Layer | Fill | Outline |
---|
Map layers
Theme | Visible | Selectable | Appearance | Zoom Range (now: 0) |
---|
Fill | Stroke |
---|---|
Collaborating Authors
Results
Abstract In this paper, a series of numerical models has been developed to study the effects of two different baffles on liquid sloshing under the six degree-of-freedom (6-DOF) excitation. The moving particle semiimplicit (MPS) method is a mesh-free method which can simulate flow with large deformation and nonlinear fragmentation of free surface effectively. The MLParticle-SJTU solver based on improved MPS method is first validated against the available experimental data for 3-D liquid sloshing in a rectangular tank under horizontal excitation. Then, the study of liquid sloshing in another tank is carried out under the 6- DOF excitation. The detailed flow field and the variation of impact pressure can be observed clearly. In addition, the effects of two different baffle types on liquid sloshing, two vertical baffle and two ring baffles, are further analyzed based on previous rectangular tank. Results show that the sloshing and pressure amplitudes are obviously reduced due to these baffles and the ring baffles are more effective on restraining sloshing. Introduction With the development of shipping industry, a mass of large vessels which load liquid specially has been constructed such as VLCC, LNGC, LPGC and so on. Because of the increase of frequency and intensity of sea condition, the navigation safety of liquid cargo vessels has been the focus of research. When external excitations are large amplitude or near resonant frequency, the liquid inside a partially filled tank will be prone to complicated and nonlinear sloshing phenomenon. The impact pressure induced by liquid sloshing may destroy the structure of tank walls and even cause more violent ship rolling. So, designing a liquid cargo rationally and predicting its motion accurately are significant for the transportation of liquid cargo. Rising interest in liquid sloshing, many studies of theory, experiment and numerical simulation have been conducted. At the outset, original researchers had to study 2D liquid sloshing by the limit of relevant knowledge and computational capabilities. Faltinsen (1978) developed a linear analytical solution for liquid sloshing in 2-D rectangular tank under horizontal excitation. Then this solution has been widely employed in a mass of numerical models. Nakayama and Washizu (1980) simulated non-linear liquid sloshing in a 2-D rectangular tank under pitch excitation by using the finite element method (FEM). Subsequently, Nakayama and Washizu (1981) also used the boundary element method (BEM) to analyze the 2-D problem of non-linear liquid sloshing. Cho and Lee (2004) simulated numerically a 2-D tank to study the large amplitude liquid sloshing. Wang and Khoo (2005) studied the 2-D liquid sloshing under random excitations by the fully non-linear wave theory. Frandsen et al. (2003, 2004) conducted a series of numerical simulations of liquid sloshing in 2-D tank under horizontal and vertical excitations in N-coordinate. Chen et al. (2004, 2005) used Navier-Stokes equations (NSE) to study 2-D viscous liquid sloshing under 3-DOF (surge, heave and pitch) excitation. Shao et al. (2012) applied another mesh-free method, smoothed particle hydrodynamics (SPH), to model 2-D liquid sloshing dynamics. Yang et al. (2015) investigated the effects of external excitation period on 2-D liquid sloshing based on MPS.
