Tong Zhu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Tong Zhu has a total of 79 co-authors. Here are the top co-authors by number of shared publications:
- Chenzhen Ji
College of Automotive and Energy Engineering, Tongji University 1 , 4800 Cao'an Road, Shanghai 201804
12 shared publications - Deng Pan
School of Mechanical Engineering, Tongji University, Shanghai 201804, China
4 shared publications Enlei Ke
College of Automotive and Energy Engineering, Tongji University 1 , 4800 Cao'an Road, Shanghai 201804
4 shared publications- Jiaying Cheng
Shanghai Marine Equipment Research Institute, Shanghai 200031, China
4 shared publications Chenzhen Ji
3 shared publicationsDeng Pan
3 shared publications
Latest Publications
- Open Access
- 23 June 2026
- Published by Elsevier in SSRN Electronic Journal
Experimental Investigation of Mode Switching Mechanisms and Control Strategies for Combustion Instabilities in a Rich-Lean Gas Water Heater
This study experimentally investigated mode switching of combustion instabilities in a richâlean gas water heater under varying thermal inputs (18 kW, 24 kW, 30 kW) and flue pipe lengths (1.3 m, 1.8 m, 2.3 m). Pressure measurements, highâspeed CH* chemiluminescence imaging, Proper Orthogonal Decomposition (POD), and thermoacoustic simulations were employed. As the excess air coefficient decreases, the system abruptly switched from separated flames with a dominant frequency around 50 Hz and weak pressure amplitude to merged flames at 75â77 Hz with intense oscillations reaching nearly 500 Pa. The convective time delay, derived from the heatârelease centroid of the first POD mode, remains stable for separated flames, increased sharply when flames merge, and then gradually into the region of maximum growth rate, triggering selfâsustained oscillations. Flue pipe length variation revealed that the highâfrequency mode is a conventional thermoacoustic mode coupled to the combustor acoustics, whereas the lowâfrequency mode is an intrinsic thermoacoustic (ITA) mode insensitive to combustor geometry. Gain experiments at different thermal inputs confirm that merged flames are more prone to instability, although sufficiently high gain could also drive separated flames into oscillation. To suppress the observed instability, a passive mitigation strategy was proposed: increasing the gap between the rich and lean flow passages to prevent flame merging. This simple geometric modification effectively eliminates intense thermoacoustic oscillations and provided a practical solution for stable operation.
- 23 June 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessExperimental Investigation of Mode Switching Mechanisms and Control Strategies for Combustion Instabilities in a Rich-Lean Gas Water Heater
This study experimentally investigated mode switching of combustion instabilities in a richâlean gas water heater under varying thermal inputs (18 kW, 24 kW, 30 kW) and flue pipe lengths (1.3 m, 1.8 m, 2.3 m). Pressure measurements, highâspeed CH* chemiluminescence imaging, Proper Orthogonal Decomposition (POD), and thermoacoustic simulations were employed. As the excess air coefficient decreases, the system abruptly switched from separated flames with a dominant frequency around 50 Hz and weak pressure amplitude to merged flames at 75â77 Hz with intense oscillations reaching nearly 500 Pa. The convective time delay, derived from the heatârelease centroid of the first POD mode, remains stable for separated flames, increased sharply when flames merge, and then gradually into the region of maximum growth rate, triggering selfâsustained oscillations. Flue pipe length variation revealed that the highâfrequency mode is a conventional thermoacoustic mode coupled to the combustor acoustics, whereas the lowâfrequency mode is an intrinsic thermoacoustic (ITA) mode insensitive to combustor geometry. Gain experiments at different thermal inputs confirm that merged flames are more prone to instability, although sufficiently high gain could also drive separated flames into oscillation. To suppress the observed instability, a passive mitigation strategy was proposed: increasing the gap between the rich and lean flow passages to prevent flame merging. This simple geometric modification effectively eliminates intense thermoacoustic oscillations and provided a practical solution for stable operation.
