Ship Stability and the DCM Framework: A Review of Advances and Integration

https://doi.org/10.65613/740883

First Author: Chunlian Luo¹; Second Author:Lieqi Zhang¹

1.Navigation College, Xiamen Ocean Vocation College, Xiamen 361000, China

Corresponding author: Chunlian Luo, clianluo@163.com

Abstract: Ship stability is a key aspect of maritime safety. Research in this area has moved from meeting basic rules to managing safety throughout a ship’s life. This paper reviews current methods for improving ship stability. It examines how stability theory has developed and proposes a framework based on design, control, and management. The design part examines hull form and reliability-based optimization. The control part covers roll damping and active anti-heeling systems. The management part discusses real-time monitoring and risk warning using digital twins and AI. The paper also points out the limits of existing methods and suggests future research directions.

Keywords: Ship Stability; Design-Control-Management Framework; Second Generation Intact Stability Criteria; Digital Twin; Risk-Based Design

1. Introduction

Ship stability is a vessel’s ability to return upright after heeling. It is a basic issue in ship design and operation. Major accidents like the sinking of the Titanic, the capsize of the Herald of Free Enterprise, and the listing of the Costa Concordia show how complex stability problems can be. Studies show that stability failures are a major cause of maritime casualties and damage^[1]. These events have changed how people think about stability. It is no longer seen as a fixed design feature but as a condition that depends on design, operation, environment, and management.

The shipping industry is changing due to globalization, decarbonization, and digitalization. These changes bring new challenges. Larger ships like ultra-large container ships and cruise ships have higher centers of gravity and larger wind areas, which reduce stability margins^[2]. New fuels such as LNG and hydrogen create free surface and phase-change issues that traditional models cannot handle well^[3]. Arctic routes also expose ships to ice, cold, and rough seas, requiring better assessment tools^[4].

In response, the IMO introduced the Second Generation Intact Stability Criteria. This shifts stability assessment from static rules to a dynamic, probabilistic approach^[5]. It requires direct simulation and probability evaluation of five failure modes. The ship is treated as a nonlinear system interacting with a random environment. This means stability improvement must go beyond design and cover control and management as well. This paper reviews modern stability technologies and proposes a Design-Control-Management framework to help improve ship safety and efficiency together.

2. Literature Review and Theoretical Foundations

2.1 Evolution of Theoretical Paradigms in Ship Stability Assessment

Stability assessment theories have developed over time in a fairly clear direction. The field has moved from simple static checks to dynamic analysis, and now to a broader view that treats safety as a property of the whole system.

Traditional stability theory uses the static righting lever curve. This curve shows how the righting moment changes as the ship heels in calm water. The main parameters are metacentric height, maximum righting lever, stability range, and area under the curve. For decades, international rules have used these parameters to set fixed standards, such as minimum area under the curve and weather criteria based on equivalent wind pressure^[6]. This approach is simple and easy to apply. It gives a clear baseline for design safety. But it has a major weakness. It turns real wave action into an equivalent static heeling moment. As a result, it cannot capture nonlinear behavior like large roll angles, changing righting levers, or dynamic instability such as pure loss of stability and broaching. Static criteria can tell us about a ship’s basic stability, but they cannot predict how safe it will be in rough seas.

In the mid-to-late 20th century, several ships capsized suddenly in heavy seas. Fishing vessels and ro-ro ships were especially affected. These incidents showed that dynamic effects matter a great deal in stability failure. Research then turned to mechanisms like parametric rolling and pure loss of stability^[7][8]. Around the same time, probabilistic methods entered damage stability assessment. The SOLAS 2009 rules, for instance, use a probabilistic approach to calculate a ship’s average survival probability across all possible damage cases. This changed stability safety from a pass-or-fail question to a question of how much risk is acceptable. It also made it easier to handle uncertainties like sea state and loading.

The IMO Second Generation Intact Stability Criteria are the current state of the art. They cover five dynamic failure modes: dead ship condition, parametric rolling, pure loss of stability, surf-riding, and excessive acceleration. The key feature is the Direct Assessment Method. This builds a detailed model of the ship and its environment and runs nonlinear time-domain simulations to compute ship motions in representative sea states. The probability of reaching a dangerous state, such as excessive roll or loss of control, is then estimated from the simulation results. The model must include many components: six-degree-of-freedom motion equations, nonlinear restoring forces, wave forces (from potential flow or CFD), rudder and propeller forces, and random wind forces. In this approach, stability becomes a risk assessment problem for a nonlinear system with random inputs. Recent reviews have summarized the numerical methods used for direct assessment of the five failure modes^[9]. These include potential flow, viscous flow, hybrid methods, and intelligent computation. Potential flow and viscous flow methods work well for most modes. Intelligent methods and hybrid approaches also show promise. This new approach has several practical implications. It requires detailed simulation of failure physics rather than empirical formulas. It treats environmental and operational randomness as part of the problem. And it requires any stability improvement to be tested within this system framework. This pushes the field away from piecemeal fixes and toward system-level integration. The DCM framework proposed in this paper is a response to this need. It turns the theoretical approach into a practical strategy covering design, control, and management.

