David W. Salter
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
David W. Salter has a total of 18 co-authors. Here are the top co-authors by number of shared publications:
Daniel Trimble
Trinity College Dublin,Department of Mechanical, Manufacturing Biomedical Engineering,Dublin,Ireland
2 shared publications- Daniel Trimble
Manufacturing Biomedical Engineering Trinity College Dublin,Department of Mechanical,Dublin,Ireland
2 shared publications Kantharuphan Annathurai
Trinity College Dublin,Department of Mechanical, Manufacturing Biomedical Engineering,Dublin,Ireland
2 shared publicationsRajesh Nimmagadda
Trinity College Dublin,Department of Mechanical, Manufacturing Biomedical Engineering,Dublin,Ireland
2 shared publicationsSeamus M. O'SHaughnessy
Trinity College Dublin,Department of Mechanical, Manufacturing Biomedical Engineering,Dublin,Ireland
2 shared publications- Seamus Michael O'Shaughnessy
Manufacturing Biomedical Engineering Trinity College Dublin,Department of Mechanical,Dublin,Ireland
2 shared publications
Latest Publications
- Open Access
- 24 April 2026
- Published by Elsevier in SSRN Electronic Journal
Novel integrated tab-and-surface cooling strategy for prismatic lithium-ion cells for high C-rate operations
Effective thermal management is critical for prismatic lithium-ion batteries operating at high C-rates, where localized heating at the tabs and strong temperature gradients across the cell can compromise performance, safety, and cycle life. This work presents an integrated cooling strategy that combines jet-assisted tab cooling with thin-film evaporation over the cell surfaces. Using a single 50 Ah prismatic lithium iron phosphate battery and Opteon SF-33 as the working fluid, three cooling configurations were experimentally investigated under 2C charging and discharging conditions: free jet impingement with non-immersed surfaces (Config_1, proposed), body immersion with free jets (Config_2), and full immersion with submerged jets (Config_3). Among these, the proposed free-jet/non-immersed configuration consistently delivered the lowest peak temperatures, with both tab and surface temperatures remaining below 30oC across the investigated flow rates. The free jet flow enhanced stagnation-point heat transfer at the tabs, while the formation of a stable evaporating liquid film over the large prismatic faces improved surface temperature uniformity and promoted lateral heat spreading. In contrast, immersion-based configurations suppressed evaporation and attenuated jet momentum, resulting in comparatively higher operating temperatures despite larger coolant inventories. Overall, the proposed configuration (Config_1) reduced peak temperatures and improved combined thermal-hydraulic performance by a factor of ~1.6-1.9 while requiring substantially less dielectric fluid compared to the fully immersion-based Config_3. These results demonstrate that combining targeted jet impingement cooling with surface-level evaporation provides an effective and scalable thermal management strategy for prismatic lithium-ion batteries operating at high C-rates in electric vehicles and stationary energy storage applications.
- 24 April 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessNovel integrated tab-and-surface cooling strategy for prismatic lithium-ion cells for high C-rate operations
Effective thermal management is critical for prismatic lithium-ion batteries operating at high C-rates, where localized heating at the tabs and strong temperature gradients across the cell can compromise performance, safety, and cycle life. This work presents an integrated cooling strategy that combines jet-assisted tab cooling with thin-film evaporation over the cell surfaces. Using a single 50 Ah prismatic lithium iron phosphate battery and Opteon SF-33 as the working fluid, three cooling configurations were experimentally investigated under 2C charging and discharging conditions: free jet impingement with non-immersed surfaces (Config_1, proposed), body immersion with free jets (Config_2), and full immersion with submerged jets (Config_3). Among these, the proposed free-jet/non-immersed configuration consistently delivered the lowest peak temperatures, with both tab and surface temperatures remaining below 30oC across the investigated flow rates. The free jet flow enhanced stagnation-point heat transfer at the tabs, while the formation of a stable evaporating liquid film over the large prismatic faces improved surface temperature uniformity and promoted lateral heat spreading. In contrast, immersion-based configurations suppressed evaporation and attenuated jet momentum, resulting in comparatively higher operating temperatures despite larger coolant inventories. Overall, the proposed configuration (Config_1) reduced peak temperatures and improved combined thermal-hydraulic performance by a factor of ~1.6-1.9 while requiring substantially less dielectric fluid compared to the fully immersion-based Config_3. These results demonstrate that combining targeted jet impingement cooling with surface-level evaporation provides an effective and scalable thermal management strategy for prismatic lithium-ion batteries operating at high C-rates in electric vehicles and stationary energy storage applications.
