Ship Motions: Six DOF in a Seaway

A ship in a seaway is a rigid body subject to six independent degrees of freedom (DOF): three translations and three rotations. Every motion has its own natural period, its own damping characteristics, and its own relationship to the wave-encounter geometry. When a wave frequency coincides with a natural frequency, the response can grow to amplitudes that damage cargo, injure crew, or threaten stability. The seakeeping discipline quantifies these responses through Response Amplitude Operators (RAOs) and uses them to define operability limits, design stabilization systems, and guide bridge watchkeeping. This article sets out the physics of each DOF, the governing equations, the key operational consequences, and the regulatory framework that links motions to intact stability requirements.

Companion calculators are available for the natural roll period and the wave encounter period; for parametric roll exposure, see the parametric roll vulnerability calculator.

The six degrees of freedom

A rigid body in three-dimensional space has six independent DOF. For a ship, the conventional naming, axis convention, and motion directions are:

The coordinate origin is conventionally at the centre of gravity, with the $x$-axis pointing forward, $y$-axis to starboard, and $z$-axis upward. Roll, pitch, and yaw angles follow the right-hand rule about the respective positive axis directions.

At small amplitudes the six motions can be treated as approximately uncoupled, which is the basis of linear seakeeping theory. The main coupling pairs that matter in practice are pitch-heave (because heave changes the submerged volume distribution along the ship length, which drives pitch excitation) and roll-yaw (because yaw changes the heading relative to the waves, shifting the effective roll excitation). Roll-pitch coupling becomes significant only at large heel angles, where the transverse metacentric height varies with the instantaneous heel angle. These couplings are preserved in the full six-DOF equations of motion used in strip theory and panel codes.

The equations of motion

The linear equations of motion for each DOF take the form:

where $M_{ij}$ is the ship mass/inertia matrix, $A_{ij}(\omega)$ is the frequency-dependent added mass matrix, $B_{ij}(\omega)$ is the frequency-dependent radiation damping matrix, $C_{ij}$ is the hydrostatic restoring matrix, $F_i^{(exc)}$ is the complex wave excitation force/moment amplitude, and $\omega_e$ is the encounter frequency. The frequency dependence of $A_{ij}$ and $B_{ij}$ distinguishes ship motions from a simple mass-spring-damper: the surrounding water adds inertia and removes energy through radiated waves, and both effects depend on how fast the hull surface is oscillating.

The solution at each frequency gives the RAO, which is the complex ratio of response amplitude to wave amplitude:

where $a$ is the wave amplitude. For heave, surge, and sway the RAO has units of m/m; for roll, pitch, and yaw it has units of deg/deg (or rad/rad). A well-designed merchant ship has heave and pitch RAOs approaching 1.0 at long wave periods (the ship follows the wave surface), falling rapidly at short wave periods (the ship doesn’t have time to respond).

Encounter frequency and wave heading

A ship moving at speed $V$ through a regular wave train of frequency $\omega$ (rad/s) and wavenumber $k = \omega^2/g$ (deep water) does not encounter waves at the wave frequency. The frequency at which the ship passes wave crests depends on its speed and heading:

where $\mu$ is the wave heading angle measured from the ship’s bow (0° is head sea, 90° is beam sea, 180° is following sea), and the deep-water dispersion relation $k = \omega^2/g$ has been used. The encounter period is $T_e = 2\pi / \omega_e$.

The heading dependence divides into three regimes:

Head sea ($\mu = 0°$): $\omega_e > \omega$. The ship moves into oncoming waves, shortening the apparent wave period. A 12-second wave encountered by a vessel doing 15 knots in head seas appears at approximately 7.5 seconds encounter period, well within the pitch resonance range for many merchant ships. Head seas drive the largest pitch and heave responses.

Beam sea ($\mu = 90°$): $\cos 90° = 0$, so $\omega_e = \omega$. Ship speed doesn’t change the encounter frequency in beam seas. The unmodified wave period acts directly on roll. This is why a vessel running along a wave front in beam seas faces the worst synchronous roll risk at the true wave period.

Following sea ($\mu = 180°$): $\omega_e = \omega - ({\omega^2}/{g})V$. The ship is running with the waves; the encounter frequency is lower than the wave frequency. At sufficient speed (surf-riding condition), $\omega_e \to 0$: the ship matches the wave speed and surfs on the crest. This is the regime where heavy weather operations includes the surf-riding and broaching hazard, because at near-zero encounter frequency the wave excitation is sustained and directional control can be lost.

