In my research, I focused on the thermal management of an electric vehicle battery pack because the performance, safety, and lifetime of lithium-ion cells are strongly affected by operating temperature. An electric vehicle battery pack generates heat during charging and discharging, and the heat generation rate depends on the current, internal resistance, electrochemical reactions, and the specific charging or discharging mode. If the heat cannot be removed efficiently, the electric vehicle battery pack can experience local overheating, poor temperature uniformity, accelerated aging, and even thermal runaway. My goal was to improve passive thermal management by coupling phase change materials with tree-shaped fins and nanomaterials, and I studied the problem through experiments and numerical simulations.
The heat generation inside a lithium-ion cell can be summarized by several contributions. The total heat generation rate is
$$Q_{\mathrm{total}} = Q_r + Q_p + Q_j + Q_s$$
where \(Q_r\) is the reaction heat, \(Q_p\) is the polarization heat, \(Q_j\) is the Joule heat, and \(Q_s\) is the side-reaction heat, which is often neglected in practical thermal analysis. The reaction heat can be approximated as
$$Q_r = \frac{Q}{3600F} I$$
where \(Q\) is the total reaction heat, \(F\) is the Faraday constant, and \(I\) is the current. The polarization and Joule heat terms are
$$Q_p = I^2 R_p$$
$$Q_j = I^2 R_j$$
where \(R_p\) and \(R_j\) are the polarization resistance and ohmic resistance, respectively. These equations show why an electric vehicle battery pack can produce more heat at high current rates and why thermal management is essential for high-power charging and discharging.
I reviewed the main cooling strategies for an electric vehicle battery pack, including air cooling, liquid cooling, heat pipe cooling, and phase change material cooling. Table 1 compares these methods from the perspective of thermal performance, complexity, safety, and suitability for an electric vehicle battery pack.
| Cooling method | Advantages | Limitations | Typical application in an electric vehicle battery pack |
|---|---|---|---|
| Air cooling | Simple, low cost, no leakage risk | Low heat capacity, weak cooling, strong dependence on flow path | Mild climate and low-power electric vehicle battery pack |
| Liquid cooling | High heat capacity, high thermal conductivity, compact cold plates | Leakage risk, pumping power, complex sealing | High-power electric vehicle battery pack |
| Heat pipe cooling | Passive operation, high effective conductivity, no external power | Complex integration, orientation sensitivity, difficult conformal contact | Local hot-spot control in an electric vehicle battery pack |
| Phase change material cooling | Near-isothermal heat absorption, large latent heat, passive thermal buffer | Low thermal conductivity, low heat transfer rate, volume expansion | Passive and hybrid thermal management of an electric vehicle battery pack |
Among these methods, phase change material cooling is attractive for an electric vehicle battery pack because it can absorb a large amount of heat while maintaining a relatively stable temperature. However, the low thermal conductivity of phase change materials limits the heat transfer rate. I therefore introduced tree-shaped fins and carbon nanotubes to enhance heat conduction and natural convection in the phase change material.
Experimental Study of Phase Change Material Coupled with Fins for an Electric Vehicle Battery Pack
I built an experimental platform for an electric vehicle battery pack thermal management study. Four 26650 lithium-ion cells were connected in series, and they were placed in an acrylic chamber with internal dimensions of \(200 \times 200 \times 65\ \mathrm{mm}\). The chamber wall thickness was \(15\ \mathrm{mm}\). The cells were arranged around replaceable fin structures. I tested three configurations: a no-fin circular tube, a rectangular fin, and a tree-shaped fin. The tree-shaped fin had four fractal levels and a branch angle of \(90^\circ\). The fin material was aluminum 6061, and the phase change material was RT35. Table 2 lists the measured thermal properties of the phase change material.
