Journal bearings: floating a shaft on a film of oil

A large pump can carry a steel shaft that weighs hundreds of kilograms, spin it all day, and still avoid metal-to-metal rubbing in normal service. That sounds impossible until you understand the bearing. A journal bearing, also called a plain hydrodynamic bearing, does not rely on balls or rollers. It relies on a moving wedge of oil that builds pressure and lifts the journal off the bearing surface.

That is why plain bearings still show up in turbines, compressors, diesel engines, gearboxes, and heavily loaded industrial fans. They can be quiet, compact, forgiving of shock, and extremely durable, but only if the oil film is allowed to form. The beginner mistake is to imagine the shaft running concentrically inside a smooth bush. In real hydrodynamic operation the shaft center shifts slightly, the clearance becomes a converging wedge, and that wedge is the whole trick.

What a journal bearing is really doing

The rotating shaft surface is the journal. Around it sits a softer bearing shell, often lined with bronze, white metal, or another bearing alloy. Between them is a small radial clearance filled with oil. When the shaft is stationary or turning too slowly, the oil does not generate enough pressure to carry the load, so the surfaces may touch in a boundary-lubrication condition. As speed rises, the rotating journal drags oil into the narrowing clearance zone. Viscosity resists that shearing motion, pressure builds inside the wedge, and the journal rides on the fluid film.

The key point is that the load is not carried by the average oil pressure in the housing. It is carried by a pressure distribution created by motion, geometry, and viscosity. No rotation means no hydrodynamic wedge. That is why startup, shutdown, oil starvation, and wrong-viscosity oil are the dangerous parts of the story.

Figure 1: In normal operation the shaft center shifts away from the bearing center, creating a converging oil wedge that generates the pressure carrying the load.

The governing physics engineers use first

You can solve journal bearings with Reynolds-equation methods and detailed charts, but the first engineering pass is simpler. Start with four quantities:

• p = W / (L d) for average bearing pressure, where W is load, L is bearing length, and d is journal diameter.
• v = pi d n / 60 for surface speed in meters per second when diameter is in meters and speed n is in rpm.
• h_min = c - e = c (1 - epsilon) for minimum film thickness, where c is radial clearance and e is the journal-center offset.
• P_loss = f W v for friction power, where f is the operating friction coefficient.

Those equations do not replace full bearing design, but they frame the real questions. Is the unit pressure reasonable for the material? Is the surface speed high enough to build film? Is the minimum film thickness comfortably larger than the roughness peaks and dirt particles in the oil? Is the friction heat something the bearing housing and oil flow can actually remove?

Viscosity sits in the middle of all of this. Thicker oil helps generate film pressure, but it also increases shear drag and heat. Thinner oil reduces drag, but the film can collapse if load is high or speed is low. That is why journal-bearing design is always a balancing act between load, speed, clearance, and oil viscosity at the actual operating temperature, not the number printed on a cold oil drum.

Why the shaft does not run in the middle

If the journal sat exactly at the center of the bearing with equal clearance all around, the oil film would also be symmetric, and there would be no net hydrodynamic pressure to support the load. Under load the journal shifts downward by a small amount. That shift makes the inlet side comparatively open and the exit side narrow. As the shaft rotates, oil is pulled into the narrowing zone, and pressure rises. The resulting pressure field pushes back against the external load.

This is why people talk about eccentricity ratio, often written as epsilon = e / c. If epsilon is small, the journal is close to the center and the film is thick everywhere. If epsilon approaches 1, the journal is nearly touching the bearing at one side and the minimum film thickness becomes very small. A bearing can still carry load in that condition, but dirt, misalignment, or temperature rise can push it into wiping or seizure.

Journal bearings survive because the oil film carries the shaft before contact happens, not because the bearing metal is comfortable rubbing forever.

Worked example 1: a pump-drive bearing load check

A process-pump shaft uses a journal bearing with diameter 60 mm and length 75 mm. The steady radial load at one bearing is 50 kN, and the shaft speed is 1450 rpm.

The average bearing pressure is:

p = W / (L d) = 50000 / (0.075 x 0.06) = 11.1 MPa

The journal surface speed is:

v = pi d n / 60 = pi x 0.06 x 1450 / 60 = 4.56 m/s

Engineers often use the pressure-speed product as a quick screening value:

p v = 11.1 x 4.56 = 50.6 MPa.m/s

That does not prove the design is safe, but it tells us the bearing is working in a serious industrial range, not a toy mechanism range. The diameter and speed are high enough for hydrodynamic film generation to be plausible, and the unit pressure tells us material choice, oil viscosity, and heat removal all deserve attention. A tiny oil groove and casual grease replacement would be completely wrong here.

Worked example 2: minimum film thickness and why dirt matters

Now consider a gearbox bearing with journal diameter 80 mm. The diametral clearance is 0.10 mm, so the radial clearance is:

c = 0.10 / 2 = 0.05 mm = 50 micrometers

Suppose operating conditions place the journal at an eccentricity ratio of epsilon = 0.78. Then the minimum film thickness is:

h_min = c (1 - epsilon) = 50 x (1 - 0.78) = 11 micrometers

Eleven micrometers is still a real fluid film, but it is not generous. If the combined surface roughness of journal and bearing is a few micrometers, and contamination particles of 8-10 micrometers are occasionally passing through the oil, the safety margin is already thin. This is the practical lesson: a bearing that looks acceptable on load and speed can still be one dirty filter away from scoring. Minimum film thickness is not an abstract number. It is your clearance budget for roughness, dirt, alignment error, and thermal distortion.

