The quest for extended driving range remains a pivotal challenge in the widespread adoption of electric vehicles (EVs). Given the current state of battery technology, where energy density improvements are incremental, systemic lightweighting offers a direct and effective pathway to enhance efficiency and mileage. Among the various subsystems of an EV, the battery system constitutes approximately 30% of the total vehicle mass. Within this system, the enclosure or pack structure, which houses and protects the battery modules, is a significant mass contributor besides the cells themselves. Therefore, focused lightweight design of the EV battery pack enclosure has emerged as a critical research area. This article presents a comprehensive study on the application of Carbon Fiber Reinforced Plastic (CFRP) to replace traditional metallic materials for EV battery pack enclosures, detailing a multi-stage optimization framework to achieve substantial mass reduction while enhancing mechanical performance.
Conventional optimization studies on metallic enclosures have employed techniques like topography optimization and genetic algorithms, achieving mass reductions typically under 20%. However, the inherent density of metals like steel and aluminum sets a fundamental limit. Advanced composite materials, particularly CFRP, present a paradigm shift. CFRP offers an exceptional strength-to-weight and stiffness-to-weight ratio, corrosion resistance, and excellent fatigue properties. Its anisotropic nature also allows for directional tailoring of stiffness and strength, which is a powerful advantage in structural design. The primary objective of this work is to leverage these properties, designing a CFRP EV battery pack enclosure that significantly outperforms its metallic counterpart in both static/dynamic performance and mass efficiency. The methodology involves detailed finite element analysis (FEA) to establish baseline performance, followed by sequential and integrated optimization of the upper cover and lower tray, culminating in a multi-objective layup optimization.
Finite Element Performance Analysis of the Composite EV Battery Pack
The foundation of any reliable optimization process is a robust numerical model that accurately predicts structural behavior. The baseline model of the EV battery pack enclosure was established with overall dimensions of 1363 mm × 1152 mm × 182 mm. The structure primarily consists of an upper cover, a lower tray (or box), and internal/external reinforcement beams made of Q235 steel for mounting and localized stiffness. For computational efficiency, minor geometric features such as small fillets, holes, and non-structural bosses were simplified. The core innovation lies in the material assignment: the upper cover and lower tray were modeled as composite laminates using CFRP properties, while the reinforcement structures remained steel. The material properties used in the analysis are summarized below.
| Material | Parameter | Value | Parameter | Value |
|---|---|---|---|---|
| CFRP (Unidirectional Ply) | Longitudinal Modulus, $E_1$ | 140 GPa | Transverse Modulus, $E_2$ | 8.4 GPa |
| Poisson’s Ratio, $\nu_{12}$ | 0.27 | Density, $\rho$ | $1.8 \times 10^3$ kg/m³ | |
| In-plane Shear Modulus, $G_{12}$ | 6.8 GPa | Out-of-plane Shear Modulus, $G_{z}$ | 5.4 GPa | |
| Longitudinal Tensile Strength, $X_t$ | 1520 MPa | Longitudinal Compressive Strength, $X_c$ | 1200 MPa | |
| Transverse Tensile Strength, $Y_t$ | 60 MPa | Transverse Compressive Strength, $Y_c$ | 193 MPa | |
| Q235 Steel | Young’s Modulus, $E$ | 210 GPa | Density, $\rho$ | $7.85 \times 10^3$ kg/m³ |
| Poisson’s Ratio, $\nu$ | 0.3 | Yield Strength | 235 MPa |
The initial laminate configuration for the CFRP parts was set as a symmetric and balanced stack of $[0^\circ/45^\circ/-45^\circ/90^\circ]_s$, with each ply group assigned an initial thickness of 0.75 mm, resulting in a total laminate thickness of 3.0 mm for both the cover and tray.
Static Performance under Typical Load Cases
The EV battery pack experiences complex loading during service. Three critical quasi-static load cases, representing combined inertial loads from maneuvering and road inputs, were defined to assess structural integrity:
- Bumpy Road & Sharp Turn: 3g vertical acceleration + 0.8g lateral acceleration.
- Bumpy Road & Emergency Braking: 3g vertical acceleration + 1.0g longitudinal acceleration.
- Bumpy Road & Hard Acceleration: 3g vertical acceleration + 0.5g longitudinal acceleration.
