Oil Whirl

Oil whirl is a self-excited rotor vibration associated with fluid-film journal bearings. Instead of acting only as damping, the oil film can generate cross-coupled forces that feed energy into lateral shaft motion. The vibration is usually subsynchronous, meaning it occurs below 1x shaft speed.

A common diagnostic quantity is whirl order:

where f_w is whirl frequency and f_{rot} is rotational frequency. Oil whirl often appears near:

This range is a practical clue, not a universal rule. Bearing geometry, preload, clearance, oil viscosity, temperature, load, speed, rotor modes and support stiffness all affect the actual instability.

Engineering Role

Oil whirl matters because it can grow even when the rotor is well balanced. A machine may show acceptable 1x vibration but still develop a dangerous subsynchronous component. Orbit plots may show forward precession, while waterfall spectra and order tracking can show a peak near a fractional order of shaft speed.

The mechanism is associated with fluid-film bearings in turbines, compressors, generators, pumps, marine machinery and high-speed rotating equipment. It is different from simple unbalance response because the excitation is not a mass eccentricity force fixed at 1x. It is also different from bearing defect frequencies in rolling-element bearings.

Oil whirl can progress into oil whip. In oil whirl, the subsynchronous frequency tends to follow a fraction of shaft speed. In oil whip, the vibration may lock near a rotor natural frequency and remain there as speed increases. Oil whip is usually more severe because the instability is coupled to a rotor mode.

Evidence and Operating Limits

Oil whirl should be treated as a stability diagnosis, not only as a spectral label. A useful first screen compares the observed subsynchronous frequency with a fractional running-speed band:

If the peak follows this band as speed changes, the evidence supports oil whirl. If the peak stops following running speed and remains close to a rotor mode, the evidence shifts toward oil whip. A simple lock-in screen is:

where f_n is the relevant natural frequency. A small M_{lock} during a growing subsynchronous response should be treated as a serious warning, especially in lightly damped machines.

Amplitude trend matters as much as order. Between two speed points, the growth ratio can be written as:

where A_w is subsynchronous amplitude in a stated unit. A growing G_A, forward-precessing orbit, rising bearing temperature or changing shaft centerline position should trigger a hold point or trip criterion before the sweep continues. The release evidence should say which response level stops the test, which level allows controlled operation and which changes require bearing inspection or oil-system review.

Worked Example: Identify Whirl Order and Oil Whip Risk

A machine with fluid-film journal bearings is tested during speed increase. The first bending natural frequency is estimated as:

At:

the rotational frequency is:

A waterfall spectrum shows a growing subsynchronous peak at:

The whirl order is:

That is consistent with oil-whirl evidence, provided the sensor, tachometer and operating condition are valid.

At:

the rotational frequency is:

If the peak moves to:

then:

The peak is still tracking a fractional order of speed, which supports an oil-whirl interpretation.

Now increase speed to:

so:

If the subsynchronous peak does not move to 36\ \text{Hz} but instead remains close to:

then it has locked near the first bending natural frequency. That behaviour suggests possible oil whip rather than ordinary oil whirl.

Engineering comment: the practical action is not to label the plot and continue operation. The engineer should review bearing load, oil temperature, clearance, eccentricity ratio, preload, alignment, shaft orbit, casing vibration, ramp rate and trip limits. A subsynchronous instability can become a machine-protection issue.

Distinction from Related Terms

Oil whirl is not unbalance response. Unbalance response is forced 1x vibration from mass eccentricity. Oil whirl is a self-excited subsynchronous instability.

Oil whirl is not an orbit plot. An orbit plot is a display. Oil whirl is one possible phenomenon that an orbit plot can help reveal.

Oil whirl is not oil whip. Oil whirl usually follows a fractional running-speed order. Oil whip tends to lock to a rotor natural frequency after the instability couples strongly to a mode.

Oil whirl is not any subsynchronous peak. Rubs, looseness, seal instabilities, rotating stall, electrical effects, control interaction and measurement artefacts can also create sub-1x components.

Validation and Common Mistakes

A defensible oil-whirl diagnosis states bearing type, oil grade, oil temperature, load, speed range, clearance, preload, sensor location, phase reference, filter settings, orbit shape, waterfall trend, order trend and operating history.

Common mistakes include:

• calling any sub-1x vibration oil whirl without checking bearing type and operating condition;
• ignoring oil temperature, viscosity, load and bearing clearance;
• confusing oil whirl with unbalance because both may appear during run-up;
• missing oil whip because the peak stops tracking fractional order and locks near a natural frequency;
• relying on a single spectrum instead of speed trend, orbit plot and repeatability;
• accepting an order estimate without checking tachometer validity, aliasing, clipping and signal-to-noise ratio;
• continuing a speed sweep through growing subsynchronous vibration without defined trip or hold criteria.