In recent years, the rapid development of electric vehicles has positioned the electric SUV as a key player in the automotive industry due to its versatility and eco-friendly attributes. As a researcher focused on automotive safety, I have investigated the low-speed frontal collision performance of a pure electric SUV based on the European Research Council for Automobile Repairs (RCAR) test protocol. This study aims to enhance the crashworthiness of the electric SUV’s front-end structure, thereby improving its insurance rating and reducing repair costs after low-speed impacts. Low-speed collisions are common in urban environments, and optimizing the vehicle’s front-end components can significantly mitigate damage to critical parts like the fan and longitudinal beams, ensuring that repairs are economical and straightforward.
The RCAR low-speed frontal collision test involves a 40% offset impact against a rigid barrier at a velocity of 16 km/h, with the barrier angled at 10 degrees to simulate real-world scenarios. According to RCAR requirements, only the bumper beam and energy-absorption boxes should deform during such collisions, while the longitudinal beams must remain largely undeformed, with plastic strain not exceeding 5%. This ensures that damage is confined to easily replaceable components, protecting more expensive parts like the cooling fan. In this study, I developed a finite element model of the electric SUV using ANSA software and performed simulations with LS-DYNA and HyperView to analyze deformation and energy absorption. The initial results revealed inadequate crashworthiness, prompting structural and parametric optimizations to enhance stiffness and strength.
To establish the finite element model, I simplified the 3D CAD geometry of the electric SUV, ensuring it accurately represented the actual vehicle with a mass of 1753 kg and a wheelbase of 2675 mm. The model consisted of 2,957,850 elements, including 795 parts, 359,921 solid elements, and 2,592,836 shell elements. Material properties were assigned based on aluminum alloys commonly used in automotive applications, such as 6005A and 6082, which influence the energy absorption characteristics. The simulation parameters were set to a collision time of 0.15 seconds and a gravitational acceleration of 9.81 m/s². Validation of the model confirmed energy conservation, with the hourglass energy ratio remaining below 0.79%, indicating high computational accuracy. The initial simulation showed that the bumper beam underwent excessive deformation, the energy-absorption boxes exhibited undesirable bending, and the longitudinal beams experienced plastic strains up to 45.41%, far exceeding the RCAR limit. This highlighted the need for optimization to improve the electric SUV’s low-speed collision performance.
The energy absorption analysis during the low-speed collision identified the bumper beam, energy-absorption boxes, and longitudinal beams as key components. The initial energy distribution showed that the bumper beam absorbed the most energy, followed by the energy-absorption boxes, but the latter did not perform optimally, leading to inefficient energy management. The total energy absorbed by the front-end structure was calculated using the equation for internal energy evolution: $$E_{internal} = \int F(s) \, ds$$ where \(F(s)\) is the force as a function of displacement \(s\). The compression distance of the energy-absorption boxes was initially 128.3 mm, which exceeded the safe limit of 110 mm—the minimum distance to the fan and its mounting plate—posing a risk of fan damage. This necessitated a multi-faceted optimization approach to enhance the crashworthiness of the electric SUV.
I first optimized the bumper beam by changing its material from 6005A to 6082 aluminum alloy, which has a higher yield strength, and altering its cross-section to a “日”-shaped design to increase stiffness. This modification reduced the bumper beam’s energy absorption from approximately 5 kJ to 2.5 kJ, allowing more energy to be transferred to the energy-absorption boxes. As a result, the energy-absorption boxes’ energy absorption increased from 2.86 kJ to 6.52 kJ, and the plastic strain in the longitudinal beams decreased from 45.41% to 35.05%. However, the compression distance of the energy-absorption boxes remained high at 111.8 mm, and the longitudinal beams’ plastic strain was still above the 5% threshold, indicating that further optimization was required for the electric SUV.
Next, I focused on the energy-absorption boxes and longitudinal beams. For the energy-absorption boxes, I added two诱导 slots at the front and a vertical rib at the rear with a thickness of 2 mm and length of 70 mm to promote symmetric folding and reduce compression distance. The longitudinal beams were also upgraded to 6082 aluminum alloy to enhance strength. After these changes, the compression distance of the energy-absorption boxes decreased to 111.8 mm, and the longitudinal beams’ plastic strain dropped to 12.52%. Although this was an improvement, it did not meet the RCAR standards, leading to a multi-objective optimization strategy for the electric SUV’s front-end structure.
