Robot Gearbox Grease: Spec Code, Life Law and Fill Quantity
Quick Answer
Robot gearbox grease is not one product, because a strain wave joint and an RV or cycloidal joint do different things to the lubricant. The first shears a thin film at high sliding speed near the wave generator; the second presses a thick film under high contact stress at low speed. The useful specification is a six-character ISO 6743-9 code plus an NLGI grade, not a list of adjectives. Life is counted in input revolutions, and one manufacturer’s own chart puts the interval a full decade lower for every 20 °C of grease temperature. Fill volume is the third number, and overfilling damages seals.
- Three reducer families, three grease jobs
- A grease spec is six characters, not eight adjectives
- The eight properties, translated into test methods
- Grease life is counted in revolutions
- A worked example: 3,000 rpm, 50 °C, 1.5x rated torque
- Fill quantity and the overfill trap
- Mixing, substitution and approval by size
- When grease stops being the answer
- Reading a failed joint
- How Greensky Power specifies the grease
- References
- FAQ
Three Reducer Families, Three Grease Jobs
“Robot gearbox grease” gets treated as a single product category. It is not. The lubricant in a robot joint is asked to do one of three different jobs, and a grease that is excellent in one is mediocre or wrong in the next. The distinction is mechanical, not commercial, and it explains why a joint that runs clean on one brand destroys grease on another.
Strain wave gearing: a thin film at high shear
In a strain wave, or harmonic, reducer the flexspline teeth engage the circular spline through an elliptical wave generator. The contact is small, the sliding speed at the tooth engagement is high, and the grease is sheared hardest in the region around the wave generator. Harmonic Drive states this directly in its catalogue: the grease deteriorates fastest where it is subjected to the greatest shear, which is near the wave generator [1].
A joint that reverses several times a second, with a small arc of tooth contact and a constantly flexing thin-walled cup, shears its lubricant far harder than it crushes it. What that joint needs is a film that survives sliding, not one that survives crushing. Extreme pressure and anti-wear behaviour matter here because the failure mode is progressive wear of a tooth profile that sets the lost motion of the whole joint.
RV and cycloidal reducers: high stress at low speed
An RV reducer carries torque through cycloid discs and pins, and a plain cycloidal reducer does the same job with fewer parts. Contact stresses are high, the output turns slowly, and the grease is expected to sit in the contact and stay there. Here the enemy is not sliding wear but film collapse under load and grease migrating away from the loaded zone.
Nabtesco’s RV catalogue is unusually explicit about the consequences. It recommends its own VIGOGREASE REO, states a standard grease replacement time of 20,000 hours, cautions that operation above 40 °C at the reducer surface calls for earlier inspection, and caps continuous surface temperature at 60 °C [2]. Those numbers assume a moderate thermal environment, and they are not valid for a joint that runs hot.
Why one grease cannot serve both
The two duties pull the formulation in opposite directions. A film thick enough to separate cycloid pins under high contact stress is also thick enough to churn and heat in a high-speed strain wave input. A low-viscosity synthetic that keeps strain wave efficiency high may simply not carry the load in a heavily loaded cycloidal stage. So both major manufacturers publish a compatibility table by frame size and ratio instead of one universal recommendation.
A Grease Spec Is Six Characters, Not Eight Adjectives
Eight properties get advertised for robot grease. Every one of them is real, and every one has a standard test behind it. A supplier can claim all eight in a brochure and still ship a grease that is wrong for your joint, because none of the eight says anything about temperature range, base oil viscosity or thickener chemistry.
