Offshore renewable energy deployment faces a fundamental materials challenge: synthetic fibre mooring lines offer an 8-fold weight reduction and tension reduction compared to steel chains, yet their certification requires predicting 20-year operational behaviour from necessarily limited experimental observations. Current approaches employ polynomial functions to represent stress-dependent material behaviour within established viscoelastic frameworks. Still, these rigid mathematical forms constrain material characterisation by forcing complex behaviour into predetermined equations rather than discovering true material physics. This research develops a neural-physics integration framework that replaces polynomial constraints with adaptive neural networks whilst preserving established Schapery-Prony physics principles. The core innovation employs functional decomposition, where neural networks learn stress-dependent nonlinearity functions (g0, g1, g2), while the proven Schapery-Prony framework handles temporal evolution, enabling the discovery of true material behaviour rather than mathematical fitting to predetermined forms.

A systematic five-phase methodology guides development from baseline optimisation through operational validation. Phase 1 employs 13-dimensional Latin Hypercube Sampling across neural architecture, training protocols, and physics parameters. Phase 2 sensitivity analysis identifies optimal methodological choices: Percentage Difference loss functions for scale-invariant learning, Multiple-Steps loading patterns for comprehensive material exercising, and enhanced relaxation spectra for improved temporal coverage. Phase 3 capability testing reveals model limitations through progressive validation scenarios. Phase 4 diagnostic analysis identifies physics model resolution as the primary constraint, leading to targeted improvement from 10 to 14 relaxation terms that address slow relaxation modes critical for long-term accuracy. Phase 5 validates robustness under realistic marine loading using industry-standard JONSWAP wave spectra. The implementation employs a shared-head neural architecture with shared backbone layers (3 layers, 96 neurons) feeding specialised heads (2 layers, 128 neurons each) for individual nonlinearity function learning. Two-stage training coordinates neural network and physics parameter optimisation through differential learning rates whilst maintaining physical consistency. The twin experiment methodology leverages validated HDPE material parameters to generate unlimited synthetic training data spanning stress ranges and loading patterns that are impossible to achieve through physical experimentation alone.Results demonstrate that the neural enhancement approach enables reliable material characterisation from limited experimental data. Temporal validation achieves 85-95% accuracy across 128-fold duration scaling, enabling decades-long predictions from short-term training data, following established principles for temporal extrapolation in physics-inspired neural networks. Marine validation under realistic JONSWAP loading conditions achieves 5.7-13.2% error across diverse sea states spanning significant wave heights 1.5-7.0m and peak periods 8-15s, demonstrating robustness comparable to validated approaches for viscoelastic modelling under complex loading conditions. The diagnostic insight that physics model completeness rather than neural network architectural sophistication determines model capability guides efficient improvement strategies applicable across computational materials science domains. The validated framework transforms synthetic fibre characterisation from an experimental constraint to a predictive capability, while establishing reproducible principles for neural physics integration across materials science applications that require improved understanding from limited observations. This enables accelerated deployment of offshore renewable energy systems whilst preserving the physical foundations necessary for reliable engineering applications.

Keywords: Neural-physics integration, Synthetic fibre characterisation, Nonlinear viscoelasticity, Schapery theory, Materials discovery, Offshore renewable energy, Computational materials science

Thin and large floating structures are increasingly used for a variety of applications, such as floating solar platforms, aquaculture, storage facilities, and even housing. For example, in the Netherlands, where thin plate-shaped floating solar platforms are being deployed in near-shore regions due to their benefits, such as uninterrupted solar irradiance and land conservation. With these deployments, it becomes essential to assess the risk posed to these structures by wave forcing in near-shore environments. Due to the large size of these flexible structures, the potential large impact of long ship-generated waves can be a key scenario of concern.

