Charles Monroy, Charles (Bureau Veritas) | Sopheak Seng, Sopheak (Bureau Veritas) | Louis Diebold, Louis (Bureau Veritas) | Alexis Benhamou, Alexis (Bureau Veritas) | Sime Malenica, Sime (Bureau Veritas)
Correct assessment of entry of a solid through a free surface is important in various hydrodynamic applications. It is especially crucial when dealing with ship motion behavior in high sea states where slamming impacts are likely to occur. There is a wide range of numerical methods designed to compute forces and pressures on the hull triggered by this phenomenon. However, from an industrial perspective, it is important to discriminate between them and find a compromise among CPU time, setup time (i.e., engineering time), and accuracy. This paper aims at comparing the merits of two different classes of methods: potential theory based on a Wagner model and computational fluid dynamics (CFD) based on a finite volume method with a volume-of-fluid (VOF) interface. The numerical results are compared against experimental data from a wave-induced loads on ships (WILS) campaign.
Wave-induced loads on ships (WILS III) was a joint industrial project conducted by the Korea Research Institute of Ships and Ocean Engineering (KRISO) with participation of several academic and industrial partners, including Bureau Veritas. In the frame of this project (see Hong et al., 2014), an experimental campaign was run in order to measure loads on different 2-D sections impacting calm water. This experimental data set is very useful in validating different numerical tools designed to assess slamming loads on a ship. Since the pioneering work of Bishop and Price (1979), the standard practice in ship hydroelastic computations involving slamming events consists of cutting the bow of the ship in several 2-D sections. 3-D computations are too time consuming and there is no adequate simplified model. An example of such a method can be found in De Lauzon, Benhamou, and Malenica (2015), which compares hydroelastic behavior of a ship against an experiment.
Potential flow remains the standard theoretical framework when it comes to assessing slamming loads on a 2-D section of a ship and integrating them in a hydroelastic computation. In De Lauzon, Benhamou, and Malenica (2015), the slamming model is based on a generalized Wagner model (GWM), and elements of this model are shown for the WILS test cases. Several other potential flow models with different levels of complexity and robustness exist (e.g., see Korobkin and Malenica (2005) for a description of the modified Logvinovich model). All the models in this class of methods are very fast, but in our opinion, the GWM model offers a good compromise among accuracy, robustness, and CPU time.
In this study, we demonstrate that friction stir processing can be used to increase the fineness of the grain structures of Zn–22Al superplastic alloy. The average grain size of 0.7 .m of the base metal was reduced to 0.3 .m within the stir zone. In addition, the results of the microstructural observations, hardness tests, and average grain size measurements are presented.
Friction Stir Welding (FSW) is a solid-state joining process that uses friction heat. This method was invented by The Welding Institute (TWI) in Cambridge, England and was patented by Thomas (1991). Because friction stir welding is particularly appropriate for aluminum alloys, which often cannot be easily joined through the use of standard welding, initial research efforts were concentrated on these materials.
In addition, FSW has generated interest because of its association with Friction Stir Processing (FSP), a new technique that employs FSW tools. FSP is currently being explored as a thermomechanical processing tool that can be used to transform a heterogeneous microstructure into a more homogenous microstructure. Mahoney et al. (2001) have emphasized that FSP can be selectively applied to a location within a conventional aluminum alloy sample to tailor its microstructure and achieve increased superplasticity. Mahoney and Lynch (2006) have reported some practical applications of FSP, including the application of FSP to Al-, Cu-, Fe-, and Ni-based alloys to improve their material properties. Some of the demonstrated beneficial effects of FSP include the doubling of the strength of cast Ni–Al–bronze, the five-fold increase in the ductility of Al alloy A356, the increased fatigue life of fusion welds, the increased corrosion resistance of a Cu–Mn alloy, and the bending of a 25-mm-thick Al alloy 2519 plate to an angle of 85° at room temperature without surface cracking (Mahoney and Lynch, 2006).
Furthermore, FSP is believed to affect the grain size of the stirred material. Nishihara (2004) has reported initial results for the FSW of superplastic Zn–22Al eutectoid alloy, which demonstrate that FSW produces a fine grain structure within the joint part.
