Glossary term

Roll Subsidence

Fast lateral-directional aircraft mode in which roll rate decays after a disturbance or aileron release, mainly governed by roll damping and inertia.

Branch: Aerospace Engineering
Glossary type: phenomenon
Content: Glossary term
Updated: Jun 24, 2026
Revision: v1.0.0 · reviewed

Definition

phenomenon

Roll subsidence is a fast lateral-directional aircraft mode in which roll rate decays after a roll disturbance or aileron release, usually dominated by aerodynamic roll damping.

Roll subsidence is commonly represented by a real, stable eigenvalue in the lateral-directional state-space model. The mode describes how quickly roll rate damps out, not the full bank-angle trajectory. It depends on roll damping, wing geometry, dynamic pressure, roll moment of inertia, Mach number, angle of attack, configuration, aileron effectiveness, actuator behavior and control-law scheduling.

Roll subsidence is the fast lateral-directional aircraft mode that describes how roll rate decays after a roll disturbance or after an aileron input is released. In most conventional small-disturbance models it is a stable, nonoscillatory real mode dominated by aerodynamic roll damping.

For a simple modal screen:

p(t)=p_0e^{\lambda_{roll}t}

where $p$ is roll rate and $\lambda_{roll}$ is the roll-subsidence eigenvalue. With the usual stable continuous-time convention, $\lambda_{roll}<0$ means the roll rate decays. A value close to zero means the aircraft keeps rolling for longer after the input or disturbance.

Engineering Role

Roll subsidence matters because it affects roll response, bank-angle capture, aileron release behavior, lateral disturbance rejection, pilot workload, autopilot tuning, actuator sizing and flight-test data interpretation. It is usually much faster than spiral mode, but it can still expose weak roll damping, high inertia, sensor filtering, actuator lag or configuration-dependent handling-quality issues.

The mode should be interpreted together with aileron control effectiveness. Aileron effectiveness estimates how much rolling moment a command can generate; roll subsidence estimates how quickly roll rate dies away once the aircraft is rolling. A strong aileron derivative does not guarantee acceptable roll response if roll damping, actuator rate or inertia are poorly represented.

Worked Example: Roll-Rate Decay Time

A lateral-directional model predicts a roll-subsidence eigenvalue:

Parameter	Value
Roll-subsidence eigenvalue, $\lambda_{roll}$	$-2.80\ \text{s}^{-1}$
Initial roll rate after release, $p_0$	$12.0^\circ/\text{s}$
Review target	roll rate below 5 percent of initial value

The time constant is:

\displaystyle \tau=\frac{1}{|\lambda_{roll}|}=\frac{1}{2.80}=0.357\ \text{s}

The roll rate after $1.0\ \text{s}$ is:

p(1)=12.0e^{-2.80(1.0)}=0.73^\circ/\text{s}

The time to reach 5 percent of the initial roll rate is:

\displaystyle t_{5\%}=\frac{-\ln(0.05)}{|\lambda_{roll}|}

\displaystyle t_{5\%}=\frac{2.996}{2.80}=1.07\ \text{s}

Now consider a configuration change that weakens roll damping and moves the eigenvalue to:

\lambda_{roll}=-0.85\ \text{s}^{-1}

The new time constant is:

\displaystyle \tau=\frac{1}{0.85}=1.18\ \text{s}

The roll rate after $1.0\ \text{s}$ is:

p(1)=12.0e^{-0.85(1.0)}=5.13^\circ/\text{s}

The 5 percent decay time becomes:

\displaystyle t_{5\%}=\frac{2.996}{0.85}=3.52\ \text{s}

Engineering comment: both eigenvalues are stable, but the second case leaves much more residual roll rate after one second. That can affect bank-angle capture, lateral tracking, pilot workload and autopilot gains. The check should be tied to the flight condition, configuration, inertia, aileron response, sensor filtering and uncertainty in roll damping.

Distinction from Dutch Roll and Spiral Mode

Roll subsidence is different from Dutch roll. Dutch roll is an oscillatory yaw-roll-sideslip mode and is usually assessed with natural frequency and damping ratio. Roll subsidence is normally a real mode that mainly describes roll-rate decay.

Roll subsidence is also different from spiral mode. Spiral mode is a slow bank-yaw-sideslip convergence or divergence that can unfold over tens of seconds. Roll subsidence is usually much faster and concerns how quickly roll rate damps after a disturbance or command.

The three modes can still interact in real aircraft. Aileron-rudder coordination, yaw damping, actuator limits, aeroelastic deformation, high angle of attack, asymmetric stores or control-law scheduling can make a clean modal separation less reliable.

What Changes Roll Subsidence

Roll-subsidence behavior depends on:

roll damping derivative and wing aerodynamic characteristics;
wing span, aspect ratio, sweep, dihedral effect and flow separation;
roll moment of inertia and fuel or payload distribution;
dynamic pressure, Mach number, altitude and configuration;
angle of attack, stall proximity, icing, stores and damage;
aileron effectiveness, adverse yaw and rudder coordination;
actuator rate limits, surface freeplay, backlash and structural flexibility;
gyro bandwidth, filtering, latency and state-estimator behavior;
control-law gains, bank-hold logic and envelope-protection schedules.

Because roll subsidence is often fast, measurement bandwidth matters. A low-rate logging system or aggressive filtering can make the response appear slower, cleaner or more damped than the aircraft actually was.

Validation and Common Mistakes

Roll subsidence can be identified from state-space eigenvalues, roll doublets, aileron release tests, bank-angle captures, system-identification maneuvers, validated simulation or frequency-response data. A defensible value states flight condition, configuration, mass properties, control-law mode, actuator state, sensor filtering, sign convention and uncertainty.

Common mistakes include:

assuming a stable roll-subsidence eigenvalue automatically means acceptable roll handling;
confusing roll subsidence with Dutch-roll damping or spiral stability;
checking aileron moment without checking roll damping and inertia;
using sensor data with insufficient bandwidth for a fast roll response;
ignoring actuator rate limits, freeplay or control-law filters;
comparing roll eigenvalues from models with different axes, state definitions or signs;
validating a nominal roll response while leaving high-angle, icing, stores or degraded-mode cases untested.

REF

Disciplines