Csaba Pakozdi

Wave-in-Deck Impact: Comparing CFD, Simple Methods, and Model Tests

The slamming of waves on the lower deck of large volume offshore platforms has received increased attention over recent years. For many existing platforms, the problem of insufficient air-gap clearance has become more acute of late due to more extreme weather conditions than was used in their original design basis and/or due to issues such as subsidence of gravity based structures. To investigate this problem, MARINTEK’s Wave Impact Loads JIP has, in one of its sub-tasks, focused towards an idealised model test setup of a rectangular block in regular waves. The block is fixed at a distance h above the calm water line. Both 2D and 3D model test experiments of the block in regular waves have been carried out in Phase 1 of the JIP (2008). This paper considers results from the 2D model test setup, and compares the measured vertical loading on the deck against two simple potential theory based methods (Baarholm, OMAE 2009-79560) and against results from a CFD code (STAR-CCM+). The results demonstrate that a second impact event closely following a first impact event can have a much flatter free-surface profile (and stronger water entry force) as a result of its interaction with the (deck) diffracted wave from the first impact event. The importance of resolving this diffracted wave in the CFD analysis is demonstrated. The paper concludes that for isolated impact events the simple potential flow based models, which do not consider the influence of one impact event on another, are adequate to predict the vertical loading on the deck. However, from a design basis criteria, if there is the strong likelihood of steep wave groupings resulting in closely following wave-in-deck impact events, then the presented simple methods may be non-conservative, and a CFD (Computational Fluid Dynamics) analysis or model test may be advisable to predict the vertical wave-in-deck loading. However the horizontal loading was significantly under-predicted in the CFD analysis compared to the measurements, so more work still needs to be done in this respect.Copyright

Green Water on FPSO Analyzed by a Coupled Potential-Flow-NS-VOF Method

The nowadays frequent use of FPSOs for offshore oil production in areas prone to green water events has increased the industrys focus on wave-induced impact loads as an important design parameter. This is a complex hydrodynamic problem where simplified engineering methods are often used in connection with model testing. Various efforts have been presented during the recent 10–15 years to establish reasonably good industry design tools, while the use of fully nonlinear methods and CFD is still in its development. The main focus of this paper is to investigate the potential of a simplified coupled method between a potential theory based Green Water engineer tool (Kinema3) and the commercial CFD tool Star-CCM+ based on its Navier-Stokes Solver (NS) and the Volume of Fluid (VOF) method. Results from a case study application on a large FPSO are validated against model test data. The case study contains analyses of the FPSO in long crested regular seas, both in fixed and in moored conditions. Three different heading directions are included. The approach for modeling green water events uses a Finite-Volume-VOF method with a complex velocity inlet boundary condition. Thus the Kinema3 engineering tool is used to generate simplified spatio-temporal inlet conditions from the relative wave elevation and wave kinematics at the bulwark, based on linear potential theory combined with nonlinear random wave kinematics. The VOF method is then used to model the detailed flow on deck, including impact forces on deck structures. Kinema3 can also generate simplified estimates for the peak water height, velocity as well as impact force values assuming an extended dam-break approach together with a simplified, local 2D deck layout, and comparisons to the CFD results show an overall fairly good agreement although flow details on deck can of course not be expected to be modeled that well. Comparisons of the above results to model test data show good agreement both for the relative wave height, water height and impact force level, in regular and irregular waves. Detailed time histories, including force rise time, from the coupled Kinema3 - Star-CCM+ CFD simulation analysis are quite similar to the measured ones. The CPU time consumption for the coupled simulation is moderate compared to a full CFD simulation of the FPSO in waves. Hence the achieved calculation time and the simplicity of the simulation setup of the numerical simulation makes this method an interesting candidate for industrial use. This work is a part of the research project “Green Water and Wave Impact on FPSO” carried out for and in cooperation with PETROBRAS.Copyright

Breaking Wave Kinematics and Resulting Slamming Pressures on a Vertical Column

A need has been identified to improve the knowledge about extreme slamming loads from breaking waves on vertical columns, such as offshore platforms and wind turbine foundations. Due to strongly nonlinear physical mechanisms and large statistical variability, more and improved experimental data are needed, as well as better qualified design procedures. In this paper, model test data and CFD simulations from a recent study with a fixed vertical column are compared and investigated in more detail. Selected individual extreme slamming events due to energetic breaking waves in 1:40 and 1:125 scaled model tests are presented and considered. Waves correspond approximately to extreme breaking wave occurrences in steep energetic sea states with 10-4 annual probability in the Norwegian sector.Slamming pressures on the column wall are measured in time and space by means of a 7 × 7 pressure sensor array covering 19m2 (full scale). Significant spatial variations are observed. When spatially averaged over the array, the observed highest pressures are typically in the range 1MPa–3MPa (full scale), while smaller measuring areas give higher values. This compares roughly to levels found from recent results in the literature; although exact comparison is difficult due to statistical uncertainty issues. Experiences obtained from parallel CFD and PIV activities are also compared to the experiments, from which free-surface particle velocities up to 25m/s (full scale) are estimated in the worst cases. Finally, a simple empirical formula for a slamming coefficient depending on the actual pressure integration area is suggested based on the results.Copyright

