Glossary term

Cross-Coupled Stiffness

Off-diagonal rotor-dynamic stiffness that creates lateral force in one direction from displacement in the orthogonal direction, often driving subsynchronous instability.

Branch: Mechanical Engineering
Glossary type: quantity
Content: Glossary term
Updated: Jun 25, 2026
Revision: v1.0.0 · reviewed

Definition

quantity

Cross-coupled stiffness is an off-diagonal stiffness coefficient that creates force in one lateral direction from displacement in the orthogonal direction.

In rotor dynamics, cross-coupled stiffness appears in bearings, seals, impellers, fluid films and other components where motion in x can produce force in y, or motion in y can produce force in x. Depending on sign convention and whirl direction, the coefficient can feed energy into a forward whirl orbit and reduce effective damping. It is one of the mechanisms behind subsynchronous instability, oil whirl, oil whip and some seal-induced vibration problems.

Cross-coupled stiffness is a rotor-dynamic coefficient that links motion in one lateral direction to force in the perpendicular direction. In a simple spring, displacement in $x$ creates force in $x$ . In a cross-coupled system, displacement in $x$ can also create force in $y$ .

This matters because the force can act in the direction of whirl motion. When that happens, the component does not simply store elastic energy. It can feed energy into the orbit and reduce the effective damping that would otherwise stabilize the rotor.

Stiffness Matrix

A linearized lateral force model can be written as:

\begin{bmatrix}F_x\\F_y\end{bmatrix}=-\begin{bmatrix}K_{xx}&K_{xy}\\K_{yx}&K_{yy}\end{bmatrix}\begin{bmatrix}x\\y\end{bmatrix}-\begin{bmatrix}C_{xx}&C_{xy}\\C_{yx}&C_{yy}\end{bmatrix}\begin{bmatrix}\dot{x}\\\dot{y}\end{bmatrix}

The direct stiffness terms are $K_{xx}$ and $K_{yy}$ . The cross-coupled stiffness terms are $K_{xy}$ and $K_{yx}$ . Their units are force per displacement, usually newtons per metre.

In rotating machinery, cross-coupled stiffness can come from:

Source	Typical mechanism	Diagnostic relevance
Fluid-film journal bearing	pressure field shifts around the journal	oil whirl or oil whip risk
Annular or labyrinth seal	circumferential fluid force follows rotor motion	subsynchronous seal instability
Impeller or compressor stage	aerodynamic or hydraulic cross force	load-dependent whirl risk
Preload or clearance error	asymmetric support force	changed stability margin

The coefficient is not automatically bad. The sign, coordinate convention, speed, direct damping, mode shape and operating condition determine whether it is destabilizing.

Distinction from Direct Stiffness and Damping

Direct stiffness resists displacement in the same coordinate direction. If the shaft moves in $x$ , a direct $K_{xx}$ term pushes mainly against $x$ . Cross-coupled stiffness is different because the restoring or destabilizing force is rotated into the other lateral coordinate. That rotation is why the coefficient can change whirl stability without looking like a simple static stiffness problem.

It is also different from damping. Damping force depends on velocity. Cross-coupled stiffness depends on displacement, but during circular whirl the displacement vector is phase shifted relative to velocity. Under some sign conventions this makes a stiffness-generated force act like negative damping. That is the practical reason engineers track cross-coupled terms during bearing, seal and impeller stability reviews.

The distinction affects engineering decisions:

raising direct stiffness may move a critical speed but may not solve a subsynchronous instability;
adding damping can improve margin if it acts in the participating mode and direction;
reducing clearance may increase direct stiffness while also changing destabilizing fluid coefficients;
changing oil temperature can modify both viscosity and effective bearing coefficients;
a balance correction may reduce 1x response while leaving cross-coupled instability unchanged.

Why It Can Destabilize Whirl

For a simplified circular forward whirl orbit:

x=r\cos(\Omega t)

y=r\sin(\Omega t)

Assume a skew-symmetric cross-coupled stiffness model:

\begin{bmatrix}F_x\\F_y\end{bmatrix}=-\begin{bmatrix}0&K_c\\-K_c&0\end{bmatrix}\begin{bmatrix}x\\y\end{bmatrix}

The resulting forces are:

F_x=-K_c r\sin(\Omega t)

F_y=K_c r\cos(\Omega t)

Because the orbit velocity is:

\dot{x}=-r\Omega\sin(\Omega t)

\dot{y}=r\Omega\cos(\Omega t)

the cross-coupled force is aligned with velocity:

\displaystyle F_x=\frac{K_c}{\Omega}\dot{x}

\displaystyle F_y=\frac{K_c}{\Omega}\dot{y}

That is equivalent to negative damping in this sign convention. A useful screening relation is:

\displaystyle C_{negative}=\frac{K_c}{\Omega}

where $C_{negative}$ has damping units. If this negative damping is larger than the stabilizing direct damping available to the mode, a subsynchronous orbit can grow.

