In the rapidly evolving field of electric vehicles, noise, vibration, and harshness (NVH) performance has become a critical development metric for manufacturers. As an engineer focused on electric SUV applications, I have encountered significant whining noise issues during full-throttle acceleration, which severely impact driving comfort and acoustic quality. This paper presents a comprehensive analysis and optimization approach for addressing whining noise in an electric SUV, utilizing a structured methodology that combines theoretical modeling, experimental testing, and practical engineering solutions. The core of this work revolves around the “noise source—structural path/airborne radiation path—in-cabin receiver” model, which allows for a systematic diagnosis and mitigation of the problem. Through this approach, I have identified key contributors to the noise, including electromagnetic forces in the motor and gear meshing in the reducer, and developed actionable strategies to reduce their effects. The insights gained from this study are not only applicable to the specific electric SUV under investigation but also serve as a valuable reference for similar electric vehicle platforms, emphasizing the importance of integrated source and path control in NVH optimization.
The electric SUV in question exhibited pronounced whining noise during full-throttle acceleration on asphalt roads, as confirmed through objective measurements and subjective evaluations. This noise was characterized by multiple distinct orders, primarily originating from the motor and reducer components. For instance, motor-related orders included the 30th order and its harmonics, stemming from electromagnetic forces between the stator and rotor, while reducer-related orders involved the 8.58th, 22nd, and their respective harmonics, resulting from gear engagement dynamics. These noise components were identified using advanced analysis tools like LMS Test.Lab, which generated waterfall plots and order curves to visualize the acoustic behavior. The subjective assessment rated the noise as unacceptable, particularly in the mid to high-speed ranges (e.g., 40–120 km/h), highlighting the urgency for intervention. To illustrate the context, consider the following representation of the noise characteristics, which underscores the multi-order nature of the issue in electric SUV applications.
To delve deeper into the whining noise mechanism, I employed a fishbone diagram analysis that categorizes the problem into sources, paths, and receivers. For the electric SUV, the primary noise sources include the permanent magnet synchronous motor and the reducer, which generate vibrations due to electromagnetic excitations and gear interactions. The structural paths involve components like mounts, high-voltage wiring, and pipelines, which transmit vibrations to the cabin, while the airborne paths relate to acoustic radiation through the vehicle’s body and sound package. The motor’s electromagnetic noise, for example, arises from factors such as air gap variations and torque fluctuations, which can be modeled using electromagnetic force equations. One fundamental equation for the force frequency is given by: $$ f_m = \frac{n \times N}{60} $$ where \( f_m \) is the frequency in Hz, \( n \) is the rotational speed in rpm, and \( N \) is the order number. Similarly, for the reducer, gear meshing frequency is critical: $$ f_g = \frac{n \times z}{60} $$ where \( z \) is the number of teeth on the gear. In the electric SUV under study, the reducer parameters are detailed in Table 1, showing a two-stage reduction setup that contributes to specific order noises. These formulas help quantify the noise sources, enabling targeted optimizations.
| Component | Primary Reduction Gear Set | Main Reduction Gear Set |
|---|---|---|
| Driving Gear Teeth | 22 | 23 |
| Driven Gear Teeth | 59 | 71 |
From the source perspective, the motor and reducer were identified as major contributors based on bench tests and near-field measurements. For instance, the motor’s 0.5 m sound pressure level test revealed that the 30th and 60th orders exceeded target levels across the 0–120 km/h speed range, as shown in Figure 5 of the original data. Similarly, the reducer’s 1 m sound pressure level test indicated that orders like the 22nd and its harmonics were problematic. The electromagnetic forces in the motor are influenced by the air gap between the stator and rotor; a smaller gap can increase magnetic flux density and force amplitudes, leading to higher noise. To mitigate this, I proposed increasing the air gap from 0.6 mm to 0.8 mm, which reduces the electromagnetic force density and, consequently, the noise. The improvement can be expressed using the relationship for magnetic force: $$ F \propto \frac{B^2}{\mu_0} $$ where \( B \) is the magnetic flux density and \( \mu_0 \) is the permeability of free space. By increasing the air gap, \( B \) decreases, thereby reducing \( F \) and the associated noise. For the reducer, reducing the pressure angle of the gears was implemented to minimize meshing impacts and transmission error, which are key drivers of whining noise in electric SUV applications. The pressure angle adjustment affects the contact ratio and load distribution, as described by: $$ \text{Contact Ratio} = \frac{\text{Path of Contact}}{\text{Base Pitch}} $$ A lower pressure angle increases the contact ratio, smoothing out engagement and reducing noise peaks.
