The rapid advancement of the electric vehicle (EV) industry has placed the safety and reliability of the power battery pack at the forefront of technological and quality concerns. As the foundational enclosure, the battery pack tray must possess high strength, lightweight characteristics, and exceptional sealing integrity to prevent the ingress of moisture, dust, and other contaminants that could compromise the battery system’s safety and longevity. Aluminum alloys, particularly the 6000 series, are favored for this application due to their favorable strength-to-weight ratio and corrosion resistance. Friction Stir Welding (FSW), a solid-state joining technique, has become the process of choice for manufacturing these trays, primarily due to its advantages of low distortion, minimal residual stress, and the virtual elimination of fusion-related defects like porosity and solidification cracking.
Despite its maturity, the mass production of EV battery pack components via FSW is not immune to defects. Process fluctuations, fixture wear, and assembly gaps can lead to imperfections that critically challenge the hermetic seal of the welded joints. Among these, lack-of-penetration defects, such as voids, tunnels, and particularly “cold weld” conditions—where insufficient thermomechanical input prevents proper metallurgical bonding—pose significant risks due to their potential to create leak paths. This study investigates a recurring leakage failure in the lower tray of a commercial EV battery pack. Through systematic failure analysis, we identify the root cause, elucidate the formation mechanism, and implement effective process optimizations to enhance production quality and reliability.
1. Materials, Component Architecture, and Initial Failure
The faulty component was the lower tray of a commercial EV battery pack, constructed by welding extruded 6005-T6 aluminum alloy profiles. The nominal thickness at the weld joint was 2.5 mm. The chemical composition and mechanical properties of the base material are detailed in Tables 1 and 2, respectively.
| Mg | Si | Fe | Cu | Zn | Cr | Ti | Mn | Al |
|---|---|---|---|---|---|---|---|---|
| 0.7 | 0.6 | 0.3 | 0.2 | 0.15 | 0.1 | 0.03 | 0.15 | Bal. |
| Material | Tensile Strength (MPa) | Yield Strength (MPa) | Elongation (%) |
|---|---|---|---|
| 6005-T6 | 285 | 230 | 16 |
The tray assembly consisted of six primary sections: two central floor panels, front and rear panels, and left and right side panels. The front and rear panels were joined to the central floor panels via FSW to form the core liquid cooling plate, which integrated internal fluid channels. This sub-assembly was then welded to the left and right side panels, again using FSW, to complete the sealed enclosure. The integrity of the entire EV battery pack housing was therefore critically dependent on the quality of these FSW joints.
The initial failure was identified during a helium leak test of the assembled EV battery pack. The test protocol involved evacuating the cooling plate, pressurizing it to 0.5 MPa with a helium-air mixture, and monitoring the leak rate. A failure was confirmed when the system could not maintain pressure, and the measured leak rate exceeded the threshold of \(4.3 \times 10^{-7} \, \text{Pa} \cdot \text{m}^3/\text{s}\). Subsequent localized testing pinpointed the leak source to a specific FSW seam on the exterior (non-coolant side) of the tray.
2. Defect Characterization and Metallurgical Analysis
Macroscopic examination of the leak location revealed a continuous, linear feature along the weld seam, superficially resembling a crack. A cross-sectional sample was extracted from this region, prepared through standard metallographic procedures (mounting, grinding, polishing, and etching with Keller’s reagent), and examined using optical microscopy (OM).
The analysis revealed a critical defect: a continuous, unbonded interface originating from the original faying surface and extending to the exterior surface of the weld. This interface acted as a direct conduit for fluid leakage from the internal channel. Crucially, this was not a classic crack but a “cold weld” or lack-of-penetration defect. The morphology indicated that during FSW, the stirring pin was severely offset from the joint line. Consequently, the original interface in the affected zone was merely plastically deformed and bent under the forging action of the tool shoulder, without undergoing the intense stirring and mixing necessary for metallurgical bonding. This created a weak, unbonded plane. The process can be conceptually described by a modified energy input equation, where effective bonding requires surpassing a critical thermomechanical energy threshold:
$$ E_{eff} = k \cdot \frac{\omega \cdot F_z}{v} \cdot \eta(T, \text{Alignment}) $$
where \(E_{eff}\) is the effective energy for bonding, \(\omega\) is the tool rotation speed, \(F_z\) is the axial force, \(v\) is the travel speed, \(k\) is a material constant, and \(\eta\) is an efficiency factor highly dependent on interface temperature \(T\) and tool alignment. In the case of severe pin offset, \(\eta \rightarrow 0\) at the original interface, leading to \(E_{eff} < E_{critical}\), resulting in a cold weld.
