Industrial Lubrication & Tribology: The Complete Engineering Guide
Every machine in a FabFlow shop — the 3-axis mill, the fibre laser, the FDM printer, the hydraulic press — is held apart by a film of oil thinner than the wavelength of green light. In a 6205 deep-groove ball bearing running at 1,500 rpm under a 500 N radial load on ISO VG 46 at 70 °C, the minimum film at the inner-race contact comes out at 147 nanometres. A human hair is about 70 µm across; the film is 1/475 of that. A red blood cell would stand about 48 times taller than the gap. And the entire fatigue life of that bearing — 5,336 hours or 21,345 hours — is set by whether that gap stays open.
That is the first thing this guide wants to change about how you think about oil. Lubrication is not a consumable you top up when the level drops. It is a hydrodynamic machine built inside the contact: viscosity is its working fluid, surface speed is its pump, and the converging geometry of the gap is its wedge. When the machine works, the metal surfaces never touch. When it stops working, they touch a few million times a minute.
The second thing is scale. A 4 µm particle — invisible, smaller than a red blood cell — is 27 times thicker than that film. This is why "the oil looks clean" means nothing, why 3 µm filtration exists, and why ISO 4406 codes are written next to pump model numbers in hydraulic specifications.
The third is money. India consumed about 5.6 billion litres of lubricant in 2025, a market of roughly USD 6.44 billion, the third-largest lubricant market in the world and one growing at about 3.5% a year on volume that is already enormous. Most of that volume goes into engines, but the industrial fraction — the hydraulic oils, gear oils, greases and metalworking fluids that keep factories running — is where the engineering decisions are made, and where a ₹200/litre fluid protects a ₹1,50,000 pump and a ₹2,00,000 day of lost production.
The guide runs in fourteen sections:
- §1–§2 — The gap and the regimes. Surface roughness, the λ ratio, the Stribeck curve, and why friction and wear are not the same problem.
- §3 — Film thickness from first principles. The Hamrock–Dowson equations, a complete 6205 calculation, and the sensitivity exponents that tell you which knob actually matters.
- §4 — Viscosity. The ISO 3448 ladder, the Walther equation, viscosity index, and how to pick the grade that exists at operating temperature.
- §5 — Base oils and additives. API Groups I–V, ZDDP, MoDTC, detergents, dispersants, and the chemistry that depletes before the oil looks dirty.
- §6 — Grease. NLGI grades, thickener families, base-oil viscosity, the 15 °C halving law, relubrication quantity, and grease compatibility.
- §7 — Bearing life. ISO 281, κ, e_C, a_ISO, and the worked example where clean lubricant multiplies life fourfold.
- §8 — Gears. ISO 12925-1, DIN 51517-3, AGMA 9005, FZG, scuffing, micropitting, and the PAG rule for worm drives.
- §9 — Hydraulics. ISO 11158, viscosity by pressure, ISO 4406 targets, filtration ratios, water, air, and fire-resistant fluids to ISO 12922.
- §10 — The machine tool. Spindle grease, linear guides, ball screws, way lube, coolant chemistry, and the lubrication map of a CNC machining centre.
- §11 — Oil analysis. What to measure, how to sample it, and the limits that actually trigger action.
- §12 — The failure atlas. Nine modes, their signatures, and the lubricant decision behind each one.
- §13–§14 — Economics and discipline. Indian market numbers, the cost model, consolidation, storage, and the one-page program.
1. The Gap: Why Surface Roughness Sets the Whole Problem
1.1 No surface is smooth at the scale that matters
A bearing raceway is the smoothest surface in a factory, and it is still a mountain range at the nanometre scale. Typical values:
Surface · Ra (µm) · Rq (µm) · Process
Superfinished bearing raceway · 0.02–0.05 · 0.025–0.06 · Lapping, honing, superfinishing
Ground shaft / gear flank · 0.2–0.8 · 0.25–1.0 · Cylindrical or form grinding
Turned / milled surface · 1.6–3.2 · 2.0–4.0 · Turning, milling, drilling
As-cast surface · 3–12 · 4–15 · Sand casting, investment casting
FDM 3D-printed surface · 8–25 · 10–30 · Layer-on-layer extrusion
The relevant number is not Ra but the composite RMS roughness of the two surfaces in contact:
and the dimensionless film parameter:
Two brand-new bearing surfaces at Sq = 0.04 µm each give Sq_c = 0.057 µm; a worn, contaminated, or poorly finished pair at 0.15 µm each gives 0.21 µm — a fourfold rougher contact for the same oil film. Nothing about the oil changed. The λ halved, and the wear rate did not halve with it; it moved into a different physical regime.
1.2 The film is built, not supplied
A lubricant's job in a concentrated contact is to be dragged into a converging wedge and pressurised until it can carry the load without the asperities meeting. The pressure is not the pump pressure — in a rolling bearing the oil enters at atmospheric pressure and leaves at 1–3 GPa, generated hydrodynamically in the inlet. That is why a 150 nm film can separate surfaces carrying 500 N: the film is not strong because the oil is thick, it is strong because the pressure stiffens it. The pressure–viscosity coefficient \alpha of a mineral oil is about 2.2 \times 10^{-8} Pa⁻¹, which means the oil at 1 GPa is roughly e^{22} — effectively a glass.
That single fact explains why the base oil, not the additive package, sets film thickness, and why an additive cannot rescue a contact that has lost its film. Additives work at the boundary — the last 10–20 nm — not in the bulk.
2. Four Regimes, One Curve
2.1 The Stribeck curve
Plot friction coefficient against the Hersey number S = \eta N/P — viscosity times speed divided by load — and every lubricated contact in the world falls on the same shape:
Regime · Film parameter · Friction coefficient · Wear rate · Where it lives
Boundary · \lambda < 1 · 0.08–0.15 · Moderate to severe · Start-up, low speed, high load, starved inlet
Mixed · 1 < \lambda < 3 · 0.04–0.10 · Mild · Most industrial bearings and gears at normal duty
Elastohydrodynamic (EHL) · \lambda \ge 3 · 0.001–0.01 · Negligible · Rolling bearings, gear flanks, cam–follower
Hydrodynamic · no asperity contact · 0.001–0.005 · Negligible · Journal bearings, thrust pads, piston skirts
The curve has a minimum, and the minimum is not where you want to operate. At low S the contact is in boundary/mixed lubrication and friction is high because asperities are shearing; as speed rises, the film builds and friction falls to the EHL plateau; at very high speed, viscous shear in a thick film raises friction again. The design target is therefore the lowest speed at which λ stays above 3, not the lowest possible friction coefficient at one operating point.
2.2 Friction and wear are different problems
A boundary lubricant can have lower friction than a mixed-film one and still wear faster, because friction is dominated by asperity shear while wear is dominated by asperity contact. That is why anti-wear additives (ZDDP) and friction modifiers (MoDTC) are separate chemistries, and why an engine oil optimised for fuel economy is not automatically the right oil for a gearbox.