A Fully Nonlinear RANS-VOF Numerical Wavetank Applied in the Analysis of Green Water on FPSO in Waves

This paper presents numerical simulations of Green Water events and wave impact on a FPSO. The simulations are performed at model scale and the results are compared against experimental model test results. The commercial Star-CCM+ CFD software is used in the simulations. The incoming waves are modeled using 5th order Stokes theory, as implemented in the CFD software. Both fixed and free floating FPSO are considered. The moving FPSO are modeled using Chimera overset mesh technology. The vessels is free to move in heave and pitch at 180 (head sea), roll and heave at 270 (beam sea), while roll, pitch and heave is released at 225 (quartering sea). The computed water height on the deck and the relative wave height in vicinity the vessel are compared against model test results at several positions. Also the impact force on load cells blocks located at the deck of the vessel is computed and compared against model test results. The comparison of the time histories of the water elevation and load histories are in reasonable agreement with the measured time series. The number of grid cells range from 7M for the simulations at head sea, where flow is assumed to be symmetric, to 21M for the simulations at quartering sea. Total wall clock simulation time was about 10days for the most computationally demanding cases, which are the quartering sea simulations. This includes simulation of 12 wave periods with the ship fixed, and thereafter 8 wave periods of the free floating vessel. The computations show that CFD tools can be used as a research tool when studying the physics of green water and wave impact events. However, due to time CPU demanding simulations, this type of CFD analysis are not yet a practical tool for parametric design studies and deck structure optimizations. This work is a part of the research project “Green Water and Wave Impact on FPSO” carried out for and in cooperation with PETROBRAS.Copyright

Viscous drift forces and responses on a semisubmersible platform in high waves

In this paper the significance of viscous effects on a moored semisubmersible platform in various sea state conditions is explored. Experimental data from a 1:50 model scale tests in a 50m × 80m wave basin are compared with numerical time domain simulations. Both regular and irregular waves are included, and some tests are also run with current. This paper presents results from a horizontally moored semisubmersible, where we focus on the hull hydrodynamics alone. The emphasis here is low frequency surge responses. Use of conventional potential theory shows large discrepancies when compared with experimental results in high sea states. For surge motion, they are believed to be due to viscous forces in the wave zone. Viscous forces and damping in the numerical model are included by the drag term of Morison’s equation using a total relative velocity approach, which is integrated up to the instantaneous free surface elevation. A common challenge is to choose a suitable Cd coefficient which provides for sufficient excitation force without introducing excessive motion damping. It is found that using a larger Cd coefficient in the wave zone gives larger excitation without influencing the total damping significantly. This way, applying the drag term of Morison’s equation can give results that compare well with measurements. Also, the coefficient is found to be lower in waves with current than in waves only.Copyright

A Numerical Study of a Focused Wave Packet Near the Surf Zone

In this paper the numerical modeling of breaking waves propagating on a gently sloping bottom in shallow water is investigated. As more and more countries look to install offshore wind farms in their coastal waters, the breaking wave impact force on wind turbine foundations has been an area of increased research. For meaningful comparisons between measurement and simulation, the numerical reconstruction of the model test breaking wave event must be fairly exact. The combination of numerical reconstruction of model tests using computational fluid dynamics has proved a valuable tool to provide insight into the physics of the breaking wave phenomenon in deep water ([1]). To refine the technique of numerically reconstructing the breaking wave in shallow water, comparisons to a series of model tests with breaking wave events near the surf zone are made. The sloping bottom is modeled with a ramp with a gradient of 2.8 degrees in a wave tank. This paper describes the numerical reproduction of a focused wave packet, for studying its shoaling and breaking. The commercial CFD tool Star-CCM+ has been used to reproduce a measured focused wave packet train in a numerical wave tank. Its RANSE physical model with VOF technique is applied for this investigation. The numerical wave generation is based on the technique presented in [1], where the measured angle of the wave makers flap and the measured free surface elevation at the flap has been used to define a transient inlet condition for the simulation. In this paper, the numerically simulated wave elevation around the breaking point is compared with the measured time series. Promising results are obtained. The numerical model reproduces, at the same location and time, the breaking events as it was observed during the model test. One conclusion from this particular case is that the time step is critical; it should be very small during the breaking events which may result in a very long simulation time. Further work is suggested to meet this challenge, as well as for more refined studies to improve the complete numerical wave tank model.Copyright