- 24 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
An Experimental Investigation of Guided Flow Liquid Cooling Methods on a Four Cell Prismatic Module
This paper aims to experimentally compare the cooling effectiveness of forced flow immersion cooling (FFIC), guided flow immersion cooling (GFIC) as well as jet cooling (JC) approaches when applied to a four-cell battery module. The effect of varying dielectric fluid (SF33) flow rate for each approach at a 0.5 C (dis)charge rate is analysed. A four-cell prismatic battery module consisting of $105 \text{Ah} \text{LiFePO}_{4}$ cells was placed in a polycarbonate chamber. Each cooling approach is tested under flow rates of $0.6 ~\mathrm{L} / \text{min}, 1 ~\mathrm{L} / \text{min}$ and $1.4 ~\mathrm{L} / \text{min}$. Results show that FFIC was the most consistent at reducing maximum temperature across all three flow rates for both charge and discharge processes. GFIC concentrates the flow at the cell terminals when the cells are fully immersed at a cost of additional pressure drop but results show that this technique is not as effective as FFIC. JC has poor cooling performance at the lowest flow rate but was proven to be the most effective cooling approach in charge and competitive in discharge at the highest flow rate of $1.4 ~\mathrm{L} / \text{min}$, showing the potential of the approach. Future work could include testing all three approaches with higher flow rates, particularly for the JC, which performed well at higher flow rates. This study incorporated single phase cooling, but future work could involve two phase cooling comparison of each of the three configurations with a view to validating future numerical studies involving this work.- 24 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
An Experimental Investigation of Guided Flow Liquid Cooling Methods on a Four Cell Prismatic Module
This paper aims to experimentally compare the cooling effectiveness of forced flow immersion cooling (FFIC), guided flow immersion cooling (GFIC) as well as jet cooling (JC) approaches when applied to a four-cell battery module. The effect of varying dielectric fluid (SF33) flow rate for each approach at a 0.5 C (dis)charge rate is analysed. A four-cell prismatic battery module consisting of $105 \text{Ah} \text{LiFePO}_{4}$ cells was placed in a polycarbonate chamber. Each cooling approach is tested under flow rates of $0.6 ~\mathrm{L} / \text{min}, 1 ~\mathrm{L} / \text{min}$ and $1.4 ~\mathrm{L} / \text{min}$. Results show that FFIC was the most consistent at reducing maximum temperature across all three flow rates for both charge and discharge processes. GFIC concentrates the flow at the cell terminals when the cells are fully immersed at a cost of additional pressure drop but results show that this technique is not as effective as FFIC. JC has poor cooling performance at the lowest flow rate but was proven to be the most effective cooling approach in charge and competitive in discharge at the highest flow rate of $1.4 ~\mathrm{L} / \text{min}$, showing the potential of the approach. Future work could include testing all three approaches with higher flow rates, particularly for the JC, which performed well at higher flow rates. This study incorporated single phase cooling, but future work could involve two phase cooling comparison of each of the three configurations with a view to validating future numerical studies involving this work.- 24 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
Numerical and Experimental Investigation of Liquid Tab Cooling of a Prismatic Battery
This paper presents an experimentally validated numerical model of liquid cooling of the terminals or tabs of a prismatic 50 Ah lithium-ion battery. The investigation is part of a wider study to attempt to reduce the amount of required working fluid compared to complete battery immersion. There have been numerous studies on direct liquid tab cooling on cylindrical and pouch type cells, but this approach remains under documented for prismatic type cells. On the experimental side, a polycarbonate flow chamber was constructed to concentrate the flow of the dielectric fluid Opteon SF33 on only the top surface, including the tabs of the battery. An infrared (IR) camera recorded the front face temperature rise of the cell during $\mathbf{1 C}$ discharging for