Torque Ripple Testing: Principles, Methods, and Applications

Torque ripple testing, also known as torque fluctuation testing, quantifies the periodic variation in output torque of electric motors, e-axles, gearboxes, and power tools — a critical indicator of smoothness, NVH performance, and control quality. Using high-bandwidth torque transducers and dynamometer test benches, accredited third-party laboratories measure peak-to-peak ripple, torque ripple coefficient, and harmonic orders to support design validation, benchmarking, and production quality control. This article explains the technical principles, measurable parameters, instrumentation, test methods, and industry applications of torque fluctuation testing.

Technical Overview of Torque Ripple Testing

Torque ripple, also referred to as torque fluctuation, is the periodic variation of instantaneous output torque about its mean value when a machine operates under nominally steady-state conditions. For rotating electrical machines, gearboxes, and power tools, this fluctuation is superimposed on the useful average torque and appears as a repeating waveform synchronized with rotor angle, electrical frequency, or gear mesh frequency. Torque ripple testing measures the amplitude, frequency content, and statistical distribution of this fluctuation to evaluate how smoothly a product delivers torque.

The importance of torque ripple has grown sharply with the spread of electrified drivetrains. In an electric vehicle traction system, periodic torque variations at the motor shaft pass through the reducer and half-shafts into the vehicle body, where they can excite resonance and generate objectionable gear whine or structure-borne noise. In servo systems, torque fluctuation causes speed ripple that degrades positioning accuracy and machining surface finish. In fastening tools, output torque inconsistency directly affects joint quality and process capability. As a result, torque ripple has become a formal acceptance criterion in many development programs and supplier specifications.

Torque ripple arises from three principal groups of mechanisms:

• Electromagnetic sources, including cogging torque caused by the interaction of rotor magnets with stator slotting, non-ideal back-EMF waveforms, current harmonics introduced by inverter switching and dead-time, and commutation ripple in six-step driven brushless DC machines.
• Mechanical sources, including rotor eccentricity, shaft misalignment, bearing runout, residual rotor unbalance, gear mesh stiffness variation, and backlash in geared drivetrains.
• Control-related sources, including limited current-loop bandwidth, current sensor gain and offset errors, and resolver or encoder angle errors that distort the injected current vector.

A complete torque ripple characterization typically distinguishes between loaded torque ripple, measured with the machine energized and producing rated torque, and cogging (detent) torque, measured with windings unenergized. Both components are usually evaluated because they respond to different design countermeasures and dominate under different operating conditions.

Testing Samples and Scope

Electric Motors and Sub-Assemblies

The most common test objects are rotating electrical machines of all types and sizes: permanent magnet synchronous motors (PMSM), brushless DC motors, AC induction motors, switched reluctance motors, stepper motors, and brushed DC motors. Samples may be submitted as bare stator-rotor pairs for early design screening, as complete motors, or as inverter-integrated drive units. Power ratings handled by accredited third-party laboratories typically range from a few watts for small appliance and instrument motors to several hundred kilowatts for traction machines, with speeds from a few revolutions per minute to above 20,000 rpm.

E-Drives, E-Axles, and Gearboxes

Integrated electric axles, e-drives with single- or multi-speed reducers, and standalone gearboxes are tested to evaluate combined motor and gear mesh torque fluctuation. In these samples, motor torque ripple, gear transmission error, and planetary mesh order components interact, so the measurement captures the resultant output ripple that the vehicle structure actually experiences. Reduction ratios up to typical axle ratios and input speeds exceeding 15,000 rpm can be accommodated on appropriate dynamometer rigs.

Power Tools and Assembly Fastening Tools

Cordless drills, pistol-grip drivers, impact wrenches, and DC fastening tools used on automated assembly lines are tested for output torque consistency and fluctuation. For these products, torque fluctuation testing overlaps with performance and conformance testing of torque output under the frameworks of ISO 5393 for rotary tools for threaded fasteners and ISO 6789 for hand torque tools, which define design conformance, quality conformance, and recalibration procedures.