| Property | RT35 phase change material |
|---|---|
| Melting onset temperature | \(35.3^\circ\mathrm{C}\) |
| Melting end temperature | \(41.0^\circ\mathrm{C}\) |
| Solidification onset temperature | \(28.3^\circ\mathrm{C}\) |
| Solidification end temperature | \(33.0^\circ\mathrm{C}\) |
| Latent heat | Approximately \(200\ \mathrm{kJ/kg}\) |
| Specific heat | Approximately \(2.2\ \mathrm{kJ/(kg\cdot^\circ C)}\) |
| Thermal conductivity | \(0.21\ \mathrm{W/(m\cdot K)}\) |
I designed the fin dimensions so that the total fin volume and total fin length were comparable among cases. Table 3 gives the fin length and thickness values used in my experiments.
| Fin type | Fin length (mm) | Fin thickness (mm) |
|---|---|---|
| No fin | 0 | 0 |
| Rectangular fin | 27.4 | 4.0 |
| Tree fin level 1 | 12.0 | 2.0 |
| Tree fin level 2 | 8.5 | 1.6 |
| Tree fin level 3 | 6.0 | 1.3 |
| Tree fin level 4 | 4.2 | 1.0 |
I used three charging modes: 1C charging, constant-current charging, and constant-current constant-voltage charging. The ambient temperatures were \(-10^\circ\mathrm{C}\), \(25^\circ\mathrm{C}\), \(40^\circ\mathrm{C}\), and \(50^\circ\mathrm{C}\). These conditions were selected to represent cold, normal, warm, and hot environments for an electric vehicle battery pack.
My experimental results showed that the tree-shaped fin coupled with phase change material gave the best thermal performance. In a low-temperature environment, the phase change material released stored latent heat and insulated the cells, so the minimum cell temperature was higher than in the no-phase-change case. The minimum cell temperature in the phase change box was up to 2.4 times that of the case without thermal management. In a normal-temperature environment, the tree-shaped fin reduced the maximum cell temperature to about 0.9 times that of the no-management case and extended the charge-discharge time by \(27.5\%\). In a high-temperature environment, the phase change box maintained the maximum cell temperature below \(40^\circ\mathrm{C}\) even when the ambient temperature was \(50^\circ\mathrm{C}\).
| Ambient temperature | Charging mode | Best fin configuration | Maximum temperature reduction | Operating time improvement |
|---|---|---|---|---|
| \(-10^\circ\mathrm{C}\) | 1C | Tree fin | Insulation effect; minimum temperature up to \(19.2^\circ\mathrm{C}\) | Longer discharge duration |
| \(25^\circ\mathrm{C}\) | 1C | Tree fin | Reduced to 0.9 times the baseline | \(27.5\%\) |
| \(40^\circ\mathrm{C}\) | Constant current | Tree fin | Up to \(22.9\%\) lower maximum temperature | Longer operating time |
| \(50^\circ\mathrm{C}\) | Constant current constant voltage | Tree fin | Maximum temperature below \(40^\circ\mathrm{C}\) | Longer operating time |
The charging and discharging modes also influenced the thermal response of the electric vehicle battery pack. Under constant-current constant-voltage charging, the cell remained near full charge for a longer time, so the phase change material had more time to absorb heat. The tree-shaped fin was especially effective in this mode because its branched structure spread heat into a larger region of phase change material. The rectangular fin also improved cooling compared with the no-fin case, but its heat transfer path was shorter and less distributed.
Visualization Experiment with a Simulated Battery for an Electric Vehicle Battery Pack
To observe the melting process directly, I replaced the 26650 cells with electric heating rods that had the same dimensions as the cells. These simulated batteries allowed me to visualize the solid-liquid interface inside the phase change material. I used heating powers of \(10\ \mathrm{W}\) and \(15\ \mathrm{W}\), and ambient temperatures of \(25^\circ\mathrm{C}\) and \(40^\circ\mathrm{C}\). The phase change material was again RT35, and I tested no-fin, rectangular-fin, and tree-fin configurations.
I recorded the liquid fraction evolution and found three heat transfer stages in the phase change material. In the first stage, heat conduction dominated and the temperature rose linearly until the melting onset temperature. In the second stage, natural convection became strong, and the temperature rose rapidly while the liquid fraction increased. In the third stage, natural convection weakened, and the temperature rise slowed. Because of gravity, the upper region of the phase change box melted earlier than the lower region. The upper simulated batteries were therefore warmer than the lower simulated batteries, while batteries at the same height had similar temperatures.
