Fretting Wear: A Long-Underestimated Failure Mode in Dexterous Hand Planetary Reducers
On May 12, 2026, at the first Hong Kong Embodied Intelligence Industry Summit and Zhiyuan Partners Conference, the Zhiyuan A3 completed a Bruce Lee-style flying kick while simultaneously demonstrating a dexterity hand ‘squeezing a tennis ball’ and single-finger and multi-finger coordinated flexion-extension. The flying kick showcases explosive power and dynamic control; squeezing a tennis ball is a contest between precision and stability — and the latter actually poses a more hidden and more sustained challenge to the joint drive system.
Today I want to discuss a failure mode that has long been underestimated in dexterity hand planetary reducers: fretting wear. Unlike tooth fracture, which causes immediate shutdown, fretting wear may ‘silently’ erode tooth flanks over thousands of hours of fine operations, eventually leading to precision degradation or even fatigue tooth fracture.
Fretting Wear: Mechanism Review in Small-Amplitude Reciprocating Slip
Fretting wear refers to wear that occurs between two contacting surfaces undergoing reciprocating relative slip at very small amplitudes — typically several microns to several tens of microns. At this scale, the conditions for ‘adequate elastohydrodynamic lubrication’ in classical lubrication theory are essentially absent — the ratio of lubricant film thickness to composite surface roughness (the Lambda value) is often less than 1, meaning dynamic pressure oil films cannot form, leaving only boundary lubrication films barely holding on.
I have conducted bench tests on a modulus 0.5 mm planetary gear: under constant contact load with 15 micrometers amplitude, 5 Hz reciprocating micro-slip applied, with hundreds of direction reversals per minute. After 200 hours, the tooth surface exhibited typical fretting wear characteristics — darkened metal surface, mirror-finish area blurred, micro-cracks distributed along the edges of the contact zone. This is not pitting, nor is it scoring — it is fretting wear. Crucially, this type of damage is very difficult to capture in conventional rotary fatigue tests.
Engineering Risk Behind the ‘Squeeze a Tennis Ball’ Action
When a dexterity hand demonstrates ‘squeezing a tennis ball,’ the audience sees fingers gently closing. But at the engineering level, this action contains a sequence of extreme conditions.
First: repeated fine force regulation. At the moment of ball contact, the control system adjusts finger position and force output at the millisecond level based on force feedback. This means the planetary reducer performs multiple forward-reverse and start-stop cycles within an extremely small angular range — generating micron-level reciprocating slip at the tooth flank contact zone, precisely the triggering condition for fretting wear.
Second: uneven loads from single-finger and multi-finger coordination. When one finger flexes, other fingers may remain stationary or under micro-tension. Multiple planetary reducers are in different load states: the active side gear bears alternating bending stress, while the stationary-side tooth flank may still generate fretting wear under vibration interference. This uneven torque distribution aggravates local damage.
Third: uncertain reaction force from elastic objects. When a tennis ball deforms under pressure, its reaction force fluctuates; to maintain stable grasping, the control system continuously adjusts force output. This causes the planetary reducer to operate in a state of ‘macroscopically stationary, microscopically vibrating’ over extended periods — tooth flanks endure high-frequency micro-slip under conditions of near-zero rotation.
It must be stated clearly: one demonstration will not immediately cause failure. The real danger lies in the dexterity hand performing such ‘fine force regulation’ operations frequently during long-term service — for example, precision assembly, compliant grasping, and long-duration stable holding. The cumulative number of fretting cycles is the key variable.
From Fretting Wear to Tooth Fracture: The Damage Chain
High-hardness oxide wear debris (such as Fe2O3) generated by fretting wear acts as an abrasive, accelerating tooth flank scratching; stress concentration at scratch sites induces micro-cracks; crack propagation forms pitting pits; pitting pits become new stress concentration sources; ultimately, fatigue cracks propagate toward the tooth root — and tooth fracture occurs under some impact load event.
This chain is more dangerous during fine operations than during heavy-load rough grasping, because it does not accompany obvious performance degradation warning signs — it belongs to ‘silent accumulation’ type failure. When operational precision degradation becomes visible to the naked eye, gear fatigue cracks often have already entered the second or even third stage.
Materials and Heat Treatment Are the Foundation, But Lubrication Is the Key Variable for Mass Production Reliability
Objectively speaking, material selection (such as carburized steel), heat treatment process, and tooth root fillet optimization for small-module gears do have a decisive effect on fatigue life. However, in a mass production context, the triggering conditions for fretting wear come from the usage conditions themselves, not from manufacturing defects. Even with extreme materials and process optimization, as long as small-amplitude reciprocating slip exists, fretting wear cannot be completely eliminated.
Therefore, grease selection becomes an important engineering variable affecting dexterity hand mass production reliability — not a universal solution, but the most direct, most economical, and most easily underestimated link at present.
Design Logic of VNOVO X500 for Fretting Wear Conditions
Taking VNOVO dexterity hand joint dedicated grease X500 as an example, the following illustrates the lubrication design logic for fretting wear conditions. This section is not an advertisement, but rather provides a referenceable technical solution sample.
X500 uses a fluorinated oil and PTFE solid lubricant composite system. In the fretting wear zone, since Lambda is typically less than 1 and dynamic pressure oil films cannot form, the PTFE solid lubricant phase can still form a stable boundary film at extremely low sliding speeds, directly blocking metal-to-metal contact and maintaining the friction coefficient at a low level of 0.05-0.08. This reduces asperity welding and tearing from the source.
In terms of extreme pressure load-bearing capacity, X500’s four-ball test weld load exceeds 800 kgf (per ASTM D2596). Planetary reducer tooth flank contact stresses are often as high as several hundred MPa; ordinary greases under high-pressure repeated action in the fretting zone easily experience oil film rupture. The fluorinated oil system combined with extreme pressure additives maintains partial oil film integrity in the micro-slip zone, delaying crack initiation caused by contact fatigue.
Wear debris management is also critical for fretting wear control. If fine hard debris generated by fretting wear remains on the tooth flank, three-body abrasion accelerates. X500 uses low-viscosity base oil (40 degC kinematic viscosity approximately 20-30 cSt); although the oil film is thin, fluidity is better, helping to carry fine debris away from the mesh zone and reduce secondary wear. Additionally, in the 100 degC, 24-hour static oil separation test, the oil separation rate is below 10%, ensuring that under long-term sealed conditions the grease does not easily dry out, avoiding lubrication failure due to base oil loss.
Furthermore, X500’s operating temperature range is -50 degC to 220 degC, covering the full temperature range from cold standby to high-load continuous operation for dexterity hands, and is also well-compatible with commonly used engineering plastics in planetary reducers (such as PEEK, PA66, etc.), without causing component swelling or cracking.
Closing Thoughts
The Bruce Lee flying kick and tennis-ball squeeze demonstrated by Zhiyuan essentially represent the challenges of two extreme operating conditions for humanoid robots: high dynamics and fine micro-operation. Both types of challenges have different natures of demands on the joint drive system, but neither can bypass the lubrication variable. The terrifying aspect of fretting wear is that it does not make noise, does not raise alarms — yet it silently erodes tooth flanks through thousands of microns-level tremors.
If you are also working on mass production reliability design for dexterity hand joints, you are welcome to share your actual fretting wear cases and data in the comments. For private discussions on specific operating condition lubrication solution selection, I am also available.