Those three things are captured by the ISO 6743-9 designation. A grease code of the form ISO-L-X B C H A 2 decodes as follows: the letters after the X give minimum operating temperature, maximum operating temperature, water resistance and corrosion protection, and whether the grease has extreme pressure properties; the final digit is the NLGI consistency grade [10].
| Position in the code | Meaning | Letter options |
|---|---|---|
| ISO-L-X | Lubricant, class L, grease family X | Fixed prefix |
| 1st letter | Minimum operating temperature | A = 0 °C, B = -20, C = -30, D = -40, E = below -40 |
| 2nd letter | Maximum operating temperature | A = 60 °C, B = 80, C = 120, D = 140, E = 160, F = 180, G = above 180 |
| 3rd letter | Water resistance and corrosion protection | A to I, combining washout resistance with rust protection |
| 4th letter | Load carrying, extreme pressure | A = no EP properties, B = EP properties |
| Final digit | NLGI consistency grade | 000 to 6 |
The NLGI grade is the only part most buyers recognise, and it says the least. It is a worked-penetration band measured at 25 °C by cone penetration, and the published bands run from 445 to 475 tenths of a millimetre at grade 000 down to 85 to 115 at grade 6 [3]. Grade 2, the default for general bearings, is the 265 to 295 band.
| NLGI grade | Worked penetration at 25 °C (0.1 mm) | Consistency | Where it turns up |
|---|---|---|---|
| 00 | 400 to 430 | Fluid | Sealed, low-speed gear housings; centralised systems |
| 0 | 355 to 385 | Very soft | Cold duty and long supply lines |
| 1 | 310 to 340 | Soft | Low temperature, high speed, pumpability |
| 2 | 265 to 295 | Soft to normal | General rolling bearings; many robot joint fills |
| 3 | 220 to 250 | Firm | Vertical shafts, leakage control |
Two greases can share grade 2 and behave nothing alike. Grade describes stiffness, not base oil viscosity, not thickener, not additive package, and not life. A specification that stops at “NLGI 2” has not specified anything you can hold a supplier to.
The Eight Properties, Translated Into Test Methods
Each of those eight properties corresponds to a laboratory method. That is the difference between a brochure and a purchase specification: a method has a number, a tolerance and a test report attached to it.
| Advertised property | What is actually measured | Method | What to ask the supplier for |
|---|---|---|---|
| Extreme pressure and anti-wear | Load at which the film fails, and wear scar under load | Four-ball EP test [4] | Weld load and load-wear index on a batch certificate |
| Viscosity and viscosity-temperature | Base oil viscosity at 40 and 100 °C, and its rate of change | Data sheet values; viscosity index | Base oil viscosity, not just the grease grade |
| Oxidation stability | Pressure drop in an oxygen bomb at elevated temperature | Oxidation stability test [5] | Pressure drop after the standard hours |
| Anti-rust and anti-corrosion | Rust on a steel specimen after water exposure | Rust-preventive test [6] | Pass or fail, and the water hardness used |
| Anti-foaming | Foam volume and collapse time after agitation | Foaming characteristics test [7] | Sequence I, II and III figures |
| Thermal stability | Not the dropping point; the grease does not fail at its drop point | Oxidation and high-temperature life tests [5] | The manufacturer’s continuous temperature limit |
| Long life | Hours to failure in a test bearing at controlled temperature | Life test in a bearing [8] | The life curve, and the temperature it was measured at |
| Low noise | Film separation and mechanical stability, not adhesion damping | Roll stability [9]; cleanliness | Roll stability figure and filtration class |
Two myths worth correcting
The first is the dropping point. A dropping point is the temperature at which a grease turns liquid in a test cup; it is a manufacturing control, not an operating limit. A grease advertised at 300 °C dropping point can oxidise and harden far below that, and its practical ceiling is usually set by base oil volatility and oxidation, not by the thickener melting.
The second is noise. Grease is often said to reduce noise through film thickness and adhesion. Adhesion helps a grease stay in place, but the noise mechanism is film separation: when the film is too thin, asperities touch and the bearing runs rough. What actually keeps a robot joint quiet is a stable, clean film of the right base oil viscosity, plus a grease that does not separate or harden in service.
Grease Life Is Counted in Revolutions
Buyers ask for a replacement interval in months. Manufacturers answer in input revolutions, because a joint that oscillates twice a second and a joint that indexes once a minute accumulate completely different amounts of shear in the same calendar period. A figure in months means nothing until you know the speed it assumed.
