Novel thermal management methods to improve the performance of the Li-ion batteries in high discharge current applications

Over the last few decades investigating the performance of thermal management in the high charge/ discharge current has been taken into consideration in many studies. In this study, a mature heat pipebased air cooling system is built to control the temperature of the lithium-ion (Li-ion) cell/module in the high current (184 A) discharging rate. The temperature of the cell/module experimentally and numerically is considered by the lack of natural convection, natural convection, forced convection, and evaporative cooling. According to the experimental results, the natural and forced convection decrease the average temperature of the cell by 6.2% and 33.7% respectively. Moreover, several numerical simulations are solved by COMSOL Multiphysics®, the commercial computational fluid dynamics (CFD) software. The simulation results are validated against experimental results at the cell level for natural and forced convection. It indicates that the evaporative cooling method is robust to enhance the current cooling systemmethod for further optimization. The results show that there is a 35.8% and 23.8% reduction in the maximum temperature of the cell and module due to the effect of the evaporative cooling method


Introduction
Nowadays, the transportation system consumes a lot of fuel, which leads to global warming, energy shortage, and environmental pollution as serious problems for human life. Thus, electric vehicles (EVs) have received worldwide attention in recent years due to the expected energy shortage and environmental pollution problems. Recently, lithium-ion (Li-ion) batteries have become more popular among all kinds of power batteries because of their low self-discharge rate, environmental friendliness, high energy density, long life cycle, and high voltage, which is an appropriate energy storage solution in EVs applications [1e5]. Nevertheless, the energy density, capacity, efficiency, and cycling life of Li-ion batteries are highly dependent on their thermal management [6,7]. Generally, Li-ion batteries in the temperature range of 25 Ce40 C have the best performance [8,9]. Conventional battery technology supports charging/discharging at relatively low current applications. However, a high current is necessary when vehicles climb the hills or during fast charging. As a noticeable amount of heat generated within the Li-ion batteries in high current applications, define and design an appropriate thermal management system is essential.
It has been investigated by many researchers that active and passive cooling systems such as liquid cooling, air cooling, phase change materials (PCMs), and heat pipe can control the temperature rising and improve the stability of electronic devices and Li-ion batteries in charging/discharging process [10e16]. Liquid coolants have a high heat transfer coefficient, which makes them an operative, efficient, quiet, and compact cooling medium. The liquid cooling system, as a promising cooling system, has been used by many researchers [17e20]. However, liquid cooling systems have a massive and complex structure with high investment costs and power consumption. Usage components like liquids, pumps/compressors, and tubing increase the weight of the system and leakage risk [21]. Air cooling is an active cooling system with a simple structure, low cost, easy maintenance, and lightweight, which is used in many cooling applications [22e24]. Nonetheless, due to the low heat transfer coefficient and heat capacity of air, such cooling systems hardly control the battery's cell/module temperature in a safe range in high current applications [25]. PCM is a passive cooling system that has been using by researchers in many cooling and energy storage applications [26e29]. As PCMs suffer from lowthermal conductivity, several studies have been done to improve PCMs thermal conductivity by adding Nanomaterial, graphite, Nanocarbon tube, aluminum mesh, etc. [30e33]. However, leaking issues, low thermal stability, high cost, and limited temperature range of phase change has influenced the usage of the PCMs. The heat pipe is an efficient passive heat transfer device in which heat can be transferred efficiently at a high rate with a very lowtemperature drop in a specified length. Many researchers have been using the heat pipe for battery cooling [34e36]. Feng et al. [36] experimentally examined the thermal and strain management for the cylindrical battery pack at a 0.5e1C rate using heat pipe and air cooling. They found a heat pipe air cooling system in the center of the battery pack is an ideal TMS through a number of tests. Behi et al. [37] numerically considered the effect of the heat pipe with copper sheets (HPCS) on the cylindrical battery module at a 1.5C discharging rate. They found the maximum module temperature in different cooling scenarios comprising forced-air cooling, heat pipe, and HPCS decreased by up to 34.5%, 42.1%, and 42.7%, respectively. Dan et al. [38] experimentally and numerically investigated the micro heat pipe array-air battery cooling for a prismatic battery pack. They found an air velocity of 3 m/s can control and preserve the heat dissipation of the battery pack under the 3C discharging rate. Zhao et al. [39] experimentally investigated the heat pipebased wet cooling system to control the Li-ion prismatic battery pack under a 3C discharging rate. They found that a combination of forced convection and wet cooling approaches can control the temperature of the battery pack in a suitable temperature range with the least cost of energy and water spray. Tran et al. [40] experimentally considered the feasibility of the flat heat pipe air cooling system for the cylindrical battery module. They found a flat heat pipe operates more efficiently compared with a conventional heat sink and kept the module temperature under 50 C. Zhao et al. [41] experimentally considered the effect of a direct evaporative cooling approach for the cylindrical battery module. They proved that the evaporative cooling approach is a capable method to control the maximum cell temperatures, temperature differences, and capacity fading of batteries.
The mentioned relevant studies shown the superior thermal performance of heat pipe based air-cooled systems in battery TMS, in module/pack level. However, all of the existing studies were carried out in low currents and normal conditions as air cooling is not practical in high current applications due to the lack of air heat transfer coefficient. In the current study, several capable and mature heat pipe-based air cooling technology is investigated to control the temperature of an LTO cell/module during an 8C discharging rate (184 A). The suggested charging/discharging current rate of the LTO cell is from 4.6 A to 92 A. Nevertheless, in the present study, the temperature distribution of the cell is studied at a184 A at the maximum current rate. The experimental tests are done to investigate the effect of natural and forced convection on the cell level. A CFD model for forced convection is validated with the experimental results and attained an acceptable agreement. As an optimization, the effect of natural and forced convection is considered on the module level consisting of the 30 cells. Moreover, the effect of evaporative cooling as a developing cooling method in  the ability to dissipate high heat flux at reasonable energy is considered. The heat generated from the cell/module is absorbed by water evaporation through the assist of heat pipes. The results show a consistent and successful cooling procedure from the natural convection to evaporative cooling at the cell/module level. The paper is prepared as follows. Section 2 represents the experimental setup. In Section 3, the experimental results of natural and forced convection are presented. The numerical and thermal modeling, validation, and optimization of the battery cell/module model are described in Section 4. The performance of the thermal management methods at the module level is presented in section 5. A comparison of the proposed cooling systems in the cell and module levels is described in section 6. Lastly, a relevant conclusion is drawn in section 7.