For rolling bearings, ISO 281 collapses this into one number — the viscosity ratio \kappa — defined in §7. For the rest of this guide, λ is the physical picture and κ is the standard's accounting of it.
3. Computing the Film: Hamrock–Dowson, Worked End to End
3.1 The dimensionless groups
For an elliptical contact between two elastic solids, Hamrock and Dowson (1977) reduced the problem to four dimensionless groups:
where \eta_0 is the oil viscosity at the contact inlet, u the entrainment velocity, E' the effective elastic modulus, R_x the radius of curvature in the rolling direction, \alpha the pressure–viscosity coefficient, F the contact load, and k the ellipticity ratio. The fully flooded, isothermal minimum and central film thicknesses are:
with h = H R_x. The exponents are the engineering content of the equation. Film thickness is almost independent of load (W^{-0.073}), strongly dependent on speed (U^{0.68}) and on the pressure–viscosity product (G^{0.49}), and only weakly dependent on geometry through the ellipticity term.
3.2 The 6205 worked example
A standard 6205 deep-groove ball bearing: 25 mm bore, 52 mm OD, 15 mm width, 38.5 mm pitch diameter, 7.94 mm balls, 9 balls, basic dynamic rating C = 14.8 kN. It runs at 1,500 rpm under a 500 N radial load, lubricated with ISO VG 46 mineral oil at 70 °C.
Input · Value · Source
Entrainment velocity u · 3.02 m/s · \pi d_m n/60
Viscosity at 70 °C · 14.9 cSt · Walther fit, §4
Dynamic viscosity \eta_0 · 0.0129 Pa·s · \nu \rho, \rho = 870 kg/m³
Effective modulus E' · 231 GPa · Steel–steel, \nu = 0.3
Radius R_x · 3.97 mm · Ball radius
Pressure–viscosity \alpha · 2.2 \times 10^{-8} Pa⁻¹ · Mineral oil
Max ball load Q · 250 N · 4.37 F_r/Z \approx 243 N
Ellipticity k · 3.0 · Ball–raceway contact
Substituting:
Now the λ ratio. On a new bearing, Sq_1 = Sq_2 \approx 0.035 µm gives Sq_c = 0.05 µm and \lambda = 147/50 = \mathbf{2.9} — mixed film, close to full separation. On a raceway that has been indented by contamination or roughened by wear to Sq = 0.11 µm each, Sq_c = 0.15 µm and \lambda = \mathbf{0.98} — boundary contact, and the wear rate rises by orders of magnitude.
Same oil. Same speed. Same load. The only difference is the surface, and the contact has crossed from "survives 20,000 hours" to "survives until the next inspection."
3.3 What the exponents actually tell you
Change · Exponent · Effect on h_{\min}
Load doubled · W^{-0.073} · −4.9%
Speed doubled · U^{0.68} · +60%
Pressure–viscosity doubled · G^{0.49} · +40%
Viscosity doubled (same speed) · via U · +60%
Ellipticity 1 → 3 · (1-e^{-0.68k}) · 0.49 → 0.87 (+78%)
The load result is the counter-intuitive one, and it is worth sitting with: doubling the load on a rolling bearing barely changes the film. A 6205 at 250 N and the same bearing at 500 N per ball differ by 5% in film thickness. This is why bearing fatigue life is so sensitive to load — L_{10} \propto (C/P)^3 — while film thickness is not, and why "the bearing is overloaded" and "the bearing is starved" are different diagnoses with different fixes.
The speed result is the other one: the same bearing at 500 rpm runs at λ = 1.4 and at 3,000 rpm at λ = 4.7. Slow machines are the ones that need the higher viscosity, the EP additive, or the oil circulation heater — not the fast ones.
3.4 Where the equation stops being true
Hamrock–Dowson assumes fully flooded, isothermal, smooth surfaces. Real contacts are none of those. Starvation (too little oil at the inlet) cuts the film by 30–70%; thermal thinning at high sliding speed reduces it further; measured films in the NASA/Koye evaluation averaged about 30% above the prediction, while starved contacts fall well below it. Treat the equation as a scaling law, not a guarantee: it tells you which variables matter and by how much, not the exact nanometres in a specific spindle.
4. Viscosity: The First Decision, and the One Most Often Made at the Wrong Temperature
4.1 The ISO 3448 ladder
Industrial oils are graded by kinematic viscosity at 40 °C, in mm²/s (cSt), with each grade a band of ±10% around a midpoint and each step roughly 50% more viscous than the last:
ISO VG · Midpoint (cSt) · Min–max (cSt) · Typical use
10 · 10 · 9.0–11.0 · Spindle oils, high-speed hydraulics
15 · 15 · 13.5–16.5 · High-speed spindles, air-line lubricators
22 · 22 · 19.8–24.2 · Hydraulics in cold shops, chain oil
32 · 32 · 28.8–35.2 · General hydraulics, light gearboxes
46 · 46 · 41.4–50.6 · The default hydraulic grade
68 · 68 · 61.2–74.8 · Way lube, medium hydraulics, ball screws
100 · 100 · 90–110 · Gearboxes, plain bearings
150 · 150 · 135–165 · Gearboxes, compressors
220 · 220 · 198–242 · Industrial gear units
320 · 320 · 288–352 · Heavy gear units, worm drives
460 · 460 · 414–506 · Worm gears, low-speed heavily loaded
680 · 680 · 612–748 · Open gears, heavily loaded worm drives
1000 · 1000 · 900–1100 · Slow, hot, heavily loaded bearings
1500 · 1500 · 1350–1650 · Open gear compounds
4.2 The Walther equation: what the grade is at operating temperature
The 40 °C number on the pail is a labelling convention, not an operating condition. A machine tool hydraulic pack runs at 50–60 °C; a gearbox sump at 60–80 °C; a spindle bearing at 50–70 °C. The viscosity that matters is the one at the contact, and ASTM D341 (the Walther equation) gives it from the two numbers every datasheet publishes:
with \nu in cSt and T in kelvin. From the 40 °C and 100 °C points:
where W = \log_{10}(\log_{10}(\nu+0.7)).
For a typical VG 46 with \nu_{40} = 46 cSt and \nu_{100} = 6.8 cSt: B = 3.684, A = 9.418. At 70 °C (343.15 K):
That is the number that belongs in the Hamrock–Dowson calculation — not 46. Using 46 cSt at 70 °C would overstate the film by a factor of three.