The manufacturer’s own chart
Harmonic Drive publishes a life chart for its greases expressed as allowable input revolutions against grease temperature, valid when average load torque does not exceed rated torque. Reading the 4B No.2 curve gives an interval of about 10 to the 10th power revolutions at 20 and at 40 °C, 10 to the 9th at 60 °C, 10 to the 8th at 80 °C, 10 to the 7th at 100 °C and 10 to the 6th at 120 °C [1].
| Grease temperature | 4B No.2: allowable input revolutions | SK-1A and SK-2 | Input hours at 3,000 rpm, 4B No.2 |
|---|---|---|---|
| 20 °C | 1 x 10^10 | 1 x 10^9 | 55,600 |
| 40 °C | 1 x 10^10 | 1 x 10^9 | 55,600 |
| 60 °C | 1 x 10^9 | 1 x 10^8 | 5,560 |
| 80 °C | 1 x 10^8 | 1 x 10^7 | 556 |
| 100 °C | 1 x 10^7 | 1 x 10^6 | 56 |
| 120 °C | 1 x 10^6 | 1 x 10^5 | 5.6 |
Between 40 and 120 °C the interval drops by one full decade for every 20 K, which is a factor of ten per 20 °C and a factor of a thousand across the range. A joint running 20 K hotter does not lose a little margin; it loses 90 percent of its interval. The catalogue’s own worked example lands where interpolation predicts: at an average grease temperature of 50 °C it quotes about 3 x 10 to the 9th input revolutions.
The torque correction is a cube
Below rated torque the chart is used as printed. Above it, the manufacturer applies a cubic correction: the interval is multiplied by the cube of the ratio of rated torque to average torque [1]. A joint averaging 1.5 times rated torque therefore keeps 1 divided by 1.5 cubed, or 0.296, of its nominal interval. Put plainly: 50 percent more torque costs about 70 percent of the interval, while 20 K more temperature costs 90 percent of it.
Published rules of thumb differ, and the difference matters before you compare two sources. A common shop-floor rule halves grease life for every 10 K; the curve above loses a decade per 20 K, which is a factor of about 3.2 per 10 K. Over a 40 K excursion the two rules disagree by a factor of six. Use the manufacturer’s curve for the joint you actually have, and treat any generic rule as an order-of-magnitude warning rather than a calculation.
Nabtesco takes the other route and publishes a flat 20,000 hours for its RV reducers, with two conditions attached: the figure is the standard replacement time due to lubricant degradation, and operation above 40 °C at the reducer surface requires earlier inspection and earlier replacement [2]. The two manufacturers are not contradicting each other; they are stating different things. One gives you a shear budget, the other gives you a thermal assumption with a maintenance trigger.
A Worked Example: 3,000 rpm, 50 °C, 1.5x Rated Torque
Take a strain wave joint on a six-axis arm. The input runs at 3,000 rpm, the measured grease temperature in steady duty is 50 °C, and the average load torque over a cycle is 1.5 times the rated torque. The grease is the 4B No.2 type from the chart above. Three steps give the real interval.
Step 1: revolutions at rated torque
At 50 °C the chart and the manufacturer’s worked example both give about 3 x 10 to the 9th input revolutions. Convert to hours by dividing by the input speed in revolutions per hour: 3,000 rpm is 180,000 revolutions per hour, so 3 x 10 to the 9th divided by 180,000 equals about 16,700 hours. That is the number a sales conversation usually stops at.
Step 2: apply the torque correction
The average torque is 1.5 times rated, so the cube correction is 1 divided by 3.375, or 0.296. The interval becomes 16,700 multiplied by 0.296, which is about 4,940 hours. Still respectable, and now it is a figure for this joint, not for a catalogue page.