Objective of the research
The temperature rise in Li-ion battery cells, especially at high current applications leads to capacity fade and thermal runaway of the cells. Using an air cooling system for high current applications need an appropriate and developed design to control the maximum battery's cell/module temperature in a safe range (25e40 C). The importance of a cooling system will be more revealed in the module when the cells are closely packed. The design of the current heat pipe based air cooling system is selected for developing a more successful cooling system based on the review of previous works. In the current study, the hybrid cooling system is investigated  experimentally and numerically. In experimental tests, the temperature of the cell is investigated in the lack and presence of natural convection. Moreover, the effect of heat pipe forced convection is considered on the cell temperature. The CFD model is developed using COMSOL Multiphysics and validates with experimental results for further consideration. The effect of natural convection, force convection, and evaporative cooling is simulated with various optimizations, such as the different inlet velocity and humidity, inlet, and outlet location at the cell/module level. The applied designs have changed air cooling as a promising cooling system to control the temperature rise of the cell/module in high current applications. The effect of the air cooling and heat pipe for the Li-ion battery cell/module is considered in various studies. However, there are a few studies to investigate the effect of aircooling, heat pipe, and evaporative cooling for high current applications. Fig. 1 shows the whole experimentation, simulation, and  optimization which are carried out in the flowchart.

