Daifen Chen
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
Daifen Chen has a total of 263 co-authors. Here are the top co-authors by number of shared publications:
- Serhiy Ivanovich Serbin
Admiral Makarov National University of Shipbuilding , Mykolaiv , Ukraine
25 shared publications - Kun Liu
School of Naval Architecture and Ocean Engineering, Jiangsu University of Science and Technology, Zhenjiang 212003, China
17 shared publications - Volodymyr Patlaichuk
Admiral Makarov National University of Shipbuilding
17 shared publications - Zongping Shao
Curtin Centre for Advanced Energy Materials and Technologies (CAEMT), Western Australian School of Mines (WASM), Curtin University, Perth, Australia
10 shared publications Liu Lu
7 shared publicationsWei Kong
7 shared publications
Latest Publications
- 2 September 2026
- Published by Elsevier in Renewable Energy
- Vol. 276
Machine-learning-assisted cathode GDL design for high-performance PEMFCs in renewable-hydrogen transportation systems
- 2 September 2026
- Published by Elsevier in Renewable Energy
- Vol. 276
Machine-learning-assisted cathode GDL design for high-performance PEMFCs in renewable-hydrogen transportation systems
- Open Access
- 26 August 2026
- Published by Elsevier in SSRN Electronic Journal
Three-layer physically zoned immersion cooling for prismatic lithium-ion batteries: Flow-heat matching and thermal uniformity enhancement
Immersion cooling is a promising thermal management strategy for lithium-ion battery, yet conventional single-layer designs suffer from a severe mismatch between coolant flow distribution and heat generation. This study proposes a novel three-layer physically zoned immersion cooling design. Horizontal partitions divide the module into three independent flow domains (upper tab zone, middle core zone, and lower bottom zone), enabling decoupled, demand-oriented coolant allocation. To comprehensively quantify intra module thermal consistency, three supplementary metrics are introduced: the standard deviations of cell maximum temperatures ("σ" _("T" _"max,mod" )), cell maximum temperature differences ("σ" _("Δ" "T" _"max,mod" )), and cell average temperatures ("σ" _("Δ" "T" _"avg,mod" )). Numerical results demonstrate that the three-layer design reduces "σ" _("T" _"max,mod" ), "σ" _("Δ" "T" _"max,mod" ), and "σ" _("Δ" "T" _"avg,mod" )by 24.6 %, 36.4 %, and 22.1 %, respectively, relative to the single-layer baseline. Further subdivision of the middle zone into four counter-flow sub-layers markedly suppresses stream wise coolant temperature rise, yielding additional reductions of approximately 60 % in "σ" _("T" _"max,mod" ) and 56 % in "σ" _("Δ" "T" _"avg,mod" ). Finally, a synergistic strategy combining outer thermal conductive aluminum plates (3.5 mm) and non-uniform channel widths (N-2-7-2 configuration) is implemented to mitigate edge-cell temperature disparities. The optimized configuration ultimately suppresses the "σ" _("T" _"max,mod" ), "σ" _("Δ" "T" _"max,mod" ), and "σ" _("Δ" "T" _"avg,mod" ) to 0.110, 0.342, and 0.202, respectively. These findings demonstrate that the proposed design achieves precise matching between coolant flow and heat dissipation demand, significantly enhancing thermal performance and temperature uniformity.
