This investigation into phase change material (PCM)-based passive thermal management systems was conducted via an experimental approach using 19.5 A h lithium iron phosphate cells with dimensions of (7.25 × 160 × 227) mm3. Trials were conducted at currents from 1 to 5C and environmental temperatures from 4 to 35 °C to simulate applications at which a Li-ion battery would be expected to perform. Based on comparisons, including an air-only control, the system consisting of PCM combined with five pores per inch (PPI) aluminum foam is the most effective at regulating average battery temperature and temperature gradient. During a 3C discharge trial at room temperature, the PCM-Al foam (5 PPI) system kept the average battery temperature and the maximum temperature difference below 28.1 and 5.2 °C, respectively, compared to the air-only control system which reached values of 48.0 and 17.2 °C, respectively. When analyzing data from trials at 4 and 35 °C, similar results are found with the PCM-Al foam systems being effective at thermal management. Thus, when compared to other systems, preliminary results show great promise in the future for the use of an PCM-Al foam passive thermal management system to effectively regulate the temperature of Li-ion batteries during use.
1 Background and Introduction
The Li-ion battery is widely considered as a viable alternative to other battery chemistries due to the battery's high-energy density and competitive cost. However, while being used during a high-power discharge, Li-ion batteries produce significant heat which can have a negative impact on battery life and create a safety hazard (Fig. 1) . Recently, the incorporation of phase change material (PCM) into passive thermal management systems for Li-ion batteries has had great success in regulating battery temperatures. Specifically, PCM serves as a heat sink for the Li-ion battery and will absorb the heat generated by the battery during use, therefore preventing the battery temperature from rising sharply [2,3]. Kizilel et al. [4,5] found that a passive thermal management system that uses PCM is more effective at dissipating heat than an active cooling system while discharging a Li-ion battery at a high current in a high ambient temperature environment. The passive cooling system used by Kizilel et al. also effectively maintained uniform cell temperature, which is important as uneven battery temperature rise significantly reduces battery life . Another important aspect that PCM-based passive cooling systems help reduce is the degradation rate of capacity of the battery. Kizilel et al.  found that a thermal management system that incorporated PCM cut the degradation rate of capacity of a Li-ion battery by more than half compared to a battery system that has no thermal management system, thus improving the effectiveness and life of the battery. PCM-based thermal management systems are also advantageous since they are considerably lighter and more compact than the typical active cooling system .
Although advantageous over active cooling systems, PCM-based passive thermal management systems are not perfect due to the low conductivity of PCM which limits the use of these systems to environments that only involve low heat transfer rates . However, multiple studies have found that incorporating foams made of materials such as copper or aluminum increased the thermal conductivity of tested systems, which in turn reduced the battery temperature rise during use and improved uniformity of temperature distribution on the surface of the battery . Two important aspects to consider when incorporating metal foams into the energy storage system for a Li-ion battery are the porosity and pore density of the metal foam. Li et al.  when using a copper foam paraffin composite found that battery surface temperature of a Li-ion battery during use increased as metal foams with greater porosity and pore density were used in the thermal management system.
Regarding application of commercially available Li-ion batteries, the batteries are primarily restricted to use in ambient temperatures that are near room temperature . When used in lower temperatures, specifically at and below 10 °C, Li-ion batteries experience a substantial loss in power capability (operating voltage) and energy (discharge capacity) due to the increased internal resistance of the battery [10,11]. Reports indicate that commercial 18,650 Li-ion batteries when tested at a temperature of −40 °C delivered only 5% of the battery energy density and only 1.25% of the battery power density when compared to values obtained from testing at 20 °C. With respect to charging and discharging at subzero temperatures, it has been found that Li-ion batteries can be normally discharged, albeit still at a reduced performance, but that charge cycles at these temperatures are particularly difficult and ineffective . On the other end of the temperature spectrum, electrochemical performance of a Li-ion battery also decreases at higher ambient temperatures . Specifically, at temperatures above 50 °C, batteries will experience a significant drop in charging efficiency and electrochemical durability .
These issues lead to a need for an energy storage system that can help Li-ion batteries perform effectively at the potential extreme temperatures that they could be used in. Therefore, the goal of this study was to test the application of PCM, octadecane specifically, and combinations of PCM with aluminum and carbon foams of varying pore densities in the insulation of Li-ion batteries at room and extreme temperatures to develop a passive thermal management system for Li-ion batteries that are effective at room and extreme temperatures. The testing metrics of the surface temperature rise of the Li-ion battery during use, temperature gradient across the surface of the battery during use, the cost, and mass of the energy storage system were all considered when determining which of the tested systems were the best option for the passive thermal management of a Li-ion battery.