Temperature · Viscosity (cSt) · Dynamic viscosity (Pa·s, ρ = 870)
20 °C · 134 · 0.117
40 °C · 46.0 · 0.040
60 °C · 20.6 · 0.0179
70 °C · 14.9 · 0.0129
80 °C · 11.1 · 0.00966
100 °C · 6.8 · 0.00592
4.3 Viscosity index
Viscosity index (VI, ASTM D2270) measures how flat that curve is, on a scale set in 1929 between a naphthenic reference (VI = 0) and a paraffinic one (VI = 100). Typical values:
Base oil · VI · Behaviour
Naphthenic Group I · 60–85 · Thin sharply when hot
Paraffinic Group I · 95–105 · Standard mineral
Group II hydrocracked · 100–115 · Better oxidation stability
Group III hydrocracked / wax-isomerised · 120–135 · Near-synthetic
PAO (Group IV) · 135–155 · Flat, low pour point
Ester / PAG (Group V) · 140–180 · Application-specific
High VI is not automatically better. A high-VI oil keeps its viscosity hot, which is good for film, but it also keeps it thick cold, which is bad for pumpability and cold start. The right question is not "what is the VI" but "what is the viscosity at my cold start and at my hot steady state, and is the ratio acceptable."
4.4 Selection in practice
Application · Grade · Reason
Hydraulics, 70–140 bar, 40–55 °C · VG 46 · Standard HM anti-wear fluid
Hydraulics, under 70 bar, cold ambient · VG 32 · Lower cold-start pressure
Hydraulics, 140–250 bar · VG 68 · Film at high pressure
Hydraulics, over 250 bar or hot · VG 100 · Film retention
Enclosed gear, 5–15 m/s pitch line · VG 150–220 · Balance churning vs film
Enclosed gear, under 5 m/s · VG 220–460 · Film at low speed
Worm gear · VG 220–460 PAG · Sliding contact, low friction
Machine-tool way lube · VG 68 · Stick-slip control, adhesive
Spindle grease base oil · VG 68–100 · High-speed film
Air compressor (reciprocating) · VG 100–150 · Hot discharge
Refrigeration compressor · VG 32/46 (POE/PVE) · Miscibility with refrigerant
The failure modes are symmetric. Too thin: λ falls, asperities touch, wear accelerates, temperature rises, viscosity falls further — a runaway. Too thick: churning losses heat the oil, the pump cavitates on cold start, filters bypass, and the extra 10 °C costs more life than the extra film buys. The sweet spot is usually the OEM grade, adjusted for actual sump temperature, not the thickest oil that fits.
5. Base Oils and Additives: The 1% That Decides the Other 99%
5.1 The API base-oil groups
Group · Sulphur · Saturates · VI · Production · Notes
I · over 0.03% · under 90% · 80–120 · Solvent refining · Cheapest, poorest oxidation stability
II · 0.03% or less · 90%+ · 80–120 · Hydroprocessing · Better stability, good solvency
III · 0.03% or less · 90%+ · 120+ · Hydrocracking / isomerisation · Near-synthetic, excellent VI
IV · — · — · — · PAO synthesis · Fully synthetic, low pour point
V · — · — · — · Esters, PAGs, naphthenics, bio · Everything else
The group is a crude proxy for performance. A Group III base stock with a good additive package will outperform a Group I with a better one on oxidation life, but a Group I often dissolves additives better and seals better in old machines. For industrial oils, Group II is now the volume default and Group III/PAO the long-life choice.
5.2 The additive classes that matter
Additive · Typical treat · Function · Failure signature when depleted
ZDDP / ZDTP · 0.08–0.15% P · Anti-wear film, antioxidant · Zn and P fall in ICP; wear rises
MoDTC · 0.02–0.1% Mo · Friction modifier (MoS₂ film) · Friction coefficient rises
Calcium/magnesium sulfonate · 0.1–0.3% Ca · Detergent, TBN reserve · TBN falls, deposits rise
Succinimide · 0.05–0.2% N · Dispersant, keeps soot suspended · Soot agglomerates, sludge
VI improver (OCP/PMA) · 0.5–2% · Raises VI · Shear-down: viscosity falls permanently
Pour-point depressant · 0.1–0.5% · Low-temperature flow · Wax crystal formation at cold start
Antifoam (silicone) · 10–100 ppm · Bubble release · Foam, cavitation, overflow
Demulsifier · 0.01–0.1% · Water separation · Emulsion, rust, sludge
EP sulphur–phosphorus · 0.5–3% · Extreme-pressure film (FeS/FePO₄) · Scuffing under shock load
5.3 ZDDP: the 100 nm film that carries the boundary regime
Zinc dialkyldithiophosphate is the most studied anti-wear additive in history, and its mechanism is worth understanding because it explains the whole logic of boundary lubrication. ZDDP adsorbs on the steel surface, and at contact temperatures around 150–200 °C it decomposes to form a glassy zinc/iron phosphate tribofilm roughly 100 nm thick. The film is soft enough to shear (low friction) and hard enough to prevent metal-to-metal contact (low wear). It also regenerates: where the film is rubbed away, fresh ZDDP reacts to rebuild it.
That is why ZDDP is sacrificial, and why its depletion is not visible. The oil stays golden; the zinc and phosphorus are gone. ICP analysis is the only routine way to see it, and declining Zn/P with rising iron is the classic "the additive package is finished, change the oil" signature.
5.4 Additive depletion is a chemical clock, not a dirt clock
Oil darkens for many reasons — oxidation, nitration, soot, dye — and colour is not a reliable indicator of remaining life. The measurable clocks are:
- TAN (total acid number): rises as acidic oxidation products form. A rise of 1.0–2.0 mg KOH/g above baseline is a change signal.
- TBN (total base number): falls as alkaline reserve is consumed. A fall of 50% or more matters.
- FTIR oxidation/nitration: absorbance increase at 1700–1720 cm⁻¹ (oxidation) and 1630 cm⁻¹ (nitration).
- Viscosity at 40 °C: rises with oxidation and soot, falls with shear-down or fuel dilution.
- Zn/P/Ca: the additive fingerprint, read by ICP.
A 15,000-hour oil that looks clean and has TAN +3 with Zn at 40% of new is not a 15,000-hour oil anymore. It is a carrier fluid with a worn-out chemistry.
6. Grease: A Sponge, Not a Liquid
6.1 What grease is
Grease is a sponge: 70–90% base oil held in a thickener matrix, plus 5–20% thickener and 1–5% additives. The thickener does not lubricate. It holds the oil in place, releases it slowly under shear, and re-absorbs it when the machine stops. The base oil viscosity — not the NLGI grade — sets the film thickness. Two NLGI 2 greases can carry base oils from VG 32 to VG 1000, and their film-forming behaviour differs by a factor of three.