Requirements
Fig. 2 displays the experimental setup of the LTO cell in lack of natural convection, presence of natural convection, and heat pipebased air cooling system. The test setup consists of an LTO cell, six flat heat pipes, a cooling fan, a data logger, eight K-type thermocouples, an anemometer, a PEC battery tester, a thermal camera, and a personal computer. The battery tester is a PEC ACT0550 which is capable of supplying up to 5 V DC and 50 A per channel with a ±0.005 accuracy. The charging/discharge cycles, voltage, and current are controlled and recorded by the battery tester while the K-type thermocouples measure the temperature data from the cell. All the thermocouples are calibrated with a maximum tolerance of ±0.2 C. It is essential to mention that the ambient room temperature is fixed at 22 C. The discharging tests are done at 8C with a high current rate of 184 A in a 446-sec current profile.
The LTO prismatic cell (L 22 mm Â H 103 mm Â W 115 mm) has a nominal capacity of 23 Ah and a rated voltage of 2.3 V, as shown in Fig. 2. The flat heat pipe (L 3.5 mm Â H 250 mm Â W 11.2 mm) is made of copper with a sintered wick structure and distilled water as a working fluid. The thermal conductivity of the heat pipes is calculated experimentally at 8212 W/m.K [42]. With the aim of a uniform and capable cooling system, six heat pipes are connected to both sides of the cell. In order to reduce the thermal contact resistance, a thin layer of gap filler with thermal conductivity of 8 W/m.K is applied at the interfaces between the cell and the heat pipes. According to Fig. 3, the heat generation inside the cell and  Table 1 The uncertainties of the experimental parameters.

Parameter
Uncertainty ±0.025 tabs in the discharging process is transferred by the evaporator sections of the heat pipes to the condenser sections and from there by natural/forced convection to the cooling medium. The condensed working fluid returns by capillary force to the evaporator section in the wick structure.
Measuring the internal resistance of the battery is the key parameter to calculate the heat generation of the battery in the CFD model of the battery. The hybrid pulse power characterization (HPPC) test is intended to measure the battery's internal resistance using a test profile that incorporates both discharge and charge pulses. The primary objective of this test is to establish, as a function of the SOC, the current rate, and the DC internal resistance of the three tested cells. The idea of this test is to apply a 10-s discharge-pulse and 10-s charge-pulse power capabilities at each given SOC and for different C-rates. A 600s-rest period is scheduled between each HPPC pulse. From the data, an algorithm will afterward determine the DC internal resistance of the cell. Fig. 4a and b depicts the cell under the HPPC test in the climate chamber at 25 C and the result of the test, respectively.
Equation (1) displays the Schultz and Cole [43,44] method to calculate the uncertainty.
Where U VI and U R are the error of each factor and total errors respectively. The measurement correctness of each factor is shown in Table 1. The maximum uncertainty is calculated as 2.01%.