The accelerations were converted into equivalent static forces distributed over the nodes on the bottom and sidewalls of the EV battery pack enclosure, simulating the inertia of the battery modules. The mounting points (e.g., brackets) were constrained in all translational degrees of freedom. The results from the linear static analysis served as a critical performance baseline.
| Load Case | Maximum Displacement (mm) | Maximum Stress (MPa) & Location |
|---|---|---|
| Bumpy Road & Sharp Turn | 7.23 | 191.0 (Mounting Bracket) |
| Bumpy Road & Emergency Brake | 7.25 | 188.8 (Mounting Bracket) |
| Bumpy Road & Hard Acceleration | 7.22 | 195.3 (Mounting Bracket) |
The analysis revealed that the maximum displacements occurred at the center region of the lower tray, while stress concentrations were consistently found at the steel mounting brackets (which remained the critical metallic parts). The stress in the composite parts was well below their strength limits, indicating initial over-design from a strength perspective but highlighting stiffness as a governing factor.
Dynamic Characteristics: Modal Analysis
To avoid resonant failure caused by road or powertrain excitations, the natural frequencies of the EV battery pack must be sufficiently separated from dominant external excitation frequencies. A modal analysis was performed with the mounting points fully constrained. The first six natural frequencies and mode shapes were extracted.
| Mode Order | Frequency (Hz) | Mode Shape Description |
|---|---|---|
| 1 | 22.17 | First bending of upper cover (single half-wave) |
| 2 | 39.91 | Second bending of upper cover (two half-waves, front-back) |
| 3 | 40.07 | Second bending of upper cover (two half-waves, front-back, orthogonal phase) |
| 4 | 47.60 | First torsional mode of upper cover |
| 5 | 67.91 | Third bending of upper cover |
| 6 | 71.86 | Complex bending mode of upper cover |
The first-order natural frequency of 22.17 Hz was identified as a critical weakness. The primary excitation sources are powertrain vibration (typically below 25 Hz) and road-induced vibration. The road excitation frequency $f_{road}$ can be estimated by:
$$f_{road} = \frac{V_{max}}{L_{min} \times 3.6}$$
where $V_{max}$ is the maximum vehicle speed (e.g., 100 km/h) and $L_{min}$ is the minimum road unevenness wavelength (e.g., 1 m for a flat road). This yields $f_{road} \approx 27.78$ Hz. To avoid resonance and meet general industry standards for battery pack durability (often requiring the first frequency > 30 Hz), the stiffness, particularly of the upper cover, needed significant improvement. This set the stage for the targeted optimization of the EV battery pack components.
Optimization of the Upper Cover Plate
The upper cover, acting as a large, thin panel, was primarily responsible for the low fundamental frequency. A two-stage optimization process was applied to enhance its bending stiffness efficiently.
Stage 1: Topography Optimization for Stiffening Concept
Topography optimization was employed to determine the optimal pattern of bead-like stiffeners on the designable region of the cover’s top surface, excluding the flanged edges. The goal was to minimize compliance (maximize stiffness) under a unit load, subject to manufacturing constraints. The key design variables for the optimization algorithm were the minimum bead width $W_{min}=30$ mm, the draw angle $\theta=75^\circ$, and a maximum bead height $H_{max}=10$ mm. After convergence, the optimization suggested a complex pattern of curvilinear beads radiating from and connecting high-stress areas. This conceptual output was then interpreted and reconstructed into a manufacturable geometry in CAD software, featuring a series of straight and curved ribs with defined cross-sections. This process added efficient bending stiffness without adding significant material thickness.
Stage 2: Size Optimization for Mass Reduction
With an improved structural form, the next step was to fine-tune the laminate thickness to achieve light weighting. The thickness of the cover laminate was defined as the design variable. The optimization problem was formulated as:
$$
\begin{aligned}
& \text{Objective:} & \min M_{cover} \\
& \text{Subject to:} & f_{1,cover} \geq 30 \text{ Hz} \\
& & 1.0 \text{ mm} \leq T_{cover} \leq 3.0 \text{ mm}
\end{aligned}
$$
where $M_{cover}$ is the mass of the cover, $f_{1,cover}$ is its first natural frequency, and $T_{cover}$ is the total laminate thickness. The optimization solver iteratively reduced the thickness until the frequency constraint was active. The result was then rounded for practicality. The final optimized cover thickness was 2.0 mm, representing a 33% reduction in laminate thickness from the initial 3.0 mm and contributing significantly to the overall light weighting of the EV battery pack. Post-optimization modal analysis confirmed the cover’s first frequency met the target.
