In the context of global energy transition and national strategic initiatives, the electrification of vehicles has emerged as a central direction in the automotive industry’s transformation. The EV battery pack, as a core component of electric vehicles, has its safety performance under constant public scrutiny. With the proliferation of high-energy-density cells and increasing demands to withstand extreme operating conditions, the sealing performance of the EV battery pack has become a critical focus during engineering development. The sealing structure of an EV battery pack serves multiple protective functions: it must block the ingress of external solid and liquid contaminants to prevent cell corrosion or short circuits, and it must balance internal pressure fluctuations to avoid electrolyte leakage or thermal runaway. Within these sealing structures, the interface between the EV battery pack upper cover and lower tray, characterized by a large contact area and high cleanliness requirements, exhibits significant quality fluctuations that can profoundly impact the sealing performance of the EV battery pack. Given the complexity of factors contributing to sealing system issues, it is essential to employ comprehensive analysis using quality control tools.
This study revolves around a leakage failure incident caused by sealant in an EV battery pack project during its development and manufacturing phase. By utilizing Quality Control (QC) tools, we successfully addressed the leakage problem occurring in the EV battery pack production process, ensuring smooth mass production. The methodologies for troubleshooting, tracking, and validating sealant-related issues can provide guidance for subsequent sealant design and development for EV battery packs.
Typically, an EV battery pack consists of an upper cover, a lower tray (or lower cover), and an underbody guard. The lower tray can be designed with an integrated cooling plate or with a cooling plate independent of the tray, depending on design intent and performance requirements. The continuous connecting flange between the upper cover and the lower tray constitutes the primary sealing interface of the EV battery pack. To meet assembly requirements for battery modules, control units, and other assemblies, as well as post-sales maintenance needs, this sealing interface is typically a functionally necessary structure. The design must ensure the mating surfaces meet “uniform” and “continuous” sealing requirements.
The interface between the upper and lower covers of an EV battery pack is usually bonded using a liquid sealant that cures in place. Based on various operational conditions and performance requirements for the EV battery pack, the sealant generally needs to possess excellent airtightness, chemical stability, electrical insulation, and ease of processing. Currently, common sealing solutions can be categorized into Cured-In-Place Gasket (CIPG) and Formed-In-Place Gasket (FIPG) technologies. CIPG involves pre-applying liquid sealant onto the sealing flange of the upper cover, followed by a thermal curing process to form a solid seal layer before component assembly. FIPG, in contrast, involves assembling the components while the liquid sealant is still in a “wet,” flowable state after application, relying on environmental conditions (e.g., temperature/humidity) or a triggering mechanism for the sealant layer to gradually cure. The actual selection of an engineering solution must consider multiple aspects such as process characteristics, cost, development timeline, manufacturing cycle time, and post-sales maintenance.
In a specific EV battery pack project, a one-component FIPG polyurethane sealant was selected as the bonding medium between the upper cover and lower tray. This sealant offers good adhesion to various substrates, excellent sealing and mechanical properties, and high-temperature resistance. The EV battery pack upper cover and lower tray are sealed jointly via external fasteners and the internal sealant. The sealant provides interfacial bonding force through chemical bonds formed upon contact with the electrophoretic painted surface of the housing, while the fasteners provide clamping force to assist the interfacial reaction.
Prior to mass production, EV battery packs for a certain platform must pass 39 validation tests at both subsystem and vehicle levels, including items for structural safety, airtightness, corrosion resistance, and others. During testing, it was discovered that after random vibration tests, the EV battery pack could not maintain pressure. Following high-pressure water spray tests, substantial water accumulation was observed inside the EV battery pack. Using helium leak detection based on the mass flow method, elevated helium concentration was detected at the seal between the upper and lower housings. This result indicated failure of the sealing structure between the covers, specifically manifested as sealant detachment from the upper cover upon disassembly. Out of 59 EV battery packs verified for testing, 2 units failed, resulting in a failure rate of 3.4%.
According to GB/T 4208-2017 “Degrees of Protection Provided by Enclosures (IP Code),” the EV battery pack in this project must simultaneously meet IP67/IP69 protection requirements against solid and liquid ingress. Due to project timeline constraints and to ensure product quality for customer delivery and the safe use of the EV battery pack throughout its lifecycle, Quality Control (QC) tools were introduced to resolve the issue.