Thickener · Dropping point · Upper service temp · Typical use
Lithium 12-hydroxystearate · 180–200 °C · 120 °C · General purpose, sealed bearings
Lithium complex · 260–280 °C · 150–160 °C · High temperature, electric motors
Calcium sulfonate complex · 280–300 °C · 150–180 °C · Water washdown, steel mills
Polyurea · 250–270 °C · 150 °C · Electric motor bearings, high speed
Clay (organo-bentonite) · none · 200 °C+ · High temperature, no melting
PTFE · none · 250 °C+ · Extreme temperature, chemicals
Silica · none · 200 °C+ · Specialty, low noise
6.2 NLGI consistency
NLGI · Worked penetration (0.1 mm) · Appearance · Typical use
000 · 445–475 · Fluid · Enclosed gearboxes, central systems
00 · 400–430 · Semi-fluid · Gearboxes, centralized lubrication
0 · 355–385 · Very soft · Low-temperature, centralized
1 · 310–340 · Soft · Cold ambient, low-speed
2 · 265–295 · Normal · The default bearing grease
3 · 220–250 · Firm · Vertical shafts, high temperature
4 · 175–205 · Very firm · Slow, sealed, long intervals
5–6 · 130–160 / 85–115 · Hard · Specialty block greases
6.3 Relubrication: the 15 °C halving law
Grease life is dominated by temperature, because oxidation and base-oil evaporation both follow Arrhenius kinetics. The SKF convention: the relubrication interval halves for every 15 °C above a 70 °C reference, and the interval is further corrected for contamination and load:
with f_T = 2^{-(T-70)/15}.
Worked example: a 6310 bearing (50 × 110 × 27 mm) with a base interval of 8,000 h at 70 °C, running at 90 °C in moderate dust with f_C = 0.5:
At 8,000 operating hours a year — two shifts, 250 days — that is a relubrication every 72 days, not every year. The same bearing at 115 °C would need grease every 8,000 × 0.125 × 0.5 = 500 h, or about 23 days, and that is before considering that many lithium greases are already at their oxidation limit.
The quantity per shot follows the SKF rule:
For the 6310: G = 0.005 \times 110 \times 27 = 14.9 g, roughly 10–15 strokes of a standard grease gun. For a sealed housing with no relief path, use half that and install a relief fitting — otherwise the excess has nowhere to go, pressure builds, and the grease is pushed past the seals.
6.4 Overgreasing is a failure mode
A bearing housing should be filled to 30–50% of free space on initial fill, and 20–30% for high-speed applications. Beyond that, the rolling elements churn the excess, temperature rises 10–20 °C, the grease oxidises faster, and the seals see pressure they were not designed for. The classic electric-motor failure — grease purged into the windings — is an overgreasing failure, not a grease failure.
6.5 Compatibility: mixing greases can destroy both
Thickener bases are not interchangeable. Lithium and polyurea, or lithium and calcium sulfonate, can soften by two or three NLGI grades when mixed, turning a firm grease into a leaking one. The practical rules:
- Same thickener type: generally compatible, but verify the base-oil viscosity and additive package.
- Different thickener, unknown compatibility: purge the housing completely and re-pack. A "top-up" of the wrong grease is worse than running the old grease longer.
- Never mix greases with different base-oil types (mineral + silicone, mineral + PFPE) — the result can be a hard, non-lubricating paste.
7. Bearing Life: ISO 281, κ, and the Multiplier Nobody Uses
7.1 Basic rating life
The basic rating life is the life reached by 90% of a population under a given load:
with p = 3 for ball bearings and 10/3 for roller bearings. A 6205 with C = 14.8 kN at P = 1.5 kN has C/P = 9.87 and L_{10} = 961 million revolutions. At 3,000 rpm — 180,000 revolutions per hour — that is 5,336 hours.
Load P (N) · C/P · L_{10} (Mrev) · Hours at 3,000 rpm
1,000 · 14.8 · 3,242 · 18,010
1,500 · 9.87 · 961 · 5,336
2,000 · 7.40 · 405 · 2,251
2,500 · 5.92 · 207 · 1,153
7.2 The modified rating life
ISO 281:2007 replaces the old additive life factors with a single integrated factor:
where a_1 accounts for reliability other than 90%, and a_{\text{ISO}} combines four interdependent effects: lubrication (κ), contamination (e_C), load level relative to the fatigue limit (C_u/P), and the material's fatigue stress limit. This is the standard's most important idea: these factors are not independent, so you cannot multiply them separately.
The viscosity ratio is:
where \nu is the actual operating viscosity and \nu_1 is the reference viscosity that would give a film equal to the composite roughness (\lambda = 1). κ = 1 is the threshold; κ ≥ 1 is adequate; κ ≥ 2 is good; κ ≥ 3 is excellent; κ above 4 is treated as 4, and κ below 0.1 invalidates the calculation.
7.3 The worked leverage
Take the same 6205 at 1.5 kN and 3,000 rpm, but change the lubricant condition:
Condition · κ · e_C · a_{\text{ISO}} (indicative) · Life (h)
Starved, contaminated, VG 32 at 90 °C · 0.4 · 0.3 · 0.4 · 2,134
Adequate mineral, clean · 1.2 · 0.6 · 1.5 · 8,004
VG 46, clean, sealed · 2.9 · 0.8 · 4.0 · 21,345
VG 68, very clean, filtered · 4.0 · 1.0 · 8–12 · 42,700–64,000
The a_{\text{ISO}} values are indicative — read them from the SKF/Schaeffler charts for the actual e_C C_u/P — but the order of magnitude is the point. The same bearing, same load, same speed, lives five to twenty times longer on the right lubricant in the right condition. No other maintenance intervention has that leverage per rupee.
7.4 What κ does not cover
ISO 281 explicitly excludes wear, corrosion and electrical erosion. A bearing can have excellent κ and still fail in six months from:
- Electrical erosion — VFD shaft currents, fluting pattern on the raceway.
- Corrosion — water ingress, acidic oil, condensation during idle periods.
- True brinelling — shock loads during mounting or transport.
- False brinelling — vibration while stationary, fretting of the raceway.
Those are failure modes of the application, not of the lubricant, and no oil analysis will fix them.
8. Gears: Where the Contact Slides as Well as Rolls
8.1 The gear contact is a different animal
In a rolling bearing, the surfaces mostly roll; in a gear mesh, they roll and slide, with specific sliding reaching 30–50% at the pitch line and much more at the tips and roots. That changes the lubrication problem in three ways:
- Scuffing — at high speed and low viscosity, the film collapses and the asperities weld, tearing metal from the flank. It is a sudden failure, not a wear-out.
- Micropitting — at high load and low λ, asperity fatigue produces a grey, matte surface and 10–30 µm pits; it is the dominant failure mode in wind-turbine and hardened industrial gears.
- Wear — slow, uniform removal of the flank, visible as a polished band and measured as a profile change.
8.2 The specifications
Specification · Scope · What it guarantees
ISO 12925-1:2024 · Enclosed gear lubricants, categories CKB–CKE · Viscosity, EP, oxidation, FZG
DIN 51517-3 (CLP) · Industrial gear oils with EP · FZG load stage, wear, corrosion
ANSI/AGMA 9005-F16 · North American gear oil selection · Viscosity by pitch line velocity
API GL-4 / GL-5 · Automotive manual transmissions · EP level (GL-5 higher)
ISO 6743-6 · Classification of gear lubricants · Family and category naming
The FZG scuffing test (DIN 51354, A/8.3/90) runs a pair of gears at increasing load stages until scuffing occurs; a CLP gear oil typically passes load stage 12 or higher, while a plain CKB oil may fail at stage 6–8. That number, not the marketing grade, is what tells you whether the oil can survive a shock load.