Experimental results in the cell level
3.1. Investigation of lack of natural convection and natural convection cooling system The first phase of the experiment is conducted under the condition of the lack of natural convection and natural convection at the discharge rate of 8C and the ambient temperature of 22 C. Natural convection has a direct effect on the operating temperatures of the battery and electronic devices. Natural convection is the initial step of temperature distributions within the battery cell/ module in determining heat losses. In natural convection, buoyancy forces cause the fluid motion inside the fluid, whereas, in forced convection, there is an external force. Fig. 5 shows the experimental test setup for LTO cell under the lack of natural convection and natural convection. It is necessary to mention that the whole surface of the cell is insulated precisely to prevent the cooling effect of natural convection.
As it is approved to preserve the best performance and to reach the full lifespan of the battery cells, the temperatures of all the cells must be kept between 25 and 40 C [8]. If the natural convection performs an efficient cooling, there is no need for other kinds of cooling strategies. Fig. 6 shows the location of thermocouples, the temperature of the cell in lack of natural convection, natural convection, and thermal image of the LTO cell during the discharging process.
It is evident in Fig. 6c that the natural convection has a slight effect on the temperature of the cell. The average temperature of the cell in the lack of natural convection reached 57.9 C (Fig. 6b). On the other hand, the average temperature of the cell in the presence of the natural convection reached 54.3 C, which shows a 6.2% reduction. In order to compute the heat transfer coefficient for natural convection, the average temperature (film temperature) of the cell and ambient air is calculated. The Rayleigh number is calculated as follows [45]: where g, b, L, a; and v are body force, volumetric thermal expansion coefficient, height, thermal diffusivity, and kinematic viscosity, respectively. By the estimation of air as an ideal gas, the Rayleigh number is calculated at 1.34 Â e6. The Nusselt number has been where the Pr stands for Prandtl number. According to equation (3), the value of the Nusselt number is calculated at 26.94. Finally, the heat transfer coefficient calculated 6.87 W/m 2 . K as follows: where k is the thermal conductivity of air at the film temperature.

Investigation of forced convection
Forced convection cooling system is one of the most commonly used active cooling methods due to its low cost of maintenance, manufacturing, and simplicity. In the current study, a forced convection cooling system is selected to control an appropriate temperature range for the LTO cell. The cell is equipped with six flat heat pipes to keep the uniform temperature distribution inside the cell. Since there is a non-uniform temperature distribution inside the cell, the location of heat pipes has been chosen in a way to reach better temperature uniformity. The ambient temperature and inlet velocity are set at 22 C and 3 m/s, respectively. The power consumption of the fan is reported 12 W at 3 m/s. Fig. 7 shows the effect of the forced convection cooling system for thermal management of an LTO cell. The average temperature of the cell equipped with heat pipes in the presence of the forced convection reached 35.97 C, which shows a 33.7% reduction compare with the natural convection cooling method.

Thermal modeling
With the purpose of thermal modeling, a combination of Matlab/Simulink with COMSOL Multiphysics is used. The internal heat source of the battery cell is generated due to ohmic resistance and is provided by Matlab/Simulink. For this purpose, an electrical model of the LTO prismatic cell is generated and validated according to the dual-polarization electric-equivalent-circuit (ECM) approach [46]. The model consists of the open-circuit voltage (V oc ) and the seriesconnected ohmic resistance R 0 as well as two parallel R//C branches which represent the time-dependent polarization processes that take place at an applied potential to the battery terminals (V batt ). Fig. 8 shows the proposed impedance model that is used in this study.
In order to obtain the parameters of the proposed ECM, several characterization experiments are conducted. Starting from the precondition and the discharge capacity tests, the capacity at various discharge current rates is obtained. The HPPC is then performed to derive the voltage response of the battery cell at various current rates. A sequence of eight current pulses (C/5, C/4, C/3, C/2, 1C, 2C, 4C, 8C) with 10-sec duration is injected into the LTO cell during both charge and discharge, over the whole operating voltage with a 5% SOC step. The mathematical model formulation in order to parameterize the proposed ECM can be found in Ref. [46]. The open-circuit voltage test is also performed to investigate the V oc at a predefined ambient temperature. The model is validated with the world harmonized light vehicle test procedure (WLTC) and the heat generation of the LTO cell is obtained by considering the Joule losses.
In the direction of the explanation of the transient thermal distribution in the LTO cell, an energy balance equation is used. According to this equation, the thermal energy which is produced inside the cell or is transferred its surrounding is formulated as equation (6): where q g and T describe the heat generation and cell temperature, respectively. C p; k; and m are the specific heat capacity, the thermal conductivity along (x,y,z)-axis, and mass of the cell respectively. Also, the external heat source of the battery cell is generated by tab domains because of the resistive loss. The transient thermal distribution of the tabs are derived as follows [42]: where I batt , R tab , r 0 , l, and A are current, electrical resistance, resistivity, length, and the cross-section of the corresponding tab respectively. q Conv represents the convective heat transfer from the cell to the ambient, which is calculated as follows: where, h, A, T bt and T amb represent the heat transfer coefficient, cross-section area of the cell, battery and ambient temperature, respectively. The temperature transient simulation of the cell in natural and forced convection is shown in Fig. 9. In the following simulation, the heat pipe is replaced by a solid region, and the effective thermal conductivity of components is used in the simulation [29,37,47,48].