The production of an EV battery pack involves multiple stages sequentially: part design, component manufacturing, assembly welding, painting and electrophoretic coating, and integration in the battery workshop. Based on this development cycle for the EV battery pack, we considered the 5M1E aspects (Man, Machine, Material, Method, Measurement, Environment) and employed a hierarchical weighting analysis. A cause-and-effect tree diagram was constructed to analyze root causes for the sealing system failure, focusing on five aspects: operational specifications, equipment and tools, material condition, sealing process, and environmental conditions. Through layered analysis, 14 terminal factors were identified and investigated concurrently.
| Aspect | Terminal Factors Identified | Investigation Method |
|---|---|---|
| Operational Specifications | Operator training level, adherence to work instructions | Supplier audit, workshop self-inspection |
| Equipment and Tools | Sealant gun temperature control, air pressure, torque tool calibration | Parameter checks, torque verification |
| Material Condition | Dimensional tolerance of covers, incoming leak test, sealant shelf-life | Sampling inspection, record review |
| Sealing Process | Design robustness, cover closing gap, fastener sequence, surface contamination | CAE analysis, gap measurement, EDS analysis, knife test |
| Environmental Conditions | Transportation packaging cleanliness, particle contamination | Logistics audit, surface inspection |
Regarding operational specifications, supplier visits confirmed that work instructions and operator training levels met requirements. Self-inspection at the battery workshop indicated that, aside from some sealant overflow, the application process showed no significant defects. The overflow areas were verified not to be leak points via helium leak detection, thus excluding operational factors.
For equipment and tools, the sealant application process strictly controlled the gun temperature at $ (45 \pm 5)^{\circ}C $ and inlet pressure at 500 kPa, complying with the sealant’s application specifications. The electric torque tools were checked; they alarmed if the dynamic torque setting was not reached. Static torque checks on randomly selected EV battery packs confirmed no torque decay, indicating proper tool function. Equipment factors were thus excluded.
Concerning material condition, the quality of the EV battery pack upper cover, lower tray, and the sealant itself was investigated. Sampling from various batches confirmed that the sealing surface profile tolerances for both covers met drawing requirements. Additionally, incoming covers undergo leak testing before packaging; sampling verified they met the specified leak rate limit at $-3.45 \text{ kPa}$. Since the sealant is time-sensitive and can cure by moisture absorption at room temperature, it must be used within 14 days after opening or replaced. The workshop maintained dedicated records for sealant opening time, replacement, and expiration, ensuring the sealant’s shelf-life was within specification. Material quality issues were therefore ruled out.
For the sealing process, we first reviewed the robustness of the initial design. To verify the structural strength of the sealant under extreme conditions, the most severe thermal runaway scenario for the EV battery pack was selected. A CAE model was established, applying an internal pressure of 30 kPa to simulate the instant when the explosion-proof valve opens during thermal runaway. The CAE results showed that the equivalent stress and nominal strain in the sealant were within its yield limits, indicating that the current geometric dimensions and mechanical properties of the sealant could meet extreme condition demands, and there was no design defect.
The cover closing gap is negatively correlated with the structural stiffness of the sealant. The engineering requirement for the EV battery pack cover closing gap is $ (6 \pm 1) \text{ mm} $. In this project, the upper cover material thickness is relatively thin at 0.7 mm, making its dimensions susceptible to process variations throughout the housing manufacturing, which can affect the upper cover’s dimensional state. Since sealant performance is strongly related to the upper cover’s dimensional condition, special attention was required for dimensional fluctuations during upper cover manufacturing. It was found that there was no fixed bolt tightening sequence during cover closing in the battery workshop; different sequences could influence the closing height. Therefore, an experiment was designed to evaluate 9 different tightening sequences, measuring the closing gap at 21 inspection points.
Statistical analysis of the closing gap at the front curved valve port area under different tightening sequences revealed significant variations. Some sequences resulted in the sealant being barely compressed, with dimensions severely out of tolerance. Consequently, standardizing the tightening sequence was necessary.
During disassembly inspection of failed EV battery packs, white deposits were observed on the electrophoretic substrate surface at the sealing interface, suspected to be residues from the supplier’s water leak testing process. Energy Dispersive X-ray Spectroscopy (EDS) in a scanning electron microscope was used to compare element types and contents between the failure surface and a standard sample, investigating potential contamination sources during the supplier’s water testing. The EDS spectrum showed significantly higher silicon (Si) content on the failure sample compared to the standard, along with impurity elements like potassium (K) and iron (Fe).