- 1 June 2026
- Published by AIP Publishing in Physics of Fluids
- Vol. 38  (6)
Numerical investigation on the control of multi-mode thermoacoustic instabilities by utilizing an embedded-necked Helmholtz resonator in a premixed swirl combustor
Lean premixed combustion is key to low-emission gas turbines and aero-engines, but its strong coupling between heat release and acoustics makes the system susceptible to self-excited thermoacoustic instabilities across various operating conditions. Helmholtz resonators are widely used to suppress multi-mode thermoacoustic instabilities but require further study. Compared with conventional designs, embedded-neck Helmholtz resonators offer a more compact solution. This study investigates the control of multi-mode thermoacoustic instabilities using parallel embedded-neck Helmholtz resonators in a premixed swirl combustor. Experiments reveal self-excited oscillations at 110 and 300 Hz under different equivalence ratio, which are accurately predicted by a three-dimensional Helmholtz solver coupled with a flame describing function. Using this validated numerical framework, the control of multi-mode thermoacoustic instabilities by parallel embedded-neck Helmholtz resonators is then systematically investigated. Acoustic and thermoacoustic simulations demonstrate that the parallel resonator achieves a broader effective bandwidth than a single resonator, enabling simultaneous suppression of both instability modes. The results further show that optimal placement of resonators at acoustic pressure antinodes is crucial for effective stabilization. Under optimal conditions, the growth rate of the dominant mode is reduced from 78.1 to â37.3 rad/s, leading to system stabilization. Furthermore, for first-order oscillations, the growth rate is significantly reduced from 1.45 to â10.36 rad/s, but only when the resonator is placed in the plenum at the pressure antinode. It is concluded that optimal control of thermoacoustic oscillations requires selecting Helmholtz resonators with eigenfrequencies tuned to the target oscillations and positioning them in regions of high acoustic pressure amplitude.
- 1 June 2026
- Published by AIP Publishing in Physics of Fluids
- Vol. 38  (6)
Numerical investigation on the control of multi-mode thermoacoustic instabilities by utilizing an embedded-necked Helmholtz resonator in a premixed swirl combustor
Lean premixed combustion is key to low-emission gas turbines and aero-engines, but its strong coupling between heat release and acoustics makes the system susceptible to self-excited thermoacoustic instabilities across various operating conditions. Helmholtz resonators are widely used to suppress multi-mode thermoacoustic instabilities but require further study. Compared with conventional designs, embedded-neck Helmholtz resonators offer a more compact solution. This study investigates the control of multi-mode thermoacoustic instabilities using parallel embedded-neck Helmholtz resonators in a premixed swirl combustor. Experiments reveal self-excited oscillations at 110 and 300 Hz under different equivalence ratio, which are accurately predicted by a three-dimensional Helmholtz solver coupled with a flame describing function. Using this validated numerical framework, the control of multi-mode thermoacoustic instabilities by parallel embedded-neck Helmholtz resonators is then systematically investigated. Acoustic and thermoacoustic simulations demonstrate that the parallel resonator achieves a broader effective bandwidth than a single resonator, enabling simultaneous suppression of both instability modes. The results further show that optimal placement of resonators at acoustic pressure antinodes is crucial for effective stabilization. Under optimal conditions, the growth rate of the dominant mode is reduced from 78.1 to â37.3 rad/s, leading to system stabilization. Furthermore, for first-order oscillations, the growth rate is significantly reduced from 1.45 to â10.36 rad/s, but only when the resonator is placed in the plenum at the pressure antinode. It is concluded that optimal control of thermoacoustic oscillations requires selecting Helmholtz resonators with eigenfrequencies tuned to the target oscillations and positioning them in regions of high acoustic pressure amplitude.