8.3 Viscosity selection by pitch line velocity
Pitch line velocity · Recommended ISO VG · Typical application
Over 20 m/s · 68–100 · High-speed turbine gears
10–20 m/s · 100–150 · High-speed industrial units
5–10 m/s · 150–220 · General gearboxes
2–5 m/s · 220–320 · Slow, heavily loaded drives
Under 2 m/s · 320–460 · Kiln drives, crushers, worm gears
The logic is the same as the bearing: at low speed the film needs more viscosity to survive; at high speed a thick oil churns and overheats.
8.4 Worm gears: the one place where base-oil chemistry is not optional
A worm drive has pure sliding contact at the mesh, with a friction coefficient 3–5 times that of a spur gear. The standard lubricant is a polyalkylene glycol (PAG) with a viscosity of VG 220–460, because:
- PAG has a high viscosity index and a low friction coefficient in sliding contact.
- Mineral oil with sulphur–phosphorus EP additive attacks the bronze worm wheel (the sulphur reacts with copper).
- PAG is not compatible with mineral oil — mixing produces a sludge, and the sump must be flushed before switching.
- PAG seals require compatible elastomers (usually FKM or NBR), and paint on the gearbox may soften.
The sump temperature of a worm gear typically runs 80–90 °C, which is why the oil life is measured in 2,000–5,000 hours rather than 10,000.
9. Hydraulics: Cleanliness Is the Specification
9.1 Fluid families
ISO 6743-4 family · Description · Anti-wear · Typical use
HL · Mineral oil, rust and oxidation inhibited · No · Low-pressure systems
HM · HL + anti-wear · Yes · The industrial default
HV · HM + VI improver · Yes · Wide temperature range, mobile
HR · HL + pour-point depressant · No · Cold climates
HG · HM + anti-stick-slip · Yes · Machine-tool ways
HS · Synthetic, fire-resistant · Varies · Special applications
ISO 11158 is the specification that defines HM and HV performance: viscosity grade, VI, anti-wear (vane pump test, ISO 20763), oxidation stability, corrosion, foam, air release, and water separation.
9.2 Viscosity by pressure
System pressure · Grade at 40–55 °C · Grade at 60–75 °C
Under 70 bar · VG 32 · VG 46
70–140 bar · VG 46 · VG 68
140–250 bar · VG 68 · VG 100
Over 250 bar · VG 100 · VG 150
These are starting points, not rules; the OEM's viscosity map for the specific pump is the authority, because a piston pump and a vane pump have different minimum-viscosity limits at the inlet.
9.3 ISO 4406: the code that decides whether the valve survives
ISO 4406 reports three numbers — particles ≥4 µm, ≥6 µm, ≥14 µm per millilitre — each as a code where one step equals a doubling of the count:
Code · Particles/mL (more than) · Up to and including
13 · 40 · 80
14 · 80 · 160
15 · 160 · 320
16 · 320 · 640
17 · 640 · 1,300
18 · 1,300 · 2,500
19 · 2,500 · 5,000
20 · 5,000 · 10,000
21 · 10,000 · 20,000
22 · 20,000 · 40,000
A code of 18/16/13 means 1,300–2,500 particles ≥4 µm, 320–640 ≥6 µm, and 40–80 ≥14 µm per mL. Put differently: a 200 L reservoir at 18/16/13 holds roughly 400 million particles larger than 4 µm, each one capable of bridging a 150 nm film.
Typical targets:
Component · Target ISO code · Why
Servo and proportional valves · 16/14/11 · Clearance 1–3 µm, spool jamming
High-pressure piston pumps · 17/15/12 · Slipper and bore wear
Standard industrial hydraulics · 18/16/13 · General reliability
Gearboxes and bearing sumps · 19/17/14 · Larger clearances
New oil from a drum · often 20/18/15 or worse · Never assume "new" is clean
That last row is the one that surprises people. New oil is not clean oil; it is oil that has not yet been filtered to the system's target. A filter cart and a 3 µm absolute element are the cheapest reliability upgrade in any hydraulic plant.
9.4 Filtration and the beta ratio
Filter efficiency is specified as a beta ratio at a given micron rating:
\beta_4(c) = 200 means 200 particles ≥4 µm upstream for every 1 downstream — 99.5% efficiency. \beta_4(c) = 1{,}000 is 99.9%. ISO 16889 defines the multi-pass test that produces these numbers. Two cautions: the rating is only meaningful with the "(c)" calibration per ISO 11171, and a filter that is bypassing (cold oil, clogged element) has an efficiency of zero regardless of its rating.
9.5 Water, air, and heat
- Water: mineral oil saturates at roughly 200–300 ppm at ambient temperature, more when hot. Beyond saturation, free water appears, emulsions form, and rust begins. Karl Fischer titration (ASTM D6304) is the measurement; 0.05% (500 ppm) is a warning, 0.1% is critical.
- Air: 5–10% entrained air is normal; it makes the oil compressible, causes cavitation and spongy response, and accelerates oxidation. Air release (ISO 9120) and foam (ASTM D892) are the tests.
- Heat: every 10 °C above 60 °C roughly doubles the oxidation rate. A reservoir running at 75 °C instead of 55 °C consumes the oil's life about four times faster. The cooler is not a comfort item.
9.6 Fire-resistant fluids: ISO 12922
Category · Composition · Water content · Notes
HFAE · Oil-in-water emulsion · 95–98% · Fire-resistant, poor lubrication, corrosion risk
HFAS · Water-based synthetic · 95%+ · Similar, better stability
HFB · Water-in-oil emulsion · 40% · Better lubrication than HFA, still fire-resistant
HFC · Water glycol · 35–55% · Common in steel and foundry; good fire resistance
HFDR · Phosphate ester · 0% · Fire-resistant, excellent lubrication, expensive, seal-sensitive
HFDU · Polyol ester / other · 0% · "Less flammable", biodegradable options
Selection is not a drop-in decision: water-containing fluids need stainless or coated reservoirs, different pump materials, and a biocide programme; phosphate esters attack many paints and elastomers; and every category has its own viscosity and seal compatibility rules.