For comparison, common FSW defects and their characteristics in the context of an EV battery pack are summarized in Table 3.
| Defect Type | Typical Morphology | Primary Cause | Impact on Sealing |
|---|---|---|---|
| Tunnel / Void | Elongated cavity within nugget | Insufficient heat, excessive travel speed, improper tool design | High (Direct leak path if connected) |
| Kissing Bond / Cold Weld | Unbonded interface, often linear | Severe tool offset, low heat input, oxide entrapment | Very High (Planar leak path) |
| Lack of Penetration | Unbonded region at root | Short pin length, excessive gap, low force | High (Root leak path) |
| S-line (Oxide Line) | Curvilinear oxide stringers | Entrapment of surface oxides, insufficient material flow | Moderate to High (Can initiate cracking) |
| Hook Defect | Upward bending of interface in lap joints | Asymmetric material flow, improper parameters | Depends on geometry |
The defect found in this EV battery pack tray was a quintessential kissing bond/cold weld initiated by tool misalignment. Secondary voids observed in the nugget were attributed to a subsequent, non-optimal Tungsten Inert Gas (TIG) repair weld attempted on the flawed FSW joint and were not the primary leak cause.
3. Root Cause Analysis and Process Optimization Strategy
The investigation conclusively identified excessive stirring pin offset as the root cause of the leakage. The production FSW equipment, while capable, relied on manual setup and fixturing for joint alignment. Cumulative errors from part fit-up, fixture wear, and tool run-out could lead to a pin deviation exceeding the permissible limit for a sound weld in this joint configuration. A misaligned pin fails to sufficiently plasticize and intermix the material at the original interface, leaving an unbonded seam. The resulting joint strength is governed not by metallurgical bonds but by frictional forces, which can be approximated as:
$$ \tau_{interface} \approx \mu \cdot \sigma_n $$
where \(\tau_{interface}\) is the shear strength at the unbonded interface, \(\mu\) is the coefficient of friction, and \(\sigma_n\) is the contact pressure from the forging force. This strength is orders of magnitude lower than that of a properly consolidated FSW joint, making it prone to forming a leak path under internal pressure.
The optimization strategy was twofold:
- Enhanced Process Control: A laser-based seam tracking system was integrated into the FSW machine. This system provides real-time feedback and dynamic tool path correction, ensuring the stirring pin remains centered on the joint line with a precision of \(\pm 0.2 \, \text{mm}\).
- Parameter Re-optimization: With the alignment issue resolved, the welding parameters were re-evaluated to ensure sufficient heat input and material flow for the now correctly aligned joint. The optimized parameters are listed in Table 4, contrasted with the nominal production parameters suspected of being used when the fault occurred.
| Parameter | Nominal (Fault Condition) | Optimized | Function & Impact |
|---|---|---|---|
| Rotation Speed, \(\omega\) (rpm) | ~1200 | 1500 | Increased heat generation, improves material plasticity. |
| Travel Speed, \(v\) (mm/min) | ~1000 | 800 | Reduced to increase heat input per unit length. |
| Axial Force, \(F_z\) (kN) | ~6.5 | 7.0 | Slightly increased to ensure forging and consolidation. |
| Pin Alignment | Manual, Uncontrolled (>0.5mm offset possible) | Laser-tracked (\(\pm 0.2\) mm) | Key Improvement: Ensures pin interacts with joint line. |
| Specific Energy Input* | ~ \(4.7 \times 10^3 \, \text{J/mm}\) | ~ \(8.8 \times 10^3 \, \text{J/mm}\) | Significantly higher, promoting complete bonding. |
*Specific Energy Input is estimated proportionally to \(\omega / v\). The optimized parameters provide a calculated increase of ~87%.
