MSc thesis supervised

Filter by master program:

Filter by topic:

Filter by year:

Sort by year:

Wave damping by large-scale offshore kelp farms - A numerical modelling framework using a porous medium approach

Joel Ruesen
Offshore and Dredging Engineering
Role: Daily supervisor
In collaboration with: Mocean Offshore

Abstract

Large-scale cultivation of seaweed presents opportunities for multiple global challenges currently at play. Cultivated seaweed can provide a sustainable source of protein for humans and cattle without competing for land, freshwater supply or the use of fertilisers. Kelp forests are known to be a solid basis for an elaborate biome that supports biodiversity in areas that have been damaged by over-fishing or rising sea temperatures. Additionally, kelp forests can lock-in large amounts of Blue Carbon, expanding the oceans’ buffering capacity to mitigate anthropogenic emissions. Furthermore, with their densely seeded lines, offshore kelp farms are found to attenuate wave amplitude, thus providing coastal protection and benefits like increased workability for offshore operations. Both academic publications and industry reviews underline the potential of this sector and significant growth in cultivation is expected in the near future. Methods currently used for quantification of the damping effects of large-scale offshore kelp farms are diverse and entail varying degrees of accuracy and computational cost. Experimental observations that support the outcomes of these methods are limited to scaled experiments in wave flumes, with various methods used to mimic vegetation. No convergence is found in the most suitable methods for application to large-scale offshore kelp farms. This research presents a novel modelling framework based upon the Finite Element method, implemented using Julia Programming Language. The effects of the vegetation on the wave climate are represented with a Darcy-Forchheimer term borrowed from porous medium flow theory, including a linear and a quadratic resistance term. The framework comprises a numerical wave tank, using the incompressible Navier Stokes equations. The single-phase model captures the free surface using the coupling of dynamic pressure with a virtual elevation variable through a linearized transpiration boundary condition. Wave energy dissipation is shown to increase significantly by moving the farm structure close to the water surface. Similarly, a decrease in relative water depth - compared to the vegetated height - increases damping potential. Wave period is found to be of strong influence on dissipation, where short waves are attenuated more. Scaling vegetation length with wave length, however, diminishes the reduction in damping of longer waves. Conversely, wave amplitude is shown to be of less influence on the transmission of amplitude through a vegetated patch. The framework presents a method that is easily scalable, flexible in application on a wide range of flows and vegetation characteristics, and at reasonable computational cost. Introduction of both the linear and quadratic terms extends applicability compared to traditional methods. The approach is verified using convergence studies, application of the model is validated by comparison to existing experimental data. It is shown that experimental set-ups can be reproduced effectively, and simulation results coincide with experimental findings. Validation of outcomes on scales larger than common wave tanks was found unfeasible due to a lack of measurement data. A theoretical case study was performed to predict wave damping of a full-scale kelp farm, demonstrating promising potential with up to 40% wave energy reduction at the local peak wave period. Further research into the establishment of the Darcy- and Forchheimer-coefficients is recommended. A preliminary range of values has been found, based upon calibration on existing experiments that represent realistic ranges of vegetation characteristics. Furthermore, the main conditions of the flow and vegetation that dictate damping potential are identified. On this basis, research into a physics-based determination of the coefficients is recommended. Additionally, full-scale measurements are advised to validate application on future kelp farm designs. Through this novel approach, the range of application is increased compared to existing methods, while straightforward setup and usage are governed, and limited computational costs allow for simulation without the need for a dedicated computer setup. The framework is shown to be robust by generating consistent simulation results. In summary, the established framework shows to be a good alternative to existing approaches to investigate the wave damping potential of large-scale offshore kelp farms.

