Cheng Cheng
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Cheng Cheng has a total of 42 co-authors. Here are the top co-authors by number of shared publications:
Xiaobing Zhang
School of Energy and Power Engineering Nanjing University of Science and Technology Nanjing 210094 China
9 shared publications- Xiaobing Zhang
School of Energy and Power Engineering, Nanjing University of Science and Technology, Nanjing, China
7 shared publications Changli Hu
3 shared publicationsChong Wang
School of Energy and Power Engineering Nanjing University of Science and Technology Nanjing 210094 China
3 shared publicationsGenghui Jiang
3 shared publicationsXi Huang
School of Energy and Power Engineering Nanjing University of Science and Technology Nanjing 210094 China
3 shared publications
Latest Publications
- Open Access
- 20 October 2026
- Published by Elsevier in SSRN Electronic Journal
Physics-Aware Deep Reduced-Order Modeling for Rapid Reconstruction of Temperature Fields in Supersonic Gas Jets
Rapid reconstruction of steady mean temperature fields in supersonic gas jets is challenging owing to nonlinear variations across nozzle-expansion regimes, practical constraints on experimental diagnostics, and the high computational cost of CFD for online monitoring. This study develops a physics-aware deep reduced-order modeling framework, termed PA-DeepROM. A physics-stratified database of 200 CFD cases covers five nozzle-expansion regimes. A proper orthogonal decomposition (POD) basis constructed exclusively from training data provides a low-dimensional representation, while wall-constrained QR pivoting selects sparse sensor locations. A deep residual network (Deep-ResNet) combines near-wall temperature observations, operating-condition features, and a regularized Gappy POD prior to learn nonlinear modal corrections. Joint selection of the POD rank and sensor budget yields a nominal configuration with 15 modes and two sensors. Across 31 independent test cases, PA-DeepROM achieves a near-field root-mean-square error (RMSE) of 36.835 K, a mean relative error (MRE) of 2.582%, and a coefficient of determination (R2) of 0.9853. Its RMSE is 19.5% and 57.7% lower than those of POD-BPNN and Gappy POD, respectively. Sensor-uncertainty tests demonstrate robustness to Gaussian measurement noise but reveal sensitivity to sensor-channel failure and position perturbation. Leave-one-regime-out tests show effective generalization to intermediate regimes but degraded performance for uncovered extreme regimes. Reconstruction of the complete 68,161-node temperature field requires an average CPU time of 1.320 ms. These findings support accurate, real-time temperature-field reconstruction within the sampled operating envelope for online thermal-state monitoring in propulsion systems.
- 20 October 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessPhysics-Aware Deep Reduced-Order Modeling for Rapid Reconstruction of Temperature Fields in Supersonic Gas Jets
Rapid reconstruction of steady mean temperature fields in supersonic gas jets is challenging owing to nonlinear variations across nozzle-expansion regimes, practical constraints on experimental diagnostics, and the high computational cost of CFD for online monitoring. This study develops a physics-aware deep reduced-order modeling framework, termed PA-DeepROM. A physics-stratified database of 200 CFD cases covers five nozzle-expansion regimes. A proper orthogonal decomposition (POD) basis constructed exclusively from training data provides a low-dimensional representation, while wall-constrained QR pivoting selects sparse sensor locations. A deep residual network (Deep-ResNet) combines near-wall temperature observations, operating-condition features, and a regularized Gappy POD prior to learn nonlinear modal corrections. Joint selection of the POD rank and sensor budget yields a nominal configuration with 15 modes and two sensors. Across 31 independent test cases, PA-DeepROM achieves a near-field root-mean-square error (RMSE) of 36.835 K, a mean relative error (MRE) of 2.582%, and a coefficient of determination (R2) of 0.9853. Its RMSE is 19.5% and 57.7% lower than those of POD-BPNN and Gappy POD, respectively. Sensor-uncertainty tests demonstrate robustness to Gaussian measurement noise but reveal sensitivity to sensor-channel failure and position perturbation. Leave-one-regime-out tests show effective generalization to intermediate regimes but degraded performance for uncovered extreme regimes. Reconstruction of the complete 68,161-node temperature field requires an average CPU time of 1.320 ms. These findings support accurate, real-time temperature-field reconstruction within the sampled operating envelope for online thermal-state monitoring in propulsion systems.