Actuators and Specialized Mechanisms

Electric power steering (EPS) column and rack actuators, aerospace control-surface actuators, camera gimbal motors, medical infusion and compressor pumps, and direct-drive appliance motors are also common samples. These applications are especially sensitive to low-frequency, low-amplitude ripple because the human operator or the controlled process perceives very small torque irregularities.

Testing Items and Measured Parameters

A standard torque ripple test report quantifies the fluctuation through a defined set of parameters, typically including:

• Mean (average) torque, maximum instantaneous torque, and minimum instantaneous torque at each steady-state operating point.
• Peak-to-peak torque ripple amplitude and RMS ripple value.
• Torque ripple coefficient, calculated as (Tmax − Tmin) / (2 × Tavg) × 100%, which normalizes the fluctuation to the useful torque and enables comparison across load points.
• Harmonic order spectrum of the torque signal, identifying ripple orders tied to the electrical frequency (for example, the sixth electrical order in three-phase machines), slot and pole combinations, and gear mesh frequencies.
• Cogging torque amplitude versus rotor angle at zero current, together with its dominant spatial harmonics.
• Ripple variation trends across the operating envelope, for example ripple versus speed at constant load and ripple versus load at constant speed.
• Coupled speed ripple or torsional vibration at the shaft, measured simultaneously with torque.
• Back-EMF waveform and its harmonic content, when indirect electromagnetic assessment is required.

Operating points are usually selected to cover rated conditions, peak torque conditions, low-speed high-torque corners, and field-weakening regions, because ripple behavior frequently changes character across the torque-speed map. Thermal state is recorded or controlled, since magnet and winding temperatures shift both mean torque and ripple amplitude.

Testing Methods and Instrumentation

Dynamometer-Based Direct Measurement

The reference method couples the device under test through a low-inertia, high-torsional-stiffness coupling to an AC regenerative or eddy-current dynamometer that holds precise, steady operating points. An inline torque flange transducer with strain-gauge telemetry is mounted between the two machines, offering measurement bandwidths from DC to several kilohertz and accuracy classes traceable to national standards. A high-resolution incremental or absolute encoder on the shaft provides the angular reference, and the data acquisition system samples torque, speed, angle, and electrical quantities synchronously at rates sufficient to capture the highest significant harmonic — as a rule of thumb, at least ten times the highest ripple frequency of interest. Rig alignment, rotor balance verification, and torsional resonance checks of the bench itself are mandatory prerequisites, because parasitic ripple from the test stand can otherwise contaminate the results.

Cogging Torque Measurement

To isolate cogging torque, the machine windings are left unenergized and the rotor is driven by an auxiliary low-ripple drive at a slow, constant speed, or the stator reaction torque is measured with the rotor locked at indexed positions. Because cogging amplitudes are often below one percent of rated torque, the setup minimizes friction disturbance using air bearings or very low-friction supports, and the torque transducer is selected for high resolution in the appropriate range. The resulting torque-versus-angle profile is Fourier-analyzed to identify the spatial harmonic orders, which directly reflect the slot-pole combination and magnet segmentation of the design.

Signal Processing and Order Analysis

Raw torque signals are resampled in the angular domain using the encoder signal so that speed variation does not smear the harmonic content. An FFT-based order spectrum then separates electrical orders (multiples of the electrical frequency), mechanical orders (multiples of rotational frequency), and gear mesh orders. Anti-aliasing filters, appropriate windowing, and averaging over multiple revolutions improve repeatability. Test procedures for rotating machines reference the IEC 60034 series and IEEE 112 for general machine test conditions, EN 62841 for motor-operated electric tools, ISO 5393 and ISO 6789 for fastening and torque tools, and ISO 21940 for the influence of rotor unbalance; vibration-related evaluation may additionally reference the ISO 20816 series.

Indirect and Complementary Methods

When direct torque measurement is impractical, complementary methods provide screening-level information. The back-EMF integration method derives expected ripple from the recorded voltage waveform; calibrated motor models estimate torque from measured phase currents; and finite element analysis results are correlated against measured data to validate electromagnetic models. Dual-transducer rigs that instrument both motor side and load side allow the ripple generated by the device under test to be separated from ripple contributed by the dynamometer.