Keywords

Numerical model; FEM; Julia Language; Porous medium; Kelp farming; Seaweed; Wave damping

How tsunami transmission is influenced by floating city design parameters

Simone Geerlings
Hydraulic Engineering
Role: Daily supervisor
In collaboration with: Blue21

Abstract

For both developed and developing nations, coastal zones form an attractive location for urban settlements. With the expected increase in the earth’s population, coastal areas will experience a further increase of inhabitants. Floating city development could therefore be an interesting alternative for land-based urban expansion on land [41]. Expanding an urban settlement towards the ocean however, will make it more susceptible to extreme forces such as a tsunami waves. By generating more knowledge on how a floating structure interacts with a tsunami wave, it can show the potential value that a floating city can bring. If a floating city limits the effect of (extreme) events such as tsunamis, it can protect the coastal zone which is located near a floating city. To provide an answer to this question, this research will focus on how a floating structure can reduce the transmission of a tsunami. With the construction of an analytical model, representing the floating structure and the tsunami wave as simplistic as possible, the system can be understood more quickly. If the problem is solvable by generic programming language, this would mean that it can solve a larger range in the spectrum of the problem. The conceptual model features two options: one where the platform has no freedom of movement, the other where the platformcan move vertically. They both assume that hydrostatic pressure holds during wave propagation and a linearization of the momentum equation describing the water particle interaction. For each option, the transmitted wave height is determined based on varying the floating structure dimensions. This gives an indication on which parameters are of influence in the transmission of the wave. First, the conceptualmodel is analysed by changing the platformdraft and the length for both the motionless and the vertically moving platform. Both options are influenced most by the length of the structure. The situation with the motionless platformshows this effect earlier, by a higher wave attenuation percentage for the same platform length. Whether the draft also has an influence is strongly dependent on the value representing the length of the platform. The difference between the two platform movement options shows that the effect of changing the allowed movement of the platformis significant. Next to the reduction in transmission, the conceptual model shows signs of resonance. The moving platform option in the model is formed by a second order differential equation. Fitting this equation, resonance is evident and therefore visible for certain combinations of the platformdimensions. In addition damping is present, ensuring that there are some parts where, despite the natural frequency pointing there, no resonance occurs. The amount of damping is strongly linked to the platformlength, with a higher level of damping for a longer platformlength. Finally, the results from the conceptual model are compared to the outcome in SWASH. This numerically based model has the possibility to simulate a tsunami wave in its development towards the coast and also features a buoyancy function for structures. This comparison serves as a validation of the conceptual model. In general, the conceptual model always results in a less reduced wave attenuation percentage and can be said to be more conservative. Due to the assumptions of leaving out the non-hydrostatic pressure and a lower level of detail, a maximum of 5% deviation in both model results occured. This however, matches with the fact that it is a less detailed model and adds to the reasoning that the conceptual model provides what it is meant for. Next to the effect of the structure itself, the positioning of the structure is also of large importance. Wave height and intensity of the wave will vary due to the surrounding local coastal features. Next to that, the local water depth is determinant in compressing the wave, therefore decreasing the wave length when the water depth decreases. The maximum wave attenuation that can be achieved according to both models is 10%, considering platform dimensions and location variations. The conceptual model appears to work for which it is intended: modelling the resulting transmission between a floating structure and a linear tsunami. It is expected that modelling programs can be expanded and/or improved, so that more realistic floating structures can be modelled. However, it will remain difficult to accurately model a tsunami as it is complex in behaviour. Yet, this research brought the field one step closer to evaluating the transmission of tsunami waves when interacting with a floating structure of certain dimensions.

Keywords

hydroelastic analysis; VLFS; floating city

Hydroelastic Analysis of a Multi-Module Very Large Floating Structure

Dorette Regout
Hydraulic Engineering
Role: Daily supervisor
In collaboration with: Blue21

Abstract

This thesis concerns a hydroelastic analysis of a multi-module very large floating structure (VLFS), analysed in the frequency domain. To this end, the fluid-structure interaction is described by a 2D model, where the VLFS is represented by four floating beams interconnected with rotational springs. The fluid is modelled as an ideal fluid, and the floating beams are modelled by the Euler-Bernoulli beam theory. The finite element method is applied to solve the governing equations of the fluid motion and the motion of the beams, where the model is built using the FE-library Gridap, written in the Julia programming language. The aim of the study is to investigate the influence of various module and connection stiffness on the behaviour of the system, with the view to obtain more insight in the complex relation between the hydroelastic response and internal loads, when the system is subject to regular waves.