- Open Access
- 14 October 2026
- Published by Elsevier in SSRN Electronic Journal
Experimental Investigation of Liquid Resurgence and Splashing under Supersonic Gas-Jet Impingement on a Free Surface
High-speed visualization and quantitative image analysis were used to investigate cavity evolution, liquid resurgence, and splashing during sustained supersonic gas-jet impingement on a free surface. Four stages were identified: impingement-driven cavity expansion, cavity necking and liquid resurgence, splash rise and nozzle wall impact, and oscillation decay. Following upper-cavity necking, surrounding liquid converged and surged upward, forming liquid tongues and crowns that fragmented, with some liquid striking the nozzle wall. Increasing gas supply pressure enhanced axial cavity penetration and lateral spreading, increasing the observed maxima of dimensionless cavity depth, cavity diameter, and splash diameter by 82.4%, 60.1%, and 17.4%, respectively, whereas maximum splash height varied non-monotonically. Increasing the initial nozzle-to-surface distance shifted the cavity from deep and narrow to shallow and wide, reducing maximum cavity depth by 38.6%. Maximum splash diameter increased, while maximum splash height first decreased and then increased. Later-stage lateral spreading continued despite decreasing splash height. Power spectral densities exhibited low-frequency dominance associated with non-stationary geometric evolution. Splash spectral peaks increased at the highest pressure and varied non-monotonically with nozzle-to-surface distance. These findings clarify the temporal relationship between cavity necking, liquid resurgence, and subsequent splashing, providing an experimental basis for investigating liquid impact on nozzle walls.
- 14 October 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessExperimental Investigation of Liquid Resurgence and Splashing under Supersonic Gas-Jet Impingement on a Free Surface
High-speed visualization and quantitative image analysis were used to investigate cavity evolution, liquid resurgence, and splashing during sustained supersonic gas-jet impingement on a free surface. Four stages were identified: impingement-driven cavity expansion, cavity necking and liquid resurgence, splash rise and nozzle wall impact, and oscillation decay. Following upper-cavity necking, surrounding liquid converged and surged upward, forming liquid tongues and crowns that fragmented, with some liquid striking the nozzle wall. Increasing gas supply pressure enhanced axial cavity penetration and lateral spreading, increasing the observed maxima of dimensionless cavity depth, cavity diameter, and splash diameter by 82.4%, 60.1%, and 17.4%, respectively, whereas maximum splash height varied non-monotonically. Increasing the initial nozzle-to-surface distance shifted the cavity from deep and narrow to shallow and wide, reducing maximum cavity depth by 38.6%. Maximum splash diameter increased, while maximum splash height first decreased and then increased. Later-stage lateral spreading continued despite decreasing splash height. Power spectral densities exhibited low-frequency dominance associated with non-stationary geometric evolution. Splash spectral peaks increased at the highest pressure and varied non-monotonically with nozzle-to-surface distance. These findings clarify the temporal relationship between cavity necking, liquid resurgence, and subsequent splashing, providing an experimental basis for investigating liquid impact on nozzle walls.
- 4 April 2026
- Published by Elsevier in Ocean Engineering
- Vol. 356
Numerical investigation on consecutive separation dynamics of multiple cylindrical bodies at water surface
- 4 April 2026
- Published by Elsevier in Ocean Engineering
- Vol. 356
Numerical investigation on consecutive separation dynamics of multiple cylindrical bodies at water surface
- Open Access
- 27 November 2025
- Published by MDPI AG in Applied Sciences
- Vol. 15 (23) , 12561
Improved Metaball Discrete Element Method with Robust Contact Algorithm for General-Shaped Particles with Rounded Features
Particle shape is a critical factor governing the macroscopic behavior of granular materials. While the Discrete Element Method (DEM) has proven to be a powerful tool for modeling these materials, the accurate representation of particles with realistic geometries remains a significant challenge. Numerous shape description techniques have been developed to address this, including sphere-clustering, polyhedral, spheropolyhedral, and superquadric approaches. However, these methods often introduce artificial surface roughness or are limited to highly regular geometries, particularly when modeling general-shaped particles with rounded features. The proposed method reformulates collision detection in the Metaball Discrete Element Method (MDEM) as a constrained optimization problem, ensuring stable and physically consistent contact handling. Validation results show excellent agreement with analytical solutions and significantly improved robustness over unconstrained MDEM approaches. Finally, the model is applied to analyze the effects of particle shape on compaction and segregation, demonstrating that the proposed MDEM effectively captures the influence of complex shapes on macroscopic behavior and offers strong potential for engineering and scientific applications.