- 1 June 2026
- Published by AIP Publishing in Physics of Fluids
- Vol. 38  (6)
Investigation on topology optimization of flow and heat transfer in liquid-cooling channels for a blade-type battery with non-uniform heat generation
Blade-type lithium-ion batteries are widely used due to their high structural integration and superior volumetric energy density. However, the high aspect ratio and meter-scale spacing between electrode tabs lead to pronounced non-uniform heat generation during charging, resulting in localized thermal accumulation and imposing stringent requirements on coolant transport. To address this issue, the present study proposes a topology-optimized (TO) liquid-cooling plate with irregular flow channels tailored to the non-uniform thermal characteristics of blade batteries. An electrochemicalâthermal coupled model is established to capture spatial heat generation under charging conditions, revealing that the heat generation near the tab regions exceeds that in the central region by approximately 30kWmâ3. A density-based TO framework is then integrated with a conjugate heat-transfer model to generate adaptive channel structures. The effects of inletâoutlet configuration and channel volume fraction on thermo-hydraulic performance are systematically investigated. The optimized structure forms a distributed branching flow network that enhances coolant redistribution toward high-heat regions and suppresses local stagnation zones. Compared with conventional designs, the temperature difference decreases by up to 68%, the maximum battery surface temperature decreases by up to 3.3 K, and the pressure drop decreases from approximately 44 Pa to about 18 Pa, corresponding to a reduction of nearly 59%. Overall, integrating realistic, non-uniform heat sources into TO significantly alters flow patterns, enabling coolant to be directed toward high-heat regions and thereby enhancing thermal management. The proposed framework provides a physics-based and energy-efficient design methodology for liquid-cooled blade batteries.
- 1 June 2026
- Published by AIP Publishing in Physics of Fluids
- Vol. 38  (6)
Investigation on topology optimization of flow and heat transfer in liquid-cooling channels for a blade-type battery with non-uniform heat generation
Blade-type lithium-ion batteries are widely used due to their high structural integration and superior volumetric energy density. However, the high aspect ratio and meter-scale spacing between electrode tabs lead to pronounced non-uniform heat generation during charging, resulting in localized thermal accumulation and imposing stringent requirements on coolant transport. To address this issue, the present study proposes a topology-optimized (TO) liquid-cooling plate with irregular flow channels tailored to the non-uniform thermal characteristics of blade batteries. An electrochemicalâthermal coupled model is established to capture spatial heat generation under charging conditions, revealing that the heat generation near the tab regions exceeds that in the central region by approximately 30kWmâ3. A density-based TO framework is then integrated with a conjugate heat-transfer model to generate adaptive channel structures. The effects of inletâoutlet configuration and channel volume fraction on thermo-hydraulic performance are systematically investigated. The optimized structure forms a distributed branching flow network that enhances coolant redistribution toward high-heat regions and suppresses local stagnation zones. Compared with conventional designs, the temperature difference decreases by up to 68%, the maximum battery surface temperature decreases by up to 3.3 K, and the pressure drop decreases from approximately 44 Pa to about 18 Pa, corresponding to a reduction of nearly 59%. Overall, integrating realistic, non-uniform heat sources into TO significantly alters flow patterns, enabling coolant to be directed toward high-heat regions and thereby enhancing thermal management. The proposed framework provides a physics-based and energy-efficient design methodology for liquid-cooled blade batteries.
- 1 January 2026
- Published by AIP Publishing in Physics of Fluids
- Vol. 38  (1)
Large eddy simulation assessment of an advanced perforated wall model for effusion cooling of gas turbine combustor
To accurately predict the effusion cooling process of a gas turbine combustor liner, a simplified heat and mass transfer model named the Advanced Perforated Wall Model (A-PWM) is developed, and turbulent flow and vortex dynamics are investigated using large eddy simulation (LES). The model is first validated for effusion cooling with a single cylindrical hole, and the prediction results match those of the conjugate heat transfer (CHT) model and experimental data. Due to the simplification of the physical cooling hole geometry, the LES-A-PWM reduces grid usage by 73% and improves computational efficiency by 62.55% compared to LES-CHT. It is also found that the local cooling effectiveness of LES-A-PWM shows deviations compared to experimental data, and LES-CHT results at the cooling hole outlet. Further analysis shows counter-rotating vortex pairs (CRVPs) form at the cooling hole outlet, enhancing mixing between cooling and hot flows and increasing Reynolds stress, thus affecting effusion cooling effectiveness. However, unlike resolved LES-CHT, the LES-A-PWM neglects the hole geometry and applies a uniform velocity, causing slight deviations near the outlet. Later, simulations on a combustor liner with 20 holes show that A-PWM agrees well with CHT and experimental data, maintaining deviations within 10%. The required mesh cells and computation time by LES-A-PWM are reduced by 35.8% and 35.9%, respectively. As the number of holes increases, turbulence interactions intensify between adjacent holes, while local heat transfer around the cooling jets weakens. These results demonstrate that A-PWM, assessed through LES, provides an accurate and efficient approach for design-level evaluation of combustor liner cooling.