2.2 Design-Level Approaches

Design is the best place to build in stability. It costs less to address stability at the design stage than to fix problems later. Modern stability optimization has moved from experience-based adjustments to more systematic, quantitative methods based on multiple disciplines and reliability analysis.

A ship’s basic stability comes from its main dimensions, hull form, and layout. The static righting lever curve helps explain how these parameters work. Ship breadth affects initial metacentric height the most. It determines how quickly the righting moment builds up at small angles. But too much breadth can slow the ship and hurt seakeeping. Moulded depth and freeboard mainly affect stability at large angles and the overall stability range. Enough freeboard gives the reserve buoyancy needed to prevent capsizing. The waterplane area coefficient and the shape of the transverse section affect how the waterplane moment changes as the ship heels, which shapes the righting lever curve. Good flare design can give more righting moment at moderate and large angles. The vertical center of gravity is the most critical parameter. Lowering it directly improves the righting lever at all angles. This is the most effective way to increase stability. So modern design pays close attention to keeping the center of gravity low through layout and material choices.

Meeting the Second Generation Criteria requires a different approach to design optimization. Traditional optimization uses fixed limits like minimum metacentric height. Reliability-based design optimization goes further. It uses the probability of stability failure, such as the chance of parametric rolling, as a constraint or target[10]. This means that time-consuming nonlinear stability simulations must be run inside the optimization loop. This has led to the use of surrogate models like Kriging and intelligent algorithms like Bayesian optimization. Stability must also be balanced against other goals that often conflict with it, such as speed, seakeeping, cargo capacity, structural strength, and cost. Multidisciplinary design optimization tries to find the best compromises in this complex space. One example is the trade-off between increasing flare for better stability and the resulting rise in wave resistance.

The design process uses CFD, potential flow theory, and time-domain simulation to satisfy dead ship stability requirements^[11]. Design variables have been extended to include bilge keels and superstructure shape. By adjusting the waterplane shape and flare carefully, an improved righting lever curve was achieved in the critical angle range. This met the dynamic stability requirements while keeping the Energy Efficiency Design Index in check. This case also shows some deeper issues. One is the safety margin paradox. Setting very high stability margins for extreme conditions may hurt the ship’s efficiency throughout its life. Another is model reliability. It is not clear how accurate the hydrodynamic models are under extreme conditions. Future work should focus on developing accurate but fast simulation tools. A closed loop connecting design, operation, and redesign should also be built, so that real operational data can inform future design improvements.

2.3 Control-Level Approaches

Once a ship is at sea, its stability keeps changing. Loading, navigation, and weather all affect it. The main job at this stage is to keep track of the ship’s state and step in when needed, using decision tools and control systems to stay within safe limits.

Modern stability management has grown from simple loading manuals to intelligent decision support systems. These systems solve optimization problems with multiple goals and limits. The decision variables cover how cargo is distributed, where it is placed, and how it is used during the voyage. The goals include stability margins, operating costs, fuel use, and port turnaround time. All this must be done while meeting stability, strength, draft, and lashing requirements. Solving this requires optimization algorithms, real-time data fusion, and accurate stability models. One recent development is Model Predictive Control. This method uses weather forecasts for the coming voyage to plan loading and ballast adjustments over a moving time window. This allows the system to maintain stability in advance. For example, it can raise the metacentric height if strong winds are expected^[12]. For navigation support in complex waters, recent reviews have summarized the relevant theories and methods^[13].

Active control systems use devices like fin stabilizers, rudders, and anti-heeling pumps to counter outside forces. These systems work at two levels. At the basic level, fin stabilizers use PID control to reduce periodic rolling, which improves comfort and safety. At the more advanced level, the goal is to prevent large heel angles. The system aims to stabilize and restore the heel angle. Strategies include combining rudder and fin control or coordinating multiple devices, as on dynamic positioning vessels. Nonlinear or optimal control algorithms are used to handle system nonlinearities and save energy^[14]. For active anti-heeling tanks, the control system must also deal with the nonlinear behavior of large fluid sloshing.