Note: in following seas $\omega_e$ can become negative if the ship speed exceeds the wave phase speed. Negative encounter frequency means the ship overtakes waves, and the apparent wave direction reverses.

The wave encounter period calculator computes $T_e$ for any combination of wave period, ship speed, and heading.

Roll motion

Roll is the dominant large-amplitude motion for most merchant ships. It is lightly damped, has a natural period in the 10 to 28-second range, and can reach angles of 30 to 45 degrees in resonant conditions. Roll drives container-securing loads, crew safety, cargo damage, and stability margins.

Natural roll period

The natural roll period determines the wave conditions that produce resonant roll. Two equivalent forms appear in the literature.

The rational form, derived from first principles, is:

where $k_{xx}$ is the transverse radius of gyration of the ship about the roll axis (including added mass) and $GM$ is the transverse metacentric height. The transverse radius of gyration is typically 0.35 to 0.42 times the ship breadth $B$ for loaded cargo ships, but varies with loading condition and cargo distribution.

The empirical Weiss approximation, widely used for quick calculations, replaces the ratio $k_{xx}/\sqrt{g}$ with the empirical coefficient $C$:

where $C$ is approximately 0.373 + 0.023$(B/d)$ $- 0.043(L/100)$ for merchant vessels, with $B$ in metres, $GM$ in metres, and $T_\phi$ in seconds. In practice $C$ ranges from about 0.73 to 0.85 for most cargo vessels in the loaded condition. The natural roll period calculator uses this formula.

Typical natural roll periods by ship type in the loaded condition:

A vessel with high $GM$ (short $T_\phi$) is called “stiff” and produces a snapping, uncomfortable roll that imposes high accelerations on cargo and crew. A vessel with low $GM$ (long $T_\phi$) is called “tender” and responds slowly but can spend more time in resonance with longer ocean swells.

Synchronous roll

Synchronous roll occurs when the encounter period $T_e$ equals the natural roll period $T_\phi$. At this condition the wave excitation frequency equals the ship’s natural roll frequency; energy input from the wave field reinforces each roll oscillation and the amplitude grows until limited by damping. In heavy weather (JONSWAP or Pierson-Moskowitz wave spectra centred near 10 to 15 seconds significant period), synchronous roll typically produces 20 to 35 degree single-amplitude roll angles for undamped or lightly damped vessels.

The operational remedy is straightforward: change heading or speed to shift $T_e$ away from $T_\phi$. IMO MSC.1/Circ.1228 (2007), Revised guidance to the master for avoiding dangerous situations in adverse weather and sea conditions, requires officers to be familiar with the vessel’s natural roll period and the conditions that produce resonance.

Parametric roll

Parametric roll is a nonlinear instability mechanism that doesn’t require synchronous excitation. It arises because in head and following seas, the wave-induced change in the submerged hull form modulates the ship’s instantaneous $GM$ periodically. When this modulation has a frequency of twice the natural roll frequency, a small disturbance in roll is amplified geometrically: each half-oscillation, the restoring moment is increased when the ship is near upright (where roll energy is highest) and decreased when the ship is heeled (where restoring moment matters less). The gain condition per half-period is:

Three additional conditions must be met simultaneously: the wave length is close to ship length (so the GM modulation is maximal), the wave height exceeds a ship-dependent threshold (roughly $H_{1/3} \geq 0.02L$ for many container ships), and the roll damping is low enough that the energy gain per cycle exceeds the damping loss.

Large bow-flare and stern-overhang hull forms, typical of modern container ships, produce the strongest GM modulation in head and following seas. This is why parametric roll has become a dominant loss mechanism for container ships rather than for tankers or bulk carriers, which have fuller midship sections.

Build-up of parametric roll is rapid: from less than 2 degrees roll to 35 degrees in as few as 5 to 8 oscillations has been documented in model tests and incidents. The APL China, a C11-class post-Panamax container ship, experienced parametric roll in October 1998 while crossing the northern Pacific during Typhoon Babs, reaching 35 to 40-degree roll angles and losing 406 containers, with roughly 1,000 more damaged. The ONE Apus on 30 November 2020 encountered 5 to 6 metre swells 1,600 nm northwest of Hawaii; the vessel rolled beyond 25 degrees in parametric resonance, losing 1,841 containers overboard with 983 more damaged (Japan Transport Safety Board investigation report). These two incidents, along with the MSC Zoe loss of 342 containers north of the Wadden Islands on 1 January 2019 (Dutch Safety Board report, June 2020), drove the accelerated development of the second-generation intact stability criteria at IMO.