Comprehensive Optimization of the Lower Tray Structure
The lower tray, bearing the direct load of the battery modules and reacting to chassis inputs, required a more intricate optimization strategy involving its laminate composition. A four-stage process was undertaken.
Stage 1: Free-Size Optimization for Thickness Distribution
Free-size optimization treats the laminate as a set of “super-plys” for each fiber orientation ($0^\circ$, $±45^\circ$, $90^\circ$). The goal is to find the optimal continuous thickness distribution of each super-ply across the panel to minimize weighted compliance (maximize global stiffness). Manufacturing constraints were included: a minimum member size of 30 mm to avoid fine, unmanufacturable features, and a balance constraint forcing the $+45^\circ$ and $-45^\circ$ ply groups to have identical thickness distributions. The optimization successfully identified areas requiring more material (like the center and around mounting points) and areas where material could be reduced. The output was a thickness map for each fiber angle, but the resulting ply shapes were highly irregular.
Stage 2: Ply-Book Interpretation and Size Optimization
The continuous thickness distributions from Stage 1 were discretized into a finite set of distinct ply patches (or “plies”) for each angle to facilitate manual layup. For example, the $0^\circ$ orientation might be represented by four separate ply patches of different shapes. These shapes were then manually “cleaned up” into more regular, manufacturable polygons. Following this, a size optimization was run with the thickness of each of these discrete ply patches as the design variables. The objective was to minimize the mass of the lower tray, subject to constraints on the first natural frequency ($>30$ Hz) and composite failure criteria (e.g., Tsai-Wu failure index < 1). This step fine-tuned the thickness of each ply patch, yielding a detailed but preliminary ply book with associated thicknesses for the EV battery pack tray.
Stage 3: Multi-Objective Optimization and Decision Making
The previous step considered mass and frequency but not directly the static performance. A more holistic multi-objective optimization (MOO) was conducted using the Isight platform. The design variables were the 16 discrete ply thicknesses from the interpreted model. The objectives were to simultaneously minimize the tray mass and maximize its first natural frequency. Constraints included the maximum displacement and stress from the three static load cases, ensuring no performance degradation. An Optimal Latin Hypercube (OLH) design of experiments was used to sample the design space and build accurate Kriging meta-models. The NSGA-II (Non-dominated Sorting Genetic Algorithm II) was then employed to find the Pareto optimal frontier—a set of solutions where one objective cannot be improved without worsening the other.
The Pareto frontier clearly showed the trade-off: a lighter tray generally had a lower frequency, and vice-versa. Selecting a single optimal solution from this set required a rational decision-making method. The Entropy-TOPSIS technique was applied. First, the entropy weight method objectively calculated the weight (importance) of each objective based on the data’s inherent information diversity. Then, the TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) method ranked all Pareto solutions by calculating their relative closeness to a hypothetical “ideal” solution (best mass, best frequency) and distance from a “negative-ideal” solution (worst mass, worst frequency). The solution with the highest relative closeness was selected as the optimal compromise. The ply thicknesses from this solution were then rounded to the nearest manufacturable ply thickness (0.125 mm). This optimized design achieved a remarkable **58.9% mass reduction** for the lower tray compared to an equivalent performing metallic design, a key breakthrough for the EV battery pack lightweighting.
| Ply Angle & ID | Optimized Thickness (mm) | Ply Angle & ID | Optimized Thickness (mm) |
|---|---|---|---|
| 0° – PLY1100 | 0.250 | -45° – PLY3100 | 0.250 |
| 0° – PLY1200 | 0.375 | -45° – PLY3200 | 0.500 |
| 0° – PLY1300 | 0.250 | -45° – PLY3300 | 0.500 |
| 0° – PLY1400 | 0.375 | -45° – PLY3400 | 0.625 |
| +45° – PLY2100 | 0.250 | 90° – PLY4100 | 0.250 |
| +45° – PLY2200 | 0.250 | 90° – PLY4200 | 0.500 |
| +45° – PLY2300 | 0.250 | 90° – PLY4300 | 0.625 |
| +45° – PLY2400 | 0.625 | 90° – PLY4400 | 0.750 |
Stage 4: Stacking Sequence Optimization
For composite laminates, the stacking sequence (the order of plies) significantly influences performance characteristics like bending stiffness, coupling effects, and impact resistance. A final stacking sequence optimization was performed on the fixed ply book from Stage 3. The optimization algorithm permuted the ply order to maximize a target (e.g., bending frequency) while enforcing practical design rules to ensure manufacturability and performance:
- Balanced Sequence: For every $+45^\circ$ ply, a $-45^\circ$ ply must exist in the stack.