- 27 November 2025
- Published by MDPI AG in Applied Sciences
- Vol. 15 (23) , 12561
Open AccessImproved Metaball Discrete Element Method with Robust Contact Algorithm for General-Shaped Particles with Rounded Features
Particle shape is a critical factor governing the macroscopic behavior of granular materials. While the Discrete Element Method (DEM) has proven to be a powerful tool for modeling these materials, the accurate representation of particles with realistic geometries remains a significant challenge. Numerous shape description techniques have been developed to address this, including sphere-clustering, polyhedral, spheropolyhedral, and superquadric approaches. However, these methods often introduce artificial surface roughness or are limited to highly regular geometries, particularly when modeling general-shaped particles with rounded features. The proposed method reformulates collision detection in the Metaball Discrete Element Method (MDEM) as a constrained optimization problem, ensuring stable and physically consistent contact handling. Validation results show excellent agreement with analytical solutions and significantly improved robustness over unconstrained MDEM approaches. Finally, the model is applied to analyze the effects of particle shape on compaction and segregation, demonstrating that the proposed MDEM effectively captures the influence of complex shapes on macroscopic behavior and offers strong potential for engineering and scientific applications.
- 1 August 2025
- Published by AIP Publishing in Physics of Fluids
- Vol. 37 (8)
The influence of crossflow velocities on hydrodynamic characteristics of the underwater vehicle with swing nozzle
Swing nozzle, as one of the underwater thrust vector control technologies, is a significant method to enhance the maneuverability and anti-interference ability of the underwater vehicles. In this paper, the Dynamic Fluid Body Interaction method coupled with Shear Stress Transport k–ω model, Volume of Fluid method, and Proportional-Derivative control method are used to simulate the underwater motion stage of the vehicle controlled by the swing nozzle under different crossflow velocities. Compared with typical experimental results, it is found that the numerical method can effectively capture the evolution of the tail cavity bubble and the thrust. As the crossflow velocity increases, the deviation of the tail cavity bubble from the symmetry axis of the underwater vehicle becomes more pronounced. Additionally, the reentrant flow exhibits a more significant deviation from the nozzle axis. This phenomenon contributes to the reduction of pressure fluctuations on the bottom of the underwater vehicle and mitigates transient impacts during the initial stage of ignition. Following ignition, the shock wave in the tail cavity bubble causes alternating positive and negative pressure differentials at the upstream and downstream surfaces of the underwater vehicle's tail section. It is an important factor causing the underwater vehicle's moment oscillation. It is also found that greater crossflow velocities can induce substantial pitch moment on the underwater vehicle, and the optimization of control parameters is required to enhance control system robustness.
- 1 August 2025
- Published by AIP Publishing in Physics of Fluids
- Vol. 37 (8)
The influence of crossflow velocities on hydrodynamic characteristics of the underwater vehicle with swing nozzle
Swing nozzle, as one of the underwater thrust vector control technologies, is a significant method to enhance the maneuverability and anti-interference ability of the underwater vehicles. In this paper, the Dynamic Fluid Body Interaction method coupled with Shear Stress Transport k–ω model, Volume of Fluid method, and Proportional-Derivative control method are used to simulate the underwater motion stage of the vehicle controlled by the swing nozzle under different crossflow velocities. Compared with typical experimental results, it is found that the numerical method can effectively capture the evolution of the tail cavity bubble and the thrust. As the crossflow velocity increases, the deviation of the tail cavity bubble from the symmetry axis of the underwater vehicle becomes more pronounced. Additionally, the reentrant flow exhibits a more significant deviation from the nozzle axis. This phenomenon contributes to the reduction of pressure fluctuations on the bottom of the underwater vehicle and mitigates transient impacts during the initial stage of ignition. Following ignition, the shock wave in the tail cavity bubble causes alternating positive and negative pressure differentials at the upstream and downstream surfaces of the underwater vehicle's tail section. It is an important factor causing the underwater vehicle's moment oscillation. It is also found that greater crossflow velocities can induce substantial pitch moment on the underwater vehicle, and the optimization of control parameters is required to enhance control system robustness.