Step 3: see what 20 K would do
Suppose the joint is mounted near a motor that runs hot, and the grease settles at 70 °C instead of 50 °C. Interpolating one decade per 20 K gives about 3 x 10 to the 8th revolutions, or 1,670 hours at rated torque. After the same torque correction the interval is about 490 hours.
| Case | Grease temperature | Input revolutions | Hours at 3,000 rpm, rated torque | Hours at 1.5x rated torque |
|---|---|---|---|---|
| As measured | 50 °C | 3 x 10^9 | 16,700 | 4,940 |
| Hot mounting | 70 °C | 3 x 10^8 | 1,670 | 494 |
| Ratio of the two | +20 K | 10x shorter | 10x shorter | 10x shorter |
Twenty kelvin of grease temperature costs ten times the interval. A 50 percent torque overload costs 3.4 times. In this example the thermal penalty is three times larger than the load penalty, and the thermal penalty is the one nobody puts in a specification. If a joint is eating grease, measure its temperature before blaming the load.
Fill Quantity and the Overfill Trap
The third number on a grease specification is how much. Manufacturers publish it per frame size, and getting it wrong in either direction causes a failure that looks like a grease quality problem but is not.
| RV frame | Grease quantity, horizontal installation | Mass |
|---|---|---|
| RV-6E | 42 cc | 38 g |
| RV-20E | 87 cc | 78 g |
| RV-40E | 195 cc | 176 g |
| RV-80E, bolt clamping | 383 cc | 345 g |
| RV-110E | 432 cc | 389 g |
| RV-320E | 1,040 cc | 936 g |
| RV-450E | 1,596 cc | 1,436 g |
Two details from the same catalogue govern the fill. First, the reducer is not greased when it ships, so the fill is an installation step, and any space between the reducer and the coupling also has to be filled. Second, about 10 percent of the internal volume must be left free. Overfilling raises internal pressure, and the catalogue names the consequence plainly: oil seal damage, a lifted lip, or leakage [2].
Underfilling fails the other way. Starved contacts run hot and wear, and because the life law is a decade per 20 K, a partially starved joint can lose most of its interval while still looking adequately greased from the outside. Both errors are invisible in a visual inspection, so the fill volume belongs on the drawing and not in a maintenance note.
Mixing, Substitution and Approval by Size
Sooner or later a plant runs out of the specified grease and asks whether it can top up with something similar. The answer depends on thickener chemistry, and it is not a matter of brand loyalty.
Nabtesco’s instruction is unambiguous: do not mix the recommended grease with any other lubricant [2]. The reason is that thickener systems are not necessarily compatible. A lithium-complex soap and a polyurea thickener can soften, harden or separate when mixed, and the result is a joint that leaks or runs hot within days. Where a mixture is unavoidable, the compatibility of the pair can be assessed by a binary-mixture test instead of guessed [11].
The second trap is subtler. Grease approval can be size-specific. Harmonic Drive publishes compatibility tables in which the recommended grease differs by frame size and by reduction ratio, with some greases listed as standard for small frames and semi-standard or long-life options for large ones [1]. A grease that is correct for a size 20 joint at 50:1 is therefore not automatically correct for a size 100 joint at the same ratio, even though the code on the tube is identical.
- Read the reducer’s own manual, not a general catalogue, and note the approved grease name and any size restriction.
- Record the current fill: product name, ISO 6743-9 code, NLGI grade and the batch on the certificate.
- If substituting, compare base oil viscosity, thickener type, EP package and temperature range, not just the NLGI number.
- If mixing is unavoidable, run a compatibility assessment, then inspect the joint after the first production shift for softening, bleeding or leakage.
When Grease Stops Being the Answer
Grease has boundaries, and two of them are published by the manufacturers themselves. Missing them turns a maintenance question into a design error.
The first is a hard size and ratio limit. Harmonic Drive specifies oil lubrication for component sets of size 50 and above when the reduction ratio is 50:1 or greater [1]. No grease selection will substitute there, because the shear and the heat removal requirement exceed what a static lubricant can do.