Validation based on experimental results
To validate the numerical thermal model of the battery, the simulation and experimental results at the 8C discharging rate are compared under the natural and forced convection cooling methods. The experimental temperatures are measured by eight Ktype thermocouples, as they are shown in Figs. 6a and 7a. The temperature comparison between the experimental and simulation results displays in Fig. 10. It can be seen there is an acceptable agreement between experimental and simulation results for thermocouples of T 1 , T 2, T 3, and T 4 .

. Optimization in cell level by evaporative cooling
A test rig at a cell level is built to validate the numerical simulation results with experimental measurements. As it is approved in Fig. 10, there is an acceptable agreement between the simulation and experimental results. The evaporative cooling method is classified as a phase change cooling that wet surface absorbs heat in order to change from the liquid state to vapor state, resulting in the reduction of heat source temperature. Evaporative cooling is an efficient, economic, and energy-saving cooling method in contrast with conventional bulky liquid cooling systems [49,50]. The vaporization latent heat of the phase change process from liquid to vapor is the main advantage of evaporative cooling compare with the forced convection cooling method [51]. Evaporative cooling is a developed knowledge that has been used in industrial cooling applications such as cooling towers and gas turbines [52]. In this technology, the temperature of the heat source(condenser) is reduced by the evaporation of water from liquid to vapor. The mathematical analysis of the evaporative cooling system is presented in the following equation [53].
where h vs , W asw , h a and K 1 represent the enthalpy of water vapor at water surface temperature, the humidity ratio of air, the enthalpy of the air, and the mass transfer coefficient respectively. Moreover, for calculation of the evaporation heat flux equation (10) can be used as follows: where K, M V ; c V and c sat represent the evaporation rate, the molar mass of water vapor, vapor, and saturation concentration respectively [54]. The offered battery TMS depends on thin flat heat pipes, which can efficiently transfer the heat from the cells to the coolant. In the proposed cooling system, the condenser section of the heat pipes is covered by a wet surface. In fact, the phase change happens inside and outside of the heat pipes. In the proposed cooling method, the relative inlet air humidity has a direct effect on cooling performance. Fig. 11 shows the maximum temperature of the cell in different relative humidity of inlet air at an ambient temperature of 22 C. It is obvious that the temperature of the cell rise by increasing the relative inlet air humidity. The reason returns to the wet surface area where the evaporative cooling is happening. When the value of relative humidity increases, it can decrease the evaporation rate from the heat pipe condenser area which rises the temperature [41,55]. Therefore, 10% of relative humidity is chosen for optimization as the best cooling performance. It is necessary to mention that 10% of relative humidity is not a standard value and requires a refrigerative and/or desiccant air dryer. Fig. 12 shows the comparison of maximum cell temperature at the end of the discharging process in different air velocities and the ambient temperature of 22 C for forced and evaporative cooling. According to Fig. 11, the relative humidity is chosen by 10% for all tests.
It can be seen that increasing the inlet velocity has a direct effect on the maximum temperature of the cell in both cooling methods. It was found that by increasing the inlet velocity the maximum temperature of the cell has been decreased from 44 C to 36.9 C and from 30.8 C to 23.6 C for forced and evaporative cooling respectively. According to the results, there is a 35.8% reduction in the maximum temperature of the cell in the inlet velocity of 3 m/s due to the effect of the evaporative cooling method. For the economical comparison maintaining the fan power at a relatively lower level at which the batteries can operate safely and reliably is more reasonable. As a result, evaporative cooling presents ideal cooling performance whereas ensuring the least amount of power consumption.   Table 2 The main properties of the module.