This paper presents a systematic numerical investigation of the water-entry problems associated with dropping triangular wedges or ship sections that uses an incompressible Immersed Boundary Method (IBM). In the IBM, the solid bodies are treated as an additional phase, and their motions are solved by a unified equation similar to those governing the air and water flows; a level-set technique is used to identify the air-water interface, and a projected Heaviside function is developed to identify the fluid-solid interface. For the purpose of comparison, a corresponding numerical simulation with or without consideration of the compressibility of the fluids is also carried out by using OpenFOAM. All results are compared with the experimental data provided by the comparative study of ISOPE 2016. The results suggest that the unified equation in the IBM can well predict the motion of the dropping bodies; the IBM can capture the entrapped air and produce an impact pressure and local and global forces that agree fairly well with the experimental data.
The water entry is a complex and nonlinear fluid-structure interaction (FSI) problem involving many physical phenomena such as air trapping, spray, and extreme free-surface deformation. A large impulsive pressure and slamming forces possibly lead to the damage of the offshore structure and are of interest for engineering purposes. Both numerical modelling and experiments have been conducted for the prediction and validation of the water-entry problems, in particular the effect of the compressibility, aeration, and hydro-elasticity (Miyamoto and Tanizawa, 1985; Okada and Sumi, 2000; Huera-Huarte et al., 2011; Ma et al., 2014, 2015; Mai et al., 2015). Against such a background, the International Hydrodynamic Committee of the International Society of Offshore and Polar Engineers has proposed a comparative numerical study of such an issue, in which the experimental data from the third iteration of the Wave Induced Loads on Ships (WILS) Joint Industry Project (MOERI, 2013) are provided. This paper presents our numerical results for the comparative study, and therefore only relevant reviews of numerical approaches have been given herein.
Great effort has been devoted to deriving analytical solutions and empirical formulas to predict the slamming forces associated with the water-entry problems, e.g., Wagner (1932), Dobrovol’skaya (1969), Armand and Cointe (1986), and Cointe (1991), which benefit the design practices although such solutions may be suitable only for simple-geometry or wedge-type bodies. However, the body shapes and impact angles, in particular the structures with a small deadrise angle, are important for the impact pressure development and free surface formation near the impact surface, as observed by the experiment (Okada and Sumi, 2000; Huera-Huarte et al., 2011). Due to the limitation of the analytical and empirical methods, the numerical methods have been utilised to solve the engineering problems with more complex geometry.
In S-lay operations, the pipeline passes over the stinger and is laid on the seabed after welding and nondestructive tests. The high combined loadings of axial tension, bending moment, roller reaction force, and the pipelay vessel motion may result in plastic deformation in the pipeline, which is difficult to accurately quantify. A real-time onboard monitoring system is recommended to guarantee the safety of the pipelaying process in challenging projects that include very large strain in the pipeline. However, most of the current onboard monitoring systems focus on the submerged section and sagbend section of the pipeline. The overbend section is ignored because of the uncertainty resulting from the combined loadings and the dynamic process during the pipe passing over the stinger. The present paper proposes a novel online monitoring method based on the roller reaction force measurement, which can monitor the dynamic overbend pipe strain fluctuations in real time. Relevant analytical equations in the model are first derived, then a large-scale hybrid model test and numerical simulation are carried out to verify the proposed monitoring system.
Submarine pipelines are often laid onto the seabed by the S-lay method schematically illustrated in Fig. 1. The name “S-lay” comes from the configuration of the pipeline suspended between the pipelay vessel and the seabed. Compared to other pipelaying methods, such as J-lay, reel lay or tow lay, S-lay is known as highly cost effective. Short sections of the pipeline are welded onto the pipelaying vessel firing line to make a continuous pipeline. It then passes through the tensioner and slides over the curved stinger downward to the seabed (Palmer and King, 2004; Heerema, 2005). In the overbend section, the pipeline bends from horizontal to the stinger curvature and then leaves the stinger at a certain departure angle depending on the water depth and pipeline dimensions. Large deformation is induced under the combined loadings of bending, axial tension, roller reaction force, and the pipelay vessel motion (Xie et al., 2015). The strain of the overbend pipeline section is relatively large compared with that of the sagbend in deepwater operations, because of the great stinger curvature and strain intensification effects at the roller support location (Torselletti et al., 2006a; Torselletti et al., 2006b; Perinet and Frazer, 2008). As the pipelay operations move to areas with harsher conditions, the risks of the overbend pipe increase. Real-time monitoring of the pipelay operation would be an effective way to guarantee the safety of the operation. Økland et al. (2008) presented a system that utilized the position of the vessel and some tension measurement for the pipe, as well as the current conditions. That system could predict the pipe catenary geometry and installed position of the pipe with good accuracy.