The solid-liquid interface moved outward from the fin surface and upward due to buoyancy. In the no-fin case, the melting front was nearly circular around the tube, but the melting rate was slow. In the rectangular-fin case, the melting front followed the fin surfaces and created additional heat flow paths. In the tree-fin case, the branched structure produced multiple heat flow paths, so the phase change material melted more uniformly and the simulated battery temperature remained lower for a longer time. Table 4 summarizes the visualization results.
| Heating power | Ambient temperature | Best configuration | Observation | Working time extension |
|---|---|---|---|---|
| \(10\ \mathrm{W}\) | \(25^\circ\mathrm{C}\) | Tree fin | Largest liquid fraction; slowest temperature rise | Up to \(54.9\%\) compared with the baseline |
| \(15\ \mathrm{W}\) | \(25^\circ\mathrm{C}\) | Tree fin | Strong convection; upper region melted first | Longer than rectangular and no-fin cases |
| \(10\ \mathrm{W}\) | \(40^\circ\mathrm{C}\) | Tree fin | High ambient temperature accelerated melting | Longer than other cases |
| \(15\ \mathrm{W}\) | \(40^\circ\mathrm{C}\) | Tree fin | Shortest working time but still best among cases | Up to \(1.1\%\) over rectangular fin |
I found that the heating power had a stronger effect on the simulated battery working time than the ambient temperature. For example, the \(40^\circ\mathrm{C}\), \(10\ \mathrm{W}\) case lasted longer than the \(25^\circ\mathrm{C}\), \(15\ \mathrm{W}\) case. This is relevant for an electric vehicle battery pack because rapid acceleration, uphill driving, and fast charging can generate high heat rates that dominate the thermal response. The tree-shaped fin was beneficial in all cases because it increased the heat transfer area and distributed heat more evenly through the phase change material.
Numerical Study of Tree Fin Fractal Level and Branch Angle for an Electric Vehicle Battery Pack
After the experiments, I performed three-dimensional numerical simulations to optimize the tree-shaped fin geometry for an electric vehicle battery pack. I designed tree fins with fractal levels 2, 3, and 4, and branch angles of \(60^\circ\), \(90^\circ\), \(120^\circ\), and \(150^\circ\). I also simulated a rectangular fin and a no-fin pure phase change material case. The battery diameter was \(26\ \mathrm{mm}\), the battery height was \(65\ \mathrm{mm}\), the shell inner diameter was \(65\ \mathrm{mm}\), and the shell thickness was \(2\ \mathrm{mm}\). The phase change material was OP44E, and the fin and shell material was aluminum. Table 5 lists the thermophysical properties used in the model.