10. The Machine Tool: A Lubrication Map of a CNC Shop
10.1 Spindle
A machining-centre spindle is the most demanding lubrication point in the shop: high speed, high precision, and no tolerance for film loss. Three architectures:
Type · Lubrication · Speed factor n d_m · Notes
Grease-packed · NLGI 2, base oil VG 68–100 · up to 500,000–700,000 · Sealed for life, 10,000–20,000 h
Air–oil · VG 32–68, 0.03–0.1 mL/min per bearing · 700,000–1,500,000 · Continuous, cooled, expensive
Oil-jet · VG 32–46, 2–5 L/min · over 1,500,000 · High speed, high cost
The speed factor is n d_m = rpm × pitch diameter in mm. A 70 mm bore spindle at 12,000 rpm has n d_m = 840,000 — beyond grease, into air–oil. Grease selection must also match the bearing's ceramic or steel ball material and the spindle's preload; the wrong grease in a high-speed spindle causes a 20–30 °C temperature rise and a preload change that shows up as taper in the part.
10.2 Linear guides and ball screws
Component · Lubricant · Interval · Failure if ignored
Linear guide carriage · NLGI 2, EP, VG 68–220 · 100–500 km travel or 3–6 months · Brinelling, uneven wear, backlash
Ball screw · NLGI 2 or VG 68 oil · 200–500 km or 3 months · Preload loss, backlash, noise
Rack and pinion · VG 220 adhesive grease · Monthly · Tooth wear, positioning error
Way (slide) surfaces · VG 68 way lube · 3–8 mL/h per axis · Stick-slip, chatter marks
The most common machine-tool lubrication failure is not the wrong grease — it is contamination from the cutting zone. Fine chips and coolant mist reach the carriage seals, the grease becomes an abrasive paste, and the rail wears in the load zone. Positive-pressure sealing and regular wiping matter more than the brand on the cartridge.
10.3 Coolant: the fluid that is not a lubricant but decides the finish
Metalworking fluid is a coolant first and a lubricant second, and its chemistry is a separate discipline:
Parameter · Target · Consequence outside the band
Concentration (refractometer × factor) · 5–10% general; 3–5% grinding; 8–15% heavy duty · Low: rust, bacteria, tool wear. High: foam, residue, skin irritation
pH · 8.8–9.3 · Below 8.5: corrosion, bacteria. Above 9.5: dermatitis, seal attack
Tramp oil · under 1–2% · Anaerobic bacteria, foam, poor finish, smoke
Bacteria · under 10⁵ CFU/mL · Odour, dermatitis, sump failure
Water hardness · 100–300 ppm as CaCO₃ · Hard water: scum; soft: foam
The refractometer reading must be multiplied by the fluid's refractive index factor (0.9–1.4 depending on type), or the concentration error is 30–40%. Top up with premix at the target ratio, not neat concentrate and not water alone; skim tramp oil weekly; and circulate the sump over weekends to prevent the anaerobic "Monday smell."
11. Oil Analysis: The Instrument That Sees Chemistry
11.1 Sampling discipline
The sample is only as good as the point it came from. Standard practice:
- Sample with the machine running and warm, from a turbulent zone — a live sample port on the pressure line, not the reservoir bottom.
- Sample upstream of the filter if you want to see what the machine is generating, downstream if you want to see what the bearings are receiving. Most programmes use upstream.
- Flush the port for 10–20 seconds before filling; use a clean, dedicated bottle; label with machine, date, hours, and oil brand.
- Take the sample at the same point and same operating condition every time — trend analysis depends on consistency more than on absolute accuracy.
11.2 The test suite
Test · What it sees · Method
ICP spectroscopy · Fe, Cu, Cr, Pb, Sn, Al, Si, Na, K, Ca, Mg, Zn, P, Mo · ASTM D5185
FTIR · Oxidation, nitration, water, glycol, fuel, soot · ASTM E2412
Kinematic viscosity at 40/100 °C · Oxidation thickening, shear-down, fuel dilution · ASTM D445
TAN / TBN · Acid build-up, alkaline reserve · ASTM D664 / D2896
Water · Dissolved and free water · ASTM D6304 (Karl Fischer)
PQ index · Ferrous wear severity · ASTM D8184
ISO 4406 particle count · Cleanliness · ISO 11500 / ISO 4406
Membrane patch / varnish potential · Insoluble varnish precursors · ASTM D7843 (MPC)
11.3 Limits that trigger action
Absolute limits are equipment-specific; trends and rates of change are more reliable. The following are indicative industrial starting points:
Parameter · Normal · Caution · Critical
Iron (gearbox) · under 50 ppm · 50–150 ppm · over 150 ppm
Copper · under 20 ppm · 20–50 ppm · over 50 ppm
Silicon (dirt ingress) · under 10 ppm · 10–25 ppm · over 25 ppm
Water · under 200 ppm · 200–500 ppm · over 500 ppm
TAN rise over baseline · under 0.5 · 0.5–1.5 · over 1.5 mg KOH/g
Viscosity change · under ±5% · ±5–10% · over ±10%
ISO code vs target · at target · one code above · two codes above
PQ index · under 20 · 20–50 · over 50
The most useful rule is the rate rule: if iron doubles between two samples at the same interval, the machine has changed, even if the absolute number is still "normal." A trend line beats a limit table every time.
12. The Failure Atlas: Nine Modes and the Lubricant Behind Each
Mode · Appearance · Mechanism · Lubricant/condition cause · Fix
Abrasive wear · Scratches, lapping marks, rising Fe/Si · Hard particles bridge the film and cut · Contamination above ISO target · Filtration, seals, breathers
Adhesive wear / scuffing · Torn, smeared flanks, sudden noise · Asperity welding at high load/speed · Film collapse, wrong viscosity, no EP · Higher VG, EP additive, cooler
Surface fatigue (pitting) · Spalls, 0.1–2 mm craters · Subsurface crack initiation · Overload, contamination, poor κ · Load reduction, clean oil, ISO 281 check
Corrosive wear · Rust, etching, discoloured races · Water/acid attack · Water ingress, TAN rise · Demulsifier, breather, oil change
Cavitation · Pitted, sponge-like surface · Bubble collapse in low-pressure zone · Air ingress, wrong inlet viscosity · Fix inlet, air release, reservoir design
Fretting · Red-brown debris at fit surfaces · Micro-slip with no oil replenishment · Starved interface, vibration · Anti-fretting grease, fit design
Micropitting · Grey, matte, 10–30 µm pits · Asperity fatigue at low λ · Low λ, high load, no anti-micropitting add · Higher VG, special additive, λ > 1
Electrical erosion · Fluting, craters, grey frosting · Shaft current through the film · VFD common-mode voltage · Ceramic bearings, grounding ring
Varnish · Amber/brown deposits, sticking spools · Oxidation products at hot spots · High temperature, additive depletion · Cooler, better oil, varnish filtration
Two of these deserve emphasis because they are systematically misdiagnosed. Micropitting looks like mild wear and is usually blamed on the gear material; it is a lubrication-film problem, and the fix is a higher λ or an anti-micropitting additive, not a harder gear. Varnish looks like dirt and is a chemical deposit; it forms at hot spots above 120–150 °C, and no filter removes it once it is on the metal — only a varnish-removal programme or an oil change with a cleaner does.