- 1 January 2026
- Published by AIP Publishing in Physics of Fluids
- Vol. 38  (1)
Large eddy simulation assessment of an advanced perforated wall model for effusion cooling of gas turbine combustor
To accurately predict the effusion cooling process of a gas turbine combustor liner, a simplified heat and mass transfer model named the Advanced Perforated Wall Model (A-PWM) is developed, and turbulent flow and vortex dynamics are investigated using large eddy simulation (LES). The model is first validated for effusion cooling with a single cylindrical hole, and the prediction results match those of the conjugate heat transfer (CHT) model and experimental data. Due to the simplification of the physical cooling hole geometry, the LES-A-PWM reduces grid usage by 73% and improves computational efficiency by 62.55% compared to LES-CHT. It is also found that the local cooling effectiveness of LES-A-PWM shows deviations compared to experimental data, and LES-CHT results at the cooling hole outlet. Further analysis shows counter-rotating vortex pairs (CRVPs) form at the cooling hole outlet, enhancing mixing between cooling and hot flows and increasing Reynolds stress, thus affecting effusion cooling effectiveness. However, unlike resolved LES-CHT, the LES-A-PWM neglects the hole geometry and applies a uniform velocity, causing slight deviations near the outlet. Later, simulations on a combustor liner with 20 holes show that A-PWM agrees well with CHT and experimental data, maintaining deviations within 10%. The required mesh cells and computation time by LES-A-PWM are reduced by 35.8% and 35.9%, respectively. As the number of holes increases, turbulence interactions intensify between adjacent holes, while local heat transfer around the cooling jets weakens. These results demonstrate that A-PWM, assessed through LES, provides an accurate and efficient approach for design-level evaluation of combustor liner cooling.
- Open Access
- 1 January 2026
- Published by National Library of Serbia in Thermal Science
- Vol. 30  (3 Part B) , 2313-2328
Pressure-driven phase change model for interfacial mass transfer in air-water systems
The Lee model uses saturation temperature as the only criterion for phase change and does not consider the dynamic influence of vapour partial pressure. This paper addresses the dry air-water system at 0-60°C by incorporating a vapour pressure gradient source term via a user-defined function within the VoF interface capture framework. This construct enhances the gas-liquid mass transfer model by dynamically updating the saturated vapour pressure via temperature-dependent expressions. The simultaneous coupling of temperature and humidity gradients enables a more comprehensive characterization of the thermodynamics of low temperature evaporation, thus broadening the applicability of the model and enabling it to better approximate real-world physical processes. Systematic investigation of the mass transfer characteristics of the phase change mass transfer coefficient, β, over five orders of magnitude (from 10âğÂı-10âğâµ) revealed the behavior of the key control mechanisms. In the high phase change mass transfer coefficient region (β > 10âğ³), the evaporation mass transfer rate is primarily controlled by the vapour pressure difference, whereas in the low phase change mass transfer coefficient region (β < 10âğ³), the evaporation mass transfer process is dominated by β. Using parameter inversion with independent experimental data, the optimal range for the phase change mass transfer coefficient, β, in low temperature evaporation simulations of dry air-water single-pore bubble systems was determined to be 10âğâ´-10âğâµ. The developed model provides an optimized framework for simulating phase change mass transfer in low temperature evaporation processes.
