Switched Reluctance Motor (SRM) Test Bench: Torque Ripple, Inductance Mapping and NVH

Why SRM Testing Requires a Different Approach

Switched reluctance motors have re-emerged as a serious contender in EV drivetrains and 48 V auxiliary systems. Unlike PMSM or induction motors, the SRM has no permanent magnets and no rotor windings — torque is generated entirely by the tendency of the rotor poles to align with energised stator poles. This simple construction brings excellent fault tolerance and low material cost, but it introduces a set of test challenges that a standard AC motor test bench is not set up to handle out of the box.

Rotor-position-dependent inductance, inherently high torque ripple, and aggressive current commutation waveforms demand specific measurement capabilities. Engineers who attempt to characterise an SRM on a test bench designed for synchronous machines will miss the most critical performance indicators.

How the SRM’s Operating Principle Changes the Test Agenda

In a PMSM, you map speed-torque-efficiency in a relatively smooth operating space. In an SRM, the machine’s behaviour is fundamentally non-linear: phase inductance varies continuously with rotor angle, and torque is a function of both current and position. Every meaningful SRM characterisation therefore starts with an inductance-versus-angle (L-θ) map.

• L-θ profiling: The test bench locks the rotor at discrete angular positions using a precision servo brake or positioning motor, applies a small AC perturbation current to each phase in turn, and measures the resulting phase inductance. The full aligned-to-unaligned inductance ratio (typically 5:1 to 10:1) defines the machine’s torque-producing capability.
• Static torque measurement: At each locked rotor position, the bench ramps phase current and records the resulting static torque via the reaction torque transducer. This builds a T(θ, i) lookup table that feeds the control unit.
• Dynamic commutation sweep: Turn-on angle (θ_on) and turn-off angle (θ_off) are swept across the operating speed range to find the optimal switching angles for efficiency and torque ripple, both at rated load and at peak torque.

Core Test Items for SRM Characterisation

1. Torque Ripple Quantification

SRM torque ripple can reach 40–100% peak-to-peak at low speed without active ripple mitigation. The test bench must capture instantaneous torque at angular resolution below 1°. A shaft-coupled high-bandwidth torque transducer (≥ 5 kHz, accuracy ≤ 0.1% FS) combined with a resolver or encoder with 4,096 PPR or better is the minimum instrumentation requirement.

Key metrics: Peak-to-peak ripple as % of mean torque; fundamental ripple frequency (= speed × rotor pole count × phase count); torque ripple across the full operating range from 10% to 110% rated load.

2. Acoustic Noise and Vibration (NVH)

The radial magnetic forces that produce torque in an SRM also deform the stator back-iron at the switching frequency harmonics. This makes SRM-driven vehicles audible in ways that PMSM-based platforms are not. NVH testing on the SRM bench typically involves a semi-anechoic chamber rating or at minimum a sound power measurement per ISO 1680 / IEC 60034-9. The microphone array is placed at 1 m from the machine surface; vibration accelerometers are bonded to the stator housing at the four cardinal positions.

Key metrics: Sound power level (dB[A]) at 1 m; dominant spectral lines vs. pole-pair switching frequency; structurally transmitted vibration velocity (mm/s RMS) at stator housing.

3. Efficiency Mapping

Because an SRM operates without back-EMF at standstill, copper loss dominates at low speeds while core loss (hysteresis and eddy current in the laminated stator) rises sharply above the base speed. An efficiency map across the full torque-speed envelope requires the bench to operate in constant-torque mode at low speeds and constant-power mode above base speed. The input electrical power must be measured on the DC bus side (upstream of the converter) to capture converter losses that form part of the system efficiency.

Key metrics: Peak system efficiency (motor + inverter); NEDC / WLTP weighted average efficiency; partial load efficiency at 10%, 25%, 50% rated torque.

4. Thermal Characterisation

SRM stators run hot at the winding end-turns, but the rotor (having no windings) remains cooler than the equivalent PMSM. Fibre-optic temperature sensors in the stator slots and a non-contact infrared probe on the rotor OD give the clearest picture. The bench should hold junction temperature within ±1°C of the target setpoint during steady-state efficiency measurement to eliminate thermal drift effects on resistance.

5. EMC / Conducted Emissions from the Power Stage

Hard-switching SR converters generate significant conducted electromagnetic interference on the DC supply rail. The test bench should include a LISN (Line Impedance Stabilisation Network) per CISPR 25 Class 3 and a 100 MHz bandwidth current probe on each phase lead to capture switching spikes. This is especially relevant for 48 V mild-hybrid BSG/ISG applications where the SRM shares a bus with sensitive ADAS electronics.

Key metrics: Conducted emission envelope vs. CISPR 25 limits; peak voltage transient per switching event (V/µs dV/dt).

Test Bench Configuration for SRM Work

A dynamometer with four-quadrant AC regenerative loading is preferred so that energy during active braking is returned to the grid rather than dissipated as heat in resistor banks — important at high duty cycles. The speed controller must handle the SRM’s highly non-sinusoidal torque output without hunting; a bandwidth of at least 200 Hz in speed control mode is typical. For high-voltage applications (400 V and above), the bench isolation transformer and safety interlock must be rated for the DC bus voltage with at least 2× safety margin.

If your programme involves switched reluctance machines — whether for automotive ISG systems, industrial servo drives, or emerging EV traction — our engineering team can specify the right measurement chain and control architecture for your motor’s pole count and power level. Talk to our engineering team about SRM bench configuration.