2 Experimental Apparatus and Procedure
2.1 Experimental Setup.
LiFePO4 pouch cells provided by smith electric vehicles were used within the energy storage systems investigated in this study. Each battery cell has a dimension of (7.25 × 160 × 227) mm3, 19.5 A h capacity, nominal voltage of 3.3 V, and energy density of 247 W h/L. The base rectangular frame used for each respective energy storage system consists of 12.5 mm thick acrylic sheets (Makrolon polycarbonate manufactured by Bayer) bolted together to form an interior volume of 240 × 155 × 22 mm3. A groove with a 3 mm width and 4 mm depth was created on the mid-width of the side and bottom walls to place the cell in the middle of the container . The air-only energy storage system was covered by foam rubber insulation (along with all other tested systems) to reduce the heat loss to the surroundings during experiments (Fig. 2(a)). The PCM-only energy storage system consisted of the entire interior volume being filled with PCM (Fig. 2(b)), thus covering the entire surface of the electrode of the battery. The PCM-foam systems consisted of conductive foam positioned to completely cover the electrode surface of the battery with the rest of the interior volume and the pores of the foam filled with PCM. Three different foams were tested in this study and are as follows: five pores per inch (PPI) aluminum (Al) foam, 40-PPI Al foam, and 45-PPI carbon foam.
Five T-type thermocouples (Omega Engineering, Inc., Norwalk, CT) with an accuracy of ±0.5 °C were used to measure temperature on the battery surface during testing. These thermocouples were attached to the surface of the battery via aluminum foil tape to ensure optimal and consistent thermocouple positioning per trial. The same positioning of thermocouples was used in this study as a previous project conducted in this lab as this positioning was designed to be able to provide an accurate measurement of both average battery temperature and battery surface temperature gradient . Two thermocouples are near the positive and negative electrodes of the battery (34.3 mm away from the top of the battery), one thermocouple is placed between the two electrodes, one thermocouple is on the center of the battery, and one thermocouple is 34.3 mm away from the bottom of the battery.
2.2 Device and Procedure Description.
Real-time data of current, battery voltage, battery surface temperature, and the temperature of the energy storage systems were collected for each trial. Temperature data were collected by the KEYSIGHT 34972A LXI Data Acquisition/Switch Unit. Voltage and current data were collected by labview code that controls and collects data from the battery discharger, a BK PRECISION-8150, and the battery charger, a TDK-Lambda ZUP6-132. The cutoff voltage for discharging and charging cycles was 2.5 V and 3.9 V, respectively. During each trial, testing would begin at the desired constant current until the target voltage was reached. The labview code would then communicate to each respective device to switch from constant current charging/discharging to constant voltage charging/discharging. Each trial would then end when the current of the respective device for each trial reached a current below 1 A (Fig. 3). This testing process was used to ensure that the Li-ion batteries were charged/discharged as close as possible to their full capacity during each trial. This in turn would result in scenarios where the Li-ion batteries would generate the most heat possible within the normal operating voltage range, thus enabling the energy storage systems to be tested to the full extent at which they might be expected to handle during use. One of the metrics used to determine if a trial was successful was the battery was charged/discharged at a constant current for at least 3/4 to 7/8 of the battery's capacity.
2.3 Room Temperature Testing.
Once assembled, each energy storage system rested at room temperature for at least 8 h to ensure that the battery and system reached thermal equilibrium with the ambient air before testing. This 8-h resting period also allowed for any PCM incorporated into a system to completely solidify in preparation for testing. Testing of the energy storage systems began with the air-only system where trials were conducted at 1, 2, 3, 4, and 5C currents at a room temperature of 20 °C to create a baseline that would be used when considering trials where the PCM-only and PCM-foam energy storage systems were used. The 8-h rest period was then enforced after each trial was completed to let the battery and the system again return to thermal equilibrium with the ambient air before another trial would begin. Upon completion of the air-only trials, testing of the PCM-only system occurred at currents of 1, 3, and 5C as it was decided that these chosen C-rates were sufficient in testing the PCM-only system. The next phase involved testing the three systems of PCM with the chosen foams. At this point, the testing current was narrowed down to only a 3C current as 5C trials resulted in too large of a voltage jump for sufficient constant current charging/discharging to occur, while 1C currents did not produce much heat generation (primary focus of this study). Regarding the entire room temperature testing process, at least two charge trials and two discharge trials were completed at each current to ensure the collection of quality data. The charge and discharge data from each of the respective two trials was then compared to ensure that similar temperature results occurred during each trial. If there was a discrepancy between data of the respective two trials, then additional trials were completed until concurring data were obtained.