- Transportation > Marine (1.00)
- Transportation > Freight & Logistics Services > Shipping > Tanker (0.53)
Abstract The improved MPS (Moving Particle Semi-implicit) method is adopted to numerically investigate the water entry of a ship bow section and three cases with different impact speeds are carried out in the present work. Firstly, the ship bow impacting water with speed U0=0.61m/s is simulated to validate the capability of the present method into water entry of a ship bow with complex shape. The vertical force and velocity of the bow section are compared with the literature data. Then, two more cases are calculated and the bow drops into water with different speeds. The comparisons of the vertical forces between the present method and literature data show that the force peak will rise up and the corresponding instant will reach earlier as increasing the initial entry speed. Introduction Water entry is a complex problem in the free surface flows and always catches numerous researchers' attentions. Greenhow and Lin (1983) investigated the free surface deformations of the circular and triangle cylinders dropping into water by using a series of experimental tests. Their works are quite significant for the numerical investigations of the water entry problems since the circular cylinder entering water has been a benchmark for different CFD (Computational Fluid Dynamics) methods. Later, many numerical investigations are carried out to study the water entry problems, such as BEM (boundary element method) (Sun and Faltinsen, 2006), CIP (Constrained Interpolation Profile) method (Zhu et al, 2007; Zhao et al, 2015), Overset grid method (Swidan et al, 2013), ALE (Arbitrary Lagrangian-Eulerian) method (Wang and Soares, 2013). In the meantime, meshless particle method is a newly CFD (Computational Fluid Dynamics) method and is quite suitable to deal with the moving boundaries and the large free surface deformation in the water entry problems due to its Lagrangian and meshless nature. Both SPH (Smoothed Particle Hydrodynamics) and MPS (Moving Particle Semi-implicit) are such particle methods. Up to now, these two meshless particle methods have also been applied into the water entry problems and the numerous nice works can be found in the literatures, i.e. WC-SPH method (Yang et al, 2014; Gong et al, 2009), incompressible SPH method (Skillen et al, 2013; Shao 2009) and MPS method (Yu and Zhang, 2013). However, the shapes of dropping objects in most of these works are circle or triangle. Here, the water entry of a ship bow section with complex shape dropping into water at different entry speeds is numerical investigated by IMPS (Improved MPS) method including nonsingularity kernel function (Zhang and Wan, 2012), mixed source term for pressure Poisson equation (Tanaka and Masunaga, 2010; Lee et al, 2011) and accuracy detection of free surface particles (Zhang et al, 2014). In addition, there are also many excellent improvements on the original MPS method such as the Laplacian model, the gradient model, the source term for pressure Poisson equation (Khayyer and Gotoh, 2011, 2013; Tsuruta et al 2013, 2015; Ikari et al 2015).
Abstract An Unsteady Actuator Line Model (UALM) is developed in this paper and applied to a 5MW floating offshore wind turbine (FOWT). This model is implemented into two- phase fluid CFD solver, naoeFOAMSJTU. The goal of the approach presented here is to investigate the interaction of the aerodynamic loads with the platform motion within acceptable time cost. A semi-submerged floating platform conceptualized in the Offshore Code Comparison Collaboration (OC4) is considered in this paper. Initially the UALM is verified by comparison with the results of a previous study. Next, two kind of fullsystem simulations with different complexity are performed: first, the wind forces are simplified into a constant thrust; second, the fully coupled dynamic analysis with wind and wave excitation is conducted by utilizing UALM. Based on the results, the aerodynamic loads and coupled responses for cases of different complexity are discussed. Introduction As with other emerging industries, the wind energy industry moves on with hesitation. According to the statistics of Chinese Wind Energy Association, wind power provided 114.6 GW for China's electricity supply at the end of 2014, which shows a booming trend. The China's offshore wind power is also developing in the fast lane and up to 229.3MW have been installed. Meanwhile, many European countries have begun to move forward towards the floating offshore wind conversion technology. Among the various floating solutions suggested, the most promising are the spar buoy, the tension-leg platform (TLP) and the semi-submersible. The present study, specially focus on a semisubmersible design. A review of current floating supporting strategies is available in Virยดe (2012). Offshore wind power has many advantages over land-based wind turbines, including large continuous areas suitable for farm deployment, stronger and more steady wind, and less wind turbulence (Musial et al, 2004). However, designing offshore wind turbine system is a challenging task. This is especially true for the floating offshore wind turbine (FOWT).because of the complex coupling effects. This coupling effects are prescribed by Sebastian and Lackner (2013). The additional Degrees of Freedom (DOFs) of the floater result in the highly unsteady properties of aerodynamics of FOWTs. Moreover, significant pitch and surge motion of the floater have been predicted in previous studies (Matha et al, 2011). Therefore, accurate simulation of the coupled dynamics of the FOWT is a substantial task, considering also the existence of very limited large-scale experimental data. The absence of acknowledged software package simulating the full system is also a restricting factor for the development of offshore wind technology. With this in mind, Task 30 OC4 is led cooperatively by the National Renewable Energy Laboratory (NREL) and the Fraunhofer Institute for Wind Energy and Energy Systems Technology (IWES). The purpose of the OC4 project is to perform a benchmarking exercise of offshore wind turbine dynamics computer codes. The project defines the load cases to be run for Phase II and the output to be reported.