IMO MSC.1/Circ.1627 (10 December 2020), Interim Guidelines on the Second Generation Intact Stability Criteria, addresses parametric roll as one of five dynamic stability failure modes. The Level 1 vulnerability check for parametric roll examines whether the ship’s $GM$ in still water differs from the minimum $GM$ in a wave equal to ship length by more than 0.1 times the mean $GM$. If it does, Level 2 requires a more detailed assessment of the time-varying $GM$ amplitude, the roll natural period, and the wave encounter frequency range in representative operational conditions.

Roll damping

Roll is the most lightly damped of the six ship motions. The components are:

Wave radiation damping: energy radiated away from the hull as outgoing waves during roll oscillation. For a ship, radiation damping is significant only at frequencies near roll resonance; at the natural roll period typical of merchant ships (10 to 25 seconds), wave radiation damping is modest compared to viscous effects.

Potential flow (form) damping: energy loss due to the pressure field generated by the oscillating hull, separate from friction. Small for typical hull forms.

Bilge keel damping: the dominant mechanism on most merchant ships. Bilge keels are longitudinal fins typically 200 to 500 mm wide and 0.3 to 0.5 times ship length long, welded along the bilge radius. They generate drag as the hull rolls, dissipating roll energy as turbulent wake. The roll decay test (free-decay test per ITTC procedure 7.5-02-05-04) measures the logarithmic decrement of successive roll amplitudes; from this the dimensionless roll damping ratio $\zeta$ is computed. Bare hull values are typically $\zeta = 0.03$ to $0.05$; with bilge keels, $\zeta = 0.07$ to $0.12$ at moderate roll angles (5 to 15 degrees), rising to $\zeta = 0.10$ to $0.18$ at larger amplitudes due to the nonlinear drag.

Active stabilization: discussed in marine stabilisers. Fin stabilisers (active hydrofoils at the bilge) can reduce bare-hull roll amplitude by 80 to 95% at operating speed. Anti-roll tanks (passive or active free-surface tanks tuned to the natural roll period) provide effectiveness at low or zero speed. Gyroscopic stabilizers are common on yachts but rare on commercial vessels.

The standard free-decay test procedure (ITTC 7.5-02-05-04) specifies the initial heel angle, release method, measurement duration (at least 5 complete oscillations), and analysis method (logarithmic decrement with the Froude method for extracting linear and quadratic damping coefficients from an amplitude-dependent curve).

Pitch motion

Natural pitch period

The natural pitch period for a merchant ship in the loaded condition is typically 6 to 12 seconds, shorter than the natural roll period for three reasons: the longitudinal radius of gyration $k_{yy}$ is smaller than the transverse $k_{xx}$ (the ship mass is more concentrated near amidships longitudinally), the longitudinal metacentric height $GM_L$ is 20 to 100 times larger than the transverse $GM$, and waterplane area moment about the transverse axis is large.

The rational form is:

For practical purposes, $GM_L \approx BM_L = I_L / V$ where $I_L$ is the second moment of the waterplane area about the transverse axis and $V$ is the displaced volume.

Pitch is excited predominantly in head and following seas. The response peaks when $T_e$ approaches $T_\theta$. Most ocean swell periods are 8 to 18 seconds; a ship in head seas shortens the encounter period, so pitch resonance is possible in moderate head-sea conditions.

Pitch-heave coupling

Pitch and heave are tightly coupled. As the bow rises in a wave trough, the change in waterplane area forward reduces the upward buoyancy restoring force and induces a bow-down pitch moment. Conversely, as heave descends, the pitch excitation is modified by the instantaneous trim. The coupled two-DOF system produces two natural modes: one where heave and pitch are in phase (the “heave mode”) and one where they are out of phase (the “pitch mode”). In strip theory the coupling appears as off-diagonal terms in both the added mass and restoring matrices.

The Salvesen-Tuck-Faltinsen (STF) strip theory, published in Transactions of SNAME vol. 78 (1970) pp. 250-287, remains the most widely used linear seakeeping method precisely because it captures this pitch-heave coupling correctly for slender hulls in head and following seas.