- 1 January 2026
- Published by National Library of Serbia in Thermal Science
- Vol. 30  (3 Part B) , 2313-2328
Open AccessPressure-driven phase change model for interfacial mass transfer in air-water systems
The Lee model uses saturation temperature as the only criterion for phase change and does not consider the dynamic influence of vapour partial pressure. This paper addresses the dry air-water system at 0-60°C by incorporating a vapour pressure gradient source term via a user-defined function within the VoF interface capture framework. This construct enhances the gas-liquid mass transfer model by dynamically updating the saturated vapour pressure via temperature-dependent expressions. The simultaneous coupling of temperature and humidity gradients enables a more comprehensive characterization of the thermodynamics of low temperature evaporation, thus broadening the applicability of the model and enabling it to better approximate real-world physical processes. Systematic investigation of the mass transfer characteristics of the phase change mass transfer coefficient, β, over five orders of magnitude (from 10âğÂı-10âğâµ) revealed the behavior of the key control mechanisms. In the high phase change mass transfer coefficient region (β > 10âğ³), the evaporation mass transfer rate is primarily controlled by the vapour pressure difference, whereas in the low phase change mass transfer coefficient region (β < 10âğ³), the evaporation mass transfer process is dominated by β. Using parameter inversion with independent experimental data, the optimal range for the phase change mass transfer coefficient, β, in low temperature evaporation simulations of dry air-water single-pore bubble systems was determined to be 10âğâ´-10âğâµ. The developed model provides an optimized framework for simulating phase change mass transfer in low temperature evaporation processes.
- 1 September 2025
- Published by AIP Publishing in Physics of Fluids
- Vol. 37  (9)
Numerical investigations on hydrogen-enriched fuelâair mixing characteristics in a single micromix nozzle for the gas turbine combustor
Efficient hydrogen-enriched fuelâair mixing in a micromix combustor is a key factor that influences the combustion process, flow dynamics, and NOx formation. This study focuses on a single nozzle within a hydrogen-enriched micromix combustor to identify optimal fuelâair mixing strategies and explore the mixing mechanism. Through detailed numerical investigations of gas flow and geometric parameters by Reynolds-averaged NavierâStoke and large eddy simulation, the influence on the cold-flow field and mixing performance is analyzed. The results indicate that the momentum ratio significantly affects the initial jet trajectory and vorticity magnitude; an optimal momentum ratio of approximately 30 yielded the best mixing characteristics. Preheating the air to 575 K was observed to increase turbulent viscosity and molecular kinetic energy, thereby promoting more effective momentum and mass diffusion rates. Increasing the number of fuel tubes and reducing the tube diameter expanded the diffusion region, improving hydrogen-enriched fuelâair interaction. In addition, higher mass flow rates resulted in deeper jet penetration, while the jet angle significantly influenced the spatial fuel distribution and vortex structure within the premixing tube. A properly adjusted jet angle of 30° led to the breakdown of large vortices into smaller structures, thereby maximizing the contact surface between hydrogen-enriched fuel and air and, thus, enhancing the mixing characteristics. Overall, the micromix combustor demonstrates strong potential for achieving efficient, small-scale fuelâair mixing, which is essential for clean, low-emission combustion.
- 1 September 2025
- Published by AIP Publishing in Physics of Fluids
- Vol. 37  (9)
Numerical investigations on hydrogen-enriched fuelâair mixing characteristics in a single micromix nozzle for the gas turbine combustor
Efficient hydrogen-enriched fuelâair mixing in a micromix combustor is a key factor that influences the combustion process, flow dynamics, and NOx formation. This study focuses on a single nozzle within a hydrogen-enriched micromix combustor to identify optimal fuelâair mixing strategies and explore the mixing mechanism. Through detailed numerical investigations of gas flow and geometric parameters by Reynolds-averaged NavierâStoke and large eddy simulation, the influence on the cold-flow field and mixing performance is analyzed. The results indicate that the momentum ratio significantly affects the initial jet trajectory and vorticity magnitude; an optimal momentum ratio of approximately 30 yielded the best mixing characteristics. Preheating the air to 575 K was observed to increase turbulent viscosity and molecular kinetic energy, thereby promoting more effective momentum and mass diffusion rates. Increasing the number of fuel tubes and reducing the tube diameter expanded the diffusion region, improving hydrogen-enriched fuelâair interaction. In addition, higher mass flow rates resulted in deeper jet penetration, while the jet angle significantly influenced the spatial fuel distribution and vortex structure within the premixing tube. A properly adjusted jet angle of 30° led to the breakdown of large vortices into smaller structures, thereby maximizing the contact surface between hydrogen-enriched fuel and air and, thus, enhancing the mixing characteristics. Overall, the micromix combustor demonstrates strong potential for achieving efficient, small-scale fuelâair mixing, which is essential for clean, low-emission combustion.