RWTH Aachen University's Chair of Production Engineering of E-Mobility Components (PEM) has published a benchmarking study in the journal Energies. The sample covers 115 traction motors from 80 electric vehicles of model years 2019 to 2025, mostly models for the Chinese market, compared against European and American motors the institute studied in 2025. The paper concludes that e-drives from the three regions show no fundamental differences in housing structure, stator and rotor geometry, or newer cooling concepts; the differences lie in material choices and component design details.
Data come from MOVO Autodatas' DARWIN benchmarking database, PEM's own teardowns and public sources. In the selected EREV and PHEV models, the electric drive does not share a housing with the combustion engine. Power density is calculated on the weight of the entire drive unit, including power electronics, reduction gear and housing, but excluding half-shafts.
The paper cites housing material and busbar design as two examples of regional differences; another is motor type: the China-focused sample contains no electrically excited synchronous motors. The authors note that the study can identify differences but not determine their causes, and the explanations offered are hypotheses.
| Item | Chinese models | European models |
|---|---|---|
| Housings and structural parts | Several automakers have switched to magnesium alloy in recent years | Aluminum alloy retained |
| Housing weight | Approx. 20%–30% lighter than aluminum housings | Baseline |
| Stator busbar | Mostly busbars with powder-coat insulation | Mostly assemblies with multiple copper busbars embedded in polymer parts; more complex structure |
| Motor type | Permanent-magnet synchronous, induction; no electrically excited synchronous | BMW and Renault have mass-produced electrically excited synchronous motors for years |
| Common ground | No fundamental differences in housing structure, stator and rotor geometry or new cooling designs. Thinner electrical steel laminations, higher speeds and finer conductor cross-sections appear in all three regions | |
| Possible causes cited in the paper | Distribution of raw material reserves and processing industries, labour costs, environmental and regulatory requirements. The paper says these are assumptions | |
| Magnesium supply | 97% of the EU's magnesium comes from China (European Commission figure, period not specified) | |
The paper argues that downsizing motors by raising speed is the main way to increase power density. In the sample, maximum speed, reduction ratio and power density all rose together; stator inner and outer diameters shrank by an average of 10 to 12 mm, rotors lost about 2.3 kg on average, and the effective core stack length was only slightly shorter.
| Item | BYD Tang L rear motor | Volkswagen ID.3 |
|---|---|---|
| Max speed | 30,500 rpm | 16,000 rpm |
| Gear ratio | 10.9 | 11.5 |
| Peak power density | 4.08 kW/kg | — |
| Interpretation of the paper | The Tang L's rear motor only gains a higher top speed, with no upward shift in the base (corner) speed; the extra rpm goes toward a higher vehicle top speed | |
Oil flows directly over the core laminations of the rotor and stator, raising sealing requirements. The paper suggests this may be why bonded laminations are becoming more common while welded and interlocked laminations are declining. Higher on-board voltages and wider use of silicon carbide devices are also increasing demands on insulation systems.
| Item | 2019–2021 model years | 2024–2025 model years |
|---|---|---|
| Cooling | Almost all water jackets | Growing use of oil splash-spray cooling combined with direct oil cooling of the rotor, stator laminations and winding end turns |
| Housing integration | Mostly low integration, with multi-section housings | Highly integrated designs dominate from 2022, with power electronics housed in the drive unit casing |
| Silicon steel lamination thickness | Thicker | 0.25 mm becomes standard |
| Stator winding | Hairpin windings increasingly common from 2021–2022 | Mainly flat-wire U-shaped hairpins from 2024 |
| Lamination fixing | Welding, interlocking | More bonded laminations |
| Figures | Scope | Source |
|---|---|---|
| 115 units; 80 models | Traction motors and models on EVs of model years 2019–2025, including BEV, EREV and PHEV | Energies 19(21), 5324, Section 4 |
| 48 units; 31 models | Sample from PEM's previous study, model years 2018–2023 | e+i 142, 312–345 abstract |
| 20%–30% | Weight saving of magnesium alloy housings vs. conventional aluminium housings, from literature cited in the paper | Energies 19(21), 5324, Section 5.1.1 |
| 1.68; 2.43; 2.7 kW/kg | Average peak power density of PMSMs: 2019, MY2024–2025, MY2025; based on total drive unit weight | Ibid., section 5.1.3 |
| 4.24; 4.08; 3.46 kW/kg | Peak power density of the NIO ET9 rear motor, BYD Tang L rear motor and Onvo L60 | Ibid., section 5.1.3 |
| 8,000–17,000; 30,500 rpm | Main maximum-speed range for MY 2020–2022; sample maximum | Ibid., section 5.2.1 |
| 9.45;11.48 | Average reduction ratio, 2019–2021 model years vs 2024–2025 model years | Ibid., section 5.2.1 |
| 0.25 mm | Standard lamination thickness for model years 2024–2025 | Ibid., section 6 |
| 17%;36% | Share of analysed inverters using silicon carbide: 4 of 23 and 16 of 44 | Ibid., section 6 |
| 10–12 mm; 2.3 kg | Average reduction in stator inner and outer diameters; average rotor weight reduction | Ibid., section 6 |