- Open Access
- 1 April 2025
- Published by KEAI Publishing in Defence Technology
- Vol. 46, 120-132
Numerical investigation on the engraving process of a pyrotechnic actuator with an improved two-phase flow model of interior ballistic
- 1 April 2025
- Published by KEAI Publishing in Defence Technology
- Vol. 46, 120-132
Open AccessNumerical investigation on the engraving process of a pyrotechnic actuator with an improved two-phase flow model of interior ballistic
- Open Access
- 27 August 2024
- Published by Wiley in International Journal of Mechanical System Dynamics
- Vol. 4 (3) , 317-330
Machine learning and numerical investigation on drag coefficient of arbitrary polygonal particles
The drag coefficient, as the most important parameter that characterizes particle dynamics in flows, has been the focus of a large number of investigations. Although good predictability is achieved for simple shapes, it is still challenging to accurately predict drag coefficient of complex‐shaped particles even under moderate Reynolds number (Re). The problem is that the small‐scale shape details of particles can still have considerable impact on the drag coefficient, but these geometrical details cannot be described by single shape factor. To address this challenge, we leverage modern deep‐learning method's ability for pattern recognition, take multiple shape factors as input to better characterize particle‐shape details, and use the drag coefficient as output. To obtain a high‐precision data set, the discrete element method coupled with an improved velocity interpolation scheme of the lattice Boltzmann method is used to simulate and analyze the sedimentation dynamics of polygonal particles. Four different machine‐learning models for predicting the drag coefficient are developed and compared. The results show that our model can well predict the drag coefficient with an average error of less than 5% for particles. These findings suggest that data‐driven models can be an attractive option for the drag‐coefficient prediction for particles with complex shapes.
- 27 August 2024
- Published by Wiley in International Journal of Mechanical System Dynamics
- Vol. 4 (3) , 317-330
Open AccessMachine learning and numerical investigation on drag coefficient of arbitrary polygonal particles
The drag coefficient, as the most important parameter that characterizes particle dynamics in flows, has been the focus of a large number of investigations. Although good predictability is achieved for simple shapes, it is still challenging to accurately predict drag coefficient of complex‐shaped particles even under moderate Reynolds number (Re). The problem is that the small‐scale shape details of particles can still have considerable impact on the drag coefficient, but these geometrical details cannot be described by single shape factor. To address this challenge, we leverage modern deep‐learning method's ability for pattern recognition, take multiple shape factors as input to better characterize particle‐shape details, and use the drag coefficient as output. To obtain a high‐precision data set, the discrete element method coupled with an improved velocity interpolation scheme of the lattice Boltzmann method is used to simulate and analyze the sedimentation dynamics of polygonal particles. Four different machine‐learning models for predicting the drag coefficient are developed and compared. The results show that our model can well predict the drag coefficient with an average error of less than 5% for particles. These findings suggest that data‐driven models can be an attractive option for the drag‐coefficient prediction for particles with complex shapes.