| Property | OP44E | Battery | Aluminum |
|---|---|---|---|
| Thermal conductivity, \(\mathrm{W/(m\cdot K)}\) | 0.18–0.20 | 32.2 | 211 |
| Density, \(\mathrm{kg/m^3}\) | 700 | 2007.7 | 2675 |
| Specific heat, \(\mathrm{J/(kg\cdot K)}\) | 2000–2350 | 837.4 | 903 |
| Latent heat, \(\mathrm{J/kg}\) | 240800 | – | – |
| Melting range, \(^\circ\mathrm{C}\) | 40.6–44.7 | – | – |
| Thermal expansion coefficient, \(1/\mathrm{K}\) | 0.00076 | – | – |
| Dynamic viscosity, \(\mathrm{kg/(m\cdot s)}\) | 0.005 | – | – |
I used the enthalpy-porosity method with the mushy zone treated as a porous medium. The continuity, momentum, and energy equations were
$$\nabla \cdot (\rho \mathbf{u}) = 0$$
$$\frac{\partial (\rho \mathbf{u})}{\partial t} + \nabla \cdot (\rho \mathbf{u}\mathbf{u}) = -\nabla p + \nabla \cdot (\mu \nabla \mathbf{u}) + \rho \mathbf{g} + \mathbf{S}$$
$$\frac{\partial (\rho H)}{\partial t} + \nabla \cdot (\rho \mathbf{u} H) = \nabla \cdot (k \nabla T)$$
The total enthalpy was
$$H = h + \Delta H$$
where
$$\Delta H = \gamma L$$
and the sensible enthalpy was
$$h = h_{\mathrm{ref}} + \int_{T_{\mathrm{ref}}}^{T} C_p \, dT$$
The liquid fraction was defined as
$$\gamma =
\begin{cases}
0, & T < T_{\mathrm{solid}} \\
\dfrac{T – T_{\mathrm{solid}}}{T_{\mathrm{liquid}} – T_{\mathrm{solid}}}, & T_{\mathrm{solid}} \le T \le T_{\mathrm{liquid}} \\
1, & T > T_{\mathrm{liquid}}
\end{cases}$$
The momentum source term in the mushy zone was
$$\mathbf{S} = \frac{(1-\gamma)^2}{\gamma^3 + \varepsilon} A_{\mathrm{mush}} \mathbf{u}$$
where \(A_{\mathrm{mush}} = 10^7\) and \(\varepsilon = 0.001\). I used the Boussinesq approximation for buoyancy:
$$\rho = \rho_0 \left[1 – \beta (T – T_m)\right]$$
I used the PISO pressure-velocity coupling scheme, PRESTO pressure discretization, and second-order upwind discretization for energy and momentum. The convergence criteria were \(10^{-4}\) for continuity, \(10^{-5}\) for momentum, and \(10^{-6}\) for energy. I verified the model against published experimental and numerical data and found average absolute deviations of \(1.9\%\) and \(2.8\%\), respectively. I also performed time-step and grid independence studies. A time step of \(0.25\ \mathrm{s}\) and a mesh size of \(0.5\ \mathrm{mm}\) were chosen.
The simulation results showed that for a fixed fractal level, a larger branch angle produced a lower average battery temperature and better temperature uniformity. For a fixed branch angle, a higher fractal level produced a lower average battery temperature and better temperature uniformity. The best case was the four-level, \(150^\circ\) tree fin, which I denote as 4-150. Table 6 compares the total melting time and average maximum velocity for representative cases.
| Case | Total melting time (min) | Average maximum velocity (mm/s) | Power (W) |
|---|---|---|---|
| No fin | 93.5 | 2.68 | 6.07 |
| Rectangular fin | 85.0 | 1.94 | 6.60 |
| 2-60 | 85.0 | 1.80 | 6.61 |
| 2-90 | 85.0 | 1.68 | 6.62 |
| 2-120 | 85.0 | 1.55 | 6.62 |
| 2-150 | 85.0 | 1.43 | 6.61 |
| 3-60 | 85.0 | 1.76 | 6.61 |
| 3-90 | 84.5 | 1.54 | 6.64 |
| 3-120 | 84.0 | 1.44 | 6.70 |
| 3-150 | 83.5 | 1.47 | 6.69 |
| 4-60 | 84.5 | 1.54 | 6.64 |
| 4-90 | 83.5 | 1.55 | 6.71 |
| 4-120 | 83.0 | 1.42 | 6.74 |
| 4-150 | 82.5 | 1.36 | 6.75 |
The 4-150 case reduced the total melting time by \(11.8\%\), reduced the battery temperature by \(4.5\%\), increased the average melting velocity by \(1.71\) times, and increased the thermal power by \(11.2\%\) compared with the no-fin case. The average battery temperature in the 4-150 case was the lowest among all cases. The temperature uniformity was also the best, with a maximum temperature difference below \(1.6^\circ\mathrm{C}\). The tree-shaped fin improved heat conduction into the phase change material and created additional natural convection paths, although the fins also suppressed some axial natural convection. This trade-off is why the no-fin case had the highest maximum phase change material velocity but the poorest overall thermal performance.