To simulate contamination, suspected pollutants from the water testing process were applied to a standard electrophoretic substrate, followed by sealant application. After curing, the adhesion was compared with that on a clean standard substrate using the GMW3368 knife test (a quick blade adhesion test). In this test, the sealant bead is manually cut at the interface with the substrate over at least 75% of its length, and the adhesion state is recorded. States are classified as Cohesive Failure (CF, failure within the sealant) or Adhesive Failure (AF, failure at the interface). Good adhesion requires CF area $\geq$ 98%, with any AF areas being non-continuous or non-penetrating. Results showed that the sealant on the silicon-contaminated panel exhibited complete detachment (100% AF), whereas the standard sample showed predominantly cohesive failure. Combined with EDS analysis, the failure mode was interpreted as hydrolysis of silicon-based sealing rings used in the tray and cover water testing fixtures, introducing free silicon into the water. Such residues on the sealing surface can react chemically with the sealant, leading to interfacial adhesion failure.
According to the adhesion force formula and CAE simulation results, the sealant bead width is strongly correlated with adhesion force. However, in leak point areas of failed EV battery packs, issues such as discontinuous bead paths, severely insufficient compressed bead width, and even broken bead were observed, falling below the design requirement for compressed width of $ (10 \pm 3) \text{ mm} $. This phenomenon was related to process variations during cover closing, necessitating process optimization.
Regarding environmental conditions, the EV battery pack manufacturing process involves multiple locations: supplier, logistics warehouse, body shop, and battery workshop, with parts frequently moving between sites. Inspection of transportation racks and packaging revealed damaged logistics packaging in the battery workshop warehouse, with dust, oil stains, and other suspected contaminants attached to the mating surfaces of lower trays and upper covers. Inadequate cleanliness control during logistics was suspected. Sealant was applied to parts from the same batch after contamination, and knife tests confirmed that surface impurities also adversely affected sealant adhesion, indicating a need for stricter control.
Based on the investigation, measurement, and experimental comparison of the 14 terminal factors, four root causes were confirmed for the EV battery pack airtightness failure problem.
| No. | Root Cause | Countermeasure | Target | Action | Location |
|---|---|---|---|---|---|
| 1 | Contamination of parts during supplier water leak test due to seal ring hydrolysis | Avoid introduction of foreign organic impurities during water testing | Pass knife adhesion test | Supplier replaces water test seal rings with compatible material | Electrophoresis supplier |
| 2 | Inadequate cleanliness of packaging and racks during transportation | Clean sealing surfaces before cover closing | Pass knife adhesion test | Wipe sealing surfaces with alcohol before closing | Battery workshop |
| 3 | Unreasonable cover closing process (fastener sequence) | Improve sealing surface tightening process | Achieve cover closing gap of $ (6 \pm 1) \text{ mm} $ | Adopt fixed tightening sequence (bolts first, then nuts, front to back) | Battery workshop |
| 4 | Insufficient compressed sealant bead width | Improve sealant application process | Achieve compressed bead width $\geq 7 \text{ mm}$ after closing | Add pressing fixture after tightening to apply uniform closing force; implement infrared inspection for bead path | Battery workshop |
For root cause 1, the supplier replaced the water test seal rings with a material not reactive with the test fluid. Compatibility was verified by placing the new seal ring against electrophoretic panels, clamping for 1 hour, moving it twice, and repeating clamping. After sealant application and 7-day curing, four sets of knife tests showed 100% Cohesive Failure (CF), indicating good compatibility and no adhesion failure to the substrate.
For root cause 2, to establish robust cleanliness control, potential contaminants (rust preventive oil, lubricating oil, dust, silicon grease, water) were applied to electrophoretic panels. Different cleaning methods (no cleaning, wiping with non-woven cloth, wiping with ethanol) were tested as control groups, followed by sealant application, clamping, and four sets of knife tests. Results demonstrated that non-woven cloth had limited cleaning efficacy, whereas ethanol effectively cleaned contaminated surfaces, with wet wiping and wet+dry wiping showing similar effectiveness. To reduce production cost, wiping the sealing surface with alcohol before cover closing was selected as the standard process.
For root cause 3, based on statistical results of the influence of tightening sequence on the cover closing gap, considering mean gap, standard deviation fluctuation, and manufacturing feasibility, tightening sequence 3 (bolts first, then nuts, tightened from front to rear) was selected as the standard production process. The battery workshop added a dedicated tightening station with torque monitoring guns following this sequence.
After process adjustment, 30 randomly selected EV battery packs from production batches were inspected for cover closing gap. Measurements using a vernier caliper around the perimeter showed both the average and extreme values of the closing gap were within the design maximum of 7 mm, meeting the $ (6 \pm 1) \text{ mm} $ requirement. The improvement can be represented by the reduction in gap variation. If the initial standard deviation was $\sigma_{\text{initial}}$, after standardization, the new standard deviation $\sigma_{\text{new}}$ decreased significantly, enhancing process capability $C_p = \frac{USL – LSL}{6\sigma}$, where $USL$ and $LSL$ are the upper and lower specification limits.