13. The Economics: ₹200/Litre Protecting a ₹1,50,000 Pump
13.1 The Indian market
Metric · Value · Source
Volume, 2025 · 5.6 billion litres · Industry benchmark
Value, 2025 · USD 6.44 billion · Market research
Forecast, 2030 · USD 7.67 billion · 3.53% CAGR
Global rank · 3rd largest · After US and China
Automotive share · 57% (2023) · Passenger and commercial vehicles
Industrial share · The rest — hydraulic, gear, metalworking, grease · Factories, mining, construction
Volume growth is slowing because drain intervals are lengthening and electrification reduces engine-oil demand; value growth is coming from higher-quality synthetics and OEM approvals. For a factory, that means the price gap between a commodity oil and a specification-compliant one is narrowing, while the cost of a failure is not.
13.2 A simple cost model
Take a 250 L hydraulic system on a 22 kW injection-moulding machine:
Item · Cost (indicative) · Notes
ISO VG 46 HM oil, 250 L · ₹45,000–55,000 · ₹180–220/L
Filter element, 10 µm · ₹3,000–8,000 · Every 1,000–2,000 h
Annual oil analysis, 4 samples · ₹8,000–20,000 · ₹2,000–5,000/sample
Servo/proportional valve · ₹40,000–1,20,000 · One contamination event
Axial piston pump · ₹60,000–1,50,000 · One wear-out
Day of lost production · ₹50,000–2,00,000 · Machine and labour
The fluid is 1–3% of the cost of one failure event. That is the entire economic argument for filtration, sampling, and buying the specified grade instead of the cheapest drum.
13.3 Consolidation and storage
Every additional lubricant in a store is a misapplication waiting to happen. A typical mid-size shop can run on 5–10 products:
- Hydraulic oil VG 46 (or 68) — HM
- Gear oil VG 220 — CLP
- Way lube VG 68
- Spindle grease NLGI 2, VG 68–100
- Bearing grease NLGI 2, lithium complex
- Water-soluble coolant
- Compressor oil VG 32/46
- Chain/Open gear adhesive grease
- Rust preventive
- Cleaner/solvent
Storage matters as much as selection: indoors, sealed, labelled with the date, decanted with dedicated containers, and never topped up from a dirty funnel. A filter cart that polishes a drum before it goes into a machine costs less than one valve.
14. The Discipline in One Page
Lubrication is a machine you design, and the design is a chain of five numbers. The film is 147 nm in a 6205 at 1,500 rpm, built by viscosity and speed and almost indifferent to load (W^{-0.073}). The roughness must be smaller than the film; λ = 2.9 on a new raceway and 0.98 on a worn one is the difference between EHL and boundary contact. The viscosity must be read at operating temperature, not at 40 °C — VG 46 is 46 cSt cold and 14.9 cSt at 70 °C, and using the label value overstates the film threefold. The chemistry must still be alive: ZDDP's 100 nm tribofilm is sacrificial, and Zn/P falling in ICP is the only warning you get. And the cleanliness must be measured in ISO 4406 codes, because a 4 µm particle is 27 times thicker than the film it is about to destroy.
Around those five numbers sit the standard's accounting systems: ISO 3448 for viscosity, ASTM D341 for the temperature curve, ISO 281 for κ and life, ISO 12925-1 and DIN 51517-3 for gears, ISO 11158 for hydraulics, ISO 12922 for fire-resistant fluids, ISO 4406 and ISO 16889 for cleanliness, NLGI and ASTM D217 for grease, and ISO 281's κ as the one number that ties lubricant selection to bearing life. The relubrication interval is not a calendar: it is t_f \times f_T \times f_C, and the 15 °C halving law turns 8,000 hours into 1,587 at 90 °C. The quantity is 0.005DB, not "until it comes out the seal."
The failures are predictable. Abrasion, scuffing, pitting, corrosion, cavitation, fretting, micropitting, electrical erosion, varnish — nine modes, each with a lubricant or contamination cause, each visible in oil analysis before it is visible in the part. The economics are not subtle: the fluid is 1–3% of the cost of the failure it prevents, and the same bearing on the right lubricant lives five to twenty times longer. The programme is five habits — select by κ, clean to the ISO code, control water, sample on a schedule, fix leaks — and it costs less than the first pump it saves.
Frequently Asked Questions
What is the difference between ISO VG 46 and SAE 10W?
They are different scales measuring different things. ISO VG is the kinematic viscosity at 40 °C (VG 46 = 41.4–50.6 cSt); SAE 10W is a low-temperature cranking/cold-pumping grade defined at sub-zero temperatures plus a 100 °C band. They are not interchangeable, and an ISO VG 46 hydraulic oil is not "SAE 10W hydraulic oil" even if a catalogue cross-reference suggests a rough equivalence.
Can I mix ISO VG 46 and ISO VG 68?
You can, physically, and you will get something in between — roughly VG 56, outside both grade bands. The problems are additive compatibility and the loss of a known specification: the mixture may not meet the OEM's anti-wear, oxidation, or seal-compatibility requirements. In an emergency, mix and change out within a few hundred hours; never run a mixed sump as a permanent solution.
How often should I grease a bearing?
Calculate it, do not inherit it. Start from the OEM's base interval at 70 °C, then apply f_T = 2^{-(T-70)/15} for temperature and f_C = 0.1–1.0 for contamination. A 6310 with an 8,000 h base, at 90 °C in moderate dust, lands at 1,587 h — 72 days at 8,000 h/year. Measure the housing temperature at relubrication and recalculate if it has drifted.
Is synthetic oil worth it?
It depends on the failure mode you are buying protection against. PAO and Group III base stocks give a higher VI (flatter viscosity curve), better oxidation stability, and lower pour point; in a hot gearbox or a cold-start application that can double oil life and reduce wear. In a clean, cool, lightly loaded system changed on schedule, the premium may never pay back. The decision is economic, not moral: compare the extended drain interval and the reduced failure rate against the price difference.
Why does my hydraulic oil turn dark?
Darkening is usually oxidation, nitration, or dissolved varnish precursors — not dirt, which a filter would remove. It can also be dye fading or contamination from a different fluid. Colour alone is not a reason to change; TAN, viscosity, FTIR oxidation, and ISO 4406 are. A dark oil with normal TAN and clean particle counts may run for thousands more hours; a golden oil with TAN +3 is finished.
What does ISO 4406 18/16/13 mean?
It is the particle count per millilitre at three sizes: 1,300–2,500 particles ≥4 µm, 320–640 ≥6 µm, and 40–80 ≥14 µm. Each code step is a doubling, so 19/17/14 is twice as contaminated as 18/16/13 at every size. Servo valves typically need 16/14/11; standard hydraulics 18/16/13; gearboxes 19/17/14.
Can I use engine oil in a hydraulic system?