I calculated the stored energy in the phase change material as
$$Q_P = \rho V \left(C_P \Delta T + \gamma L\right)$$
where \(\rho\) is the density, \(V\) is the volume, \(C_P\) is the average specific heat, \(\Delta T\) is the temperature rise, \(\gamma\) is the liquid fraction, and \(L\) is the latent heat. The power of the thermal management system was
$$P = \frac{Q}{t}$$
where \(Q\) is the stored energy and \(t\) is the total melting time. I also defined the average maximum velocity as
$$V_{\mathrm{avg,max}} = \frac{\int_{0}^{t_{\mathrm{total}}} V_{\max} \, dt}{t_{\mathrm{total}}}$$
These metrics helped me compare the thermal response of each electric vehicle battery pack thermal management design.
Numerical Study of Fin Quantity and Nano-Phase Change Material for an Electric Vehicle Battery Pack
Based on the previous results, I selected the four-level, \(150^\circ\) tree fin and studied the effect of fin quantity and multi-walled carbon nanotube concentration. I tested no fin, rectangular fin, two tree fins, three tree fins, and four tree fins. I also added multi-walled carbon nanotubes at mass fractions of \(0\%\), \(3\%\), and \(6\%\). The base phase change material was still OP44E, and the fin material was aluminum. Table 7 lists the properties of the multi-walled carbon nanotubes and the base phase change material.
| Property | MWCNT | Base PCM |
|---|---|---|
| Density, \(\mathrm{kg/m^3}\) | 1600 | 700 |
| Specific heat, \(\mathrm{J/(kg\cdot K)}\) | 796 | 2000–2350 |
| Thermal conductivity, \(\mathrm{W/(m\cdot K)}\) | 3000 | 0.18–0.20 |
| Dynamic viscosity, \(\mathrm{Pa\cdot s}\) | – | 0.005 |
| Thermal expansion coefficient, \(1/\mathrm{K}\) | 0.000019 | 0.00076 |
| Latent heat, \(\mathrm{J/kg}\) | – | 240800 |
I calculated the nano-phase change material properties using mixture relations. The density was
$$\rho_{\mathrm{N-PCM}} = \phi \rho_{\mathrm{MWCNT}} + (1-\phi) \rho_{\mathrm{PCM}}$$
The specific heat was
$$(\rho C_P)_{\mathrm{N-PCM}} = \phi (\rho C_P)_{\mathrm{MWCNT}} + (1-\phi)(\rho C_P)_{\mathrm{PCM}}$$
The thermal conductivity was estimated by
$$k_{\mathrm{N-PCM}} = k_{\mathrm{PCM}} \left[ \frac{k_{\mathrm{MWCNT}} + 2k_{\mathrm{PCM}} – 2\phi(k_{\mathrm{PCM}} – k_{\mathrm{MWCNT}})}{k_{\mathrm{MWCNT}} + 2k_{\mathrm{PCM}} + \phi(k_{\mathrm{PCM}} – k_{\mathrm{MWCNT}})} \right]$$
The thermal expansion coefficient was
$$(\rho \beta)_{\mathrm{N-PCM}} = \phi (\rho \beta)_{\mathrm{MWCNT}} + (1-\phi)(\rho \beta)_{\mathrm{PCM}}$$
The viscosity was
$$\mu_{\mathrm{N-PCM}} = \frac{\mu_{\mathrm{PCM}}}{(1-\phi)^{2.5}}$$
and the latent heat was
$$(\rho L)_{\mathrm{N-PCM}} = (1-\phi)(\rho L)_{\mathrm{PCM}}$$
I found that increasing the nanotube mass fraction reduced the maximum battery temperature and improved temperature uniformity, but it also reduced the total stored energy and the average maximum phase change material velocity. The best combination was \(6\%\) MWCNT with the four-fin, \(150^\circ\) tree fin. Table 8 summarizes the key thermal results for different nanotube concentrations and fin configurations.
