with a proximal policy optimization (PPO) algorithm in deep reinforcement learning is constructed, and a policy transfer strategy which is trained in two-dimensional (2D) environment and migrated to three-dimensional (3D) environment is proposed and analyzed. By policy, we mean...

To enhance the power gain of vortex-induced vibration of a circular cylinder, the active control method of pulsed blowing jets located at θ = 90° is utilized to intensify its oscillation with the two-dimensional simulation of Reynolds-averaged Navier–Stokes at 2.0 × 10 ⁴ ≤ Re ≤ 9.6 × 10 ⁴ . Different from traditional continuous jets, the blowing je...

In this paper, deep reinforcement learning (DRL) is applied on forced convection control of conjugate heat transfer systems governed by the coupled Navier-Stokes and heat transport equations. A novel value-based deep Q-network (DQN) integrating with three advanced techniques is utilized for identifying underlying heat and mass transfer mechanism in...

This study applies deep-reinforcement-learning algorithms to integrated guidance and control for three-dimensional, high-maneuverability missile-target interception. Dynamic environment, reward functions concerning multi-factors, agents based on the deep-deterministic-policy-gradient algorithm, and action signals with pitch and yaw fins as control...

This paper presents a novel reduced-order model for internal and external flow field estimations based on a sparse convolution neural network. Since traditional convolution neural network requires “rectangular” matrixes as input, the convolutional operation is computationally inefficient when applied to these problems with input matrix having spars...

The Richtmyer–Meshkov instability (RMI) is caused by an incident planar shock wave impinging on the heavy-gas-density interface. We have numerically investigated the RMI controlled by different transverse magnetic-field strengths based on the ideal compressible magnetohydrodynamics (MHD) equations. The MHD equations are solved by the corner transpo...

In this paper, the Richtmyer-Meshkov instability has been studied numerically by using the high resolution Roe scheme based on the two-dimensional unsteady Euler equation, which was caused by the interaction between shock wave and the helium circular light gas cylinder with different component distributions. The numerical results further discuss th...

The hollow projectile is a new type of projectile that has complex water entry hydrodynamics characteristics and has attracted significant attention in recent years. As such, it is important to investigate the effects of different entry velocities and aperture diameters on the cavity morphology, cavitation, dynamics, and motion characteristics of h...

The oblique water entry of hollow cylinders with four different apertures is investigated experimentally at low velocities. The evolution characteristics of cavities, jets and splashes with different initial conditions and apertures are disclosed and discussed. The results show that the transition of cavity closure patterns with increasing water-en...

It is of great significance to study the hydrodynamic process of trans-media projectiles passing through ice-water mixture at high speeds for the application of weapons in low-temperature environments of high-latitude water areas during the winter. Based on the volume of fluid (VOF) method and the Schnerr-Sauer cavitation model, the processes of a...

We propose a steady-state aerodynamic data-driven method to predict the incompressible flow around airfoils of NACA (National Advisory Committee for Aeronautics) 0012-series. Using the Signed Distance Function (SDF) to parameterize the geometric and flow condition setups, the prediction core of the method is constructed essentially by a consecutive...

Based on the compressible Navier-Stokes equations combined with the fifth-order weighted essentially non-oscillatory scheme, this paper discusses the interaction of a planar shock wave with an annular SF 6 cylinder. The influence of the position and radius of inner cylinder on the evolution of the annular cylinder is examined in detail. Numerical s...

Natural convection of nanofluids in annular pipes has been investigated in many studies due to its high occurrence in heat transfer systems. Solving natural convection problems numerically itself requires numerical discretization and iteration of differential equations, and introduction of nanofluids makes the solution process more complicated. In...

Synthetic aperture radar (SAR) ship detection has been the focus of many previous studies. Traditional SAR ship detectors face challenges in complex environments due to the limitations of manual feature extraction. With the rise of deep learning (DL) techniques, SAR ship detection based on convolutional neural networks (CNNs) has achieved significa...

In this paper, the fully developed turbulent channel flow with sinusoidal deformation controlled by the space-dependent electromagnetic forces is studied numerically at Re = 4000. The expanded Fourier-Chebyshev spectral method with coordinate transformation is applied to solve the turbulent flow with the Direct Numerical Simulation (DNS) method. Th...