three different flow rates. On the numerical side, a Newman's Pseudo-2D (P2D) electrochemical model of the experiment was developed in ANSYS Fluent 2024 R1. The voltage/current profiles as well as the cell front face temperature rises were compared. Results show an average voltage difference of $\mathbf{1 0 0 ~ m V}$ between the two approaches and that both numerical and experimental temperature rise plots captured the vertical temperature gradient in the cell. Deviations in maximum temperature of $0.3 \%, 13.9 \%$ and 17.9 % were reported for $0.21 ~\mathrm{L} / {\min}, 0.81 ~\mathrm{L} / {\min}$ and 1.28 L/min coolant flow rates respectively, demonstrating a reasonably strong agreement between the numerical model and the experimental IR thermography, therefore validating the model. The temperature gradients were excessive, requiring a hybridized approach with cell face air cooling which the model could test.- 24 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
Numerical and Experimental Investigation of Liquid Tab Cooling of a Prismatic Battery
This paper presents an experimentally validated numerical model of liquid cooling of the terminals or tabs of a prismatic 50 Ah lithium-ion battery. The investigation is part of a wider study to attempt to reduce the amount of required working fluid compared to complete battery immersion. There have been numerous studies on direct liquid tab cooling on cylindrical and pouch type cells, but this approach remains under documented for prismatic type cells. On the experimental side, a polycarbonate flow chamber was constructed to concentrate the flow of the dielectric fluid Opteon SF33 on only the top surface, including the tabs of the battery. An infrared (IR) camera recorded the front face temperature rise of the cell during $\mathbf{1 C}$ discharging for three different flow rates. On the numerical side, a Newman's Pseudo-2D (P2D) electrochemical model of the experiment was developed in ANSYS Fluent 2024 R1. The voltage/current profiles as well as the cell front face temperature rises were compared. Results show an average voltage difference of $\mathbf{1 0 0 ~ m V}$ between the two approaches and that both numerical and experimental temperature rise plots captured the vertical temperature gradient in the cell. Deviations in maximum temperature of $0.3 \%, 13.9 \%$ and 17.9 % were reported for $0.21 ~\mathrm{L} / {\min}, 0.81 ~\mathrm{L} / {\min}$ and 1.28 L/min coolant flow rates respectively, demonstrating a reasonably strong agreement between the numerical model and the experimental IR thermography, therefore validating the model. The temperature gradients were excessive, requiring a hybridized approach with cell face air cooling which the model could test.- 24 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
Numerical and Experimental Comparative Study on Cold Plate Cooling of a 5.6kW Tesla Battery Module
This paper aims to numerically investigate and experimentally validate the performance of two 5.6 kWh Tesla battery modules subjected to indirect liquid cooling. The battery modules are charged while connected in series at a low C-rate and discharged individually at 0.5 C and 0.8 C while subjected to $0.3 ~\mathrm{L} / {\min}$ and $0.6 ~\mathrm{L} / {\min}$ coolant flow rates. A numerical model of the Tesla battery modules is then modelled on SolidWorks using heat generation rates of an NMC-111 18650 cell obtained through an in-house numerical study previously completed using Newman's Pseudo 2-D (P2D) model. The experimental charge and discharge process were post-processed using on-board sensors. Charging the modules from it OCV (38.5 V) to cut off (48.5 V) resulted in a 15° C temperature increase. Charging the modules from 38.5 V to 48.5 V resulted in a 15° C temperature increase. Charging the modules from 38.5 V to 48.5 V resulted in a 15° C temperature increase. The numerical model demonstrated excellent agreement with the 0.5 C discharge rate, but showed a deviation at the 0.8 C rate.- 24 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
Numerical and Experimental Comparative Study on Cold Plate Cooling of a 5.6kW Tesla Battery Module