- 18 June 2024
- Published by Springer Nature
- p. 3833-3852
Comparison of the Flow Field and Thrust Characteristics of Solid Rocket Motor with Conventional Nozzle and Canted Nozzle
- 18 June 2024
- Published by Springer Nature
- p. 3833-3852
Comparison of the Flow Field and Thrust Characteristics of Solid Rocket Motor with Conventional Nozzle and Canted Nozzle
- 21 November 2022
- Published by Wiley in Propellants, Explosives, Pyrotechnics
- Vol. 48 (2)
Numerical Investigation on the Gas‐Solid Flow in Blast Flow Fields Induced by a High‐Speed Body
As one of the critical phenomena during the firing of guns, propellant combustion creates an expanding gas plume with smoke containing the combustion residues. This paper aims to clarify the influence of gunshot residue particles on muzzle flow. Firstly, a two‐dimensional gas‐solid two‐phase model is established to describe the muzzle flow with gunshot residue for a small caliber rarefaction wave gun. And a discrete phase model is used to track particle trajectories for the dilute gas‐solid flow. Then the different development processes are discussed between the pure‐gas and the gas‐solid flow. Finally, the particle diameter and the mass fraction of particles are investigated to show the effects on the development process of the muzzle flow with gunshot resides. The results show that particles greatly influence the main flow parameters in the muzzle flow with gunshot residue. The gas velocity decreases due to the interaction between the particles and the gas. This paper provides a valuable reference for the study of muzzle flow.
- 21 November 2022
- Published by Wiley in Propellants, Explosives, Pyrotechnics
- Vol. 48 (2)
Numerical Investigation on the Gas‐Solid Flow in Blast Flow Fields Induced by a High‐Speed Body
As one of the critical phenomena during the firing of guns, propellant combustion creates an expanding gas plume with smoke containing the combustion residues. This paper aims to clarify the influence of gunshot residue particles on muzzle flow. Firstly, a two‐dimensional gas‐solid two‐phase model is established to describe the muzzle flow with gunshot residue for a small caliber rarefaction wave gun. And a discrete phase model is used to track particle trajectories for the dilute gas‐solid flow. Then the different development processes are discussed between the pure‐gas and the gas‐solid flow. Finally, the particle diameter and the mass fraction of particles are investigated to show the effects on the development process of the muzzle flow with gunshot resides. The results show that particles greatly influence the main flow parameters in the muzzle flow with gunshot residue. The gas velocity decreases due to the interaction between the particles and the gas. This paper provides a valuable reference for the study of muzzle flow.
- 8 November 2022
- Published by Wiley in Propellants, Explosives, Pyrotechnics
- Vol. 47 (12)
A New Method to Correct Interior Ballistic Characteristics of Multi‐Projectile System Based on Rarefaction Wave Theory
One of the most significant challenges in designing a hyper‐firing‐rate weapon is to balance the interior ballistics behaviors of a multi‐projectile system. This paper proposes a multi‐projectile system with a front nozzle to correct interior ballistic characteristics based on rarefaction wave theory. Firstly, a two‐dimensional two‐phase flow model for the hyper‐firing‐rate weapon is established. Then the application to a standard virtual gun as a benchmark for interior ballistic codes is used to validate the accuracy. In addition, the interior ballistics behaviors of the three‐projectile system without and with the front nozzle are investigated. Finally, the interior ballistics behaviors of the three‐projectile system are corrected by changing the open time of the front nozzle. The results show that the new correction method for the multi‐projectile system reduces the recoil and the pressure while balancing the muzzle velocity. The finding guides future ballistic research on the multi‐projectile system.
- 8 November 2022
- Published by Wiley in Propellants, Explosives, Pyrotechnics
- Vol. 47 (12)
A New Method to Correct Interior Ballistic Characteristics of Multi‐Projectile System Based on Rarefaction Wave Theory
One of the most significant challenges in designing a hyper‐firing‐rate weapon is to balance the interior ballistics behaviors of a multi‐projectile system. This paper proposes a multi‐projectile system with a front nozzle to correct interior ballistic characteristics based on rarefaction wave theory. Firstly, a two‐dimensional two‐phase flow model for the hyper‐firing‐rate weapon is established. Then the application to a standard virtual gun as a benchmark for interior ballistic codes is used to validate the accuracy. In addition, the interior ballistics behaviors of the three‐projectile system without and with the front nozzle are investigated. Finally, the interior ballistics behaviors of the three‐projectile system are corrected by changing the open time of the front nozzle. The results show that the new correction method for the multi‐projectile system reduces the recoil and the pressure while balancing the muzzle velocity. The finding guides future ballistic research on the multi‐projectile system.