- 26 August 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessThree-layer physically zoned immersion cooling for prismatic lithium-ion batteries: Flow-heat matching and thermal uniformity enhancement
Immersion cooling is a promising thermal management strategy for lithium-ion battery, yet conventional single-layer designs suffer from a severe mismatch between coolant flow distribution and heat generation. This study proposes a novel three-layer physically zoned immersion cooling design. Horizontal partitions divide the module into three independent flow domains (upper tab zone, middle core zone, and lower bottom zone), enabling decoupled, demand-oriented coolant allocation. To comprehensively quantify intra module thermal consistency, three supplementary metrics are introduced: the standard deviations of cell maximum temperatures ("σ" _("T" _"max,mod" )), cell maximum temperature differences ("σ" _("Δ" "T" _"max,mod" )), and cell average temperatures ("σ" _("Δ" "T" _"avg,mod" )). Numerical results demonstrate that the three-layer design reduces "σ" _("T" _"max,mod" ), "σ" _("Δ" "T" _"max,mod" ), and "σ" _("Δ" "T" _"avg,mod" )by 24.6 %, 36.4 %, and 22.1 %, respectively, relative to the single-layer baseline. Further subdivision of the middle zone into four counter-flow sub-layers markedly suppresses stream wise coolant temperature rise, yielding additional reductions of approximately 60 % in "σ" _("T" _"max,mod" ) and 56 % in "σ" _("Δ" "T" _"avg,mod" ). Finally, a synergistic strategy combining outer thermal conductive aluminum plates (3.5 mm) and non-uniform channel widths (N-2-7-2 configuration) is implemented to mitigate edge-cell temperature disparities. The optimized configuration ultimately suppresses the "σ" _("T" _"max,mod" ), "σ" _("Δ" "T" _"max,mod" ), and "σ" _("Δ" "T" _"avg,mod" ) to 0.110, 0.342, and 0.202, respectively. These findings demonstrate that the proposed design achieves precise matching between coolant flow and heat dissipation demand, significantly enhancing thermal performance and temperature uniformity.
- Open Access
- 7 July 2026
- Published by Elsevier in SSRN Electronic Journal
Localized Hotspot Cooling Performance of a Two-Phase Jet Cold Plate Under Intra-Die Nonuniform Heat Flux
The continuous increase in power density and heat flux of high-power GPU chips has positioned two-phase cooling as an attractive solution for data-center thermal management. However, localized intra-die hotspots still pose a critical challenge to temperature uniformity and cooling reliability. This study proposes a two-phase jet cold plate integrating central impingement and surrounding skived fins for localized hotspot cooling, and thus investigates its performance through combined numerical and experimental methods. Firstly, uniform-heating simulations are conducted to clarify the basic cooling behavior. The average temperature in the skived-fin region adjacent to the impingement zone Tfin and the average temperature in the impingement region Timp both decrease with increasing flow rate, while maintaining Timp < Tfin , and ΔTfi remains within 0.4–1.3 °C. Secondly, nonuniform-hotspot simulations are performed to evaluate hotspot suppression. At a local hotspot heat flux of about 300 W/cm², the maximum hotspot temperature is maintained at 72–74.5 °C, with a local temperature rise of 3.6–4.5 °C. Finally, steady uniform-heating experiments using R245fa are carried out for validation. The deviations of Tfin and Timp are within 5.3 °C and 5.5 °C, respectively. The overall thermal resistance Rt remains within 0.028–0.033 °C/W, and the pressure drop ΔP stays below 0.3 bar. Meanwhile, the evaporation pressure Pev and temperature Tev fluctuate only slightly, and the boiling curves further indicate stable phase-change heat transfer over the tested heat flux range. Overall, this work provides a practical reference for localized hotspot suppression and high heat flux thermal management in next-generation high-power GPU chips.
- 7 July 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessLocalized Hotspot Cooling Performance of a Two-Phase Jet Cold Plate Under Intra-Die Nonuniform Heat Flux
The continuous increase in power density and heat flux of high-power GPU chips has positioned two-phase cooling as an attractive solution for data-center thermal management. However, localized intra-die hotspots still pose a critical challenge to temperature uniformity and cooling reliability. This study proposes a two-phase jet cold plate integrating central impingement and surrounding skived fins for localized hotspot cooling, and thus investigates its performance through combined numerical and experimental methods. Firstly, uniform-heating simulations are conducted to clarify the basic cooling behavior. The average temperature in the skived-fin region adjacent to the impingement zone Tfin and the average temperature in the impingement region Timp both decrease with increasing flow rate, while maintaining Timp < Tfin , and ΔTfi remains within 0.4–1.3 °C. Secondly, nonuniform-hotspot simulations are performed to evaluate hotspot suppression. At a local hotspot heat flux of about 300 W/cm², the maximum hotspot temperature is maintained at 72–74.5 °C, with a local temperature rise of 3.6–4.5 °C. Finally, steady uniform-heating experiments using R245fa are carried out for validation. The deviations of Tfin and Timp are within 5.3 °C and 5.5 °C, respectively. The overall thermal resistance Rt remains within 0.028–0.033 °C/W, and the pressure drop ΔP stays below 0.3 bar. Meanwhile, the evaporation pressure Pev and temperature Tev fluctuate only slightly, and the boiling curves further indicate stable phase-change heat transfer over the tested heat flux range. Overall, this work provides a practical reference for localized hotspot suppression and high heat flux thermal management in next-generation high-power GPU chips.