. Configuration design of the module
In this section, a module consisting of 30 cells is designed to characterize the thermal performance of the heat pipe-based air cooling system. For this aim, a 3D thermal model is developed and solved numerically by COMSOL Multiphysics® to predict the transient temperature rise of cells in the actual battery module. Battery module thermal performance is diff erent from a single battery cell, as cells are influenced by each other during the heat transfer process. The temperature of the cells in the module is significantly affected by the cell arrangement, charge/discharge strategy, and TMS. One of the most common types of battery thermal management systems is air cooling, which has a low cost and simple structure. Several companies such as Toyota Prius, Honda Insight, Nissan, and GM, use forced convection as a thermal management system [56,57]. Air cooling owing to the low heat transfer coefficient of air is not recommended in high current and stressful conditions [58,59]. However, in the current study, the mentioned problem has been reduced using heat pipes. The main specifications of the module are given in Table 2.  Fig . 13a and b shows the geometry and mesh distribution of the LTO prismatic battery module, respectively. Moreover, Fig. 13c displays the inlet and outlet air velocity of the module. In this design, the module combined with heat pipes whiles the evaporator section (115 mm) of heat pipes connected to the cells and condenser section (135 mm) of heat pipes are cooled by a cooling medium.

Boundary condition and mesh grid independency for module level
As it is mentioned in the cell level, the initial temperature of the battery cells, heat pipes, and the coolant are set to 22 C. Moreover, coolant inlet velocity is set to 3 m/s, and the outlet is assumed as the ambient pressure. The heat transfer through radiation is assumed to be negligible. The condenser section of the heat pipes is assumed to be wet for the evaporative cooling method. The 3Dthermal model is developed and solved numerically by COMSOL Multiphysics® using the heat transfer module of a commercial finite element solver, which is computed using the GMRES solver. The time step size is set on the 1 s. Besides, the grid-independent test is performed to refine the grid number while the results are not influenced by further refinement of the mesh, and the results error is kept within 5% [60]. The mesh is generated by default Physics-controlled mesh, which is an unstructured tetrahedral mesh. Fig. 14 displays the maximum simulated temperature of the  The module temperature should be studied in natural convection conditions, to compare with the heat transfer efficiency of the heat pipe-based air cooling system and evaporative cooling method. The temperature distribution inside the module has been considered in the same initial boundary condition as the cell level. Fig. 15 shows the module which cooled through natural convection and experiences a maximum temperature of 58.8 C. The main inability of the cooling method is the low cooling surface area and the limited thermal heat transfer of the natural air convection. The temperature distribution grows non-uniform into the module over the whole surface of the cells with more concentration in the center and top region of the cells.

Effect of the forced convection on the module
With the implementation of a heat pipe-based air cooling system, the maximum temperature of the module significantly decreased. At the 8C discharging rate, the maximum temperature rise decreases from 58.8 C to 49.1 C compared with natural convection. As shown in Fig. 16 when the module is subjected to cool by forced convection and heat pipes, the internal temperature of the battery module is controlled to below 49 C at the end of the discharging process. This suggests that the present cooling system decreased the temperature by 16.4% however, it roughly meets the requirements of battery thermal management. To further improve the performance of the present cooling system and to meet the tight requirement on maintaining the module temperature in a safe range [61] several scenarios like evaporative cooling, increasing the flow rate, or decreasing the airflow temperature could be implemented.