2.4 Extreme Temperature Testing.
Regarding extreme temperature testing, the rooms that were used consisted of a hot room set at 35 °C and a cold room set at 4 °C. These temperatures were decided upon primarily as they created environments where a Li-ion battery could be reasonably tested at higher constant currents while still exposing the batteries to non-ideal temperature conditions. Testing at extreme temperatures followed the same guidelines as described in Sec. 2.3 with the 8-h rest period after system setup and trials, and repeating trials until quality data were collected. Regarding testing the energy storage systems at extreme temperatures, only the PCM-Al foam (40 PPI) system and the air-only system were tested. As was previously done for testing the PCM-foam systems at room temperature, a current of 3C was used at extreme temperatures. Trials of the PCM-Al foam (40 PPI) system at a 3C current were completed successfully with at least two charge trials and two discharge trials completed for each temperature. However, testing the air-only system at 3C currents led to issues as the batteries would experience large changes in voltage due to the extreme temperatures negatively affecting performance. This would cause the testing system to switch to constant voltage charging/discharging much quicker than desired. To counter this issue, testing was then lowered down to 1C currents for both the PCM-Al foam (40 PPI) and the air-only systems to ensure that each trial would consist primarily of constant current charging/discharging followed by a short period of constant voltage charging/discharging as was desired when considering successful trials. After this change in testing current, consistent results were then collected for both considered systems.
3 Model Simulations
Upon completion of the experimental portion of this study, computational fluid dynamics (CFD) models were developed using Star-CCM++ (version 13.04.011). The assembly file from SolidWorks was imported into Star-CCM++ to generate high-quality volume mesh for CFD simulations of heat and mass transfer within the whole computational domain (including battery, coolant, acrylic frame, and foam insulation). The model was run on a workstation with Xeon(R) Gold 6240 CPU and 192 GB memory.
3.1 Governing Equations.
Simulating the heat and mass transfer of PCM and PCM-foam materials requires the coupling of mass and energy between solid and liquid during phase change [15,16]. This study focuses on effects of foam materials on the thermal management of batteries and simplified the simulation by comparing the air, water, and water-foam composite materials as coolants. The simplified model solves the following governing equations on heat and mass transfer:
Heat is generated only in the battery region through the generation rates measured by our previous experiments. Since the heat generation rate is dependent on the depth of discharge or state of charge, the equation fitted from experiments  are used to generate a table with time-dependent heat generation rates, , as the heat input to the model.
3.2 Initial and Boundary Conditions.
The model solves the above governing equations together with the following initial and boundary conditions. The initial and boundary conditions are consistent with our experimental setup:
The initial temperature of the battery, coolant, container, and insulation materials is 20 °C.
No movement boundary conditions apply to all walls.
Natural convection boundary conditions are applied to all surfaces (mainly insulating materials surfaces) directly contacting with the environment to consider the overall heat loss to the environment. The natural convection coefficients are estimated using empirical coefficients  of heated plates with various sizes facing different orientations. Natural convection coefficients are between 3.9 and 7.5 W/m2/K in this study.
3.3 Thermal Physical Properties.
The model applies thermal physical properties summarized in Table 1 to different regions of the model. It should be noted that properties of Al and graphite are used as properties of bulk Al foam and carbon foam. The porosity of foams is set as 0.9.
|Battery||Acrylic frame||Rubber insulation||Air||Water||Graphite||Al|
|Specific heat (kJ/kg/K)||1.28||1.47||1.88||1.01||4.18||0.71||0.90|
|Thermal conductivity (W/m/K)||2.37||0.3||0.17||0.026||0.62||24||237.0|
|Viscosity (Pa · s)||–||–||–||1.8 × 10−5||8.9 × 10−4||–||–|
|Battery||Acrylic frame||Rubber insulation||Air||Water||Graphite||Al|
|Specific heat (kJ/kg/K)||1.28||1.47||1.88||1.01||4.18||0.71||0.90|
|Thermal conductivity (W/m/K)||2.37||0.3||0.17||0.026||0.62||24||237.0|
|Viscosity (Pa · s)||–||–||–||1.8 × 10−5||8.9 × 10−4||–||–|
3.4 Model Validations.
For validation purposes, model results of average battery surface temperature change with time at 3C discharge were then compared to the experimental data collected in this study to confirm the reliability of the CFD models. Additionally, as the experimental test apparatus focused on in this study is symmetric in nature, the CFD models consisted of “a quarter” of the experimental test setup to reduce computational time and costs.
