A multi-roller traction drive replaces the helical gear reducer in high-speed EV drive units — delivering bearing-grade quiet operation, 100 000+ RPM capability, and a 15:1 ratio in a single planetary stage.
High-speed EV motors operate at 20 000+ RPM. Conventional helical gear reducers struggle with the NVH, tooth loading, and efficiency loss that comes with these speeds. Traction drives eliminate the root cause.
No gear mesh frequencies. A traction drive operates with bearing-like acoustics — no tooth-to-tooth impact excitation at any speed.
NASA Lewis testing on the Nasvytis multi-roller drive demonstrated operation at 73 000 RPM. The physics support 100 000+ RPM operation.
The compound stepped-planet architecture achieves what would require multiple gear meshes in a conventional reducer — in one compact stage.
Every ball bearing is a traction drive: the inner race is a sun, the outer race is the ring, and the balls are the planets. Traction drives simply engineer those same rolling contacts to transmit useful power.
Torque is transferred through elastohydrodynamic (EHL) traction film — a pressurized lubricant that solidifies under load, shearing to transmit tangential force without metal-to-metal contact. Life is predicted using the same Lundberg-Palmgren L10 methods used for rolling bearings.
The Nasvytis architecture takes this further with compound stepped planets: each planet set has a large outer roller (P1) contacting the sun and a smaller inner roller (P2) on the same shaft contacting an outer P3 roller. This two-stage compounding inside a single ring achieves ratios of 14–15:1 with 20 shared EHL contact zones.
| Attribute | Helical Gear | Multi-Roller TD |
|---|---|---|
| Noise (NVH) | Gear mesh excites housing | Bearing-grade quiet |
| Max speed | Limited by tooth loading | 100 000+ RPM |
| Ratio / stage | Typically 4–6:1 | 15:1 single stage |
| Efficiency | 97–98% peak | ~98% (comparable) |
| Life method | Gear tooth fatigue | L10 bearing methods |
| Contact surface | Precision ground teeth | Bearing-grade rollers |
| Packaging | Offset layouts | Coaxial, cylindrical |
The multi-roller traction drive is not speculative. Algirdas Nasvytis at TRW developed and patented the compound stepped-planet architecture in the 1960s–70s. NASA Lewis (now Glenn) conducted extensive test programs validating the technology at power and speed levels that still exceed automotive requirements today.
NASA Lewis test program validated the Nasvytis drive at 73 000 RPM sun speed — far above any current EV motor requirement.
Peak power transmission demonstrated during testing with 1970s-era traction fluids. Modern fluids (CoF doubled to ~0.12) would yield higher capacity.
Measured efficiency at rated conditions using the test hardware documented in NASA Technical Papers 1710 and 2027.
Lundberg-Palmgren bearing life methods applied directly to traction contacts — the same tools bearing engineers already use.
The Nasvytis drive was shelved in the 1980s — not because the technology failed, but because the application didn't exist yet. Three changes have made the timing right.
Modern EHL traction fluids achieve a coefficient of friction of ~0.12, up from ~0.06 in the NASA test era. Higher CoF means higher torque capacity for the same contact load.
EV motor speeds have increased dramatically. This demands high-ratio, single-stage reducers — exactly the operating regime where traction drives outperform gears.
Unlike ICE vehicles, EVs don't need variable ratio. The CVT complexity that historically burdened traction drives is simply unnecessary. A fixed-ratio stage avoids all those problems.
Traction Drive LLC is led by Joe Kliewer, a bearing and traction drive engineer with over 20 years of experience in rolling element contact design, specializing in the intersection of bearing technology and power transmission.
Joe's work on the multi-roller traction drive builds on a deep background in traction drive and CVT technology, including 11 issued U.S. patents spanning continuously variable transmissions, hydraulic ratio-shifting mechanisms, torsional damping systems, and integrated EV propulsion architectures. This prior art foundation directly informs the fixed-ratio multi-roller design.
Experience in bearing design for next-generation advanced reactor programs, including sodium-immersed bearing design basis, friction estimation, and material selection — Stellite alloys, Inconel superalloys, and ceramic bearing components for extreme temperature and corrosive service.
Hertz contact, EHL film thickness, Lundberg-Palmgren life prediction — applied to traction drives and conventional rolling bearings alike.
Whether you're evaluating traction drive technology for an EV platform or need bearing design expertise for an advanced reactor program, we'd like to hear from you.
Bloomfield Hills, Michigan