Sunteți pe pagina 1din 7

Karman 1929 Simple method/ not true/ stationary free surface/ used for DRA >30

Wagner 1931 More realistic/ added mass theory/ the free surface rise at the interface/ used DRA<30
 Objectives:
 Validation of existing high quality experimental data of predicting 2D slamming loads on high speed catamarans using CFD
 Understanding the mesh quality required for accurate simulation of flows with a high rate of change, such as catamaran impacting
with water.
 Develop a new method by the aid of CFD to calculate the slamming loads on asymmetric large catamarans in a cost effective manner.

In CFD, the cornerstone is the fundamental governing equations of fluid dynamics:

Continuity, momentum and energy equations. These equations are mathematical statements of physical principles upon which the whole fluid
dynamics is based:

1. mass is conserved

2. Newton's second law

3. energy is conserved

The equations can be obtained in various forms, but for most hydrodynamics theory, the particular form of the equations used makes little
difference. The equations are the same whether the flow is around a ship, over a submarine, through a platform, past a cylinder, past a car or an
airplane. What separates the cases, are the flow fields determined by the boundary conditions. These conditions dictate the particular solution to
be obtained from the governing equations.

 Current methods of predicting the vertical motions high speed catamarans:

 The main techniques used in CFD analyses are:

-BEM, FEM, FDM and FVM

Method Theory Application Advantages Disadvantages


COMET Applies Berenoulli (calculate mass) Azcueta (2001) Ensures that both Past work results, predicted higher
(FVM) and RANSE (momentum Drop tests for V-shaped mass and slamming loads than full scale
Geometry using conservation of a cell) by integrating hulls ‘2D’ momentum are trials.
(ANSYSSOLID4 across the cell, then averaging for the Angles of roll of 0 and 5 conserved without
5) centre of the cell. degrees. inducing errors.
2D strip theory Nonlinear 2D
3D Solves the fluid through the domain
2D CFD Coupled with strip theory method Free falling wedge impacting Improving the
simulation water overall accuracy of
the experimental
results
CFX Able to model
unsteady free
surface flows
RANSE CFD. Predicting the forces on the hull by Predicting the forces on the Most Accurate Is not economically viable with
simplified form of Navier stokes hull Results current computational power,
equation (Caponnetto 2003) complex to setup, therefore ill
suited to running a number of
different hulls in the preliminary
design stage.
Hybrid method Accurate numerical models Better crew comfort at rough Better
(simon 2010) seas and safer understanding of
boat motions

Remarks:

1.Domain:

In 1995 Arai found that the suitable dimensions for the computational space were 6 times the model breadth and 6.6 times vertically downwards
from the keel.
 Using…… CFD to model slamming of a 2D catamaran hull form.
 Solver to determine the motions of a body during slam event
 Validating …. CFD results against past work drop test data obtained by Whelan (2004)

Multiple Moving Grid Scenario:

ANSYS cells (vertex numbers) are not distributed in a well-known order. So trying this method using COMET, is complex, and inaccurate, as,
the primary grid is expanded and contracted to fit around the moving secondary grid (model).

Body Fixed Scenario:

In this method the fluid moves while the ship remains stationary in the domain. The inlet boundary is either the inlet velocity or pressure
depending on the more important to be independent variable and the outlet boundary is a pressure boundary.

Single Moving Grid Scenario:

Both bottom and top boundaries are specified as pressure boundaries; however the bottom one varies according to the variation in the pressure
head above the bottom (ship movement) in order to maintain a zero inlet velocity to the fluid. Ship moves in the space, so no distortions happens
to the primary grid.

Comparisons:

With Whelan Tank results using acceleration and peak pressures

Chine at centre bow piercer, Flap Control system (simple and effective increase in the pitch damping), Al-Kahera demihull shape at the centre bow,

Experiments were carried out on two scale models in regular and irregular sea states by D’Arcy et al. (2006), and data for this study is presented
on regular wave motions. The experiments conducted provide a valuable source of data on high speed planning craft. The experimental results
were then compared with the motion predictions of a 2D numerical model. The results presented indicate that the numerical model adopted gives
a good representation of the motions of the craft. The full analysis of the emiprical strip theory is presented in appendix B. These experimental
results are used for the validation of the 2D CFD strip method in chapter 9.

Overview of Current CFD Methods


The correct choice of mathematical model for the problem is an important prerequisite. Identifying the best trade-off between modelling
complexity and speed is of considerable economic importance as well. Synthesising CFD, experimental methods and theoretical analysis will, in
most cases, give the best overall result (Anderson, 1995).

1. There are Three Main Methods within CFD, each solving fluid models of increasing complexity:
a. Panel method:
 Assumptions: inviscid, incompressible and irrotational potential flow
b. Euler Equation:
 Assumptions: Inviscid flow, eliminating a second derivative.

c. Navier stokes Equations:


 Relate to: Conservation of momentum equations, coupled with continuity equation and other applicable equations to allow the
system to be solved. True fluid behaviour represented by a full set of equations for either Compressible or In-Compressible fluids.
However, compressibility can reduce the stiffness of the numerical system and often improves the speed of convergence for iterative
solution techniques (ANSYS, 2009).
 The problem is complicated further when introducing multiphase flow. In that case, the governing equations for both incompressible
(equations 3.9) and compressible (equations 3.2) flow are solved for each fluid, with the appropriate boundary conditions applied to
the fluid interface. These are usually continuity of pressure and shear stress as well as ideally interfluid surface tension.
2. Mesh generation and quality (finite volume mesh generation):
 Calculations can be made directly in the physical plane on a non-uniform mesh (unstructured mesh) and of any arbitrary shape.
 Mesh aspect ratio values greater than the suggested acceptable maximum of 100 will lead to round-off errors and associated
difficulties converging the discretised equations.

There are a number of commercial, academic and open source solve s available including ANSYS CFX, ANSYS Fluent, CD-Adapco, FreeCFD
and OpenFOAM. For this project, ANSYS CFX was selected for the following reasons:

• CFX is a commercial code, and has a good user interface as well as customer support. The ease of use and availability of support reduces the time spent
learning how to use the software.

• CFX is a validated code, and this removes the need to develop or verify a new code.

• The University of Southampton has 32 parallel CFX licenses available.

• CFX has the ability to allow user defined functions to be written in FORTRAN.
SPH:

Preliminary work by Shahraki et al. (2011) has shown the potential of SPH to model the slamming of multihulls. It is proposed that SPH
presents significant advantages over other CFD approaches since it has the inherent capacity to capture free surfaces, sprays and complex
kinematics.

S-ar putea să vă placă și