The mesh and time step independences of model results were also checked by varying the number of mesh from 2,127,772 to 48,583 and the time step from 0.01 s to 1 s. When the number of mesh is larger than 48,583 and time step is less than 0.1 s, the change on simulated average battery surface temperature at 900 s changes was less than 1%.
4 Results and Discussion
4.1 Room Temperature Tests.
For comparing the performance of each energy storage system, the temperature gradient on the surface of the battery and the overall average temperature of the surface of the battery were calculated to compare each system. Based on data collected from room temperature trials, the most effective energy storage system was the PCM-Al foam (5 PPI) system which was then followed by the PCM-Al foam (40 PPI) system, the PCM-only system, the PCM-carbon foam (45 PPI) system, and lastly the air-only system was the least effective energy storage system (Fig. 4). Comparing with the air-only system, PCM can reduce the temperature difference as well as the average temperature increase on the battery surface mainly because of the high heat capacity and the latent heat of PCM during phase change. Considering the relatively low thermal conductivity of PCM (0.24 W/m/K), developing composite materials by integrating PCM with materials with high thermal conductive (such as Al foams) could further improve the thermal management property. At a 3C current, the PCM-Al foam (5 PPI) system held the average battery temperature to a maximum of 28.1 °C compared to 48 °C temperature allowed by the air-only system. Additionally, at a 3C current, the PCM-Al foam (5 PPI) system allowed only a maximum temperature difference of 5.2 °C across the battery surface compared to the 17.2 °C difference allowed by the air-only system. As such, it can be noted that the PCM-Al foam (5 PPI) system was quite effective in both reducing the average temperature change during battery use and kept the temperature gradient of the battery below the 10 °C boundary.
Figure 4 also indicates that integrating PCM with carbon foam is not an effective approach for battery thermal management. Since the carbon foam has relatively low thermal conductivity (0.033–0.050 W/m),2 the PCM-carbon foam composite material has even lower effective thermal conductivity with pure PCM. In addition, the carbon foam suppressed the natural convection of melted PCM during the heating process and further slowed down the heat transfer process. Therefore, the PCM-carbon foam composition material is not a viable option, comparing with pure PCM, as passive coolant for the battery. Integrating PCM with aluminum foams, however, will significantly increase the effective thermal conductivity because the thermal conductivity of the aluminum foam (5.8 W/m/K) is more than an order of magnitude higher than that of PCM (∼0.24 W/m/K). Comparing with the battery cooled by pure PCM, the maximum temperature difference as well as the average temperature increase on the surface of the battery was both significantly decreased using PCM-Al foam composite materials.
In addition, the PCM-Al foam (5 PPI) showed slightly better cooling effect than PCM-Al foam (40 PPI) mainly because the Al foam with larger pore size will have much less permeability (flow resistance) so that the natural convection of PCM is not significantly reduced. In addition, the specific heat of Al (0.9 J/g/K) is much less than that of pure PCM (>2.5 J/g/K). The 5-PPI Al foam (0.91) has slightly higher porosity than the 40-PPI Al foam (0.87). Please note that the porosities of Al foams are measured in lab using the Archimedes principle. Therefore, the effective specific heat of the PCM-Al foam (5 PPI) composite material is slightly higher than that of PCM-Al foam (40 PPI).
4.2 Extreme Temperature Tests.
Regarding extreme temperature testing, the focus was on comparing how the air-only and PCM-Al foam composites performed at each of the three test temperatures. As such, when comparing average change in temperature for a system at the three test temperatures, it was necessary to consider the average change from the starting temperature to normalize the data for comparison (e.g., 20 °C was subtracted from all the temperature data collected at room temperature).
When air was the coolant, the average battery surface temperature increase (Fig. 5(a)) could be between 5.8 °C (35 °C environmental temperature) and 12.5 °C (4 °C environmental temperature). Even at 1C discharge rate, the temperature increase is relatively high because air has very low thermal conductivity and low heat capacity. Similarly, the maximum temperature difference on the battery surface was also high: 2.7 °C (4 °C environmental temperature), 3.5 °C (35 °C environmental temperature), and 5.3 °C (20 °C environmental temperature).