Keywords

hydroelastic analysis; VLFS; floating city

Perforation of Monopiles to Reduce Hydrodynamic Loads and Enable use in Deep Waters

Jorne van der Ploeg
Sustainable Energy Technology
Role: Daily supervisor
In collaboration with: SIF Offshore Foundations, Delft Offshore Turbine (DOT)

Abstract

The offshore wind industry in Europe has experienced significant growth in the past decade, with wind farm development mostly focusing on the shallow area's in the North Sea. Naturally, the market is driven to reduce the Levelised Cost of Electricity (LCoE) to become more competitive with fossil fuels and less dependent on government subsidies. A transition in wind farm development towards deeper waters is expected and already observed in the market, driven by decreasing availability of shallow area's and higher wind resource at far offshore locations. The majority of the northern part of the North Sea is between 60 - 120 meter deep, currently the jacket foundation is deemed as the foundation of choice for this water depth range. Despite several technological advantages of the jacket, the main downsides are the large engineering effort and welds required to produce such a foundation resulting in difficult series production and high costs. This does not align with the industry's ambition to lower the LCoE. As such, the need for a technologically viable and economically attractive foundation concept for waters between 60 - 120 meter deep arises. The goal of this research can be divided in to two parts. The first part is to determine the potential of conventional monopiles in this water range and identify the main limiting factors. To do so, a monopile is dimensioned at a selected reference location for three turbines representing the current, near future and future outlook of the market. The designed monopiles are tested for manufacturabiliy, Ultimate Limit State (ULS) and Fatigue Limit State (FLS) to identify the technical showstoppers. Next, in the second part, a novel monopile design is introduced and analysed to work around the identified limits. To dimension the monopiles for the three reference turbines a parametric dimensioning script is developed. The monopile geometry is dimensioned to have a selected first natural frequency of 0.20 Hz, based on the relevant frequency diagrams. Next, these geometries are tested against mudline ULS failure for the power production and parked condition load cases. Hereafter, an FLS check for the B1, C1 and D S-N curves is conducted based on the obtained scatter tables for site conditions. It was found that D-curve fatigue damage for non grinded butt welds is the main limiting factor for all dimensioned monopiles in deep water. However, industry experts believe that all welds can be grinded in the production process, eradicating the need to assess the D-curve. When assuming this statement to be true the newly obtained limits become ULS failure during parked conditions for the 15 MW reference turbine and manufacturability constraints for the 20 MW reference turbine. The Haliade X showed no limits within the specified water depth range when neglecting the D curve fatigue damage. A perforated monopile concept with reduced available area for wave loading is introduced. A Computational Fluid Dynamics (CFD) model based on the 2003 Menter Shear Stress Transport turbulence model is constructed for a perforated monopile to gain insights into how waves propagate through the structure and the forces associated with this. The CFD model is verified against experimental wave flume data before being used for further analysis showing a root mean square error of 0.0192 between model results and experiments. The CFD model is used to assess three geometries with different perforations and levels of porosity. No increased drag around the first natural frequency caused by the perforations, hinting to favourable dynamics, was found in any of the test cases. As such, the dynamic response of the three perforated monopiles was found to be unchanged when compared to a reference pile without perforations. Despite this, a significant reduction of lifetime fatigue damage was observed caused by the reduced forces acting on the structure resulting from the smaller frontal surface. Next, the mudline stresses are recalculated and a structural finite element model to assess the stress concentrations around the perforations is set up to verify the maximum allowable stress level threshold is not exceeded. A geometry was found which shows a 35.5% reduction of lifetime fatigue damage whilst stresses remain below the maximum threshold, hereby showing the potential of the perforated monopile. Implementing this perforation allows the use of monopiles up to 87 meter deep, limited by D curve fatigue. Reference piles without perforations were found to be infeasible for all assessed water depths, also limited by D curve fatigue. It is shown that the perforation concept can either be implemented to push the monopile foundation to deeper waters, or can be used to realise steel reduction at current water depths.