The second is motion. Nabtesco notes that when the output rotation angle is small, around 10 degrees or less, service life can fall because lubrication is poor and the internal parts see a concentrated load [2]. A joint that only ever moves through a few degrees never carries fresh grease into the loaded zone, and it never distributes wear. This is a duty problem, not a lubricant problem, and it belongs in the design review, not in the first failure report.
Three other conditions need a conversation with the supplier, not a catalogue pick: continuous operation below the grease’s minimum temperature, where cold-start torque can exceed the motor’s capability; cleanroom or vacuum duty, where volatile base oil matters; and food-contact areas, where the lubricant needs a registration the standard range does not have. None of these is solved by a higher NLGI grade.
Reading a Failed Joint
Grease that has degraded tells you how it degraded, if you look at it before wiping the joint clean. The pattern is usually more informative than the interval.
| Symptom in the joint | Most likely cause | What to check first | Action |
|---|---|---|---|
| Grease darkened, acidic smell | Oxidation from sustained high temperature | Measured grease temperature against the life chart | Reduce heat or shorten the interval; raise the temperature class |
| Oil pooling beside the grease | Bleeding, often from storage or thermal cycling | Storage temperature and shelf life; fill date | Use fresh stock; check the bleed specification |
| Grease hard, crumbly | Thickener breakdown or wrong grease mixed in | Top-up records and thickener type | Full clean-out and refill; never top up blindly |
| Leakage at the output seal | Overfill, blocked breather, damaged seal lip | Fill volume against the table; free space | Correct the fill to about 90 percent; replace the seal |
| Grey or black streaks on teeth | Fretting wear from micro-motion at reversal points | Duty cycle and reversing frequency | Specify a grease with demonstrated fretting resistance |
| Whine that grows over weeks | Film thinning, or contamination by particles | Grease cleanliness and filter class | Clean out, refill, and check the seal path for ingress |
Two of these rows are thermal, two are contamination or handling, one is a seal and one is a duty problem. Only one is a genuine lubricant quality issue. A better grease rarely fixes the other five.
How Greensky Power Specifies the Grease
We supply cycloidal reducers and planetary gearboxes for robot and automation duty, and the lubricant is part of the delivery specification. Three numbers leave our factory with the reducer, not one.
- Fill volume in cubic centimetres, per mounting orientation, with the free space stated, so the installer cannot overfill.
- The approved grease and its ISO 6743-9 code plus NLGI grade, with any size or ratio restriction made explicit.
- The replacement interval in input revolutions, with the grease temperature and torque ratio the figure assumes, so you can correct it to your own duty.
If a joint is already in service and eating grease, the useful first question is what temperature the joint actually reaches, not which grease to buy. Send us the duty cycle, the measured surface temperature and the current fill, and we will return the corrected interval. Background on the reducers themselves is on the cycloidal reducer page, the high-precision planetary gearbox page and the gearbox selection guide. For matching a motor to the reducer, see how to choose a gear reducer and selecting a reduction gearbox for a motor.
Robot platforms that combine a reducer with a direct-drive axis are covered in high-torque BLDC motors under 100 mm and robot and AGV motor solutions, and the gear motor side of the same question in applications of precision planetary gearboxes. For the intralogistics case see motors for logistics robots and DC motors for robots; where the joint, the winding or the mounting flange has to change, start from custom electric motors.