In the interdisciplinary field of data-driven models and computational fluid mechanics, the reduced-order model for flow field prediction is mainly constructed by a convolutional neural network (CNN) in recent years. However, the standard CNN is only applicable to data with Euclidean spatial structure, while data with non-Euclidean properties can o...

By direct numerical simulations, the non-isothermal effects on turbulence anisotropy and growth rate are investigated in the three-dimensional spatially developing supersonic mixing layers with high convective Mach numbers ( M c > 0.6). Hot air is blown in the high-speed stream, and cold air is added on the low-speed side. Two non-isothermal simula...

The two-dimensional Rayleigh-Taylor instability (RTI) in compressible flow with intermolecular interactions is probed via the discrete Boltzmann method. The effects of interfacial tension, viscosity, and heat conduction are investigated. It is found that the influences of interfacial tension on the perturbation amplitude, bubble velocity, and two k...

In this paper, an experimental method is firstly applied to study the effect of time interval on cavity evolution characteristics of two tandem spheres entering water vertically. The water-entry process of is analyzed for different cases by defining the Froude number Fr, which is determined by velocity and time interval. It is found that there are...

The ice floating on water surfaces during winter significantly influences the water entry process of a cylinder. Therefore, it is crucial to evaluate the influence mechanism of the ice-water mixture during the water entry process. In this study, we used computational fluid dynamics (CFD) to simulate the process of a cylinder vertically entering the...

Deep reinforcement learning (DRL) has gradually emerged as an effective and novel method to achieve active flow control with outstanding performance. This paper focuses on exploring the strategy of improving learning efficiency and control performance of a new task using existing control experience. More specifically, the proximal policy optimizati...

Underwater supersonic gas jet flows are sophisticated physical phenomena that widely exist in underwater propulsion, such as underwater rocket, high-speed torpedo, and water-piercing missile launcher (WPML). The present study aims to explore the characteristics of underwater supersonic jets. This research carries out two-dimensional and three-dimen...

In this paper, an intelligent guidance law based on Deep Q Network (DQN) algorithm is proposed, for enabling the missile to intercept different maneuvering targets following the idea of the parallel-approach method. In specific, we propose the inverse ratio of the absolute value of line-of-sight (LOS) angle rate as the shaping reward function which...

The Rayleigh-Taylor Instability (RTI) in compressible flow with inter-molecular interactions is probed via the Discrete Boltzmann Method (DBM). The effects of interfacial tension, viscosity and heat conduction are investigated. It is found that the influences of interfacial tension on the perturbation amplitude, bubble velocity, and two kinds of en...

This paper applies deep reinforcement learning (DRL) on the synthetic jet control of flows over an NACA (National Advisory Committee for Aeronautics) 0012 airfoil under weak turbulent condition. Based on the proximal policy optimization method, the appropriate strategy for controlling the mass rate of a synthetic jet is successfully obtained at [Fo...

Based on Navier-Stokes equations, combined with fifth-order weighted essentially non-oscillatory scheme and adaptive structured grid refinement technique, interactions between converging shock and annular SF6 layers with different initial perturbation amplitudes and thickness are numerically investigated. The evolution mechanism of shock structures...

In this paper, the modes of vortex shedding in the wake of a stationary sphere with the Lorentz force, which can be generated by the actuators in weakly conductive fluids and is parallel to the sphere surface, are numerically investigated at Re = 300. The relations among the wake structures, the vorticity distribution, and the motions of the rear s...

We develop a deep neural network-based reduced-order model (ROM) for rapid prediction of the steady-state velocity field with arbitrary geometry and various boundary conditions. The input matrix of the network is composed of the nearest wall signed distance function (NWSDF), which contains more physical information than the signed distance function...

Heat convection is one of the main mechanisms of heat transfer, and it involves both heat conduction and heat transportation by fluid flow; as a result, it usually requires numerical simulation for solving heat convection problems. Although the derivation of governing equations is not difficult, the solution process can be complicated and usually r...

This paper proposes a novel guidance law for intercepting a high-speed maneuvering target based on deep reinforcement learning, which mainly includes the interceptor–target relative motion model and value function approximation model based on deep Q-Network (DQN) with prioritized experience replay. First, a method called prioritized experience repl...

A data-driven model for rapid prediction of the steady-state heat conduction of a hot object with arbitrary geometry is developed. Mathematically, the steady-state heat conduction can be described by the Laplace's equation, where a heat (spatial) diffusion term dominates the governing equation. As the intensity of the heat diffusion only depends on...