For root cause 4, the battery workshop added a pressing fixture applied for 2 minutes after bolt and nut tightening to maintain uniform closing force on the sealing surface, improving sealant curing. Simultaneously, infrared inspection equipment was introduced to ensure proper sealant application path and completeness. After fixture implementation, 20 EV battery packs were disassembled, and the compressed bead width was measured. The average and minimum values of the narrowest width satisfied the requirement of being $\geq 7 \text{ mm}$.
Since these root causes are interrelated, implementing all countermeasures was necessary to achieve the goal of zero failures for the EV battery pack sealing system. Based on the identified causes, the team optimized several standards, including EV battery pack housing cleanliness control measures, sealant application standard operating procedures, rework operation instructions, and mass production bead width control, providing reference for subsequent new EV battery pack projects.
To verify countermeasure effectiveness, 13,000 EV battery packs in the mass production phase were tracked continuously, confirming no airtightness failures. All leakage rates were within the specified limit of 10 mL/min (at standard conditions), demonstrating effective implementation. Through this problem resolution, the leakage rate was reduced from 3.4% to 0%, preventing significant part scrap losses post-mass-production. For this platform’s various vehicle models, estimated cost savings from avoided scrap amounted to approximately 8.9 million yuan. The financial benefit can be expressed as:
$$ \text{Cost Savings} = N \times P_f \times C_{\text{scrap}} $$
where $N$ is the total production volume, $P_f$ is the initial failure probability, and $C_{\text{scrap}}$ is the cost per scrapped EV battery pack. With $N$ large, even a small reduction in $P_f$ yields substantial savings.
Furthermore, the sealant adhesion strength improvement can be quantified. The adhesion force $F_{\text{adh}}$ is proportional to the effective bond area $A_{\text{bond}}$ and the interfacial shear strength $\tau_{\text{int}}$:
$$ F_{\text{adh}} = \tau_{\text{int}} \times A_{\text{bond}} $$
By increasing the compressed bead width $w$ and ensuring continuous application, the bond area for a perimeter length $L$ is $A_{\text{bond}} = w \times L$. Process optimizations increased $w$ from below 7 mm to consistently above 7 mm, thereby boosting $F_{\text{adh}}$. The interfacial shear strength $\tau_{\text{int}}$ is also enhanced by eliminating surface contaminants, which can be modeled as a reduction in a contamination factor $\beta$ (where $0 \leq \beta \leq 1$) that reduces effective strength: $\tau_{\text{int}} = \tau_{\text{max}} (1 – \beta)$. Cleaning with alcohol reduces $\beta$ to near zero.
In addressing the leakage issue in the EV battery pack production process, due to the multitude of influencing factors and the complexity of the problem, we employed Quality Control (QC) tools to conduct a comprehensive analysis from the perspectives of “Man, Machine, Material, Method, Environment.” Based on the identified root causes, four effective countermeasures were formulated, resolving the EV battery pack airtightness failure and insufficient sealant adhesion strength issues, achieving zero airtightness failures off the production line. Corresponding robust control measures and guidance standards were established, closing the problem loop. The experience and consolidated measures from this activity can be extended to the development and production processes of subsequent EV battery pack projects, playing a significant role in enhancing the overall product quality of EV battery packs.
The successful resolution underscores the importance of a systematic approach in EV battery pack manufacturing. Future work could involve developing predictive models for sealant performance under dynamic loads. For instance, the stress on the sealant under internal pressure $P$ in an EV battery pack can be estimated using a simplified model for a rectangular seal bead:
$$ \sigma = \frac{P \cdot A_{\text{internal}}}{A_{\text{bead}}} $$
where $\sigma$ is the stress on the sealant, $A_{\text{internal}}$ is the internal area of the EV battery pack subject to pressure, and $A_{\text{bead}}$ is the cross-sectional area of the sealant bead. Ensuring $\sigma$ remains below the sealant’s yield strength $\sigma_y$ is crucial for long-term integrity of the EV battery pack.
Moreover, the study highlights the criticality of supply chain and inter-process cleanliness for EV battery pack reliability. Implementing statistical process control (SPC) charts for key parameters like cover closing gap and bead width can further enhance quality assurance. For example, monitoring the mean $\bar{x}$ and range $R$ of bead width measurements from samples of size $n$ can help maintain process stability for EV battery pack production.
In conclusion, the integration of QC methodologies, combined with technical investigations and process refinements, proved effective in solving a critical sealing issue in EV battery pack assembly. The lessons learned are broadly applicable to sealing system design and manufacturing quality control in the rapidly evolving field of electric vehicles, ensuring the safety and durability of EV battery packs under diverse operating conditions.