| Configuration | Maximum battery temperature (\(^\circ\mathrm{C}\)) | Total melting time (min) | Average battery temperature uniformity | Power (W) |
|---|---|---|---|---|
| 0% MWCNT, no fin | 68.4 | 93.5 | 2.18 | 6.07 |
| 3% MWCNT, no fin | 67.8 | 91.0 | 1.85 | 6.08 |
| 6% MWCNT, no fin | 67.2 | 88.5 | 1.81 | 6.09 |
| 0% MWCNT, rectangular fin | 61.2 | 85.0 | 1.04 | 6.60 |
| 3% MWCNT, rectangular fin | 60.1 | 83.0 | 1.02 | 6.59 |
| 6% MWCNT, rectangular fin | 59.5 | 81.0 | 1.00 | 6.58 |
| 0% MWCNT, 150°-2 | 60.5 | 87.0 | 1.42 | 6.51 |
| 3% MWCNT, 150°-2 | 59.8 | 85.0 | 1.40 | 6.51 |
| 6% MWCNT, 150°-2 | 59.2 | 83.0 | 1.37 | 6.49 |
| 0% MWCNT, 150°-3 | 59.0 | 85.0 | 1.17 | 6.63 |
| 3% MWCNT, 150°-3 | 58.5 | 83.0 | 1.15 | 6.61 |
| 6% MWCNT, 150°-3 | 58.0 | 81.0 | 1.12 | 6.60 |
| 0% MWCNT, 150°-4 | 58.5 | 82.5 | 0.96 | 6.75 |
| 3% MWCNT, 150°-4 | 58.0 | 80.5 | 0.94 | 6.74 |
| 6% MWCNT, 150°-4 | 57.9 | 79.0 | 0.92 | 6.70 |
The \(6\%\) MWCNT and 150°-4 case reduced the maximum battery temperature by up to \(15.3\%\) compared with the baseline. The phase change material temperature remained below \(56^\circ\mathrm{C}\). The total melting time was \(79\ \mathrm{min}\), which was \(15.5\%\) shorter than the baseline. The average battery temperature uniformity was \(0.92\), the lowest among all cases. The power was \(6.70\ \mathrm{W}\), and relative to the 0% MWCNT no-fin case, the power increase was \(1.11\) times. I also observed multiple peaks in the maximum phase change material velocity for the 150°-2 case, with values of \(3.81\ \mathrm{mm/s}\) at \(3\%\) MWCNT and \(3.72\ \mathrm{mm/s}\) at \(6\%\) MWCNT. These peaks occurred between 40 and 60 minutes.
My numerical results confirmed that the combination of tree-shaped fins and carbon nanotubes is an effective way to improve the thermal performance of an electric vehicle battery pack. The fins increase the heat transfer area and create preferential heat conduction paths, while the nanoparticles increase the effective thermal conductivity of the phase change material. However, adding nanoparticles also increases viscosity and reduces the latent heat per unit mass, so there is an optimum concentration. In my study, \(6\%\) MWCNT gave the best temperature reduction and the shortest melting time, while \(0\%\) MWCNT gave the highest stored energy and the highest average maximum velocity.
Comparison of Heat Transfer Mechanisms in the Electric Vehicle Battery Pack
The heat transfer in my phase change material system involved conduction, weak convection, and strong convection. During early melting, conduction dominated because the liquid fraction was low. The fin structure provided a high-conductivity path from the battery to the phase change material. As melting progressed, natural convection became important, especially in the upper region because of buoyancy. The tree-shaped fin distributed the heat more uniformly than the rectangular fin, so the temperature gradients were smaller and the melting front advanced more evenly.
The melting process can be described by the energy equation and the liquid fraction. The local heat transfer rate was influenced by the fin surface area, the thermal conductivity ratio between the fin and phase change material, and the Rayleigh number. The Rayleigh number can be expressed as
$$Ra = \frac{g \beta \Delta T L^3}{\nu \alpha}$$
where \(g\) is gravitational acceleration, \(\beta\) is the thermal expansion coefficient, \(\Delta T\) is the temperature difference, \(L\) is a characteristic length, \(\nu\) is the kinematic viscosity, and \(\alpha\) is the thermal diffusivity. A higher Rayleigh number indicates stronger natural convection. In my system, the tree-shaped fin increased the local temperature difference near the fin but reduced the overall temperature difference in the phase change material, which improved uniformity.