Abstract In the present study, the process of free-falling wedge impacting on water is numerically studied by our in house solver based on Moving Particle Semi-Implicit (MPS) method. Some improved schemes are used in this solver to suppress numerically unphysical pressure oscillation in traditional MPS method. For validation purpose, computational results of wedge with different tilting angles are compared against experimental results from the Wave Induced Loads on Ships Joint Industry Project III (WILS JIP-III). Numerical pressures, free surface elevations and velocities of wedge show agreement with experimental data. Introduction During ship sailing at rough sea, slamming occurs when the forefoot of ship hull rises above the water surface and then drops into water with high vertical velocity (Southall et al., 2014). Periodical and short duration impact loads can cause serious damage to ship structure. Hence, the slamming problem is important for ship design and operation. Over the past decades, the slamming problem was commonly investigated as the similar flow of wedge entry into water (Yang and Qiu, 2012), and firstly studied by von Karman (1929) and Wagner (1932). Among the early established methods, theoretical approaches were much popular to solve this problem. However, these methods are hard to describe the complex nonlinear free surface flow. On the contrary, kinds of numerical methods based on the Navier-Stokes equation are developed and show the capability to solve the water entry problem. Among these approaches, Lagrangian particle methods are more and more popular to free surface flow problems in the near few years. The Smoothed Particle Hydrodynamics (SPH) and Moving Particle Semi-implicit (MPS) methods are the representative Lagrangian type mesh-less methods. The SPH method is originally developed for compressible flows by Monaghan (1994). By choosing a sufficiently high speed of sound and a much small size of time step, it can be employed to solve water entry problem (Oger et al., 2006; Shao 2009; Liu et al., 2012; Koukouvinis et al., 2013; Ma and Liu, 2014; Amicarelli et al., 2015). Compared to SPH method, MPS method was originally proposed by Koshizuka and Oka (1996) for incompressible flow. Since the pressure of fluid is computed by a semi-implicit algorithm, a relatively large size of time step can be used in MPS method. Recently, several applications of the water entry problem based on MPS method were published. For example, Lee et al. (2010) employed the MPS method to calculate the impact loads by falling flat plate with incident angles. Yokoyama et al. (2014) numerically studied the water entry of spheres by MPS method and discussed the influence of the surface conditions of the solids falling into the water on the formation of the splashes. Sun et al. (2015) proposed a MPS and modal superposition coupled method to study the 2D flexible symmetric wedge dropping into water problem. Hwang et al. (2015) developed a MPS-FEM coupled method to simulate a wet drop with deformable wedge.
Comparative studies of 3-D LNG tank sloshing based on the VOF and IMPS methods
Wang, Jianhua (Shanghai Jiao Tong University) | Wan, Decheng (Collaborative Innovation Center for Advanced Ship and Deep-Sea Exploration) | Chen, Gang (Shanghai Jiao Tong University) | Huang, Wenhua (Collaborative Innovation Center for Advanced Ship and Deep-Sea Exploration)
Abstract The present work is focused on the comparative study of two numerical methods, i.e. MPS method and VOF method, for the prediction of sloshing in LNG tank. Numerical simulations are carried out by an in house meshless solver MLParticle-SJTU based on improved Moving Particle Semi-Implicit (IMPS) method, and VOF based CFD solver naoe-FOAM-SJTU developed on the open source platform OpenFOAM. Several sloshing conditions with different kinds of tanks are applied to validate the present two numerical methods. The time histories of impact pressure and flow patterns are presented and compared to the experimental data. For the rectangular tank, sway motion with different excitation periods are taken into account to nvestigate the sloshing performance and validate the two simulation methods. For the membrane-type LNG tank, pitch motion with different excitation periods are simulated. According to the numerical results, the two methods can both predict the impact pressure compared with the experiment data. The VOF method and the MPS method show different flow patterns when encountering with breaking waves. Introduction Generally, sloshing is the motion of fluid with free surface in partially filled tanks and is of significant importance in the field of ship and ocean engineering. Sloshing flow is a highly nonlinear problem, which may involve complicated phenomena, such as breaking wave, highspeed impact on tank wall and overturning of free surface. Violent liquid sloshing in an oil or liquefied natural gas (LNG) ship can cause local breakage and global instability to the ship hull, and even lead to leakage of oil, and capsizing of ship (Shao et al., 2012). The influence of sloshing depends on the amplitude and excitation period of the tank motion, liquid-fill depth, liquid properties and tank geometry. When the excitation frequency is close to the highest natural frequency of the liquid, sloshing will