Keeping stability under control throughout a voyage needs a control-management setup with two layers. The decision layer uses mission plans and forecasts to set daily loading and stability targets. The execution layer carries out closed-loop control based on real-time sensor data, second by second, to keep the ship within safe limits. The main challenges are standardizing data exchange between systems, switching smoothly between automatic and manual control, and defining who does what in abnormal situations. Too much automation may reduce crew awareness and skills, which can create new risks.

The Ship Energy Efficiency and Stability Intelligent Management and Control System used by COSCO SHIPPING is an example of this approach^[15]. It uses weather routing to adjust ballast and optimize metacentric height and roll period. This creates a loop of advance planning and real-time monitoring. This case also raises new safety questions for intelligent shipping. It is not clear whether highly integrated software systems go through the same rigorous checks for reliability, cybersecurity, and failure modes as hardware systems. This is a basic issue that needs attention as stability assurance becomes more digitalized.

2.4 Management-Level Approaches

Ship stability management is moving away from simple monitoring and alerting toward predictive health management. The basic idea is to use sensor data, models, and AI to build a digital copy of the physical ship. This digital twin runs in real time and can simulate what might happen next. It helps detect risks early and supports better decision-making.

The digital twin is an important tool for stability risk management. It has three layers. The data layer collects real-time data from different sources, such as motion sensors, tank levels, stress gauges, weather data, and equipment status. The model layer is the core. It brings together models from different areas, including geometry, hydrostatics, ship motion, hydrodynamics, structural response, and control systems. A set of models is built to balance accuracy and speed for different tasks, such as routine monitoring and extreme condition simulation. The service layer provides functions like stability display, data analysis, scenario simulation, and risk warnings. A major challenge is keeping the models up to date. An online calibration method based on data assimilation is needed. This would let the digital twin learn from real measurements and track how the ship’s performance changes over time, making it more accurate and reliable^[16].

AI provides useful tools for handling the complex data generated by digital twins. Supervised learning can link key stability parameters to easy-to-measure motion variables, allowing soft sensing of these parameters. Unsupervised learning and anomaly detection can spot hidden risks such as cargo shifting or flooding by finding deviations from normal behavior. Reinforcement learning can learn decision-making strategies for anti-heeling control or route planning in uncertain conditions. Physics-informed neural networks include physical laws in the training process, which helps them make more consistent and reliable predictions, even when data is limited^[17]. This approach is gaining interest in ocean engineering^[18].

Warning systems should move beyond simple on-off alerts and provide probabilistic risk information. Uncertainty analysis can show how input uncertainties, like sea state or model errors, affect output risks such as capsize probability. Safety limits should not be fixed numbers. They should be probability contours that change with sea state, loading, and speed. The system should assess in real time how close the ship is to high-risk zones and give graded warnings with confidence levels.

The SAVE CUBE system from Deltamarin and Rolls-Royce provides probabilistic warnings. It uses radar data to estimate sea state and runs short-term motion simulations^[19]. An intelligent platform from the Shanghai Ship and Shipping Research Institute uses knowledge graphs to diagnose specific risks^[20]. These examples also show a basic problem. Generating a warning is one thing, but making sure it leads to action is another. False alarms can make crews ignore warnings. Too much complex information can overwhelm crews in an emergency. Future system design should focus on human-machine interaction. Through clear displays, simple alarm strategies, and clear responsibilities, humans should remain the most capable and ultimately responsible part of the safety system.

3. 3. Conclusion

This review has brought together work on ship stability around a central point. Stability is not something you check once at the design stage and then forget. It keeps changing during the ship’s life and needs ongoing attention through design, control, and management. The review also points out trade-offs at each stage. Changes that improve stability often reduce efficiency. More control systems add complexity. More automation can reduce the crew’s involvement. And models never fully match what happens at sea. Future work should look at better theory for extreme conditions, multi-physics control, certification of AI systems, lifecycle resilience, and ship-shore integration. The DCM framework can serve as a common reference for regulators, designers, operators, and technology developers. If these groups work together and take this integrated view, a cycle can be created where design learns from operations, and operations inform better design. That is how ship safety systems can become more robust and adaptable to the uncertainties of the sea.