Keywords

Monopiles; Offshore Wind; Perforation; Renewable Energy; Wind. Sif; DOT

Exploratory research on application Multi-Level Multi-Fidelity Monte Carlo in fluid dynamics topics: study on flow past a porous cylinder

Jan Modderman
Offshore and Dredging Engineering
Role: Daily supervisor

Abstract

In offshore engineering complex simulation models are constructed for design optimization using Monte Carlo methods. These models incur large computational costs. Multi-Level Multi-Fidelity Monte Carlo is proposed as a method to reduce the computational cost of these simulations. In addition, research is conducted on the use of porous media as passive damping systems. Hence, an analysis on the effect of porosity on the vortex shedding frequency is conducted. This thesis is an exploratory investigation on the application of Multi-Level Multi-Fidelity Monte Carlo in fluid dynamics topics and its particular use for analysis of the effect of porosity on the vortex shedding frequency on a porous circular cylinder. Three case studies are conducted. Firstly, applying Multi-Level Multi-Fidelity Monte Carlo on a solid circular cylinder case, which is deemed as a successful application, based on the estimated quantity of interest, variance reduction and computational cost reduction. Furthermore, two parametric studies are conducted: 1) to discover empirical relationships (low-fidelity models) and 2) forward uncertainty propagation with Multi-Level Multi-Fidelity Monte Carlo using a uniform input distribution. Both parametric studies consist of a number of equally distributed points of porosity on a case setup of flow past a porous circular cylinder. The parametric studies use a frequency detection algorithm, which approximates the vortex shedding frequency using the frequency of lift force oscillation. The results of the first parametric study indicate there is a drop in vortex shedding frequency as experienced by the cylinder for increasing porosity. The hypothesis is that for increasing porosity the formation length of vortex shedding increases. Two empirical relationships are derived from the results by curve fitting the Strouhal number (dimensionless form of the vortex shedding frequency) versus porosity. These empirical relationships are incorporated in the Multi-Level Multi-Fidelity Monte Carlo method and applied to a similar parametric study on the effect of porosity on the vortex shedding frequency. The results indicate the presence of systemic errors in the high-fidelity model. The conjecture is that the major influence on these errors is due to the resolution of the frequency detection algorithm being too low. For this reason, no clear conclusion on the validity of the empirical relationships is obtained and further research is required.

Keywords

Monte Carlo; Porous Media; Multi-Level Multi-Fidelity; Strouhal number; vortex shedding

Design of a suction pile multiline anchor system for floating offshore wind turbines

Koen van Koten
Offshore and Dredging Engineering
Role: Daily supervisor
In collaboration with: SPT Offshore

Abstract

In this thesis I will show the realisation of a multiline anchor system using a suction pile anchor (SPA). From the literature research it can be concluded that the SPA is a suitable anchor for a multiline anchor system and that the most probable mooring configurations will be either the original single-line, the 3-line or the 6-line system. Furthermore it could be concluded that when using a catenary mooring solution for the 3-line system, that the anchor perceives a lower horizontal net force due to anchor lines coming from different directions and cancelling parts of each other. This would, in theory, make the SPA for the 3-line multiline system smaller and also there would be 67% less anchors needed in a wind farm array. Furthermore the 6-line anchor does seem to have a bigger horizontal net force which requires the SPAs to be bigger but 83% less anchors are needed in an array of wind turbines. The looked at FOWT-system was the University of Maine VolturnUS-S reference floating offshore wind turbine semi-submersible, which supports the IEA-15-240-RWT turbine with a turbine rating of 15MW. With the original mooring configuration being a catenary one, a comparison has been made with a taut system and its implementation into a multiline anchor system has been researched. Lastly three different soil profiles have been chosen: normally consolidated clay, slightly consolidated clay and loose sand. In Chapter 4 Suction anchor concept design, the forces on the anchor are set up using the data set-up by NREL and the University of Maine and the C. Fontana papers. A SPA has more bearing capacity when the mooring-anchor connection point is below the mudline. Because of this feature the mooring line attaches to the SPA at an angle and in this thesis a range of angles of approach are chosen to be investigated: 25 and 35 degrees for the chain and 45 degrees for the taut line. Because of the angle there is a vertical force introduced to the anchor which need to be added together for each line. Because of this the total net force on the anchor is increased for the 3-line and 6-line anchors compared to the single-line anchor. With the mean and maximum forces the calculations for the holding capacity/pull-out capacity can be set up by looking at the Ultimate Limit State (ULS) or maximum forces in the system. These calculations are taken from the DNV guidelines which are the industry standard. To set-up the first parameters estimation a embedment (h/D-ratio) starting value has to be chosen and from literature and in discussion with SPT offshore the starting values were set at 5 for the clay profile and 1.5 for the sand profile. The weight of the SPA must be defined which was done as the mean vertical force applied to the anchors as such they will not be pulled out over time. The installation and removal calculations are an important step in the design and are also set up. Here the under pressures required to fully install the SPAs are calculated. Furthermore, structural failure due to buckling is checked for and different soil failures are analysed. Lastly the removal pressure of each concept is checked which allows for complete removal of the suction anchor. By lowering the embedment ratios of the different concepts the pressures inside the anchors can be minimised and problems can be averted. In the detailed design the full design of a in use SPA is shown and each part is defined. Furthermore, the one-line-broken criterium is discussed and it can be concluded that in case this happens the 6-line multiline system is very dangerous because a chain reaction can be started which can take out large parts of a wind farm array. Also a weight estimation of each SPA is made from which the extra needed ballast is calculated. Subsequently a cost estimation of each anchor concept can be made by calculating the cost of each system from the structural weight, the ballast and the mooring line lengths. At a depth of 200m, at which this study is situated, the taut multiline system cannot be set-up but the single-line taut system can be compared to the catenary single line system. Lastly a parametric analysis is done where changes in different parameters are compared to each other. What can be concluded from this thesis research is that a multiline system is technically feasible for a 15MW floating offshore wind turbines using SPAs. The 3-line and 6-line systems both have larger anchors than the single-line system although they need less anchors in a system. When including the mooring line costs together with these anchor costs it can be concluded that the 3-line anchor is more economically viable but the 6-line anchor is not. What can also be concluded from these mooring line costs and what is discussed in the parametric analysis is that the system works better if the wind turbines are closer together because the lines will be shorter. This distance is dictated by the wake recovery and an optimisation study is recommended for the 15MW wind turbine but it is also recommended that smaller turbines and deeper depths are looked at.