Breaking Wave Impact on a Platform Column: An Introductory CFD Study

The slamming of breaking waves on the legs of large volume offshore platforms has received increased attention over recent years. To investigate this problem, MARINTEK’s Wave Impact Loads JIP has, in one of its sub-tasks, focused towards an idealised model test setup of a rectangular cylinder in breaking waves. The model consists of a vertical column with a fragment of a horizontal deck attached. The model is fixed at a distance L ahead of the wave maker. Physical scale model test experiments of the block in regular waves and in wave groups have been carried out in Phase 1 of the JIP (2008). The objective of this study is the CFD simulation of a long crested breaking wave and its impact on the aforementioned cylinder and deck structure in order to find out the feasibility of the numerical reconstruction of such events. The commercial CFD tool Star-CCM+ V5.03.0056 (www.cd-adapco.com) is used in this study. This paper considers results from the test setup, and compares the measured wave elevation against results from the CFD code. The position of the cylinder in relation to the breaking wave front is investigated in the numerical simulation in order to analyze its effect on the slamming force. Use of an unsteady wave boundary condition, matching the exact motion history of the wave-maker with the measured free surface elevation at the wave maker gives an almost exact matching between the computed wave profile and the measured wave profile. The improvement in the numerical tool of Star-CCM+ which makes it possible to use higher order time integration scheme for VOF significantly decreases the numerical diffusion of the wave propagation. This new scheme also enables the use of a time step 10 times larger than the first order scheme which reduces the computational time. Because a large time step can be chosen it is important that the time step is small enough to capture the correct time evolution of the physical phenomena of interest. Capturing the pressure evolution at a slamming event demands very high spatial resolution. Spatially averaged slamming pressures look fairly similar to the model test observations, while further work is needed for a more detailed comparison.Copyright

Contribution of the Mooring System to the Low-Frequency Motions of a Semisubmersible in Combined Wave and Current

This paper presents the results of an experiment carried out on a semi-submersible model to measure the steady drift force and low frequency surge motions. In the experiments, the influence of mooring systems was also investigated in different combinations of current and sea state. The measurements were carried out with a 1/50 scale model which was moored using horizontal springs and catenary mooring lines. A comparative study of the mean values of steady drift motions and the standard deviation of the low frequency motion amplitudes is presented. In addition, the effect of current on the damping ratio is discussed. It is found that for both horizontal and catenary moorings, the presence of a current increases the damping ratio of the system. For the catenary mooring system, as expected, the presence of mooring lines and their interaction with waves and current increases the damping compared to the damping of the horizontal mooring system. The measured mean values of the surge motions in a wave–current field are compared to the superposed values of those obtained from waves and current separately. For the horizontal mooring, it is found that there is good agreement in moderate sea states, while in higher sea states the measured motion responses are larger. In the wave-current field, the standard deviation of the surge motion amplitudes is found to be less than that obtained in waves alone. This can be explained by the increased magnitude of the damping ratio. Only in the cases of high sea states with the horizontal mooring system, was it found that the standard deviation of the surge motions is slightly larger than those obtained for waves and current separately. This may be explained by the absence of catenary mooring line damping.Copyright

Using a Simplified Smoothed Particle Hydrodynamics Model to Simulate Green Water on the Deck

Severe storms have gained more attention in recent years. Improved metocean data have led to new insight into severe wave conditions for marine design. Therefore, there exists an industrial demand for fast and accurate numerical tools to estimate the hydrodynamic loads during e.g. green water events.Model tests generally play an important role in these studies. In the recent past, several practical engineering tools have also been developed, based on the experience from the experimental data bases in combination with simplified but still theoretical formulations. One such tool is Kinema2, which is based on non-linear random wave modeling combined with 3D linear diffraction theory to initially identify green water events, and then finally apply a simplified water-on-deck and slamming load estimation. This forms the background for the work presented in this paper which shows the feasibility of a new technique based on the Smoothed Particle Hydrodynamics (SPH). This method can give more detailed forecast of the hydrodynamics on the deck than the simplified water-on-deck estimation. SPH uses a Lagrangian framework (particles) to describe the fluid dynamics. The water propagation and kinematics of the green water events are, in this introductory stage of the study, reproduced by using a SPH inlet condition where particles are injected with given velocity from a curved rectangular area against the deck and the deckhouse. The relative wave height and water particle velocities found from KINEMA2. Numerical results for water elevation and velocity on deck are compared against model test time series and previous results from other numerical simulation methods. The present Lagrangian nature (compared to traditional Eulerian-VOF methods) can in principe significantly reduce the CPU demand and increase the simulation speed. Slamming pressures can then be calculated e.g. from simple slamming formula calculations. In principle, pressures can also be found directly from the SPH calculations, while this would demand a significantly larger number of particles which increases CPU demand of the SPH method.Copyright