In terms of path control, the structural transfer through mounts and the airborne radiation via acoustic packages were critical areas for optimization. The mounts’ vibration isolation performance was assessed through experimental tests, revealing that the original mounts had insufficient isolation, with transmissibility values below 20 dB in the 1000–5000 rpm range. To address this, I reduced the mount stiffness by changing the rubber angle from 48° to 40°, which improved the isolation rate. The transmissibility \( T \) of a mount can be modeled as: $$ T = \frac{1}{\sqrt{1 + \left( \frac{2 \zeta \omega}{\omega_n} \right)^2}} $$ where \( \omega \) is the excitation frequency, \( \omega_n \) is the natural frequency, and \( \zeta \) is the damping ratio. By lowering stiffness, \( \omega_n \) decreases, shifting the system away from resonance and enhancing isolation. Additionally, for airborne noise, I introduced an acoustic wrap around the motor, which acts as a barrier to sound radiation. The insertion loss of such a wrap can be approximated by: $$ \text{IL} = 10 \log_{10} \left( \frac{1}{\tau} \right) $$ where \( \tau \) is the transmission coefficient. This wrap reduced noise levels by 5–10 dB(A) at speeds above 70 km/h, demonstrating its effectiveness in electric SUV noise control. The combined impact of these path optimizations is summarized in Table 2, which compares the subjective and objective improvements.
| Optimization Measure | Effect on Low Speed (0–40 km/h) | Effect on Medium Speed (40–80 km/h) | Effect on High Speed (80–120 km/h) | Overall Subjective Improvement |
|---|---|---|---|---|
| Reducer Pressure Angle Reduction and Motor Air Gap Increase | +1 | -1 | +1 | +1 |
| Mount Stiffness Reduction | +1 | +1 | +0.5 | +1 |
| Motor Acoustic Wrap | 0 | +0.5 | +1 | +1 |
The validation of these optimizations involved both objective testing and subjective evaluations. For the electric SUV, the motor air gap increase and reducer pressure angle reduction led to a noticeable reduction in the 60th order noise, with peak levels dropping by approximately 5 dB(A) in the 80–100 km/h range. The mount stiffness modification resulted in improved transmissibility, with most values exceeding 18 dB and approaching the target of 20 dB. The acoustic wrap further attenuated airborne noise, as confirmed through in-cabin measurements. Subjectively, the noise was rated using a 10-point scale, where higher scores indicate better performance. The original state scored 5 across speed ranges, but after optimizations, scores improved by up to 1 point, reflecting enhanced acoustic comfort. This holistic approach ensures that the electric SUV meets NVH targets while maintaining practical engineering feasibility. The success of these measures underscores the importance of a balanced strategy that addresses both source excitations and path transmissions in electric vehicle design.
In conclusion, the whining noise issue in the electric SUV was effectively mitigated through a systematic analysis and optimization process. By leveraging the “noise source—structural path/airborne radiation path—in-cabin receiver” model, I identified key contributors and implemented solutions such as increasing the motor air gap, reducing reducer pressure angles, lowering mount stiffness, and adding acoustic wraps. These interventions resulted in significant noise reductions, validated through objective data and subjective feedback. The methodologies and results presented here provide a robust framework for addressing similar challenges in other electric SUV models, emphasizing the value of integrated NVH strategies. As electric vehicles continue to evolve, such approaches will be crucial for enhancing passenger comfort and advancing the overall driving experience. Future work could explore advanced materials and control algorithms to further refine noise performance in electric SUV applications.