Keywords

Anchor; anchor mooring; anchor points; multiline; SPA; suction pile anchor; FOWT; Floating offshore wind turbine; SPT Offshore

Hydroelastic wave deformation of Very Flexible Floating Structures: A performance study of a monolithic finite element model

Sjoerd van Hoof
Hydraulic Engineering
Role: Daily supervisor
In collaboration with: Mocean Offhsore

Abstract

At the moment, the world is at the verge of an energy transition. One of the most promising green resources is solar energy, which is a rapidly growing market. However, to fully use its potential of economy of scale, the application of offshore floating solar should be explored. A promising option is the use of a flexible type of Very Large Floating Structures (VLFSs), which are called Very Flexible Floating Structures (VFFSs). They are characterised by their large length to height ratio compared to rigid bodies and depending on their material properties have a hydroelastic response to the incident wave. In the late 1990s, a lot of research has been done on VLFSs by Tsubogo and Okada (1998) who derived an analytical dispersion relation assuming a zero­draught structure. However, only recently, Schreier and Jacobi (2020b) did experimental research on VFFSs in a towing tank at the Delft University of Technology, as little is still known about flexible structures. This report focuses on a numerical alter­ native that covers both VLFSs as well as VFFSs using a Finite Element Method (FEM) Fluid Structure Interaction (FSI) model which has been built based on potential flow to model the fluid and a dynamic Euler­Bernoulli beam that represents the floating structure using the Julia package Gridap. One of the main advantages is that the zero­draught assumption is not necessary and, therefore, structures with larger draughts can also be modelled. Next to this, the numerical model is able to cope with irregular shapes, for which no analytical method yet exists. The model is built such that it can handle 2D as well as 3D domains. A 2D analysis has been made to understand the influence of hydroelastic wave deformation of the incident wave, in terms of wavelength dispersion as well as amplitude dispersion on floating structures. To verify the model, the numerical results were compared to the analytical solution and experimental research in a towing tank, which showed accurate results. Test runs were set up that mimicked the towing tank set­up and a full­scale solar park. Furthermore, a sensitivity study was executed that shows the limits of the flexible domain and to see in which cases significant (>1%) hydroelastic wave deformation would occur using governing mean and extreme ocean waves, as well as a typical lake wave. Finally, the influence of the draught of the structure was examined. This report provides a good overview of when wave deformation should be accounted for in terms of bending rigidity and density. Confirming existing theory, it was found that the stiffness of the VFFS causes wave stretching and the draught of the structure influences the extent of wave shortening. It was also found that significant wave deformation will not occur for ocean waves as the required stiff­ ness is beyond existing materials. For extreme ocean waves, there is even no dispersion at all. As the wave frequency increases, the hydroelastic interaction gets stronger. The typical lake wave showed to be well within the flexible regime and also showed significant dispersion with realistic material parame­ ters. The numerical model is able to cope with large draught scenarios which lead to wave shortening, which in its turn leads to wave focusing. Ultimately, the numerical model showed to be a good alternative to existing methods to investigate the behaviour of VLFSs outside the floating solar domain, where one could think of ice floes, floating islands or floating airports.