Slamming

Slamming is the impulsive bottom impact that occurs when the bow emerges from the water during a large pitch-heave cycle and re-enters with high relative velocity. Ochi (1964) established the two-condition criterion that is still the standard: (i) the relative vertical motion at a forward station (typically 0.1L from the bow) must exceed the local draft, and (ii) the relative vertical velocity at re-entry must exceed a threshold value $v_{th}$, typically 0.093$(gL)^{0.5}$ m/s for conventional displacement vessels.

Impact pressures in severe slamming reach 10 bar or more in full-scale measurements, causing plastic deformation of forward bottom plating, frame cracking, and damage to bow thruster structures. The probability of slamming per unit time is computed from the joint probability distribution of relative displacement and velocity, using the RAOs for heave and pitch. The slamming probability rises steeply above Beaufort 7 in head seas for vessels with fine bow forms; masters typically begin reducing speed at $P_{slam} > 0.03$ per encounter minute, consistent with the NORDFORSK 1987 operability criterion.

Heave motion

Heave is the vertical translation of the ship’s centre of gravity. In moderate seas, heave is the dominant contribution to vertical motion amidships; at the bow and stern, pitch acceleration adds algebraically to the heave acceleration to give the total vertical acceleration at any deck point.

The natural heave period depends on the waterplane area $A_{wp}$ and the ship displacement $\Delta$:

where $A_{33}(\omega_z)$ is the frequency-dependent added mass in the heave direction and $\rho$ is water density. For most merchant ships, $T_z$ is 6 to 12 seconds, similar to the natural pitch period.

Heave resonance is generally less dangerous than roll resonance for two reasons: heave damping from wave radiation is much larger (the flat bottom of a ship radiates energy efficiently), and the heave amplitude at resonance is self-limiting because it cannot exceed the wave amplitude at the resonant frequency. Still, large heave amplitudes in shallow water reduce the underkeel clearance dynamically, compounding the static squat effect calculated at the vessel’s forward speed.

Yaw motion

Yaw is the rotation about the vertical axis. In a seaway it is driven by the asymmetric wave forces on the hull when waves arrive at an oblique angle. For a vessel on autopilot, yaw is continuously corrected; for a vessel with manual steering in confused seas, yaw excursions of 5 to 15 degrees are normal and reduce the effective service speed by lengthening the track.

The critical yaw-related failure mode is broaching. In a following sea, if the vessel surfs on a wave crest and the wave phase speed exceeds the vessel’s steering speed, the rudder loses authority and the hull yaws uncontrollably. The bow swings broadside to the wave direction, converting the following-sea geometry to an effective beam sea with the vessel heeled and the wave at maximum roll-excitation angle. Broaching is the mechanism behind most following-sea capsizes of small vessels; for large merchant ships it is less common but documented for vessels caught in extreme following seas at high speed.

IMO MSC.1/Circ.1627 addresses surf-riding and broaching as one of the five second-generation stability failure modes, alongside pure loss of stability, parametric roll, dead ship condition, and excessive acceleration. The Level 1 surf-riding check compares the Froude number to a threshold of 0.3; above this, Level 2 requires calculation of the surf-riding probability in a prescribed sea state.

Sway and surge

Sway (transverse translation) and surge (longitudinal translation) are the two motions with no hydrostatic restoring force: there is no buoyancy mechanism that returns the ship to a neutral sway or surge position. In the open sea this means sway and surge are driven purely by wave excitation and decay only through radiation damping, giving large low-frequency responses in beam and following seas respectively.

For seakeeping analysis of a ship underway, sway and surge are generally the smallest contributors to crew or cargo risk because: ship speed dominates the surge signal, and sway is counteracted by steering. They become critical in two contexts: moored vessels (where sway and surge are the principal load components on mooring lines and fenders, and no speed damping exists) and offshore vessels maintaining dynamic positioning (where sway and surge drive the thruster demand and fuel consumption).

The Response Amplitude Operator

The RAO is the key output of a seakeeping analysis. It is a complex transfer function: its modulus gives the amplitude ratio and its argument gives the phase of the response relative to the wave crest at the ship’s centre of gravity.

