1999: Nonlinear Wave Radiation Forces

Engineers and designers are commonly faced with problems of fluid/structure interactions. These problems affect not only naval architects and ocean engineers such as ISOPE conference attendees, but also professionals in various other industrial fields such as automotive design, aeronautics, spacecraft design, internal combustion engines and turbo-machinery. To validate their solutions before bringing their project to the production stage, researchers generally resort to model simulation. Depending on the kind of problem being considered, the choice is open between physical models or numerical models. Historically, there was no choice; the experimental approach was obviously the only solution to visualizing the fluid flowing around bodies. In aeronautics, for instance, wind tunnels were systematically used for the design of aircraft up to the Sixties, and the knowledge of the phenomena involved was essentially acquired through experiments. The equivalent of a wind tunnel in naval and offshore applications is the wave basin or the towing tank.

As in the case of aeronautics, where computer codes have progressively replaced the wind tunnels, researchers are developing numerical wave tanks (NWT). These are intended to reproduce as closely as possible the flow around marine structures such as ships or offshore rigs as they are excited by ocean waves and currents. For historical and technical reasons, these computer codes are less advanced in hydrodynamics than in aerodynamics. The presence of the moving free surface, which adds one complexity level compared to the air/body interaction problem, is in large part responsible. Physical wave tanks are thus still necessary. This is clearly proven by the large number of facilities built during the last decade, and by the number of ongoing projects.

All around the world, several hydrodynamics teams are actively involved in research aimed at developing NWTs. Because they used to meet and discuss related technical matters at the annual ISOPE conference, it was proposed by Prof. C.H. Kim (Texas A&M Univ., USA) at The ISOPE Hague Conference in 1995 to establish an informal group that would gather all the participants interested in this topic. The proposal was immediately supported by the attending researchers. In 1996, Dr. A.H. Clément (Ecole Centrale de Nantes, France) was elected to head the NWT group. He initiated the annual NWT Workshop session at the ISOPE 1997 conference, in Honolulu. The main purpose of this Workshop, which takes place every year during a special session of the ISOPE Conference, is to offer the participants an opportunity to share their experience on a single topic related to NWT techniques. Every year, a technical point is discussed among the participants to be selected as the next topic to be examined. A common benchmark test case is then proposed and researched by all those interested during the year. The results, analyzed by the NWT group leader, are presented at the forthcoming conference — and the process continues. The numerical results from the contributors are stored on a free-access data bank on the Internet, where they are available to everyone, inside or outside the group.

1998: Numerical Wave Absorption. For obvious reasons, physical wave basins cannot be of infinite extent, as is the ocean they are intended to model. Large modern basins are typically 50 m to 100 m in length. The size of a numerical wave tank is closely related to the amount of computer memory it needs to run; thus, it is also limited. The infinite "numerical ocean" must be truncated at some distance as well. In physical wave basins, artificial beaches or more sophisticated devices are placed along the end walls in order to avoid the reflection of the waves, which would otherwise propagate backwards to the model and spoil the experiments. Several techniques have been devised to do the same in NWTs. Among them are numerical beaches, sponge layers, active piston absorbers and Orlanski’s conditions. The first benchmark topic of the NWT group was designed to test these techniques. The free surface of a numerical basin (either 2-D or 3-D) is given an initial deformed shape and released at t = 0. The polychromatic waves thus generated are supposed to exit the domain through the non-reflecting (absorbing) boundary. Other tests with monochromatic wave groups were also proposed. Five contributors have participated in this first benchmark series. The results were presented to the 40 attendees of the NWT workshop session at the ISOPE-98 Conference in Montréal, and published in the 1999 Brest Conference proceedings. Data files are freely available by anonymous ftp connection to ftp://ftp.ec-nantes.fr/NWT/1998.

1999: Nonlinear Wave Radiation Forces. The second benchmark topic addressed the problem of wave radiation by bodies moving at the free surface of a basin. It has been defined by the present leader of the NWT group, Dr. K. Tanizawa (Ship Res. Inst., Tokyo, Japan). This problem was solved a long time ago in the framework of linear potential flow theory, in which one assumes sinusoidal waves of infinitesimal amplitude. But in the real world, waves are far from such ideal conditions. Thus, in order to refine the design of marine structures that have to survive extreme sea-state conditions, engineers and researchers must be able to take into account the non-linearities arising from large amplitude waves, large motions of the bodies, and so on. This is precisely the aim of the Numerical Wave Tank approach.

For this second benchmark test series, a simple test case has been defined to compare the output of various computer programs and techniques. A 2-D wedge-shaped body is forced to oscillate vertically around its equilibrium position at the free surface of the fluid. Various amplitude and frequencies, which are the governing parameters of the problem, were proposed. Several physical quantities such as: wave elevation, dynamic pressure and total force on the body, are probed at every time step. Contributions from at least 7 different research teams from Japan, France and The Netherlands were compared and analyzed during the special NWT Workshop session of the 9th ISOPE Conference in Brest on Thursday, June 3. A report is available online at the following web address: http://www.srimot.go.jp/dyn/member/tanizawa/nwtws1999/index.htm

2000: Nonlinear (2-D) Wave Diffraction. A fixed free-surface piercing body is attacked by incoming regular waves generated at one end of the basin. The diffracted waves must be absorbed simultaneously in order to reach a steady state. Wave forces and runup will be computed in the time domain. This will be the topic of the next series, to be proposed for the coming year and discussed at the ISOPE-2000 Seattle Conference. The details of the benchmark will soon be specified by Dr. K. Tanizawa and posted in a specific page on the ISOPE Web site (http://www.isope.org).

Dr. A.H. Clément, ISOPE member

Laboratoire de Mécanique des Fluides (CNRS UMR6598)

(http://www.ec-nantes.fr/dhn) École Centrale de Nantes (France)