Keywords

Very Large Floating Structures; Fluid Structure Interaction; Very Flexible Floating Structures; FEM; potential flow; Floating Solar

Workability optimisation of the Stella Synergy: during monopile installation with a motion-compensated gripper and using its DP-system

Jap Verbruggen
Offshore and Dredging Engineering
Role: External committee member

Abstract

Using a floating vessel operating on its DP-system and using a motion-compensated pile gripper to install monopiles could be the installation method of the future. Therefore, this thesis focuses on this method. The main objective of this thesis project is to build a model that accurately describes the motions of the Stella Synergy, the monopile, and the motion-compensated gripper, depending on the environmental conditions. This model is built in Anysim, which is a time-domain simulation software program of MARIN based on the RK2 numerical method. The model considers the early pile driving phase because this phase is governing in terms of risk. The monopile acts as an inverted pendulum in this phase, and the motion-compensated pile gripper must guarantee the stability of the monopile. The vessel uses its DP-system for station keeping. The DP-system contains a position reference system, a filter, a control system, and a thruster allocation algorithm. The vessel describes the wind, current and wave forces on the monopile and vessel. The environmental conditions are assumed to be collinear, and wave spreading is added to the model for some simulations. The wave forces on the vessel are determined with diffraction calculations in Ansys AQWA. The diffraction calculation for the vessel is verified with a diffraction calculation of MARIN, and the diffraction calculation for the monopile considers the shielding effect and is verified with a calculation with the Morison equation. A motion-compensated pile gripper with two PD-controllers is built in Python. The gripper considers static and dynamic friction forces and a maximum delta force per numeric timestep to model the pressure build-up time of the hydraulic cylinders. Multiple 3-hour simulations are run to generate results. These simulations, which considers each a different sea condition, are tested by the six limitations of the model. First, the preferable incoming angle of environmental conditions is determined. The workability of the Stella Synergy is calculated operating at the North Sea using this preferable incoming angle of attack. Then, two adaptations to the model are tested to increase the workability. Using fast-rotating thrusters or changing the DP-gains result in the workability of 96.4%. The governing limitation is the pitch motion of the vessel. It is tested if using mooring lines in combination with the DP-system results in a footprint reduction. It is concluded that adding mooring lines could result in a footprint reduction, but it is crucial to gain insight into the optimal axial stiffness of the mooring lines. The monopile's influence on the vessel's motion is also tested. It is concluded that the vessel's surge, sway, roll and yaw motion increases significantly due to the environmental forces on the monopile, which are passed through the gripper to the vessel. Finally, the workability of the vessel during the worst-case single failure is determined. After improving the DP-gains for particular sea conditions, the workability for the worst-case single failure was 96.0%. The failure results thus in a minor difference in the workability.

Keywords

Workability; DP-system; Motion-compensated gripper; monopile; installation

Bayesian system identification for structures considering spatial and temporal dependencies