Formally, for a linear time-invariant system:

The response spectrum for any motion is obtained by convolving the RAO with the encounter wave spectrum:

The significant single-amplitude response (analogous to significant wave height) is $2\sqrt{m_0}$ where $m_0 = \int_0^\infty S_{response}(\omega_e) \, d\omega_e$ is the zeroth spectral moment. Higher spectral moments give the mean zero-crossing period, mean crest period, and bandwidth of the response.

In practice, RAOs are computed either by strip theory (adequate for most slender merchant ships in head and following seas) or by 3D panel methods (better for full-form hulls, beam seas, and off-resonance behaviour). For the forward-speed problem, the Ogilvie-Tuck correction to the STF theory improves accuracy at moderate Froude numbers. CFD with free-surface modelling (volume-of-fluid Reynolds-averaged Navier-Stokes) is used for research and for extreme-motion predictions (parametric roll, slamming, green water) where linear theory breaks down.

ITTC Recommended Procedures and Guidelines 7.5-02-05-04 (seakeeping experiments) specify the model test procedure for measuring RAOs: the test matrix covers at least five headings (0°, 45°, 90°, 135°, 180°) and at least eight wave frequencies spanning 0.3 to 2.0 $\omega_\phi$, with regular wave tests for RAO measurement and irregular wave tests for response spectrum validation.

Motion-induced effects

Cargo and lashing loads

The combined heave, roll, and pitch accelerations at any point on the ship produce inertia forces on cargo. The Cargo Securing Manual (CSM), required by SOLAS Regulation VI/5.6 for cargo ships of 500 GT and above, is calibrated to the vessel’s design motions. The standard design accelerations from IACS Unified Requirement S11 (2003 edition) are:

• Transverse: $a_t = 0.98(1.5/\sqrt{L} + V/\sqrt{L^3} + 0.6)\sin(\phi)$ m/s$^2$
• Longitudinal: $a_l = 0.70 \cdot g / L^{0.5}$ m/s$^2$
• Vertical: $a_v = 9.81 \pm g(0.7 L^{-0.5})$ m/s$^2$

Container ships operate with prescribed stack weights, stack heights, and lashing configurations that collectively define the approved stability and securing range. Exceeding the design acceleration envelope, as occurred in the MSC Zoe on 1 January 2019 in the 10-metre wave conditions north of the Wadden Islands, can cause simultaneous lashing failures across multiple bays.

Green water on deck

Green water occurs when a wave crest reaches above the deck edge, depositing a solid water mass on the exposed deck. This is distinct from spray: a green-water event deposits hundreds of tonnes on the foredeck in a few seconds, producing dynamic loads on hatch covers, forward deck equipment, and container stacks. The green-water probability is computed analogously to slamming probability, from the joint distribution of relative wave surface elevation and freeboard; the NORDFORSK 1987 criterion specifies a threshold of 0.05 green-water occurrences per encounter minute for cargo ships.

Freeboard is the primary design defence against green water; see freeboard and reserve buoyancy for the ICLL 1966 calculation. Bow form also matters: flared bow sections shed water more effectively than plumb stems.

Propeller emergence

When the stern rises in heavy pitching, the propeller shaft may emerge from the water surface. During emersion the thrust drops to near zero and the propeller over-speeds; on re-immersion the sudden reloading produces a torque spike at the crankshaft. Repeated emergence-submergence cycles cause fatigue damage to the propeller, shaft, gearbox, and engine couplings. For typical container ships and bulk carriers, significant propeller emergence risk begins at significant wave heights of 4 to 5 metres in head seas at or near design service speed. The marine propeller article covers the design constraints on propeller tip submergence.

Motion sickness incidence

The motion sickness incidence (MSI) is the percentage of a naive (unacclimatized) crew or passenger population expected to vomit within a 2-hour exposure to the vessel’s motions. O’Hanlon and McCauley (1974) established the empirical model in Aerospace Medicine vol. 45 pp. 366-369: MSI peaks at vertical acceleration frequencies of 0.167 Hz (10 cycles per minute), which falls squarely within the heave period range of most ships. At 0.167 Hz the threshold vertical acceleration for 10% MSI is approximately 0.03 g RMS; at 0.05 g RMS the predicted MSI exceeds 30%.

ISO 6954:2000 addresses the related problem of structural and habitability vibration (1 to 80 Hz frequency range, covering propeller, engine, and hydrodynamic excitation), specifying frequency-weighted RMS velocity limits: for passenger cabin areas, adverse comments are probable above 4 mm/s RMS and not probable below 2 mm/s RMS. Ship motions in waves (below 0.5 Hz) fall outside the ISO 6954 frequency range; the NORDFORSK 1987 criteria cover this low-frequency regime.