A hierarchical wave interaction theory is reviewed as an innovative idea for treating hydrodynamic interactions among a great number of bodies rigorously in the framework of linear potential theory. We also introduced experimental results that were obtained using a structure consisting of 64 truncated vertical circular cylinders arranged in a periodical array of 4 rows and 16 columns. From the observation of measured results and their comparison with computed results, the effects of multiple-body interactions on the wave elevation and local steady forces are noted with respect to the wave frequency and the spatial position inside the structure. It is observed that the characteristics of wave interactions clearly change depending on whether the wave frequency is below or above the near trapped-mode frequency. Overall agreement between measured and computed results is very good, although slight differences attributed to viscous effects are observed.
In free-surface hydrodynamic problems of the ship and ocean engineering, there exist a number of examples in which hydrodynamic interactions among multiple bodies are of critical importance. Some of the examples in ship hydrodynamics are the wave interactions between demihulls of a catamaran (Kashiwagi, 1993) and the side-wall interference effects on a ship advancing in waves in a waterway with vertical side walls (Kashiwagi and Ohkusu, 1991). The latter problem can be analyzed by the method of mirror images reflected in both of the parallel side walls, and thus the wave interactions among an infinite number of bodies must be considered. However, since the pattern of ship-generated waves may change drastically depending on the ship’s forward speed U and oscillation frequency ω in waves, it is said that the effect of wave interactions becomes less important when the value of parameter τ=Uω/g (where g is the gravitational acceleration) is larger than 1/4.
On the other hand, looking at ocean-engineering problems, the forward speed of an ocean structure is normally zero or very small. Thus the effects of wave interactions must be taken into account in the wave-related hydrodynamic analyses for a structure with several surface-piercing columns that support an upper deck (e.g. Newman, 2001). However, when the number of interacting bodies is not large, no special theoretical treatment is needed, and a conventional free-surface Green-function method, for instance, can be applied directly to the entirety of interacting bodies as a single structure.
A two-phase Smoothed Particle Hydrodynamics (SPH) method is used to simulate the early stage of the water entry of a 2-D wedge or a ship-section structure. From a comparison of the numerical results and the measured data, it is found that good agreement can be achieved. The variation of the velocity field, the pressure distribution, and the total hydrodynamic loads on the wedge are presented and discussed. The later stage of the cavity evolution for the wedge water entry and the formation of the entrapped air cavity for the ship-section water entry are simulated well by the two-phase SPH method.
Slamming on ships and offshore structures can induce local and global structural responses (Faltinsen, 2015). The mortal slamming force can damage the ships or offshore structures. A better understanding of the slamming force on the structure and the pressure distribution on the structure is important for the design and operation of ships and offshore platforms. The slamming stage is the initial stage of the water-entry problems, which includes the water landing of the spacecraft, aircraft ditching, and other applications in naval architecture and ocean engineering (Streckwall et al., 2007). The period of the slamming stage is extremely short, basically in the order of milliseconds, so the modelling slamming process requires pressure sensors with high-frequency responses in experiments and the high temporal resolution model in the numerical simulation.
In marine hydrodynamics, the entrapped air at the impact belongs to one of the important hydrodynamic phenomena associated with the water-entry-induced slamming of the ship, the sloshing-induced slamming in a liquefied natural gas (LNG) tank, and the steep-wave-induced impact on a vertical wall (Gao et al., 2012; Lugni et al., 2010a, 2010b; Gong et al., 2011). In recent years, mesh-free methods have played an important role in modelling hydrodynamic flows with the free surface. The Smoothed Particle Hydrodynamics (SPH) method is one of the mesh-free methods and has advantages in dealing with large deformation and breaking of the free surface (Oger et al., 2006). It was first utilized by Lucy (1977) to solve astrophysical problems. The applications of the SPH are mainly focused on fluid-dynamics-related areas including heat transfer, mass flow (Cleary, 1998), and multi-phase flows (Monaghan and Kocharyan, 1995). The SPH method has been developed into a competitive approach dealing with impulsive loading and large deformation events. Gong et al. (2009) adopted the SPH method with an improved approach for computing the pressure of the particles on the wall boundary condition.