References
- Harmonic Drive Systems. Speed Reducers for Precision Motion Control: gear catalogue, lubrication chapter. Harmonic Drive LLC. https://www.harmonicdrive.net/_hd/content/documents1/reducer%20catalog_US.pdf
- Nabtesco Corporation. (2021). Precision Reduction Gear RV: E Series, C Series, Original Series, catalogue and operation manual. Nabtesco. https://www.abi.nl/assets/uploads/Downloads/Aandrijvingen/Precisievertragingen/Nabtesco%20cycloide%20vertragingen/RV%20serie/RV_Rev4.1_EN_no_drw__nabtesco.pdf
- ASTM International. (2021). ASTM D217-21a, Standard Test Methods for Cone Penetration of Lubricating Grease. ASTM. https://www.astm.org/standards/d217.htm
- ASTM International. ASTM D2596, Standard Test Method for Measurement of Extreme-Pressure Properties of Lubricating Grease (Four-Ball Method). ASTM. https://www.astm.org/standards/d2596.htm
- ASTM International. ASTM D942, Standard Test Methods for Oxidation Stability of Lubricating Greases by the Oxygen Pressure Vessel Method. ASTM. https://www.astm.org/standards/d942.htm
- ASTM International. ASTM D1743, Standard Test Method for Determining Corrosion Preventive Properties of Lubricating Greases. ASTM. https://www.astm.org/standards/d1743.htm
- ASTM International. ASTM D892, Standard Test Method for Foaming Characteristics of Lubricating Oils. ASTM. https://www.astm.org/standards/d892.htm
- ASTM International. ASTM D3336, Standard Test Method for Life of Lubricating Greases in Ball Bearings at Elevated Temperatures. ASTM. https://www.astm.org/standards/d3336.htm
- ASTM International. ASTM D1831, Standard Test Method for Roll Stability of Lubricating Grease. ASTM. https://www.astm.org/standards/d1831.htm
- FUCHS Lubricants. ISO 6743-9: identification of lubricating greases. FUCHS SE. https://www.fuchs.com/es/en/products/product-program/lubricating-greases/iso-6743-9-identification-of-lubricating-greases
- ASTM International. ASTM D6185, Standard Practice for Evaluating Compatibility of Binary Mixtures of Lubricating Greases. ASTM. https://www.astm.org/standards/d6185.htm
FAQ
Q: What is robot gearbox grease? It is the lubricant filled into a robot joint reducer, and it is application specific, not universal. Strain wave reducers shear a thin film at high sliding speed near the wave generator, while RV and cycloidal reducers hold a film under high contact stress at low output speed. The two duties want different base oils and different thickener systems, so manufacturers publish approval tables by frame size and ratio [1].
Q: How often should robot gearbox grease be replaced? Ask for the interval in input revolutions, not months. Harmonic Drive charts allowable input revolutions against grease temperature, and that interval falls by a decade for every 20 °C above about 40 °C [1]. Nabtesco publishes 20,000 hours for its RV reducers and asks for earlier inspection once the reducer surface exceeds 40 °C [2]. Both figures assume rated torque; above it, multiply by the cube of rated divided by average torque.
Q: Can I mix two different robot greases? Not safely. Nabtesco instructs users not to mix its recommended grease with any other lubricant, because thickener systems are not universally compatible and a mixture can soften, harden or separate [2]. Where mixing cannot be avoided, the compatibility of the specific pair can be evaluated with a binary-mixture test instead of assumed [11], and the joint should be inspected after the first production shift.
Q: What does an ISO 6743-9 grease code mean? It is a short code that states the minimum operating temperature, the maximum operating temperature, water resistance and corrosion protection, whether the grease carries extreme pressure additives, and the NLGI consistency grade [10]. A code such as ISO-L-X B C H A 2 means a grease usable from -20 to 120 °C, with good water and corrosion behaviour, no EP package, and NLGI grade 2 consistency.
Q: Does a higher dropping point mean a better grease for a robot joint? No. Dropping point is the temperature at which grease turns liquid in a test cup, and it is a manufacturing control, not an operating limit. Oxidation and base oil volatility usually end a grease’s service life well below its dropping point, so the practical ceiling comes from a life test or from the reducer maker’s temperature limit, not from the dropping point figure.
Q: How much grease goes into a robot reducer? The quantity is published per frame size and per mounting orientation, and about 10 percent of the internal volume must be left free. Overfilling raises internal pressure and can damage the oil seal, while underfilling starves the loaded contacts [2]. Nabtesco RV frames, for example, take from 42 cc on an RV-6E to 1,596 cc on an RV-450E, and those reducers are shipped dry, so the fill is an installation step.