Voluntary speed reduction

Masters routinely reduce speed in heavy weather to keep motions within acceptable limits: to prevent slamming damage, to protect deck cargo, to maintain crew effectiveness, and to satisfy charter party requirements. The speed reduction is typically 30 to 60% of calm-water service speed. For a 15-knot container ship, a 40% reduction to 9 knots in Beaufort 8 is representative.

The consequences cascade. A voyage delayed by heavy weather at reduced speed adds fuel burn (the engine runs longer), misses port windows, and reduces the annual DWT-miles denominator in the CII calculation. For CII compliance, the operational speed reduction is a real cost: MARPOL Annex VI Regulation 28 (the CII mechanism, in force 1 January 2023) measures the ratio of annual CO2 emissions to annual transport work; a week’s speed reduction in a 12-knot North Atlantic crossing in winter affects the annual CII rating.

Seakeeping criteria and operability

NORDFORSK 1987 thresholds

The NORDFORSK 1987 Assessment of Ship Performance in a Seaway (Nordic Co-operative Project, 1987) provides the most widely used operability criteria for merchant ships. The general limiting criteria for cargo ships are:

Exceeding any criterion doesn’t mean the vessel is unsafe, but it does mean the human or cargo system associated with that criterion is operating outside acceptable conditions. In practice, vessels with operability below 70% of sailing days on a given trade route for the seasonal sea state distribution are considered to have inadequate seakeeping for that service.

Second-generation intact stability criteria

IMO MSC.1/Circ.1627 (10 December 2020) represents the current interim framework for the five dynamic stability failure modes. Each mode is assessed at three levels of increasing rigor:

• Level 1: simple vulnerability check using hull form parameters. Fast, conservative. A vessel passing Level 1 needs no further assessment for that mode.
• Level 2: intermediate check using wave-based criteria, requiring calculation of the time-averaged $GM$ variation in regular waves or numerical integration over a wave scatter diagram.
• Direct Stability Assessment (DSA): time-domain simulation in irregular waves, the highest accuracy tool but requiring specialist software and significant computational time.

The five failure modes and their governing physics:

The Interim Guidelines remain voluntary pending full adoption. Mandatory application is under discussion at IMO’s Sub-Committee on Ship Design and Construction (SDC). Several flag states, including the Marshall Islands and Liberia, have issued advisory circulars encouraging voluntary application for new designs.

Seakeeping in design and class rules

IACS Common Structural Rules for Bulk Carriers and Oil Tankers (CSR BC&OT, 2024 edition) specify design wave-induced loads for structural sizing using North Atlantic scatter diagrams and a 25-year return period. The longitudinal wave bending moment, shear force, and torsional moment used for scantling calculation all depend on the ship’s motion RAOs in the design wave conditions. DNV Rules for Classification (2024 edition), Part 5, give the equivalent procedure for vessels not covered by CSR.

Calculation methods

Strip theory

Strip theory divides the hull into transverse strips (typically 11 to 21 stations between forward and aft perpendiculars) and solves the 2D potential flow problem for each strip to obtain the frequency-dependent added mass and damping coefficients. These are integrated along the ship length, with corrections for forward speed and end effects, to give the 3D motion equations. The STF method (Salvesen, Tuck, and Faltinsen, 1970) remains the industry benchmark for slender displacement hulls in head and following seas up to moderate wave heights. It is accurate for heave and pitch RAOs to within 10 to 15% for Froude numbers below 0.35; for roll, the additional viscous damping from bilge keels must be added as a supplementary empirical term because potential flow alone underestimates roll damping.

Software implementing strip theory includes NAPA Motions, SHIPMO (BMT Fluid Mechanics), and SEAKEEPER (Bentley/Maxsurf). The open-source code NEMOH implements the 3D frequency-domain boundary element method (panel code) and has been validated against WAMIT for offshore structures.

3D panel methods

For vessels with fuller hull forms (block coefficient $C_b > 0.75$), for beam seas, or for operation at low Froude numbers where the ship’s waterplane becomes important, 3D panel methods provide better accuracy. These methods solve the linearized free-surface potential flow problem on a panelled representation of the hull (typically 500 to 3,000 panels). WAMIT (version 7.x) and ANSYS Aqwa are the most widely used commercial codes. Computation time is minutes to hours, depending on panel count and frequency range.