The Nusselt number can be used to evaluate convective heat transfer:
$$Nu = \frac{h L}{k}$$
where \(h\) is the convective heat transfer coefficient and \(k\) is the thermal conductivity. The fin-enhanced system had a higher effective Nusselt number because the fin increased the surface area and disrupted the thermal boundary layer. The carbon nanotubes further increased the effective thermal conductivity, which increased the conduction component of heat transfer.
Design Implications for an Electric Vehicle Battery Pack
Based on my experiments and simulations, I can draw several design implications for an electric vehicle battery pack. First, a tree-shaped fin with four fractal levels and a branch angle of \(150^\circ\) provides the best balance between heat conduction and natural convection. Second, adding multi-walled carbon nanotubes at \(6\%\) mass fraction improves the effective thermal conductivity and reduces the maximum battery temperature, but it reduces the total latent heat storage. Third, the phase change material should be selected so that its melting range is close to the desired operating temperature of the electric vehicle battery pack. Fourth, the thermal management system should be designed to accommodate volume expansion during melting and to avoid leakage. Fifth, the fin structure should be integrated with the battery pack casing to minimize thermal contact resistance.
I also found that the ambient temperature and heat generation rate strongly affect the performance of the electric vehicle battery pack. In cold environments, the phase change material can provide insulation and keep the battery warmer. In hot environments, the phase change material absorbs heat and limits the temperature rise. In normal environments, the phase change material reduces temperature fluctuations during charging and discharging. The tree-shaped fin and carbon nanotubes enhance these benefits by improving heat transfer in all directions.
Summary of Key Findings
| Study | Main variable | Best configuration | Key result |
|---|---|---|---|
| Cell experiment | Fin type, ambient temperature, charging mode | Tree fin | Maximum temperature reduced to 0.9 times baseline; charge-discharge time increased by \(27.5\%\) |
| Visualization experiment | Heating power, ambient temperature, fin type | Tree fin | Working time extended by up to \(54.9\%\) |
| Numerical fin optimization | Fractal level and branch angle | 4-150 | Melting time reduced by \(11.8\%\); battery temperature reduced by \(4.5\%\); power increased by \(11.2\%\) |
| Numerical nano-PCM study | Fin quantity and MWCNT concentration | 150°-4 with \(6\%\) MWCNT | Battery temperature reduced by up to \(15.3\%\); melting time reduced by \(15.5\%\); uniformity \(0.92\) |
Conclusions
I studied the thermal management of an electric vehicle battery pack using phase change materials coupled with tree-shaped fins and carbon nanotubes. My experiments showed that tree-shaped fins improve heat transfer and temperature uniformity compared with rectangular fins and no-fin cases. The phase change material provided thermal buffering in cold, normal, and hot environments. My visualization experiments revealed the solid-liquid interface evolution and the importance of buoyancy-driven convection. My numerical simulations identified the optimal tree fin geometry as four fractal levels with a \(150^\circ\) branch angle. Adding \(6\%\) multi-walled carbon nanotubes further improved the thermal performance by increasing the effective thermal conductivity, reducing the maximum battery temperature, and shortening the melting time.
The main conclusions are as follows. The tree-shaped fin significantly enhanced heat transfer in the electric vehicle battery pack. The four-level, \(150^\circ\) tree fin gave the lowest battery temperature and the best temperature uniformity. The carbon nanotube additive improved thermal conductivity but reduced the latent heat and increased viscosity, so the concentration must be optimized. The combination of tree-shaped fins and nano-phase change material is a promising passive thermal management strategy for an electric vehicle battery pack. Future work should consider different phase change materials, larger battery packs, real driving cycles, and topology optimization of the fin structure. I believe that these findings can help improve the safety, lifetime, and performance of an electric vehicle battery pack.