I formulated a multi-objective optimization problem with six design variables: the thicknesses of the bumper beam (\(t_1\)), energy-absorption box (\(t_2\)), longitudinal beam (\(t_3\)), and their respective reinforcement ribs (\(t_4\), \(t_5\), \(t_6\)). The objective was to maximize the total energy absorption (\(E\)) of the bumper beam and energy-absorption boxes while minimizing their total mass (\(M\)), subject to constraints on the compression distance (\(\delta \leq 110\) mm) and the sectional force (\(F \leq 104\) kN). The mathematical model is defined as:
$$
\begin{aligned}
\text{Maximize} \quad & E(t_1, t_2, t_3, t_4, t_5, t_6) \\
\text{Minimize} \quad & M(t_1, t_2, t_3, t_4, t_5, t_6) \\
\text{Subject to} \quad & F \leq 104 \\
& \delta \leq 110 \\
& 2.4 \leq t_1, t_2 \leq 3.2 \\
& 2.8 \leq t_3 \leq 3.2 \\
& 1.6 \leq t_4, t_5 \leq 2.4 \\
& 2.3 \leq t_6 \leq 2.7
\end{aligned}
$$
Using orthogonal experimental design in Isight software, I generated 36 sample points and performed simulations to obtain response values for energy absorption, mass, compression distance, and sectional force. A response surface methodology (RSM) was employed to create approximate models for each response. The second-order polynomial equations for the responses are as follows:
For energy absorption \(E\):
$$E = -68.55t_1 – 14.45t_2 + 418.97t_3 + 158.55t_4 – 231.06t_5 – 48.11t_6 – 3.10t_1^2 – 8.13t_2^2 – 70.56t_3^2 – 3.78t_4^2 + 21.67t_5^2 + 1.54t_6^2 – 11.89t_1 t_2 + 27.14t_1 t_3 + 25.69t_1 t_4 – 25.41t_1 t_5 + 15.18t_1 t_6 + 17.08t_2 t_3 – 13.96t_2 t_4 + 23.66t_2 t_5 – 0.79t_2 t_6 – 32.74t_3 t_4 + 12.36t_3 t_5 – 31.37t_3 t_6 – 11.17t_4 t_5 – 23.20t_4 t_6 + 55.55t_5 t_6$$
For mass \(M\):
$$M = 6.33t_1 + 28.15t_2 – 50.96t_3 – 19.88t_4 + 3.42t_5 – 12.13t_6 – 0.47t_1^2 – 1.34t_2^2 + 7.36t_3^2 + 0.37t_4^2 – 1.57t_5^2 – 0.95t_6^2 + 1.68t_1 t_2 – 2.82t_1 t_3 – 1.13t_1 t_4 + 0.74t_1 t_5 + 0.69t_1 t_6 – 3.91t_2 t_3 – 0.82t_2 t_4 – 2.50t_2 t_5 – 2.27t_2 t_6 + 4.91t_3 t_4 + 1.23t_3 t_5 + 5.86t_3 t_6 + 1.73t_4 t_5 + 2.41t_4 t_6 + 0.24t_5 t_6$$
For compression distance \(\delta\):
$$\delta = 953.25t_1 – 770.93t_2 – 146.38t_3 + 506.53t_4 + 501.11t_5 – 2839.34t_6 – 23.50t_1^2 + 67.97t_2^2 + 88.44t_3^2 – 135.23t_4^2 – 118.35t_5^2 + 461.12t_6^2 – 59.59t_1 t_2 – 251.22t_1 t_3 – 97.69t_1 t_4 + 70.17t_1 t_5 + 75.91t_1 t_6 + 76.87t_2 t_3 + 43.69t_2 t_4 + 67.32t_2 t_5 + 12.07t_2 t_6 + 61.47t_3 t_4 – 135.98t_3 t_5 – 118.74t_3 t_6 + 41.79t_4 t_5 – 33.75t_4 t_6 – 22.67t_5 t_6$$
For sectional force \(F\):
$$F = 21.67t_1 + 41.57t_2 + 27.13t_3 – 58.24t_4 – 38.30t_5 + 200.95t_6 – 12.67t_1^2 – 7.79t_2^2 – 33.72t_3^2 + 15.63t_4^2 + 11.11t_5^2 – 55.22t_6^2 – 15.95t_1 t_2 + 30.87t_1 t_3 + 15.87t_1 t_4 – 29.04t_1 t_5 + 11.93t_1 t_6 + 21.19t_2 t_3 – 6.42t_2 t_4 – 2.53t_2 t_5 + 2.83t_2 t_6 – 0.46t_3 t_4 + 15.39t_3 t_5 – 2.55t_3 t_6 – 11.02t_4 t_5 – 4.17t_4 t_6 + 23.90t_5 t_6$$
The accuracy of these models was verified using the coefficient of determination (\(R^2\)) and root mean square error (RMSE). As shown in the table below, all \(R^2\) values exceeded 0.9, and RMSE values were below 0.2, confirming the models’ reliability for optimization.