Jan Koune
European Wind Energy
Role: External committee member

Abstract

Bayesian system identification, including parameter estimation and model selection, is widely used to infer partially known, unobservable parameters of the models of physical systems when measurement data is available. A common assumption in the Bayesian system identification literature is that the discrepancy between model predictions and measurements can be described as independent, identically distributed realizations from a univariate Gaussian distribution. However, the decreasing cost of sensors and monitoring systems leads to more frequent structural measurements in close proximity to each other (e.g. fiber optics and strain gauges). In such cases, dependency in modeling uncertainty could be significant, both in space and time, and the assumption of uncorrelated Gaussian error may lead to inaccurate parameter estimation. The aim of this thesis is to explore how Bayesian system identification can be feasibly performed using large datasets when spatial and/or temporal dependence might be present and to assess the impact of considering this dependence. A pool of models, each assuming a different correlation structure, is defined and Bayesian inference is performed. In particular, stress measurements obtained on a steel road bridge are used to update the parameters of the corresponding FE model and the parameters of the correlation structure. The results are compared to a reference model where only measurements of the response peaks are used under the assumption of independence. Nested sampling is utilized to compute the evidence under each model and Bayesian model selection is applied. The question of efficiently performing system identification for large datasets (N > 102 for temporal dependencies and N > 103 for combined spatial and temporal dependencies) is investigated, and a novel approach for efficiently calculating the exact log-likelihood is derived. An approximation based on the Fisher information matrix is used to efficiently calculate the information content of measurements. It is found that the choice of correlation function can significantly affect the posterior distribution of the model prediction uncertainty. Additionally, it is shown that using large datasets and considering dependence makes it possible to perform system identification for a larger number of parameters compared to the reference model. The results of the case study indicate that using measurements from multiple sensors under combined spatial and temporal dependence and additive model prediction error yields reduced uncertainty in the posterior and up to 29% reduction of the posterior predictive credible interval range compared to the reference case. Furthermore, the efficiency of the proposed likelihood evaluation method is assessed. Using this method, exact calculation of the log-likelihood can be performed for >106 points in under a second in the case of correlation in one dimension. For combined spatial and temporal correlation it is shown to be approximately 900 times faster than naive evaluation for a 64 by 64 grid of observations. The results of the case study indicate that the described approach can be feasibly applied to real-world structures and can potentially improve parameter estimation and reduce prediction uncertainty. These findings suggest that further research into the approach could yield improvements over current methods.

Keywords

Bayesian Inference; Bayesian statistics; System Identification; Structural Health Monitoring (SHM)

QUAD Lift Workability Analysis: Evaluating the Dynamic Interaction Between Two Dual crane Vessels using a Frequency Domain Modelling Technique

Niels van Duijn
Offhsore and Dredging Engineering
Role: Daily supervisor
In collaboration with: Jumbo Maritime

Abstract

A QUAD lift is a new lifting method in which dual crane vessels combine their vessel capability to increase their offshore lift performance. The use of the Jumbo J-Class vessels in a QUAD lift creates the opportunity to increase the offshore lift capacity and to install structures with larger dimensions. When floating vessels are close to each other in an offshore environment, their motion will be different than in the freely-floating situation because of hydrodynamic coupling and wave diffraction forces. The main objective of this thesis is to create a model of the QUAD Lift method which predicts the vessel and load motions and evaluate the workability such a lift. Both potential solvers AQWA and OrcaWave are used to assess the hydrodynamic parameters of the interacting vessels. The gap between the vessels is 40 m and the vessel configuration is such that the cranes are parallel to each other. In between the vessels, transversal wave resonance induces peaks in the frequency dependent radiation forces of the vessels. An additional damping lid in between the vessels effectively reduces the resonance behaviour, which is overestimated by potential solvers. The damping lid has an negligible effect on the final workability of the QUAD lift. A 18-DoF linear Matlab model is created which includes the mechanical connection between the vessels and the load. The cranes and the cables are modelled as linear springs. The natural frequencies and eigenvectors show large coupling between the vessel roll and the load sway motion. Tugger lines between the vessel and load are added to shift the natural frequencies of the system and to decrease the large horizontal responses of the load. A parametric study is done on the effect of the load mass, cable lengths and wave directions on the system motion in the most probable wave condition in the Central North Sea. An increase of the mass of the load leads to larger vessel and load motions. The shorter the cable length, the larger the vessel and load motions. The motions are most severe in beam and quarter waves. Depending on the stiffness of the tugger lines the workability can be improved up to 85, 55 and 24 % in respectively head, quarter and beam waves. The limiting factor for the workability is the off-lead angle of the cranes. Broadening of the off-lead angle limit of the crane shows great potential to further increase the workability.

Keywords

Diffraction/radiation model; Coupled Multibody and Hydrodynamics; Vessel motions; Frequency Domain; Offshore Installation; Offshore Heavy Lifting; Workability