Effect of the evaporative cooling on the module
This section represents a transient simulation to estimate the temperature rise of the battery module under the 8C discharging rate and with inlet air temperature and velocity of 22 C and 3 m/s, respectively. According to Fig. 17, the maximum temperature of the module reached 44.8 C. As compared to the temperature of the module with the natural cooling system, the overall surface temperature of the module is reduced by 23.8% using the evaporative cooling method. The simulation results further confirm that evaporative cooling is a possible cooling method for the development of the Li-ion battery module/pack TMS. 5.2.3.1. Investigation of temperature uniformity in the evaporative cooling on the module. In the last section, evaporative cooling is introduced as a possible cooling method that keeps the maximum temperature of the module at 44.8 C. However, the temperature uniformity is not suitable and the cell temperatures increase from the inlet to the outlet. Therefore, the optimization of the inlet and outlet of the airflow is done to improve temperature uniformity and reduce maximum module temperature. Table 3 shows the battery TMS with inlet and outlet optimization in three different cases with an inlet velocity of 3 m/s. Case 1 represents the existing model which suffers from temperature uniformity due to the position of the inlet and outlet and a large number of cells in the module. Case 2 displays the inlet in front of the module with two outlets from the sides. It can be seen, the maximum temperature has been decreased. However, there is a temperature increase in the middle of the module owing to the tendency of the airflow to the outlets. In case 3 the inlet is located at the top and the outlet located on the downside of the module. The maximum temperature of the module has been decreased by 26.8 C which shows a 45.5% reduction. Moreover, there is an acceptable temperature uniformity inside the module. The thermal performance of Li-ion batteries varies with time at the cell and module level. Therefore, the ability of a TMS to preserve the cell/module temperature under different working conditions is vital. In order to investigate the efficiency of the evaporative cooling method and compare the results with forced and natural convection, the average temperature of the cell in different time steps is shown in Fig. 18. In the baseline condition, the single-cell experienced dramatic temperature growth and reached 53.66 C. The temperature of the cell has been decreased significantly by equipping the heat pipe-based air cooling system and reached 36.88 C. The evaporative cooling system shows a perfect cooling performance and maintains the temperature of the cell in the range of 22 C. It is obvious that at the primary stages of discharge, the temperature of the cell cooled and experiences slight temperature drops owing to the evaporation effect which improves convective heat dissipation.

Module-level
According to Fig. 19, for natural convection, the module level experienced a more severe temperature increase and reached 56.6 C. The average temperature of the module increased from 22 C to 56.6 C which shows a 61.1% increase. To control and decrease the heat accumulation of the module the heat pipe-based air cooling system is carried out on the battery module. The temperature of the module has been decreased to 45.61 C by the mentioned cooling system. The evaporative cooling system shows a reasonable cooling performance and decreased the average temperature of the module to 37.92 C. As it was proved the forced convection and evaporative cooling are successful cooling systems for the module level in the 8C discharging rate. Therefore, the investigation of different inlet velocity effects on the maximum temperature of the module is an important item. Fig. 20 shows that increasing the inlet velocity from 1 m/s to 7 m/s has a direct consequence on the maximum temperature of the module in both cooling methods. The trend of the maximum temperature of the module has been decreased from 51 C to 46.7 C and from 47.6 C to 40.3 C for forced and evaporative cooling respectively. According to the results, there is an average of 10% reduction in the maximum temperature of the module due to the effect of the evaporative cooling method. The present study shows that the evaporative cooling method can preserve the maximum temperature of the module in a safe range (25e40 C) with an inlet velocity of 7 m/s.

Conclusion
The current study experimentally and numerically investigated a heat pipe-based air cooling TMS for the LTO battery cell/module. A number of optimization scenarios were developed. In the first scenario, the temperature of the cell was considered experimentally in the lack and presence of natural convection. In the second scenario, the effect of the heat pipe-based air cooling system was investigated. It was found that the average cell temperature reached from 54.3 C to 35.9 C, which showed a 33.7% reduction.  Finally, in the third scenario, the effect of evaporative cooling was considered for the cell/module level. According to the results, there was a 35.8% reduction in the maximum temperature of the cell due to the effect of the evaporative cooling method. Moreover, the maximum temperature of the battery module was reduced by 23.8% compared with the natural convection.

Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.