This paper aims to numerically investigate and experimentally validate the performance of two 5.6 kWh Tesla battery modules subjected to indirect liquid cooling. The battery modules are charged while connected in series at a low C-rate and discharged individually at 0.5 C and 0.8 C while subjected to $0.3 ~\mathrm{L} / {\min}$ and $0.6 ~\mathrm{L} / {\min}$ coolant flow rates. A numerical model of the Tesla battery modules is then modelled on SolidWorks using heat generation rates of an NMC-111 18650 cell obtained through an in-house numerical study previously completed using Newman's Pseudo 2-D (P2D) model. The experimental charge and discharge process were post-processed using on-board sensors. Charging the modules from it OCV (38.5 V) to cut off (48.5 V) resulted in a 15° C temperature increase. Charging the modules from 38.5 V to 48.5 V resulted in a 15° C temperature increase. Charging the modules from 38.5 V to 48.5 V resulted in a 15° C temperature increase. The numerical model demonstrated excellent agreement with the 0.5 C discharge rate, but showed a deviation at the 0.8 C rate.- 24 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
Manufacturing and Design Optimisation of an Air Amplifier for Data Centre Server Cooling Applications
This paper presents an experimentally validated CFD optimisation of the cooling potential of three differing air amplifier geometries identified in the literature as part of a wider study to improve the overall efficiency of next generation data centres. This is the first combined manufacturing and design optimisation for air amplifiers in server cooling applications. The aim of this paper is to conduct a design and manufacturing optimisation workflow to ascertain the suitability of various air amplifier geometry approaches. The numerical optimisation includes numerous independent design points in ANSYS Fluent, investigating the influence of nozzle gap width and angle on the outlet flow rate for three different geometries including an air amplifier, an air knife and a Coandă ejector. The numerical results highlight surface plots for a Coandă adherence ratio as well as overall performance metric efficiency and the air amplifier results were tested experimentally. 3D printed manufacturing was utilised, along with a polycarbonate flow channel emulating the dimensions of a data centre server, to test cooling approaches on a heated foil with an IR camera placed underneath. A good, medium and poor set of numerical Coandă adherence ratio result points yielded average foil surface heat transfer coefficients of 164 W/m2 K, 127 W/m2 K and 116 W/m2K respectively. This demonstrates a strong agreement between the CFD and the experiments, therefore validating the numerical, design and manufacturing optimisation approaches highlighted in this work.
- 24 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
Manufacturing and Design Optimisation of an Air Amplifier for Data Centre Server Cooling Applications
This paper presents an experimentally validated CFD optimisation of the cooling potential of three differing air amplifier geometries identified in the literature as part of a wider study to improve the overall efficiency of next generation data centres. This is the first combined manufacturing and design optimisation for air amplifiers in server cooling applications. The aim of this paper is to conduct a design and manufacturing optimisation workflow to ascertain the suitability of various air amplifier geometry approaches. The numerical optimisation includes numerous independent design points in ANSYS Fluent, investigating the influence of nozzle gap width and angle on the outlet flow rate for three different geometries including an air amplifier, an air knife and a Coandă ejector. The numerical results highlight surface plots for a Coandă adherence ratio as well as overall performance metric efficiency and the air amplifier results were tested experimentally. 3D printed manufacturing was utilised, along with a polycarbonate flow channel emulating the dimensions of a data centre server, to test cooling approaches on a heated foil with an IR camera placed underneath. A good, medium and poor set of numerical Coandă adherence ratio result points yielded average foil surface heat transfer coefficients of 164 W/m2 K, 127 W/m2 K and 116 W/m2K respectively. This demonstrates a strong agreement between the CFD and the experiments, therefore validating the numerical, design and manufacturing optimisation approaches highlighted in this work.