be significantly violent, which is called resonance phenomenon. As mentioned above, it is essential to avoid the ship first natural frequency being too close to the dominant frequency of the environment condition to achieve a good motion performance. Since sloshing can be a significant factor for the safety and stabilization of ship, many researchers have conducted a lot of corresponding research work. Early studies on sloshing are usually theoretical method based strong hypothesis (Faltinsen, 1978), where flow is irrational and geometry of tank is simple. Thus, analytical solution is invalid for sloshing in membrane-type LNG tank, especially when the tank is oscillated in resonance frequency. Traditionally, the experimental researches for sloshing problems are widely used (Akyildiz and รnal, 2005; Bulian et al., 2014; Kim et al., 2015; Lugni et al., 2006) and experimental results can validate the numerical solutions.
Abstract Resistance prediction for a ship is one of the vital tasks at the design stage. In this paper, the in-house multifunction solver naoe-FOAMSJTU based and developed on the open source code OpenFOAM is applied to study the resistance, sinkage & trim characteristics and the local flow around the stern of Japan Bulk Carrier (JBC) with and without an energy saving device sailing in calm water. A VศฆV (verification and validation) method is used to determine the errors and uncertainties. The numerical results of the presented paper agree very well with the measurement data of model test. Introduction Resistance predictions for a ship is one of the most important tasks at the design stage in order to ensure that ship can sail at a desired speed with the installed engine capacity and fulfill the mandatory regulations imposed by IMO such as Energy Effiency Design Index(EEDI). Since new concerns on environment and efficiency have risen in recent decades, predictions are getting more important and as a result the interests on Energy Saving Device (ESD) increased significantly. There are three different ways to predict resistance. Empirical methods are the simplest and fastest among them. Another tool for predictions is model testing which is the most reliable and accurate method. Alternative to model tests and the last method of prediction is numerical simulation. Among them, CFD method shows a huge advantage and has gained popularity in the past decades due to a more physics-based modeling, capability of handling non-linear free-surface, especially for detail presentation of flow fields, which is important to study the effect of energy saving devices. Wake equalizing duct (WED), studied in this paper, is one of the most commonly used energy saving devices for improving the propulsion performance of a ship and reducing the propeller-excited vibrations and viscous resistance forces, especially suitable for vessels with fat stern such as bulk carrier and oil tanker. Wake equalizing duct is fitted in stern before propeller, the general function of it is set to improve uniformity of propeller's inflow field, thus homogenizing the wake and improving hull efficiency. Also, the wake equalizing duct can accelerate the flow by means of the lift created by the aerofoil shape of the duct cross-section.
- Energy (0.87)
- Transportation > Marine (0.73)
- Transportation > Freight & Logistics Services > Shipping (0.73)
Abstract The objective of this study is to predict and analyze the viscous flow and the ship-ship interaction between two different tankers KVLCC2 and Aframax advancing in shallow water with same speed and with a fixed separation distance by solving the unsteady RANS equations in combination with the k-ฯ SST turbulence model. The computational results of the resistance, lateral force, yawing moment, as well as wave height measured by the wave gauge are validated against EFD conducted in Flanders Hydraulics Research (FHR) towing tank. Though the error for case A is not so satisfactory by up to 73%EFD, the tendency is agreed well with EFD data. Moreover, the error case B is much batter by less than 6.25% for both ships. For better understanding of the ship-ship interactions, the wave pattern of the free surface, surface pressure distribution of the ship hull, the asymmetric ship wake and vortex system are also given. Introduction With the birth of the very large crude carriers (VLCC) and ultra large crude carriers (ULCC), which have been proven to be one of the best solution to satisfy the demands of oil transportation, a new problem arose. Many ports, which may not be deep enough or have narrow entrances or small births, are not suited to receive ships of such big seizes. On the other hand, a commercial ship which is not under navigating is just a very expensive warehouse. Therefore, the time spent in the harbor, which is mainly determined by the time needed to unload and re-load the oil, must be as short as possible. Both the two problems can be solved by the lightering operation. However, manoeuvring such large vessels without the assistance of tugs at a precision of meters is highly difficult. On top of these difficulties, hydrodynamic interaction forces take place, which influence greatly the relative motion of the vessels. This can result in accidents with important consequences as oil spills or severe damage to the vessels.