Figure 2: Journal-bearing performance depends as much on oil delivery, surface finish, and inspection discipline as on the shell material itself.

Worked example 3: friction heat and oil-flow reality

A blower bearing carries 18 kN at a journal surface speed of 6.3 m/s. In full-film operation, assume the friction coefficient is about 0.0025. The friction power is:

P_loss = f W v = 0.0025 x 18000 x 6.3 = 283.5 W

That means this one bearing is turning nearly 0.28 kW into heat continuously. Suppose the oil flow through the bearing is 3 L/min. Taking oil density as 860 kg/m^3, the mass flow is about:

m_dot = 0.003 / 60 x 860 = 0.043 kg/s

If the oil specific heat is roughly 2000 J/kg.K, the ideal bulk temperature rise needed to absorb that friction heat is:

Delta T = P_loss / (m_dot c_p) = 283.5 / (0.043 x 2000) = 3.3 C

That is a useful sanity check. A few degrees of rise sounds manageable, but only if the oil really reaches the loaded zone, the housing can shed heat, and the bearing is not also being heated by nearby process equipment. If the oil flow falls, viscosity changes, or the bearing runs partially mixed instead of fully hydrodynamic, the heat balance worsens quickly.

Assumptions and where the simple picture breaks down

The beginner explanation assumes a perfectly round shaft, a perfectly round bearing, steady load, Newtonian oil behavior, no cavitation trouble, and negligible edge effects. Real machines violate most of those assumptions. Misalignment pushes load toward one edge. Flexible shafts wobble and shift the load zone. Dynamic loads from gears, blades, or reciprocating forces change the film thickness every revolution. Temperature changes viscosity faster than many people expect.

Length-to-diameter ratio matters too. A very short bearing has less projected area and stronger edge effects. A very long bearing may not share load evenly because the shaft and housing bend. That is why experienced designers do not stop at one pressure number. They check geometry, oil path, expected operating temperature, startup regime, and whether the bearing is being asked to locate the shaft axially in a job better left to a thrust bearing.

Common failure modes and how they appear

• Oil starvation: the film collapses because not enough oil reaches the inlet. The bearing overheats rapidly, and the surface can wipe or seize.
• Wiping: the soft lining smears in the loaded zone after local overheating or direct contact. The shiny polished patch becomes torn or dragged.
• Scoring: hard dirt particles or a damaged shaft surface cut grooves along the direction of motion.
• Edge loading: misalignment concentrates load near one end, often leaving one side darkened or worn while the other still looks acceptable.
• Cavitation and erosion: low-pressure zones at the oil-film exit can create vapor pockets and pitting in severe service.

Notice that none of these failures are fixed by saying, "Use a stronger bronze." Journal bearings fail as a lubrication system, a geometry system, and a thermal system. Material matters, but film formation matters more.

Figure 3: Healthy plain bearings look smooth and evenly polished; wiped, scored, or edge-loaded shells tell you the oil film or alignment broke down.

Practical rules of thumb

• For many industrial journal bearings, a first-pass L / d ratio between about 0.5 and 1.0 is common, but the final value depends on stability, space, and allowable pressure.
• Diametral clearance often lands on the order of 0.001 d as a starting point, not a law. Final clearance must reflect speed, oil viscosity, thermal growth, and manufacturing capability.
• If operating speed is low and load is high, hydrodynamic film formation becomes harder. That is the regime where startup wear and mixed lubrication need real attention.
• Cleaner oil often extends bearing life more effectively than switching to a more exotic lining alloy.
• A plain bearing that runs cool and stable is usually telling you the oil system is healthy; a rising metal temperature is an early warning, not a nuisance signal to ignore.

How standards treat journal bearings

Rolling bearings often get discussed through life-rating standards such as ISO 281, but journal bearings are treated differently. Hydrodynamic plain-bearing work is commonly organized around standards in the ISO 7902 family, which focus on the calculated operating characteristics of oil-lubricated journal bearings: film pressure, minimum film thickness, oil flow, friction, and temperature behavior. Material-focused standards such as those in the ISO 4386 family help classify and inspect metallic plain-bearing layers.

The practical lesson from those standards is important: plain bearings are not chosen only by bore size and a catalog load number. Standards push engineers to think in terms of operating viscosity, clearance, thermal balance, surface finish, alignment, and minimum film thickness. In other words, the standards treat the bearing as a fluid-film machine element, not just a shaped piece of metal.

Engineering judgment: respect the oil film

A journal bearing can be wonderfully simple in hardware and surprisingly sophisticated in operation. The shell may look like a plain sleeve, but the real load-carrying element is the oil wedge created by motion, viscosity, and geometry. If speed, clearance, and oil supply are right, the shaft floats. If any of those go wrong, metal contact returns fast.

The best beginner habit is to ask four questions every time: what is the unit pressure, what is the surface speed, how thick is the minimum film likely to be, and where does the heat go? If those answers make sense, you are thinking like a bearing designer instead of a parts buyer. For more machine-design explainers in the same style, continue through the EnggTools articles library and the related shaft and bearing topics.