References

• Angelou, M., Bačkalov, I., Begović, E., et al. Future of ship stability[C]. Proceedings of the International Conference on the Stability and Safety of Ships and Ocean Vehicles (STAB&S 2024), Wuxi, China, October 13-18, 2024: 832-851.
• Fuentetaja-Merino, M., Silva-Campillo, A., Herreros-Sierra, M.A., et al. Structural analysis of the geometric alternatives of double-bottom floor plates of a Panamax-class container ship[J]. Applied Sciences, 2024, 14(22): 10684.
• Yang, Y., Liu, C., Wang, Z., et al. Real-time monitoring-based stability analysis of an extra-large LNG tank roof under construction[J]. Sensors, 2025, 25(8): 2498.
• Deggim, H. The international code for ships operating in polar waters (Polar Code)[M]//Encyclopedia of Maritime and Offshore Engineering. Springer, 2018: 15-35.
• Francescutto, A., Spyrou, K. J., Themelis, N. Pros & cons of second generation intact stability criteria[C]. Proceedings of the International Congress of the International Maritime Association of the Mediterranean (IMAM 2025), Crete, Greece, September 28 – October 2, 2025.
• Sheng, Z., Liu, Y. Principles of Naval Architecture (Vol. II)[M]. Shanghai Jiao Tong University Press, 2004. (in Chinese)
• Ahmed, S., Li, T., Huang, et al. Dynamic and quantitative risk assessment of cruise ship pod propulsion system failure: An integrated Type-2 Fuzzy-Bayesian approach[J]. Ocean Engineering, 2023, 279: 114601.
• Chung, J., Shin, D. M., Kim, W.-D., et al. Current status of the 2nd generation of intact stability: Investigation of the pure loss of stability and parametric roll mode[J]. Journal of the Society of Naval Architects of Korea, 2020, 57(2): 95-106.
• Gu, M., Chu, J. L., Zhang, P. J., et al. State-of-the-art and prospects of direct stability assessment methods for second-generation intact stability criteria[J]. Chinese Journal of Ship Research, 2026, 21(1): 3-11.
• Mert, A. C., Pham, T. A., Guo, X. Multi-objective reliability-based design optimization of piles considering soil-structure interaction and soil spatial variability[J]. Computers and Geotechnics, 2026, 192: 107909.
• Suner, M., Bas, M. A new approach to narrow waterways traffic routing with potential flow theory and CFD[J]. Ocean Engineering, 2022, 261: 112158.
• Vergara, D., Johansen, T. A., Sorensen, A. J. A machine learning based Bayesian decision support system for efficient navigation of double-ended ferries[J]. Journal of Ocean Engineering and Science, 2024, 9(6): 605-615.
• Chu, X. M., Zheng, F. J., Liu, C. G. Digital-intelligent navigation architecture and evolution paradigm for autonomous ships in key inland waterway areas[J]. Chinese Journal of Ship Research, 2026, 21(4): 1-16. (in Chinese)
• Perez, T., Blanke, M. Ship roll damping control[J]. Annual Reviews in Control, 2012, 36(1): 129-147.
• Kougiatsos, N., Scheffers, E. L., van Benten, M. C., et al. An intelligent agent-based resilient framework for marine vessel mission adaptations[J]. IEEE Open Journal of Intelligent Transportation Systems, 2025, 6: 184-203.
• Coraddu, A., Oneto, L., Baldi, F., et al. Vessel fuel consumption forecasting with a dual-step approach[J]. Transportation Research Part C: Emerging Technologies, 2019, 107: 310-328.
• Raissi, M., Perdikaris, P., Karniadakis, G. E. Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations[J]. Journal of Computational Physics, 2019, 378: 686-707.
• Koshizuka, T., Sato, I. Understanding generalization in physics informed models through affine variety dimensions[C]. Advances in Neural Information Processing Systems 38 (NeurIPS 2025), 2025: 50177-50215.
• Wang, D., Liang, S., Li, M. Power output estimation of a two-body hinged raft wave energy converter using HF radar measured representative sea states at Wave Hub in the UK[J]. Renewable Energy, 2023, 202: 103-115.
• Chen, K., Chen, S. Y., Zhang, D., et al. Question answering system in the field of ship based on knowledge graph[J]. Chinese Journal of Engineering Mathematics, 2026, 43(1): 183-198.

* Project supported by foundations:2024 Teaching Reform Research and Practice Project (Research and Practical Exploration on Digital Transformation and Intelligent Trend of Ship Inspection Major) and Course Ideological and Political Education Demonstration Project (Ship Structure and Drawing), Xiamen Ocean Vocational College (2024–2026).