Okawa, Teppei (Nippon Steel & Sumitomo Metal Corporation) | Shirahata, Hiroyuki (Nippon Steel & Sumitomo Metal Corporation) | Nakashima, Kiyotaka (Nippon Steel & Sumitomo Metal Corporation) | Yanagita, Kazuhisa (Nippon Steel & Sumitomo Metal Corporation) | Inoue, Takehiro (Nippon Steel & Sumitomo Metal Corporation)
To develop a simplified evaluation method for brittle crack arrest toughness in a heavy-thick plate, the correlation between large-scale tests (such as the ESSO test) and small-scale tests (such as the Naval Research Laboratory (NRL) drop-weight test and the Charpy V-notch impact test) was investigated. It was found that the Nil-Ductility Transition Temperature (NDTT) obtained by the NRL drop-weight test at the surface layer and the fracture appearance transition temperature (vTrs) obtained by the Charpy V-notch impact test at the quarter thickness position have a high correlation with brittle crack arrest toughness. A simplified evaluation equation for arrest toughness was suggested by the combination of the results of the small-scale tests. It was confirmed that the developed equation can be applied to various steels independent of the chemical composition and the manufacturing process.
To prevent severe damage by brittle crack propagation in ship structures, brittle crack arrest toughness is required in heavy-thick steel plates applied to large container ships (Nippon Kaiji Kyokai, 2009). Usually, evaluations of the brittle crack arrest toughness are performed by large-scale tests such as the ESSO test. Recently, the evaluation test methods of the brittle crack arrest toughness of materials were carefully investigated by a committee of The Japan Welding Engineering Society (Kawabata et al., 2014; Shimada et al., 2014; Kaneko et al., 2014; Handa et al., 2014), and a standard of the ESSO test method was established (The Japan Welding Engineering Society, 2014). However, these tests may not be suitable for quality assurance tests for mass-produced steel plates because the testing cost is considerably high and these tests are tedious. Additionally, a large-scale experimental facility possessing more than 1,000-ton loading capacity is required to conduct these tests. Therefore, an alternative simple test method to evaluate the brittle crack arrest toughness in heavy-thick steel plates needs to be developed.
Kim, Kyong-Hwan (Korea Research Institute of Ships and Ocean Engineering) | Choi, Young-Myung (Korea Research Institute of Ships and Ocean Engineering) | Hong, Sa Young (Korea Research Institute of Ships and Ocean Engineering)
There are many experimental studies of the measurement of the water impact load. Pressure sensors are widely used to measure the water impact load, but there are few papers handling the characteristics of the pressure sensor thoroughly. The present study investigates the characteristics of the pressure sensor for the measurement of the water impact load. For this purpose, seven pressure sensors are attached to a 2D wedge surface, and the water impact pressure is measured during the free falling of the wedge. Those pressure sensors have different types, sensing areas, and sensitivities. The peak pressures, rise times, and pressure impulses measured by the different types of pressure sensors are compared with one another. On the basis of these results, the characteristics and reliability of the pressure sensor for the measurement of the water impact load are discussed.
The measurement of pressure is a major concern in water impact problems, e.g., water-entry, ship-slamming, green water, and sloshing problems. A pressure sensor, a so-called pressure transducer, was widely used to measure the water impact load in previous experimental studies. The measured pressure was compared with that of the analytic and numerical solutions, and reasonable agreements were reported in the previous studies. However, it is not easy to find research cases that consider the reliability and accuracy of the pressure sensor itself.
WILS JIP-III (Wave Induced Loads on Ships Joint Industry Project-III) was carried out to investigate the hydroelastic behaviors and ship-slamming phenomena of an ultra-large containership (Hong et al., 2014). In the WILS JIP-III, drop tests of the 2D wedge and various ship sections were carried out. In the ship section drop test, two different types of pressure sensors were used (piezoelectric with 9.5 mm diameter and piezoresistive with 5.5 mm diameter), and a discrepancy of the measured peak pressure was observed. Figure 1 shows the measured pressure signals during the ship section drop test of WILS JIP-III. The piezoelectric and piezoresistive pressure sensors show the different peak pressure and decaying pressure. A similar tendency is also observed in the previous study (Kim et al., 2015). Kim et al. (2015) presented that the sloshing impact pressure can be different depending on the pressure sensor (see Fig. 2). The measured pressures of the Integrated Circuit Piezoelectric (ICP) sensors show a large difference in the magnitude although the pressure sensors are the same type and have the same sensing area.
