In the rapidly evolving landscape of the automotive industry, the shift towards electric vehicle cars has brought unprecedented focus on passenger comfort. One of the most critical challenges in achieving superior cabin refinement in these quiet vehicles is managing Noise, Vibration, and Harshness (NVH) from ancillary systems. The electric compressor, a core component of the thermal management system for cabin air conditioning and battery cooling, has emerged as a significant NVH source. Unlike in internal combustion engine vehicles where its noise is often masked by the engine, the serene acoustic environment of an electric vehicle car makes compressor-related noises distinctly audible and potentially objectionable to occupants.
My analysis centers on a specific NVH issue encountered during the development of an electric vehicle car: pronounced knocking noises and steering wheel judder during the compressor’s start and stop sequences. These transient events, occurring every time the climate control is activated or deactivated, present a direct threat to perceived quality. This paper details my first-person engineering investigation into the root causes of these start-stop noises and presents a multi-faceted improvement strategy encompassing both software control logic and mechanical design modifications.
Problem Characterization and Data Analysis
The subjective complaint was clear: an unacceptable “thudding” or “knocking” sound accompanied by palpable steering wheel vibration during compressor engagement and disengagement. To quantify the issue, I conducted detailed vibration measurements on the compressor assembly and its mounting points. The time-domain data and frequency-domain colormap plots revealed distinct signatures correlated to the transient events.
During the start-up phase, the time-domain signal showed a continuous, high-amplitude vibration lasting approximately 3 seconds. The colormap translation of this data pinpointed two key phenomena: a persistent low-frequency disturbance around 31 Hz during the initial “open-loop” control phase, and prominent 24th-order noise (related to the motor’s 8 poles and 3 phases: 8 * 3 = 24) throughout the ramp-up. The shutdown event was marked by a single, sharp vibration impulse, also exhibiting strong energy at 31 Hz. These data points provided the first clues, indicating issues related to structural resonance (31 Hz) and excitation from motor control strategies.
Fundamental Noise Source: Mechanical Architecture
The root of the problem lies in the inherent mechanical design of the scroll-type electric compressor commonly used in electric vehicle cars. The core driveline consists of the motor rotor, a drive shaft, an eccentric sleeve (crank pin), and the orbiting scroll. A critical design feature is the flexible connection between the drive shaft and the eccentric sleeve. This clearance fit is essential to compensate for manufacturing tolerances, thermal expansion, and operational deformations, ensuring smooth engagement of the scroll set under various loads.
However, this flexibility becomes a liability during transient conditions. At the moment of start or stop, significant gas pressure forces and resistance torques act on the orbiting scroll. The dynamics can be modeled by considering the torque balance and impact mechanics. The equation governing the motion of the eccentric sleeve relative to the drive shaft can be simplified as:
$$ I_{e} \ddot{\theta}_{e} + c(\dot{\theta}_{e} – \dot{\theta}_{s}) + k(\theta_{e} – \theta_{s}) = T_{gas} – T_{friction} $$
Where:
$I_{e}$ is the inertia of the eccentric sleeve and orbiting scroll assembly.
$\theta_{e}$ and $\theta_{s}$ are the angular positions of the eccentric sleeve and drive shaft, respectively.
$c$ and $k$ represent the effective damping and stiffness in the flexible connection (initially just clearance).
$T_{gas}$ is the resisting torque from gas compression forces.
$T_{friction}$ is the system friction torque.
During a sudden start, the shaft accelerates ($\dot{\theta}_{s}$ increases rapidly). If the motor’s starting torque is high, the shaft velocity outpaces the sleeve’s ability to follow due to the resisting $T_{gas}$. The clearance is taken up violently, resulting in an impact. The impact force $F_{impact}$ is related to the relative velocity at collision:
$$ F_{impact} \approx \frac{m_{e} \Delta v}{\Delta t} $$
where $m_{e}$ is the effective mass of the eccentric assembly and $\Delta v$ is the relative velocity change across the clearance. This impact force excites the natural modes of the compressor housing and its mounting system, which includes the 31 Hz resonance identified earlier. The resulting structure-borne noise is efficiently transmitted through the brackets and body of the electric vehicle car into the cabin, manifesting as the objectionable knock and steering wheel shake.
Comprehensive Optimization Strategy
My improvement strategy attacked the problem from two angles: first, by minimizing the excitation source through refined motor control algorithms (Software Strategy), and second, by modifying the mechanical interface to attenuate any remaining impact energy (Hardware Strategy).
Software Strategy Optimization: Precision Control of Transients
The initial open-loop start-up phase, where the motor is driven with a fixed current without rotor position feedback, was a primary culprit. The original calibration used a high current ($I_{start}$) for a prolonged duration ($t_{open-loop}$) to ensure reliable starting under all possible system pressure conditions. This, however, created excessive and prolonged torque, exacerbating the shaft-sleeve impact. I systematically tuned these parameters, balancing start reliability with NVH performance. The optimized start significantly reduced the impulsive excitation. The relationship between start current, acceleration, and impact energy can be conceptualized as:
$$ E_{impact} \propto \frac{1}{2} I_{e} (\omega_{s} – \omega_{e})^2 \approx \frac{1}{2} I_{e} \left( \frac{k_t I_{start} – T_{load}}{I_{e}} \cdot t_{open-loop} \right)^2 $$
where $k_t$ is the motor torque constant and $T_{load}$ is the lumped load torque. Reducing $I_{start}$ and $t_{open-loop}$ directly reduces the energy $E_{impact}$ available at impact.
