Electric Power Steering Motor Test Bench: 토크 리플, Acoustics, and Functional Safety Validation

Why Electric Power Steering Motors Demand a Dedicated Test Approach

Electric power steering (EPS) motors sit at the intersection of three demanding disciplines: precision motion control, cabin acoustics, and automotive functional safety. A passenger feeling vibration through the steering wheel traces directly to torque ripple from the assist motor. A buzzing noise at low speed—often called “EPS whine” 또는 “motor moan”—originates in the same device. And because the system is ISO 26262 ASIL-C or ASIL-D rated, a latent fault in the motor or its drive electronics must not result in hazardous loss of assist.

These requirements push far beyond what a generic motor dynamometer provides. The test bench must resolve torque to fractions of a Newton-meter at low speed, measure acoustic noise below 50 데시벨(에이), and execute fault injection sequences while monitoring safety-relevant signals in real time. This guide covers the key test items, required bench architecture, and common pitfalls when setting up EPS motor validation.

EPS Motor Types and Their Test Implications

Three topologies dominate the EPS landscape today:

• Column-EPS (C-EPS): Motor and reduction gear sit on the steering column; power ratings typically 400–900 W. Acoustic requirements are strictest because the motor is closest to the driver.
• Pinion-EPS (P-EPS) and Rack-EPS (R-EPS): Motor acts at the rack or pinion; power up to 2.5 kW for larger vehicles. Torque ripple translates into lateral force perturbations on the rack and front wheels.
• Steer-by-Wire (SbW) road-feel actuator: No mechanical fallback; ISO 26262 ASIL-D hardware architecture with redundant motor windings and dual motor control units (MCUs). Testing must validate failover transitions within the fault-tolerant time interval (FTTI) specified in the safety case.

핵심 테스트 항목

1. Torque Ripple and Cogging Characterisation

EPS torque ripple drives the dominant steering-feel complaint. The test bench rotates the motor shaft at speeds from 10 에게 3,000 rpm and measures instantaneous torque with a non-contact rotary torque transducer (full-scale typically 20–50 Nm, class 0.1 accuracy per IEC 60034-1). Ripple is expressed as peak-to-peak amplitude and as spectral harmonics (6th, 12th for a 6-phase or 12-slot machine). Cogging torque—the zero-current component—is measured separately under de-energised rotation.

주요 지표: peak-to-peak ripple < 0.5% 정격 토크의; 코깅 < 1% of peak torque; 6일차 고조파 < −40 dB relative to fundamental.

2. 음향 소음 및 진동 (NVH)

EPS NVH testing requires a semi-anechoic cell or an anechoic box coupled to the dynamometer. Sound pressure level (SPL) is measured with a calibrated microphone at 0.1 m or 1 m from the motor housing, while a laser Doppler vibrometer or accelerometer captures surface vibration. Order-tracked spectra isolate electromagnetic noise (슬롯 고조파, switching frequency sidebands) from mechanical noise (bearing frequency, gear mesh). Pass/fail thresholds are OEM-specific but commonly require SPL < 45 데시벨(에이) at rated assist torque and 300 rpm shaft speed.

3. Efficiency and Thermal Performance

Input electrical power (DC bus voltage × current, measured at 0.1% 정확성) versus shaft mechanical power (torque × angular velocity) yields efficiency maps across speed-torque grid points. Thermal imaging and embedded thermocouples (굴곡, magnet, housing) confirm that the motor operates within its thermal limits under the sustained assist duty cycle. For C-EPS motors with tight packaging inside the column shroud, a recirculating oil or coolant jacket is sometimes used; the bench must replicate actual cooling flow rate and inlet temperature.

4. Functional Safety and Fault Injection

ISO 26262 part 5 and part 6 require verification that the motor and its electronic control unit (ECU) respond correctly to fault conditions. The test bench acts as a controllable fault injector: it can interrupt individual phase windings, short phases together, inject phase-voltage offsets, or simulate sensor failures (resolver, GMR position sensor). The safety controller monitors whether the system transitions to the defined safe state (reduced-assist or no-assist mode) within the FTTI—typically 50–200 ms depending on speed and vehicle dynamic requirements. Hardware-in-the-loop (힐) integration connects the real ECU to a vehicle dynamics model so that steering column torque, vehicle speed, and lateral acceleration signals are all present during fault injection.

5. Endurance and Durability

EPS motors must survive 15 years or 250,000 km of mixed driving. The endurance profile combines high-cycle assist events (parking manoeuvres generate peak torque at near-zero speed), sustained assist at highway speed (낮은 토크, moderate speed), and thermal cycles. The bench runs an automated accelerated life test (ALT) schedule per IEC 60068-2 and OEM-specific profiles, logging cumulative winding temperature and insulation resistance at intervals.

6. EMC Pre-Compliance

EPS motors operate at switching frequencies of 8–20 kHz. Radiated and conducted emissions are pre-screened on the test bench before full CISPR 25 chamber testing, with a near-field probe array mapping harmonic hot spots on the motor housing and cable harness.

Test Bench Architecture for EPS Motor Validation

A complete EPS motor test bench integrates several subsystems. The dynamometer is typically a small-frame AC or hysteresis unit rated 5–20 Nm at 0–6,000 rpm, with torque accuracy better than 0.1% FS to resolve the small ripple signals. The drive system provides a programmable DC bus (8–60 V for 12 V systems, 24–60 V for 48 V systems) with bidirectional regeneration. A dSPACE, NI, or similar real-time platform runs the closed-loop speed/torque controller and the HIL safety logic. The acoustic enclosure is lined with absorptive foam and sits on anti-vibration mounts to isolate structure-borne noise from the building slab.

테스트 벤치 선택에 있어 이것이 의미하는 바

Selecting a test bench for EPS motor development requires attention to torque transducer resolution (not just full-scale range), acoustic integration capability, and the software framework for fault injection scripting. A bench that excels at large traction motor efficiency mapping is not automatically suitable for 900 W EPS NVH characterisation. If your programme is transitioning from C-EPS to SbW architectures, the bench must also support redundant signal paths and dual-ECU configurations.

Our engineering team has configured EPS validation benches for both column and rack architectures across 12 V and 48 V platforms. To discuss test requirements and bench specifications for your programme, 우리 엔지니어링 팀에 문의하세요.