- Outer Surface Ply: The outermost plies should be $\pm45^\circ$ to improve impact resistance and handling.
- Contiguity Constraint: No more than two plies of the same orientation should be stacked together to minimize matrix cracking risk.
The final, fully optimized sequence ensured the EV battery pack tray laminate was not only light and stiff but also robust and practical to manufacture.
Verification of the Fully Optimized EV Battery Pack Enclosure
The optimized upper cover and lower tray were assembled into a complete EV battery pack enclosure model. Comprehensive FEA was performed again to verify that all performance criteria were met or exceeded.
Static Strength Verification
The static load cases were re-run on the optimized model. The results demonstrated a dramatic improvement in structural efficiency.
| Load Case | Baseline Max. Displacement (mm) | Optimized Max. Displacement (mm) | Reduction | Baseline Max. Stress* (MPa) | Optimized Max. Stress* (MPa) |
|---|---|---|---|---|---|
| Bumpy Road & Sharp Turn | 7.23 | 2.39 | 66.9% | 191.0 | 107.8 |
| Bumpy Road & Emergency Brake | 7.25 | 2.12 | 70.8% | 188.8 | 110.5 |
| Bumpy Road & Hard Acceleration | 7.22 | 2.40 | 66.8% | 195.3 | 106.8 |
*Maximum stress typically located on steel mounting brackets. Stress in CFRP components was significantly lower. The displacements were drastically reduced, indicating a much stiffer overall structure for the EV battery pack. Stresses, though still highest at steel brackets, were also reduced and remained well within the yield limit of the steel and the strength limits of the CFRP.
Dynamic Performance Verification
A final modal analysis of the complete, optimized assembly confirmed the success of the dynamic stiffness improvement.
| Mode Order | Baseline Frequency (Hz) | Optimized Frequency (Hz) | Improvement |
|---|---|---|---|
| 1 | 22.17 | 50.63 | +128.4% |
| 2 | 39.91 | 74.21 | +85.9% |
| 3 | 40.07 | 82.52 | +106.0% |
| 4 | 47.60 | 121.46 | +155.2% |
| 5 | 67.91 | 128.34 | +89.0% |
| 6 | 71.86 | 145.66 | +102.7% |
The first natural frequency of the EV battery pack enclosure soared from a deficient 22.17 Hz to a robust 50.63 Hz. This is well above the estimated road excitation frequency of 28 Hz and the common industry target of 30 Hz, effectively eliminating the risk of resonance and providing a large safety margin. This substantial improvement validates the effectiveness of the topography optimization on the cover and the multi-objective layup optimization on the tray.
Conclusion
This study successfully demonstrates a systematic and highly effective methodology for the lightweight design of an EV battery pack enclosure using Carbon Fiber Reinforced Plastic. By transitioning from isotropic metal to anisotropic CFRP, the design space was expanded to allow for directional property tailoring. The process began with establishing accurate performance baselines via finite element analysis, identifying stiffness (particularly modal frequency) as a key constraint for the EV battery pack.
The upper cover was efficiently optimized through a sequential topography and size optimization, enhancing its bending stiffness through intelligent ribbing and then reducing its mass to an optimal thickness. The lower tray underwent a more sophisticated four-stage optimization: free-size optimization for conceptual thickness distribution, ply-book interpretation and sizing, multi-objective optimization trading off mass and frequency, and finally stacking sequence optimization. The use of the Entropy-TOPSIS decision-making method provided an objective and rational way to select the best compromise solution from the Pareto frontier.
The final results are compelling. The optimized CFRP EV battery pack enclosure achieved a drastic improvement in dynamic performance, with the first natural frequency increased by 128% to 50.63 Hz. Simultaneously, static stiffness was greatly enhanced, with maximum displacements under critical loads reduced by over 66%. Most importantly, these performance gains were accompanied by exceptional lightweighting. The lower tray alone achieved a 58.9% mass reduction compared to a conventional metallic design, contributing to a significant overall reduction in the mass of the EV battery pack system. This work conclusively shows that through integrated material substitution and advanced, multi-disciplinary design optimization, CFRP can enable the next generation of lightweight, high-performance, and range-extending EV battery pack enclosures.