- Transportation > Marine (1.00)
- Energy > Oil & Gas (1.00)
- Transportation > Freight & Logistics Services > Shipping > Tanker (0.53)
Abstract This paper presents a numerical study of flow over a circular cylinder at high Reynolds number using detached-eddy simulation (DES). The k-C Shear Stress Transport (SST) based DES model is selected for handling massively separated flow. The computations are performed with an Otype grid and the instantaneous and statistical characteristics of the turbulent wake flow behind the cylinder is extensively studied. The results are compared with the available experimental data, as well as the previously published numerical data obtained from Reynoldsaveraged Navier-Stokes (RANS) and large-eddy simulation (LES). The incompressible flow assumption and finite volume discretization is adopted. All the works are carried out with the use of OpenFOAM toolbox. Introduction The Reynolds-Averaged Navier-Stokes (RANS) equations are widely used to model turbulence in industrial applications for its cheap costs. RANS do not mean to resolve any turbulent flow structures at any scale, but to model the time-averaged turbulent properties using varies kinds of mathematical formulations. All turbulent fluctuations are eliminated during the time averaging. Therefore, it is inaccurate to predict massively separated flows which contain detached eddies at different length scales. While the unsteady RANS (URANS) attempts to solve the problem but with little efforts. The large-eddy simulation (LES) and direct numerical simulation (DNS), on the other hand, is capable for accurately predicting unsteady flows since most turbulence eddies are resolved, but the cost is expensive mainly because the dense grid resolution at boundary layer and small time steps. Detached-eddy simulation (DES), as a hybrid RANS/LES method, combines the best part of the two worlds. The basic idea of DES is to model the attached flow near wall and to resolve the detached and free shear flows in the other regions. The original version of DES was proposed by Spalart et al. (1997). The version, referred as DES97 here, is a modification of the one-equation Spalart-Allmaras eddy viscosity model (Spalart and Allmaras, 1994). DES97 was soon discovered to be sensitive to wallparallel grids in the boundary layers. It is caused by the simple and crude split of RANS and LES region. The improper arrangement of wall-parallel grids spacing results in the wrong regions for RANS and LES. The region in outer part of boundary layers which should be modeled by RANS turns to be resolved by LES. While the grid refinement is not enough to resolve all the eddy viscosities. A new version called delayed DES (DDES) was proposed by Spalart et al. (2006) to address that issue. Based on the basic idea of the hybrid model, Menter el al. (2003) also proposed the SST based DES model by modifying the two-equation SST model (Menter, 1994). The SST version of DES provides a "shield" which can prevent grid induced separation (GIS) and address model stress depletion (MSD).