- 10 June 2026
- Published by Springer Nature in International Journal of Minerals, Metallurgy and Materials
- Vol. 33 (6) , 1711-1733
Comprehensive review of current trends and future directions of metal-supported solid oxide fuel cell: From materials to applications
- 10 June 2026
- Published by Springer Nature in International Journal of Minerals, Metallurgy and Materials
- Vol. 33 (6) , 1711-1733
Comprehensive review of current trends and future directions of metal-supported solid oxide fuel cell: From materials to applications
- 22 May 2026
- Published by Admiral Makarov National University of Shipbuilding in Collection of Scientific Publications NUS
- Vol. 1 (1 (503)) , 132-140
CHARACTERISTICS OF A GAS TURBINE COMBUSTION CHAMBER WHEN USING HYDROGEN-AMMONIA FUEL
- 22 May 2026
- Published by Admiral Makarov National University of Shipbuilding in Collection of Scientific Publications NUS
- Vol. 1 (1 (503)) , 132-140
CHARACTERISTICS OF A GAS TURBINE COMBUSTION CHAMBER WHEN USING HYDROGEN-AMMONIA FUEL
- Open Access
- 26 March 2026
- Published by Elsevier in SSRN Electronic Journal
Research on Flow Field-Dependent Optimization of Three-Dimensional Porosity Distribution in the Cathode Gas Diffusion Layer of PEMFC
In response to the challenges of oxygen transport limitations and liquid water accumulation in the gas diffusion layer (GDL) of proton exchange membrane fuel cells (PEMFCs) under high current density, this study develops a three-dimensional multi-physics numerical model that integrates flow, mass transfer, electrochemical reactions, and two-phase transport. A porosity optimization method using a gradient boosting decision tree (GBDT) surrogate model is proposed. This approach constructs a nonlinear mapping between porosity distribution and cell performance, forming a "predict—validate—update" optimization framework for efficient identification of optimal porosity distributions across various flow fields. The findings show that this method significantly reduces the computational cost of high-fidelity simulations, enabling rapid identification of optimal porosity structures. It also reveals the coupling mechanisms between porosity optimization and flow field structures. The results demonstrate that the optimal porosity distribution varies along the main flow, rib arrangement, and thickness directions, with the main flow and thickness directions generally showing monotonic decreases, while the rib arrangement is more flow field-dependent. These results suggest that porosity optimization is a multidirectional process driven by the mass transport demands induced by the flow field.
- 26 March 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessResearch on Flow Field-Dependent Optimization of Three-Dimensional Porosity Distribution in the Cathode Gas Diffusion Layer of PEMFC
In response to the challenges of oxygen transport limitations and liquid water accumulation in the gas diffusion layer (GDL) of proton exchange membrane fuel cells (PEMFCs) under high current density, this study develops a three-dimensional multi-physics numerical model that integrates flow, mass transfer, electrochemical reactions, and two-phase transport. A porosity optimization method using a gradient boosting decision tree (GBDT) surrogate model is proposed. This approach constructs a nonlinear mapping between porosity distribution and cell performance, forming a "predict—validate—update" optimization framework for efficient identification of optimal porosity distributions across various flow fields. The findings show that this method significantly reduces the computational cost of high-fidelity simulations, enabling rapid identification of optimal porosity structures. It also reveals the coupling mechanisms between porosity optimization and flow field structures. The results demonstrate that the optimal porosity distribution varies along the main flow, rib arrangement, and thickness directions, with the main flow and thickness directions generally showing monotonic decreases, while the rib arrangement is more flow field-dependent. These results suggest that porosity optimization is a multidirectional process driven by the mass transport demands induced by the flow field.