4. Validation of Optimized Process
The effectiveness of the implemented solutions was validated through metallographic examination and mechanical testing of joints produced with the optimized setup.
Metallurgical Quality: Cross-sectional analysis of the optimized weld confirmed the complete elimination of the cold weld defect. The weld nugget showed a sound, defect-free structure with a well-defined thermomechanically affected zone (TMAZ) and heat-affected zone (HAZ). The original joint line was fully consumed within the dynamically recrystallized nugget zone.
Mechanical Performance: Transverse tensile tests were conducted on weld coupons extracted from optimized production samples. The results, summarized in Table 5, demonstrate a significant recovery in joint strength.
| Sample | Ultimate Tensile Strength (MPa) | Failure Location | Joint Efficiency (%)* |
|---|---|---|---|
| 1 | 211 | HAZ/TMAZ boundary | 74.0 |
| 2 | 208 | HAZ | 73.0 |
| 3 | 210 | HAZ | 73.7 |
| 4 | 207 | HAZ | 72.6 |
| 5 | 209 | HAZ/TMAZ boundary | 73.3 |
| Average | 209 ± 1.6 | — | 73.3 ± 0.5 |
*Joint Efficiency = (Avg. Weld UTS / Base Metal UTS) * 100%. Base Metal UTS = 285 MPa.
The average tensile strength of 209 MPa represents approximately 73.3% of the base material strength, which is typical and acceptable for 6005-T6 FSW joints where failure commonly occurs in the softened HAZ. Fractographic analysis via Scanning Electron Microscopy (SEM) revealed a dimpled rupture surface, confirming a ductile failure mode, which is indicative of a sound metallurgical joint free from planar defects.
Production Quality Metrics: The optimized process was implemented in full-scale production of the EV battery pack trays. Batch monitoring over several production cycles showed a dramatic improvement in leak test yield. The first-pass acceptance rate stabilized above 99.8%, effectively eliminating the costly failure mode associated with weld leakage.
5. Conclusion and Broader Implications
This investigation successfully diagnosed and remedied a critical leakage failure in an EV battery pack lower tray manufactured via FSW. The primary conclusions are:
- The leakage was caused by a “cold weld” or kissing bond defect, resulting from severe misalignment of the FSW tool stirring pin relative to the joint line. This misalignment prevented the thermomechanical conditioning necessary for metallurgical bonding at the original faying surface, creating a continuous, unbonded planar defect that served as a leak path.
- The integration of a laser seam tracking system to maintain pin alignment within \(\pm 0.2 \, \text{mm}\) was the pivotal corrective action. This was supplemented by an optimized parameter set (1500 rpm, 800 mm/min) to ensure adequate energy input \(E_{eff}\). The relationship underscores that both precise geometric control (\( \eta(T, \text{Alignment}) \rightarrow 1 \)) and sufficient parametric input are non-negotiable for robust weld integrity in EV battery pack applications.
- The optimized process eliminated the defect, as confirmed by metallography, and produced joints with consistent mechanical properties (average UTS ~209 MPa, 73% joint efficiency) and a ductile fracture mode. Most importantly, it restored and enhanced production reliability, achieving a sustained leak-test pass rate >99.8%.
This case study highlights that in high-volume, quality-critical applications like EV battery pack manufacturing, advanced process control systems are essential to mitigate the risks posed by inherent production variabilities. The findings reinforce that tool alignment is as critical a parameter as rotation speed, travel speed, and force in the FSW process window. For future designs, incorporating self-locating joint geometries or specifying slightly oversized (dwell-initiating) pin features for challenging fit-ups could provide additional robustness. The methodology and solutions presented herein offer a validated framework for troubleshooting and enhancing the quality of FSW joints in safety-critical automotive components.