Numerical Prediction of KCS Self-Propulsion in Shallow Water
Wang, Jianhua (Shanghai Jiao Tong University) | Liu, Xiaojian (Collaborative Innovation Center for Advanced Ship and Deep-Sea Exploration) | Wan, Decheng (Shanghai Jiao Tong University) | Chen, Gang (Collaborative Innovation Center for Advanced Ship and Deep-Sea Exploration)
Abstract The present work is focused on the numerical prediction of ship selfpropulsion in different shallow water conditions: H/T=2.0 and H/T=1.2 (H is the depth of water and T is draft of ship). The KRISO Container Ship (KCS) model is used in the present simulation. Numerical computations are carried out by our own solver naoe-FOAM-SJTU, which is developed on the open source platform OpenFOAM and mainly composed of a dynamic overset grid module and a full 6DoF motion module with a hierarchy of bodies. A proportional-integral (PI) controller is applied to adjust the rotational speed of the propeller to achieve the desired ship speed. The simulated results, i.e. the rate of revolution of propeller n, propulsion coefficients, are compared to the experimental data provided by Flanders Hydraulics Research (FHR) in SIMMAN 2014. Good agreements show that the present approach is applicable for self-propulsive prediction in shallow water. Introduction Self-propulsion, which is closely bound up with energy consumption, is one of the most important characters in ship performance. However, due to the economic slump, the world shipping industry has now been experiencing difficulties. And the coming out of EEDI proposed by IMO also makes it essential to accurately estimate ship self-propulsion in its design stage. Up to the present, the main method for predicting ship self-propulsion is still model scale experiments in a conventional towing tank. With the performance of computers boosting in the past few decades, tremendous advances have been made in the development of Computational Fluid Dynamics (CFD) on ship hydrodynamics. Thus, direct simulation of ship resistance and seakeeping becomes feasible. However, great challenges remain in the area of ship hydrodynamics, especially for fully appended model with rotating propellers and rudders, i.e. self-propulsion and free maneuvering computations. The dynamic overset grid method, including a hierarchy of bodies that enable computation of full 6DoF and moving components (rudders, propellers), makes it possible for computing complex motions, including problems like ship motion in large-amplitude waves, selfpropulsion and free maneuvering with rotating propellers and moving rudders.
- Transportation > Marine (1.00)
- Energy > Oil & Gas > Upstream (1.00)
- Information Technology > Mathematics of Computing (0.48)
- Information Technology > Software (0.34)
Abstract A practical multi-objective optimization tool, named OPTShip-SJTU, is utilized here to optimize both of the resistance and seakeeping performance for a surface combatant DTMB Model 5415. During the optimization procedure, free-form deformation (FFD) method is used as parametric hull surface modification technique to produce a series of alternative hull forms subjected to geometric constrains. The Neumann-Michell (NM) theory and an extension of Bales seakeeping ranking method are implemented in the evaluation module of the tool, and to predict the wave-making drag and Bales seakeeping rank factor R respectively. An optimized Latin hypercube sampling (OLHS) method is employed to generate 60 samples within the design space, and an approximation model is established in terms of these samples and corresponding predictions. A muti-objective genetic algorithm, NSGA-??, is adopted to produce pareto-optimal front. The numerical optimization results are analyzed, and the availability of the OPTShip- SJTU is confirmed by this application. Introduction The optimization of hull forms to improve the hydrodynamic performances has attracted attention in the recent years. In the past, a large number of alternative hulls were proposed according to the experience of ship designers and tested by model experiments before the final design was obtained, obviously, this method is in low effect and uneconomical. At present, with the rapid development of numerical methods and optimization algorithms, advanced modification methods, evaluation tools and optimizers have been proposed and integrated together to form various numerical optimization platforms, and they have been presented in a huge body of literature. Kim et al. (2010) used two approaches including shifting method and radial basis function interpolation to modify the hull surface, and a practical CFD tool and Bales seakeeping ranking method were used to predict the objective functions associated with resistance and seakeeping. Eventually, valid results were obtained using the MOGA algorithm for optimization. Tahara et al. (2011) took a fast catamaran as the initial design and the numerical optimization was performed based on an URANS solver, a potential flow solver and global optimization (GO) algorithms. The HSSL-B geometry was successfully optimized and an experimental campaign was carried out for validation. Zhang et al. (2015) studied the minimum total resistance hull form design method based on potential flow theory of wave-making resistance and considering the effects of tail viscous separation, and the nonlinear programming method was chosen as the optimization scheme. Campana et al. (2006) present the fundamental elements of a SBD environment for shape optimization. Both of the CFDSHIP-Iowa and MGShip were used as simulation tools, and CAD-free and CAD-based techniques had been adopted for shape grid manipulation. Additionally, GA approach was utilized in the computations and approximation model management optimization (AMMO) was employed to reduce the computational efforts. Based on above techniques, our in-house hydrodynamic optimization tool, OPTShip-SJTU, was developed for numerical multi-objective optimization of ship hull (Wu, Liu and Wan, 2015).
- Transportation > Marine (0.60)
- Shipbuilding (0.60)