The Bristol Channel has one of the largest tidal ranges in the world. A key cause for this is the resonance with the dominant semidiurnal tides. In this paper we use numerical simulations to investigate this resonance. We first vary the frequency on the boundary of the model and examine at which frequency the model is excited. Second, we apply a disturbance to the model and analyse the frequency at which it resonates. We examine the sensitivity of these results, finding them sensitive to the bed friction used (with possible implications for energy extraction) but insensitive to small changes in the tidal amplitude on the boundary or the mean-water level.
The Bristol Channel and Severn Estuary constitute one of the largest, semienclosed water basins in the United Kingdom. The Bristol Channel is located in the southwest coast of Great Britain. The Severn Estuary is situated at the upper reaches of the Bristol Channel, which has the second-largest semidiurnal tidal ranges worldwide. The typical mean spring tidal range is 12.2 m, with the high spring tidal range approaching 14 m at the Severn mouth. The large tidal ranges observed in the Bristol Channel and the Severn Estuary are driven by two main mechanisms (Robinson, 1980; Xia et al., 2012; Serhadlıo˘glu, 2014). One is the funnelling effect at the upper reaches of the Bristol Channel due to its wedge-shaped geometry and shallow bathymetry. However, it has long been pointed out by Marmer (1922) that this effect is not enough to produce the observed tidal range. The other mechanism is the quarter wavelength resonance of the Bristol Channel with the incident North Atlantic tidal wave (Fong and Heaps, 1978).
Despite a number of previous model studies having been undertaken for the Bristol Channel, its complex hydrodynamic system is not yet fully understood, particularly given its resonant nature. Resonant systems are typically very sensitive to small changes, and these responses are highly site dependent (Adcock et al., 2015). In this study we seek to improve the understanding of the resonance in the Bristol Channel. A simplified 2-D model has been developed from the model of Serhadlıoglu et al. (2013) to investigate the resonances in the Bristol Channel.
In this paper, the model equations and the model parameters used for the Bristol Channel region are first considered. Then, the model is tested by comparing its results with previous model studies and observations. Two methods have been used to determine the resonant periods of the Bristol Channel. A frequency sweep is used by varying the forcing frequency on the open boundary of the model to find the peak response of the semidiurnal tidal amplitude. Next, the key properties that influence the resonances are investigated. Finally, wind disturbances are applied to examine the oscillation periods of surge response.
Kwon, Chang Seop (Samsung Heavy Industries Co. Ltd.) | Kim , Hyun Joe (Samsung Heavy Industries Co. Ltd.) | Jung, Dong Woo (Korea Research Institute of Ships and Ocean Engineering) | Lee, Sung Wook (Korea Maritime and Ocean University)
The purpose of this study is to provide a guideline to estimate the damping coefficient for a box-shaped Floating Production Storage and Offloading (FPSO) under various loading conditions and bilge keel heights through model tests and Computational Fluid Dynamics (CFD) analysis. A series of free roll decay model tests is carried out under various conditions for parameters such as the draft, metacentric height (GM), radius of gyration, and bilge keel height. 3D CFD simulations are carried out and 6DOF motion of the FPSO is realized through the employment of the overset mesh technique. The effects of the loading condition and bilge keel height on the roll damping performance of a box-shaped FPSO and a prediction method of roll damping through the use of CFD simulations are discussed in detail.
As FPSOs are operated under various loading conditions, the motion responses for each loading condition should be investigated. The roll damping is essential for the accurate prediction of the motions at initial and detailed design phases. Ikeda (1976) and Himeno (1981) extensively investigated the empirical roll damping prediction method. Prediction based on a database from previous projects may be a practical solution, but the available database for box-shaped FPSOs is quite limited. Recently, Computational Fluid Dynamics (CFD) solvers have been applied to investigate the roll damping. Atluri (2009) computed hydrodynamic coefficients of oscillating bodies by CFD and validated the method on a flat plate. The roll damping for sharp and rounded bilges of a 2D rolling hull section was studied by Jaouen (2011). Veer and Fathi (2011) investigated the roll damping of a converted FPSO with riser balcony and bilge keels through CFD analysis. Additionally, Veer et al. (2012) investigated a validated methodology to calculate the oscillatory loads on bilge keels of ships operating in irregular sea states through the numerical and experimental studies. Yan (2013) studied the effect of bilge keel tip configuration on the normal force acting on the bilge keel through CFD. Thilleul (2013) investigated the turbulence model effect on the drag force acting on a circular cylinder in oscillatory flow.