No. Engine oils are formulated for combustion by-products, soot dispersion, and a different viscosity curve; they lack the anti-wear and demulsibility package a hydraulic system needs, and their detergents can cause foaming and seal swell. Use an ISO 11158 HM or HV fluid.
Why does grease turn hard or cake in a bearing?
Two mechanisms: base-oil evaporation/oxidation (the oil leaves, the thickener remains) and thickener degradation at temperature. A grease that has caked has no oil left to release, so the bearing runs dry even though it looks "greased." The fix is the right base-oil viscosity and thickener for the temperature, the correct fill quantity, and a relubrication interval based on f_T, not a calendar.
How clean should new oil be before it goes into a machine?
Cleaner than the machine's target, ideally by one ISO code. New oil from a drum is often 20/18/15 or worse — dirtier than a servo system's 16/14/11 target. Filter it with a 3 µm absolute filter cart before transferring, and never use the same container for different fluids.
Previous guides in this series: Abrasive Machining & Precision Grinding · Springs & Spring Design · Industrial Wastewater Treatment · O-Rings, Gaskets & Mechanical Seals · Machine Safety & Guarding · PCB Fabrication & Bare Board Manufacturing · Rolling-Element Bearings · Gears & Gearboxes · Hydraulics · Pneumatics & Compressed Air · Industrial Pumps · Industrial Valves.
References
[1] B. J. Hamrock and D. Dowson, "Isothermal Elastohydrodynamic Lubrication of Point Contacts, Part III — Fully Flooded Results," Journal of Lubrication Technology, 99(2), 1977, pp. 264–276; and NASA TN D-8447 / NASA TP 1342, Minimum Film Thickness in Elliptical Contacts for Different Regimes of Fluid-Film Lubrication (1978) — minimum-film formula H_{\min} = 3.63U^{0.68}G^{0.49}W^{-0.073}(1-e^{-0.68k}) and central-film formula H_c = 2.69U^{0.67}G^{0.53}W^{-0.067}(1-0.61e^{-0.73k}). [2] D. Dowson and G. R. Higginson, Elasto-Hydrodynamic Lubrication, Pergamon, 1966/1977 — line-contact theory and the pressure–viscosity mechanism. [3] ISO 281:2007, Rolling bearings — Dynamic load ratings and rating life — basic rating life L_{10} = (C/P)^p, modified rating life L_{nm} = a_1 a_{\text{ISO}} L_{10}, viscosity ratio κ, contamination factor e_C, fatigue load limit C_u. [4] STLE Tribology & Lubrication Technology, "ISO 281:2007 bearing-life standard — and the answer is?" (July 2010) — history of the Ioannides–Harris model and the integrated a_{\text{ISO}} approach. [5] Schaeffler medias, "Load carrying capacity and life" — κ definition, e_C bands, a_{\text{ISO}} limits (κ capped at 4, a_{\text{ISO}} capped at 50, invalid below κ = 0.1). [6] SKF, Rolling Bearings catalogue and SKF life modification factor a_{\text{SKF}} guidance — κ, e_C, P_u/P and the a_{\text{SKF}} charts. [7] ISO 3448:1992, Industrial liquid lubricants — ISO viscosity classification — 18 grades, ±10% bands, 40 °C reference. [8] ASTM D341 / ASTM D2270 (ISO 2909) — Walther equation \log_{10}(\log_{10}(\nu+0.7)) = A - B\log_{10}T and viscosity index. [9] API 1509 Annex E, API Base Oil Interchangeability Guidelines — Group I–V definitions (saturates, sulphur, VI). [10] NLGI, Grease Glossary; ASTM D217 / ISO 2137 — NLGI consistency numbers 000–6 and worked penetration ranges. [11] SKF, Grease lubrication and relubrication guidance — base interval t_f, f_T halving per 15 °C above 70 °C, f_C contamination factor, quantity G = 0.005DB, 30–50% initial fill. [12] ISO 12925-1:2024, Lubricants for enclosed gear systems — categories CKB, CKC, CKD, CKE and the FZG test requirements. [13] DIN 51517-3 (CLP) and ANSI/AGMA 9005-F16 — industrial gear-oil specifications and viscosity selection by pitch line velocity. [14] ISO 11158, Hydraulic fluids — HM and HV; ISO 6743-4, Family H classification — HL, HM, HV, HR, HG, HS families. [15] ISO 4406:2021, Hydraulic fluid power — Solid contaminant code — three-size code (4/6/14 µm(c)), doubling per step. [16] ISO 16889, Multi-pass method for evaluating filtration performance — beta ratio \beta_x(c). [17] ISO 12922:2020, Fire-resistant and less flammable hydraulic fluids — HFAE, HFAS, HFB, HFC, HFDR, HFDU categories. [18] ISO 20763 (vane pump anti-wear) and ISO 20623 (four-ball EP) — hydraulic and gear lubricant performance tests. [19] K. A. Koye, An experimental evaluation of the Hamrock and Dowson minimum film thickness equation for fully flooded EHD point contacts, NASA Technical Memorandum (1981) — measured films averaging about 30% above the prediction. [20] ASTM D5185 (ICP), ASTM E2412 (FTIR), ASTM D445 (viscosity), ASTM D664/D2896 (TAN/TBN), ASTM D6304 (Karl Fischer water), ASTM D8184 (PQ index), ASTM D7843 (MPC varnish) — oil-analysis methods referenced in §11. [21] India lubricant market data: Ken Research, India Lubricants Market 2026–2032 (USD 6,445 million in 2025; 5.60 billion litres; 3.53% CAGR); MarketsandMarkets, India Lubricants Market (USD 6,444.8 million in 2025, USD 7,666.5 million by 2030); ResearchAndMarkets, India Lubricants Industry Research 2024–2030 (automotive 57% share in 2023; USD 7.19 billion in 2023, USD 9.70 billion by 2030 at 4.4% CAGR). [22] Metalworking-fluid practice: ASTM E2275 (bioresistance), refractometer refractive-index factors, pH 8.8–9.3, tramp oil under 1–2%, bacteria under 10⁵ CFU/mL — field guidance from coolant-management literature.
Note on sources: the governing equations are Hamrock–Dowson for film thickness and ISO 281 for life; the specifications are ISO 3448, ISO 4406, ISO 11158, ISO 12922, ISO 12925-1, DIN 51517-3, AGMA 9005, NLGI and ASTM D341/D217. Where vendor guidance differs by a few percent — κ bands, e_C values, grease base intervals — this guide prints the working number and the standard behind it, because a lubricant decision is a contract between a specific oil, a specific surface and a specific duty cycle, and no catalogue has seen all three. Measure the temperature, count the particles, sample the oil, and calculate the film before choosing the pail.
FabFlow connects factories with the manufacturing capacity that keeps them running — precision machining, sheet metal, casting, additive manufacturing and industrial fabrication. Post a job at fabflow.app and get quotes from verified fabricators.