Simple Stability Screen

A practical dimensionless screen is:

\displaystyle S_{cc}=\frac{K_c}{C_d\Omega}

where $C_d$ is the direct stabilizing damping for the relevant mode or bearing direction and $\Omega$ is whirl angular frequency. Values well below 1 suggest damping dominates the cross-coupled effect. Values near or above 1 indicate that a more serious stability review is needed.

Example: a rotor shows a subsynchronous component near $70\ \text{Hz}$ during a speed sweep. The angular frequency is:

\Omega=2\pi(70)=439.8\ \text{rad/s}

The estimated direct damping is:

C_d=800\ \text{N s/m}

The cross-coupled stiffness is:

K_c=420000\ \text{N/m}

The stabilizing damping stiffness scale is:

C_d\Omega=800(439.8)=351840\ \text{N/m}

The cross-coupling screen is:

\displaystyle S_{cc}=\frac{420000}{351840}=1.19

This result does not prove the full machine is unstable, but it explains why a forward subsynchronous orbit could grow. If the same machine is modified so that effective direct damping rises to $1100\ \text{N s/m}$ , then:

\displaystyle S_{cc}=\frac{420000}{1100(439.8)}=0.87

The stability margin is still not generous, but the direction of change is beneficial.

Design and Troubleshooting Use

In design, cross-coupled stiffness is used to decide whether a rotor-bearing-seal system has enough stability margin across the operating envelope. The question is not only “where are the critical speeds?” but also “will the forward modes have enough damping when fluid forces are included?” This is especially important for compressors, turbines, pumps, high-speed fans, turbochargers, marine shafting support equipment and machines with tight seals or fluid-film bearings.

In troubleshooting, the term helps separate forced response from self-excited response. A high 1x vector that changes coherently with a trial weight points toward unbalance response. A growing sub-1x component with forward orbit precession, temperature sensitivity or load dependence points toward a stability mechanism. Cross-coupled stiffness is one candidate mechanism, but the evidence must still be checked against rubs, looseness, process pulsation and measurement artifacts.

Typical actions include reviewing bearing preload, pad geometry, oil supply, seal clearance, pressure ratio, balance state, operating load, ramp rate and trip logic. A useful report states which change is expected to reduce the destabilizing coefficient, which change is expected to add damping and how the result will be verified during a repeat sweep.

Field Evidence

Cross-coupled stiffness is inferred from a combination of model and measurement evidence. Useful field indicators include:

a subsynchronous peak that grows with speed or load;
a forward-precessing orbit from proximity probes;
a waterfall spectrum showing the component across a speed sweep;
shaft centerline movement consistent with a changed bearing or seal operating point;
sensitivity to oil temperature, pressure, clearance, preload, load or seal condition;
poor explanation by simple 1x unbalance response.

The evidence should be interpreted with a tachometer or once-per-revolution reference when phase and order matter. Without a reliable reference, a fixed-frequency component, aliasing problem or spectral leakage can be mistaken for a rotor instability.

Common Mistakes

Do not treat cross-coupled stiffness as ordinary direct stiffness. Increasing a direct stiffness may raise a natural frequency, but increasing a destabilizing cross-coupled coefficient can reduce stability even if the machine looks stiffer in a static sense.

Do not use cross-coupled stiffness as a label for every subsynchronous vibration. Rubs, looseness, aerodynamic forcing, rolling-element bearing faults, electrical runout, sampling problems and process pulsation can also create non-1x components. The diagnosis should connect the frequency trend, orbit direction, operating condition and machine physics.

Another mistake is balancing the rotor first because the vibration plot looks large. Cross-coupled stiffness usually drives self-excited subsynchronous motion. Balance correction targets synchronous 1x unbalance response, not a destabilizing fluid-force mechanism.

Validation Limits

Cross-coupled stiffness coefficients are usually linearized around an operating point. They can change with speed, load, oil viscosity, pressure ratio, clearance, eccentricity ratio, seal wear, thermal state and shaft position. A coefficient that appears stable in one operating condition may be unsafe after a temperature change or load shift.

A defensible stability review should state the coordinate convention, units, operating point, source of coefficients, damping assumption, measurement chain and acceptance criterion. For a field troubleshooting decision, the minimum evidence package should include orbit plots, waterfall or order tracking, probe calibration, runout review, shaft centerline trend and bearing or seal operating data. For design release, a validated rotor-dynamic model and OEM or project-specific stability criteria are normally required.

REF

Disciplines