From the results of the average temperature comparison in Fig. 6(a), it can be noted that the PCM-Al foam (40 PPI) system did not experience an average temperature change greater than 7 °C when considering a 1C discharge at any of the three testing temperatures. Additionally, to provide reference with the air-only control, at the 4 °C temperature, the PCM-Al foam (40 PPI) system only allowed the average battery temperature to increase by 6.7 °C compared to the air-only system that allowed a 12.4 °C increase. While this difference in temperature may not seem large, this is because of the testing being done at a 1C current and this gap between temperatures would likely be much larger at greater currents (e.g., see prior results at room temperature testing with a 3C current). The primary revelation from the testing done at a 1C current at these three temperatures is that the PCM-Al foam (40 PPI) system was able to hold the temperature gradient across the battery surface below the 4.2 °C boundary at all three temperatures (Fig. 6(b)). This is an especially important factor when considering the degradation Li-ion batteries already experience during use at non-room temperatures. It should also be noted that the melting temperature of the PCM is around 28 °C. The PCM were completely melted as liquid in all tests under 35 °C environmental temperature. The temperature difference on the battery surface could result in natural convection of liquid PCM to facilitate the heat exchange between the battery and the coolant. As a result, both the average temperature increase and the maximum temperature difference on the battery surface under 35 °C environmental temperature were the lowest among all three environmental temperatures during tests using PCM-Al foam. Overall, it can thus be concluded that the use of PCM combined with conductive foams is an effective manner in regulating battery temperature at not only room temperature but at more extreme temperatures as well.
4.3 Model Simulations.
This thermal model applies uniform heat generation rate within the whole battery domain and cannot capture the non-uniform current density distribution and heat distribution within the battery. As a result, the highest temperature on the battery surface is near the center of the battery, while the highest temperature is observed near the two terminals on the top of the battery using thermal camera. Nevertheless, the simulated overall average temperature increase on the battery surface when the battery is discharged at 3C matches with experimental data tested with air and water-Al foam as the coolant (Fig. 7). In addition, the maximum temperature difference on the battery surface is consistent with experimental observations (Fig. 8): the temperature difference is the highest using air as the coolant and is the lowest when water-Al foam is used as the coolant. The integration of water with Al foam increases the overall thermal conductivity of the coolant, while the natural convection of the fluid is suppressed by the foam material. The volume-average velocity of pure water is 8.80 × 10−4 m/s, while the volume-average velocity of water in Al foam is only 1.93 × 10−5 m/s. The significant increase of the thermal conductivity (and thermal diffusivity) and the decrease of the natural convection by using Al foam still facilitate the thermal management of the battery due to the thin coolant layer and relatively high permeability of the foam.
More thorough model simulations will be carried out in the future to simulate the heat and mass transfer within PCM and PCM-foam composite materials. The accuracy of volume-average CFD models depends heavily on the accuracy of the effective material properties (porosity, effective thermal conductivity, permeability, etc.) . The model in this study used simplified empirical correlations to estimate these properties. However, these empirical correlations are only accurate when they are applied to porous materials with the right pore size, pore geometry, and porosity. Detailed pore-scale simulations  that can accurately predict effective material properties of porous media will be integrated with volume-average CFD models to elucidate battery thermal management in great details. Furthermore, the thermal model of the coolant also needs to be integrated with the thermal-electrical model of the battery to consider the non-uniform reactions within the battery and the synergistic interactions between local temperature and local reaction rate.
This study has investigated various coolants (air, water, PCM, PCM-carbon foam composite, and PCM-Al foam composites) for passive thermal management of Li-ion cells. The 19.5 A h LiFePO4 cells were charged and discharged under different C-rates (1–5) and different environmental temperature (4, 20, and 35 °C). The experimental comparisons and model simulation reach the following conclusions:
The average battery surface temperature can increase up to 25 °C at 3C discharge rates. In addition, there are huge temperature differences on the battery surface (up to 17 °C), and the temperature difference within the battery is likely even higher.
Integrating PCM with thermally conductive foams is an effective manner in regulating battery temperature. The PCM-Al foam can keep the temperature increase below 7 °C while temperature difference is within 5 °C at 3C discharge.
The PCM-metal foam composites are effective when the battery is operated in all environmental temperatures (4, 20, and 35 °C) tested in this study.
Metal foams with larger pore size (or higher permeability) are preferred to increase the overall thermal conductivity of the composite material without significantly suppressing the natural convection.
Using coolant with high capacity (or latent heat) is critical for short-term, high C-rates operations.
The authors want to thank the financial support from New Faculty General Research Fund provided by the University of Kansas (Funder ID: 10.13039/100007859) and Li-ion batteries provided by Smith Electric Vehicles. DB wants to thank funding support from NASA EPSCoR (80NSSC18M0030; Funder ID: 10.13039/100005714) and the University of Kansas Center (Funder ID: 10.13039/100006727) for Undergraduate Research.