This thesis explores this scenario using a hybrid model, which couples a weakly nonlinear 2D finite element-based Boussinesq wave-generation model (FEBOUSS) with a 3D linear monolithic fluid-structure interaction (FSI) model. A key contribution of this work is the implementation of a novel, one-way partitioned fluid-fluid coupling algorithm, which efficiently transfers wave-information from the FEBOUSS model to the FSI model. The coupling is achieved by (1) enforcing the normal wave particle velocity from FEBOUSS as a boundary condition at the vertical inlet of the FSI model, and (2) using a damping zone that not only absorbs reflected waves from the membrane but also gradually enforces wave-elevation and normal wave particle velocity at a subsection of the free surface boundary of the FSI model. The accuracy of the coupling algorithm can be adjusted by refining the partition of feeding particles in the overlapping zone while simultaneously controlling the length of the damping zone. This approach ensures both computational efficiency, in terms of total simulation time and data storage requirements, as well as high physical accuracy at the coupling interface. The scenario of ship-induced waves entering a narrow harbour housing a floating structure is demonstrated, focusing on how the velocity of the moving vessel, represented non-dimensionally by the depth-Froude number, influences the hydroelastic response of the membrane. The analysis revealed that for all tested depth-Froude numbers, the membrane initially follows the long ship-induced waves, with subsequent responses that may be either (significantly) amplified or dampened with respect to the incoming waves, depending on the specific case under consideration.

The work presented in this thesis makes a significant contribution to offshore engineering literature by introducing a novel fluid-fluid coupling algorithm that can be extended to even more complex scenarios. Both the FEBOUSS wave-generation model and the FSI model are highly versatile, making them suitable for studying a broad range of fluid-structure interaction problems. Future research could explore the effects of variable bathymetry, irregularly shaped floating structures, and other relevant factors, expanding the scope of this study to include a broader range of physical phenomena.

The Netherlands has agreed, by signing the Paris Agreement, to be climate-neural by 2050. To reach this goal, many large-scale renewable energy projects need to be developed in the coming years using for example wind, geothermal or solar energy. The problem with these projects is that they require large amounts of land and are especially needed in densely populated areas where the energy demand is high. For solar energy, a solution is to move it to bodies of water, creating a floating photovoltaic (FPV) system. Besides not requiring valuable land, FPV offers other benefits; The water has a cooling effect on the panels, increasing the efficiency of the solar panels. There is less water evaporation, which is desired in for example drink water basins. FPV systems have a high potential for system integration with nearshore and offshore wind turbines, optimizing space utilization and making cable pooling possible.

Multiple companies have already developed an FPV system, each designed for specific water categories and having its own (dis-)advantages. This thesis analyses different types of FPV systems and offers a tool for project developers to help decide what kind of system to select in a preliminary phase of project development.

As input to this decision tool, the environmental loads acting on FPV systems are calculated. The forces due to the wind, current and waves are found for three different types of systems. These systems are Floating Solar, Zimfloat and OceanSun. The forces are first found on a small section of the structures and then scaled up to a size of one hectare including sheltering factors. The results show that the membrane-type structure of OceanSun has significantly lower environmental forces than the Floating Solar and Zimfloat systems. These two are both built up of high-density polyethylene floaters and a metal frame holding the PV panels. The wind force is the highest on the Floating solar structure due to the larger tilt angle. The wave forces are found to be the highest for the Zimfloat structure which can be explained by the floaters. These are square blocks instead of circular tubes, making them less aerodynamic and having a larger volume which increases the inertia forces.

The decision tool uses a multi-criteria analysis that takes twelve important aspects of an FPV system
into account including the environmental loads. Examples of these criteria are energy production, costs, energy density and also less quantifiable criteria such as ecological impact, safety and the technical readiness level of the systems. The tool compares six different types of systems and can easily be adjusted to a specific location by changing the weight factors. With these weight factors, the tool provides a clear overview of the most important aspects of an FPV system and helps to decide in an early phase of project development whether a system is suitable for the desired location or not.

The decision tool is used for the case study ’Haringvliet’. Haringvliet is a former estuary in the Netherlands and is chosen as a potential location for an FPV park. The calculations mentioned above are performed using the site conditions found at Haringvliet. The conclusion of the decision tool is that the structure of Floating Solar is the most suitable for the Haringvliet. This is partly because Haringvliet is an ecologically protected area and the floating solar structure is very open, letting sunlight pass through the system into the water.