To ensure that the missile is safely separated from the internal weapons bay, the jet is used to control the process of missile separation, which is mounted on the front edge of the bay. The length-to-depth ratio of the bay was L/D=8, the diameter of the missile was d1 =0.178 m, the diameter of the jet was d2 =0.05 m . The FLUENT software was combi...

The modifications of thermodynamic fluctuations by inertial particles are investigated in decaying compressible isotropic turbulence with an initial turbulent Mach number of 1.2 through direct numerical simulations. The particles interact with turbulence through two-way coupling under the Eulerian–Lagrangian point-source framework. Five simulations...

The development of lattice gas model to discrete Boltzmann method is briefly introduced for modeling multiphase complex fluid systems. Based on the basic principles of statistical physics, the Boltzmann equation is given through the idea of coarse-grained modeling. The physical images of progressively refined measurements contained in the Chapman-E...

The vertical water-entry processes of a hollow cylinder with three different velocities are investigated both numerically and experimentally. Three different closure patterns and flow characteristics of the air-entraining cavities induced by water-entry processes are discovered and discussed. Our results show that three different closure patterns a...

In this paper, based on neural networks, we develop a data-driven model for extremely fast prediction of steady-state heat convection of a hot object with arbitrary complex geometry in a two-dimensional space. According to the governing equations, the steady-state heat convection is dominated by convection and thermal diffusion terms, thus the dist...

The evolution of the Kelvin-Helmholtz (KH) instability in the presence of classical hydrodynamics and magneto-hydro-dynamics is investigated numerically by using the magneto-hydro-dynamic (MHD) equations. The MHD equations are solved with the corner transport upwind plus constrained transport algorithm that guarantees the divergence-free constraint...

In this article, we propose an unsteady data-driven reduced order model (ROM) (surrogate model) for predicting the velocity field around an airfoil. The network model applies a convolutional neural network (CNN) as the encoder and a deconvolutional neural network (DCNN) as the decoder. The model constructs a mapping function between temporal evolut...

In this paper, we propose a neural network based reduced-order model for predicting the unsteady flow field over single/multiple cylinders. The neural network model constructs a mapping function between the temporal evolution of the pressure signal on the cylinder surface and the surrounding velocity field, where Convolutional Neural Network (CNN)...

The Richtmyer-Meshkov instability (RMI) generated by the shock wave interacts with a density interface in the presence of a transverse magnetic field is investigated numerically using the ideal compressible Magneto-hydro-dynamic (MHD) equations. The MHD equations are solved with CTU+CT (Corner Transport Upwind + Constrained Transport) algorithm whi...

Numerical simulation of steady-state heat conduction is common for thermal engineering. The simulation process usually involves mathematical formulation, numerical discretization and iteration of discretized ordinary or partial differential equations depending on complexity of problems. In current work, we develop a data-driven model for extremely...

Previous research studies indicate that the proportion of Magnus force at the spinning projectile tail position is very high. Meanwhile, the large mutations of aerodynamic characteristics are found after adding boattail structures. In order to study the influences of boattail structures on aerodynamics of a spinning projectile, a 6.37-diameter long...

The shallow-angle oblique water-entry of a high-speed supercavitating projectile is studied by using the high-speed photography technology. The characteristics of the ballistic trajectory, splash formation and underwater pressure wave propagation during the initial water-entry impact at different sideslip angles are observed and analyzed from the u...

Based on the ideal magneto-hydro-dynamic equations (MHD) and adopting the CTU (Corner Transport Upwind) + CT (Constrained Transport) algorithm, the Richtmyer-Meshkov (RM) instability of the Gaussian distribution R22 gas column generated by the interaction of a planar shock wave with the gas column in the presence of magnetic field is investigated n...

Fluid‐structure interactions (FSI) of rigid and flexible bodies are simulated in this article. For the fluid flow, multi‐direct forcing immersed boundary method (IBM) is adopted to capture the moving boundary and Lattice Boltzmann Method (LBM) is used to evolve the flow field. Compared to our previous no‐penetration IBM, less iterations are require...

The vertical water-entry of a low-speed hollow cylinder is studied numerically and experimentally in this work. The special flow characteristics of the hollow cylinder during the water entry have been revealed and discussed. Different from water entry of typical solid bodies, a through-hole jet with a bubble attached at the top forms at the impact...