In the rapidly evolving landscape of new energy vehicles, the electric vehicle (EV) battery pack serves as a critical component, directly influencing range, safety, and overall performance. As market penetration increases, automakers are intensifying efforts to optimize chassis space and structural integrity. One innovative approach involves mounting the EV battery pack not only on traditional body rails but also on chassis structural parts, such as torsion beam mounting brackets. This configuration enhances space utilization and improves installation stiffness, but it introduces complex dimensional chains that combine body and chassis tolerances. During small-batch production validation, we observed intermittent assembly interference risks, such as misaligned holes between the EV battery pack and torsion beam brackets, with a defect rate of approximately 30%. To address this, we employed 3DCS software for virtual assembly deviation analysis, aiming to identify root causes and propose effective optimization strategies. This study details our methodology, simulation model, sensitivity analysis, and implemented solutions, emphasizing the role of dimensional engineering in ensuring robust EV battery pack assembly.
Dimensional engineering is fundamental to managing manufacturing variations in automotive assembly. Traditional methods include the Worst Case (WC) approach, which assumes all tolerances are at their extreme limits, leading to conservative and often costly designs. The Root Sum Squares (RSS) method, based on probability theory, offers a more realistic analysis by combining tolerances statistically: $$ \Delta_{total} = \sqrt{\sum_{i=1}^{n} (\Delta_i)^2} $$ where $\Delta_{total}$ is the total deviation and $\Delta_i$ are individual tolerances. However, RSS assumes normal distributions and struggles with non-linear tolerances like form or composite deviations. In contrast, Monte Carlo simulation, implemented in tools like 3DCS, randomly samples tolerance distributions over thousands of virtual assemblies, accurately modeling real-world production scenarios. This method handles complex geometries and non-normal distributions, enabling precise prediction of assembly outcomes and identification of key contributing factors. For the EV battery pack assembly, we leveraged 3DCS to simulate the cumulative effects of part tolerances, assembly sequences, and positioning strategies, ensuring a comprehensive deviation analysis.
The assembly process for the EV battery pack involves multiple stages, integrating the battery pack with the underbody and chassis components. First, the torsion beam mounting brackets are assembled to the underbody via locating holes, forming a combined underbody-chassis subassembly. Second, the EV battery pack is positioned onto this subassembly using dedicated locating holes on the underbody. The EV battery pack features through-holes that align with threaded inserts on both body rails and torsion beam brackets, secured by flange bolts. This two-step process introduces potential stack-up errors, particularly at the interface between the EV battery pack and torsion beam brackets, where dimensional chains include both body and chassis tolerances.
Positioning strategies are crucial for minimizing variation. For the torsion beam bracket assembly, we defined a 6-degree-of-freedom constraint system: the primary locating hole controls translations along Z, X, and Y axes; the secondary locating hole controls rotation about the Z-axis; and the outer mounting holes on left and right brackets each control translation along the Z-axis. Similarly, for the EV battery pack assembly, three planar surfaces around locating holes control Z-translation, while primary and secondary holes control X and Y translations and Z-rotation, respectively. These schemes aim to stabilize the EV battery pack during assembly, but inherent tolerances can lead to misalignment. The table below summarizes key tolerances derived from GD&T specifications, which were input into the 3DCS model.
| Component | Tolerance Name | Tolerance Value (mm) |
|---|---|---|
| Torsion Beam Bracket | Primary Locating Hole Position | ±1.0 |
| Torsion Beam Bracket | Secondary Locating Hole Position | ±1.0 |
| Torsion Beam Bracket | Left Outer Mounting Hole Profile | ±0.4 |
| Torsion Beam Bracket | Right Outer Mounting Hole Profile | ±0.4 |
| Torsion Beam Bracket | EV Battery Pack Mounting Hole Position | ±0.5 |
| EV Battery Pack | Primary Locating Hole Position | ±1.0 |
| EV Battery Pack | Secondary Locating Hole Position | ±1.0 |
| EV Battery Pack | Locating Surface Profile | ±0.3 |
| EV Battery Pack | Mounting Hole Position (to Bracket) | ±0.5 |
| Underbody | Locating Hole Position | ±0.5 |
| Tooling | Primary Pin Float (Body-Tool) | ±0.15 |
| Tooling | Secondary Pin Float (Body-Tool) | ±0.15 |
| Tooling | Primary Pin Float (Battery-Tool) | ±0.20 |
| Tooling | Secondary Pin Float (Battery-Tool) | ±0.20 |
| Assembly | Pin Float (Bracket to Underbody) | ±2.0 |
We developed a 3DCS simulation model to replicate the EV battery pack assembly process. The model incorporated part geometries, assembly sequences, and tolerance definitions from the GD&T data. Measurement points were created at the EV battery pack mounting locations on the torsion beam brackets to assess alignment deviations. The simulation involved 50,000 Monte Carlo iterations, each representing a virtual assembly under production variations. The initial results revealed a minimum edge distance between the EV battery pack hole and torsion beam bracket hole of -4.61 mm, indicating severe interference. The out-of-specification rate was 32.47%, closely matching the observed 30% defect rate in production, validating the model’s accuracy. This confirmed that the EV battery pack assembly was prone to misalignment due to tolerance stack-up.