For shutdown noise, the strategy shifted from a “freewheel” stop to a “pulse-controlled braking” stop. In the freewheel method, all power electronics are shut off, allowing the compressor to coast down under its own mechanical and gas loads, often resulting in a final, high-impact collision as the shaft stops. The new strategy implements a 1:9 “point-brake” control after crossing a critical low-speed threshold. In this phase, the lower IGBTs are pulsed on briefly (10% duty cycle), using the motor’s back-EMF to generate a gentle braking torque that brings the assembly to a more controlled halt, minimizing the final relative velocity $\Delta v$.
Hardware Strategy Optimization: Damping the Source Impact
While software optimization brought substantial improvement, the fundamental flexible connection remained a potential source for residual noise or future issues under extreme conditions. Therefore, I explored a hardware modification to the eccentric sleeve assembly. The concept involved adding a resilient damping element within the clearance space to absorb impact energy. A pin-and-socket design was developed, incorporating a specially formulated elastomeric bushing. The improved impact dynamics with the bushing can be modeled by modifying the earlier equation to include nonlinear damping and stiffness from the bushing material:
$$ F_{impact, damped} = \int c_{bush}(\dot{x}) \dot{x} , dt + \int k_{bush}(x) x , dt $$
where $x$ is the relative displacement across the clearance, and $c_{bush}$ and $k_{bush}$ are the damping and stiffness functions of the bushing material, which are typically viscoelastic and non-linear. This design transforms a high-impact, high-frequency excitation into a softer, more damped transient, significantly reducing the input force to the housing structure. The selection of the elastomer material is paramount, as it must withstand the harsh internal environment of the compressor, including high pressure, wide temperature swings (-30°C to 130°C), and exposure to refrigerant and lubrication oil. Materials like hydrogenated acrylonitrile butadiene rubber (HNBR) or specially compounded fluorosilicones are primary candidates due to their excellent resistance to heat, chemicals, and compression set.
Quantitative Results and Comparative Analysis
The effectiveness of the implemented strategies was validated through rigorous bench testing and in-vehicle assessments on the electric vehicle car platform. The following tables summarize the key parameter changes and the resulting NVH performance metrics.
| Parameter | Initial Calibration | Optimized Calibration | NVH Effect |
|---|---|---|---|
| Open-loop Start Current ($I_{start}$) | 9.5 A | 7.5 A | Reduced initial torque peak, lower impact energy. |
| Open-loop Duration ($t_{open-loop}$) | 2800 ms | 1400 ms | Shortened high-torque phase, less time for sustained excitation. |
| 24th Order Noise Level | Prominent | Subdued to negligible | Reduced magnetic force excitation from lower current. |
| 31 Hz Vibration Duration | Continuous ~3s | Single, short impulse | Elimination of resonant “ringing”. |
| Shutdown Phase | Freewheel Strategy | 1:9 Point-Brake Strategy | Mechanism & Outcome |
|---|---|---|---|
| Controlled Deceleration | Normal ramp-down | Normal ramp-down | Identical initial phase. |
| Final Stop Control (< Critical Speed) | IGBTs OFF (Freewheel) | Lower IGBTs pulsed at 10% duty cycle | Back-EMF braking provides controlled deceleration. |
| Final Impact Energy | High | Low | Final relative velocity ($\Delta v$) is minimized. |
| 31 Hz Impulse Magnitude | High amplitude peak | Low amplitude peak | Significant reduction in structure-borne excitation. |
| Improvement Pathway | Key Action | Advantage | Consideration / Challenge |
|---|---|---|---|
| Software Strategy Tuning | Reduce start current & duration; Implement brake-stop. | Highly effective, no cost or mass increase, reversible. | Must be balanced against start reliability under high system pressure. |
| Mechanical Design Modification | Add elastomeric bushing in eccentric assembly. | Addresses root cause, robust against varied conditions. | Adds cost and complexity; Critical material selection for durability. |
| Combined Approach | Implement both software tuning and hardware damping. | Provides maximum NVH margin and robustness for premium electric vehicle cars. | Highest development effort and unit cost. |
Conclusion and Forward-Looking Perspectives
This investigation into the start-stop noise of an electric compressor for an electric vehicle car underscores the intricate interplay between mechanical design, control software, and system-level NVH performance. The flexible connection, while mechanically necessary, creates a fundamental vulnerability to impact-induced noise during transients. My work demonstrates that a systematic approach—first refining the electromagnetic excitation profile through careful calibration of start/stop algorithms, and second, considering source-path attenuation via mechanical damping—can effectively solve this challenging issue.
The software-based solution offers a powerful and immediate tool for engineers developing thermal management systems for electric vehicle cars. It highlights the importance of viewing the compressor not just as a mechanical pump but as a mechatronic system whose acoustic output is profoundly influenced by its electronic control unit (ECU) software. The hardware modification, though involving greater design effort, provides a more fundamental solution and is a valuable consideration for next-generation compressor designs aimed at the most stringent NVH standards.
Looking ahead, the lessons learned are directly applicable to the broader field of electrified mobility. As the electric vehicle car market matures and consumer expectations for quietness continue to rise, the NVH refinement of all ancillary systems will remain a critical competitive differentiator. Future research could integrate real-time system pressure estimation to dynamically adapt start-up torque, or explore advanced active vibration control algorithms that use the motor itself to cancel specific vibration orders. Ultimately, achieving silent and seamless operation of components like the electric compressor is a essential step in delivering the tranquil and refined driving experience that defines the modern electric vehicle car.