| Response Variable | \(R^2\) | RMSE |
|---|---|---|
| Energy Absorption (\(E\)) | 0.9073 | 0.0707 |
| Mass (\(M\)) | 0.9991 | 0.0065 |
| Compression Distance (\(\delta\)) | 0.9107 | 0.0631 |
| Sectional Force (\(F\)) | 0.9236 | 0.0608 |
I applied the NSGA-II algorithm in Isight with a population size of 20, 50 generations, a crossover probability of 0.9, and a maximum of 1000 iterations to find Pareto-optimal solutions. The optimal design variables were selected and slightly adjusted to practical values, as summarized in the following table:
| Design Variable | Initial Value (mm) | Optimized Value (mm) | Adjusted Value (mm) |
|---|---|---|---|
| \(t_1\) (Bumper Beam Thickness) | 3.0 | 2.620 | 2.6 |
| \(t_2\) (Energy-Absorption Box Thickness) | 3.0 | 3.026 | 3.0 |
| \(t_3\) (Longitudinal Beam Thickness) | 3.0 | 2.805 | 2.8 |
| \(t_4\) (Bumper Beam Rib Thickness) | 2.0 | 1.721 | 1.7 |
| \(t_5\) (Energy-Absorption Box Rib Thickness) | 2.0 | 2.010 | 2.0 |
| \(t_6\) (Longitudinal Beam Rib Thickness) | 2.5 | 2.442 | 2.4 |
After implementing these optimized values in the finite element model, I re-ran the simulation to evaluate the performance of the electric SUV. The results demonstrated significant improvements: the compression distance of the energy-absorption boxes decreased to 105.44 mm, below the 110 mm limit, effectively protecting the fan. The sectional force reduced from 103.11 kN to 96.41 kN, and the plastic strain in the longitudinal beams dropped to 5.04%, meeting the RCAR requirement. The total energy absorption increased from 9923 J to 10128 J, a 2.06% improvement in crashworthiness, while the total mass decreased from 7.670 kg to 7.054 kg, achieving an 8.03% reduction in weight. This highlights the success of the multi-objective optimization in enhancing the low-speed collision performance of the electric SUV while promoting lightweight design.
The discussion of these findings emphasizes that the optimized front-end structure of the electric SUV now effectively manages energy absorption and deformation, minimizing damage to critical components. The use of high-strength aluminum alloys and strategic reinforcements allowed for a balance between crashworthiness and weight reduction. However, it is important to note that this study focused solely on low-speed collisions; future work should investigate the impact of these optimizations on high-speed collision performance to ensure comprehensive safety for the electric SUV.
In conclusion, this research successfully optimized the front-end structure of a pure electric SUV for RCAR low-speed frontal collisions through finite element simulation and multi-objective optimization. The improvements in energy absorption, compression distance, and plastic strain not only meet RCAR standards but also enhance the vehicle’s repairability and insurance rating. The methodologies developed here can be applied to other electric SUV models, contributing to safer and more economical vehicles in the evolving automotive landscape.