Time-domain and CFD methods

For large-amplitude motions, parametric roll, green water, and slamming, frequency-domain linear methods fail because the problem is nonlinear. Time-domain methods, from Rankine panel codes with the nonlinear body boundary condition to full RANS CFD with free-surface capturing (volume-of-fluid methods), are used for these problems. Full RANS seakeeping simulations for a container ship in irregular following seas require 500 to 5,000 CPU-hours per simulation; this limits routine use to research and high-value design studies.

Limitations

The material in this article describes idealized and linear-theory behaviour. Several important limitations apply in practice:

Linear RAO validity: RAOs are strictly valid only for small-amplitude responses where the superposition principle holds. For roll amplitudes above approximately 10 degrees, nonlinear restoring (variation of $GM$ with heel) and nonlinear damping (drag from bilge keels is quadratic in velocity) become significant. RAO-based predictions overestimate roll resonance amplitudes for stiff ships and may underestimate them for tender ships where large-angle stability reduction is relevant.

Wave spectrum assumptions: the Pierson-Moskowitz and JONSWAP spectra assume wind-generated seas in equilibrium or fetch-limited conditions. Swell components, bimodal spectra (wind sea plus crossing swell), and directional spreading all affect the motion response in ways that are not captured by a single-peaked unimodal spectrum.

Strip theory for blunt hulls: strip theory loses accuracy for hull forms with $C_b > 0.78$, for vessels at Froude numbers above 0.40, and in beam and quartering sea conditions where 3D effects are strongest. Full-form tankers and bulk carriers in beam seas require 3D panel corrections.

Shallow water: all formulas in this article assume deep water ($kd \gg 1$). In water depths less than half the wave length, the wave speed and phase velocity change, the dispersion relation becomes $\omega^2 = gk\tanh(kd)$, and the encounter frequency formula must use the depth-dependent wavenumber. Shallow-water seakeeping is important for vessels navigating estuaries, rivers, and port approaches.

Parametric roll in confused seas: the encounter frequency condition for parametric roll is easily met in irregular seas where the spectral peak has a range of frequencies. Whether the roll actually builds to large amplitudes depends on the duration of the resonant condition and the degree of spectral concentration near the critical frequency. Time-domain simulation in irregular seas is more conservative than the deterministic encounter-frequency check alone.

IMO criteria development: MSC.1/Circ.1627 explicitly states that the Interim Guidelines carry “a certain degree of uncertainty in the recommendations developed.” The DSA procedures in Part B of the circular are still subject to refinement at IMO SDC sessions. Designers applying the criteria to novel hull forms (e.g., ultra-large container ships above 400 m, trimaran ferries, unconventional stern forms) should treat Level 1 and Level 2 outcomes with engineering judgment rather than as hard pass/fail gates.

See also

Related wiki articles:

• Seakeeping (operability analysis methods, polar diagrams, and route-specific performance)
• Marine stabilisers (fin stabilisers, anti-roll tanks, gyroscopes)
• Heavy weather operations (bridge procedures, avoidance criteria)
• Intact stability (the 2008 IS Code criteria, weather criterion, stability booklet)
• Metacentric height (GM calculation from keel to metacentre)
• GZ curve and righting arm (large-angle static stability)
• Freeboard and reserve buoyancy (ICLL 1966 freeboard assignments)
• Cargo securing manual (lashing design loads from ship motions)
• Squat effect (dynamic underkeel clearance compound with heave)
• Ship resistance and powering (added resistance in waves, Markov-type R_AW models)
• Marine propeller (propeller submergence limits and emergence loads)
• Container ship (parametric roll susceptibility of large container ships)
• Bulk carrier (roll characteristics and cargo shifting)
• LNG carrier (sloshing interaction with ship motions)
• What is CII (speed reduction effects on annual CII rating)
• Weather routing (routing systems that minimize motions and fuel)
• Damage stability (residual stability after flooding affects roll period)

Calculators:

• Natural roll period calculator
• Wave encounter period calculator
• Parametric roll vulnerability calculator
• Deck wetness calculator
• Heave RAO calculator
• Seakeeping criterion check
• Metacentric height (GM) calculator
• GZ curve calculator
• STAWAVE-2 added resistance in waves