Sensitivity analysis identified the primary contributors to the interference. Using 3DCS, we calculated the contribution percentage of each tolerance to the overall deviation in the edge distance. The results are summarized in the table below, highlighting factors such as pin float tolerances and bracket hole positions. These insights guided our optimization efforts, focusing on reducing the normal distribution bandwidth (6σ) and adjusting critical dimensions.
| Rank | Tolerance Name | Contribution (%) |
|---|---|---|
| 1 | Pin Float (Bracket Primary Hole to Underbody) | 31.57 |
| 2 | Pin Float (Bracket Secondary Hole to Underbody) | 28.42 |
| 3 | Torsion Beam Bracket Primary Hole Position | 7.67 |
| 4 | Torsion Beam Bracket Secondary Hole Position | 6.88 |
| 5 | EV Battery Pack Mounting Hole Position | 5.21 |
To mitigate interference, we proposed two optimization strategies. First, we enhanced the assembly positioning robustness of the torsion beam bracket by adopting a single-side double-limit pin design. This replaced the original single-pin setup with dual pins on one side, improving constraint stability and reducing the 6σ value. The optimized positioning scheme controls translations and rotations more effectively, minimizing bracket movement during assembly. The 6σ bandwidth, calculated as $$ 6\sigma = 6 \times \text{standard deviation} $$ decreased from 7.29 mm to 3.27 mm, a 55% reduction. However, simulation still showed a minimum gap of -0.90 mm, indicating residual risk for the EV battery pack installation.
Second, we adjusted the EV battery pack mounting hole diameter. Initially, the hole was ϕ16 mm for M10 bolts. By increasing it to ϕ20 mm, we provided additional clearance to accommodate tolerance stack-up. To compensate for reduced bolt contact area, we specified larger washers (ϕ36 to ϕ40 mm). This modification directly addresses the edge distance issue without tightening part tolerances, which could increase manufacturing costs. The combination of optimized positioning and enlarged holes was simulated again in 3DCS. Results showed a minimum positive clearance of 1.11 mm, eliminating interference risks over 50,000 iterations. The table below compares key metrics before and after optimization.
| Parameter | Initial Design | After Positioning Optimization | After Hole Diameter Optimization |
|---|---|---|---|
| 6σ Bandwidth (mm) | 7.29 | 3.27 | 3.27 |
| Min Edge Distance (mm) | -4.61 | -0.90 | 1.11 |
| Out-of-Spec Rate (%) | 32.47 | ~15 | 0 |
| EV Battery Pack Hole Diameter | ϕ16 mm | ϕ16 mm | ϕ20 mm |
The optimization strategies were implemented in production, and we monitored approximately 1,000 vehicles post-implementation. No instances of EV battery pack installation misalignment or interference were reported, confirming the effectiveness of the changes. The dual-pin design significantly improved assembly repeatability, while the enlarged hole diameter provided a safe margin for tolerance accumulation. These measures demonstrate how 3DCS-based deviation analysis can guide practical solutions for complex assemblies like the EV battery pack, balancing performance, cost, and manufacturability.
In conclusion, mounting the EV battery pack on chassis structures offers space and stiffness benefits but introduces dimensional challenges. Through 3DCS simulation, we accurately modeled assembly deviations and identified key tolerance contributors. By optimizing positioning with single-side double-limit pins and increasing the EV battery pack mounting hole diameter, we eliminated interference risks and enhanced production robustness. This approach underscores the value of virtual tolerance analysis in automotive manufacturing, particularly for critical components like the EV battery pack. Future work could explore dynamic effects or thermal expansions on EV battery pack assembly, further refining dimensional control strategies.