Calibration and Traceability

Torque transducers are calibrated in accordance with recognized static torque calibration procedures such as DIN 51309, with results traceable to national metrology institutes. Measurement uncertainty budgets account for transducer class, amplifier bandwidth, temperature drift, coupling parasitics, and sampling resolution, so that reported ripple amplitudes carry defensible uncertainty statements suitable for supplier qualification and homologation support.

Applications and Industry Use Cases

Torque ripple testing is applied across the full product lifecycle, from early design iteration to end-of-line quality control. Representative use cases include:

• Electric vehicle traction motors and e-axles, where measured ripple feeds NVH simulation, validates skewing, pole-slot selection, and magnet shaping, and verifies harmonic current injection strategies in the inverter control software.
• Electric power steering systems, where low-speed micro-ripple is correlated with steering feel assessments and audible noise complaints.
• Industrial servo motors for machine tools and robotics, where speed ripple derived from torque fluctuation determines surface finish quality and motion smoothness.
• Household appliances such as direct-drive washing machine motors and inverter-driven compressors, where ripple reduction is a primary lever for meeting noise specifications.
• Power tools and automated fastening equipment, where torque output consistency governs joint integrity and process capability indices.
• Aerospace and defense actuators, where micro-ripple affects pointing stability and control-surface smoothness.
• Production end-of-line screening, in which simplified ripple signatures are compared against golden-sample limits to detect winding faults, magnet defects, and assembly errors.
• Model validation and benchmarking, in which measured waveforms confirm finite element analysis and analytical predictions or quantify competitor performance.

In regulated industries, test reports issued by accredited third-party laboratories are routinely used as objective evidence in design verification plans, supplier PPAP-style submissions, and customer-specific specification audits.

Frequently Asked Questions

What is the difference between torque ripple and cogging torque?

Cogging torque is the ripple component that exists with zero winding current, caused purely by the magnetic reluctance variation between rotor magnets and stator slots. Torque ripple in the broader sense is the total fluctuation observed under load, which combines cogging with commutation effects, current harmonics, back-EMF imperfections, and mechanical contributors. The two are measured under different test conditions and address different design countermeasures.

How is the torque ripple coefficient calculated?

The most common definition is the torque ripple coefficient equal to (Tmax − Tmin) divided by twice the average torque, expressed as a percentage. Some specifications instead use peak-to-peak ripple divided by the mean torque without the factor of two, so the definition in the applicable specification document must always be confirmed before comparing results between suppliers or reports.

What torque ripple level is considered acceptable?

Acceptance limits are application specific. High-grade servo and EV traction machines are frequently specified in the range of roughly one to five percent ripple coefficient at rated conditions, while general-purpose industrial machines may tolerate higher values. Cogging torque targets are commonly expressed in millinewton-meters or as a small fraction of rated torque. The governing limit should always be taken from the customer or product specification rather than a universal threshold.

What sampling rate and transducer bandwidth are required?

The measurement chain must capture the highest ripple order of interest. For a traction machine running at high speed with significant slot harmonics, significant components can extend to several kilohertz, which motivates torque flange bandwidths of several kilohertz and sampling rates of at least ten times the highest frequency of interest, combined with appropriate anti-alias filtering. Lower-speed samples can be served by proportionally lower-bandwidth setups.

Why do measured ripple values sometimes differ between laboratories?

Differences typically stem from test bench torsional dynamics, coupling inertia and stiffness, dynamometer control quality, transducer mounting, thermal state of the sample, and the signal processing definitions applied. Reputable laboratories therefore document operating point control, angular resampling, averaging, and uncertainty, and follow calibration and procedure standards such as DIN 51309, IEC 60034 series guidance, and ISO 21940 balance checks to keep results comparable.

How can torque ripple be reduced?

Design-side measures include rotor or stator skewing, optimized slot-pole combinations, magnet shaping and segmentation, auxiliary grooves, and fractional-slot winding layouts. Control-side measures include sinusoidal current shaping, harmonic current injection, and higher current-loop bandwidth with accurate angle measurement. Mechanical measures include rotor balancing, alignment control, and gear profile corrections. Torque ripple testing quantifies the effectiveness of each measure before and after implementation.