Shipborne wind measurements corrected for airflow distortion by computational fluid dynamics

During the FETCH (Flux, Etat de la mer et Télédetection en Condition de fetcH variable or Flux, sea statE and remoTe sensing in Condition of variable fetcH) experiment, conducted in March-April 1998 in the Mediterranean Sea (Golf du Lion), several platforms were deployed. Two research aircraft (Merlin-IV and Fokker-27), one research vessel (L'Atalante), one moored plate-form (ASIS, Air-Sea Interaction Spar) and two moored buoys provided atmospheric measurements.

To obtain turbulent flux paramerisation and comparison from different platforms and methods, turbulent sensors and motion packages were deployed over the moored ASIS buoy and on the ship L'Atalante. One aircraft, the Merlin-IV, was also equipped for turbulent measurements.

Well forward in the bow of this ship, a special pylon was erected and several instruments were mounted on its top, 19 meters above the sea surface. In particular, the bow pylon was instrumented for fast measurements with a sonic anemometer, a refractometer and a motion sensor.

Turbulent fluxes were calculated using both the direct eddy correlation method and the inertial dissipation method. The data were recorded for 5 weeks, including Mistral/Tramontagne wind events. Results may be compared platform to plaform and method to method.

Airflow distortion caused by the ship and around its hull, decks, mast and pylon is taken into account and is simulated in three dimensions by a computational fluid dynamics software. The modeling code used is a finite volume one developped by Fluent Inc.

To use the code, the chosen body must be paneled. The full ship model of L'Atalante, from stem to stern, is so meshed and is enclosed in a large volume of measurements to simulate a wind tunnel. The simulation allows the relative wind direction and speed to be specified in a large range of values. The model takes count of the thermodynamical characteristics of the sea surface and of the materials of the ship. Some motions of roll and pitch are even tested to know the limits of the model.

In this paper, the complete comparison for the momentum flux are not presented but only the errors caused by airflow distortion around the L'Atalante are discussed. For example, the pressure coefficient of the airflow calculated by the model in the volume of measurement of each of the sensors, is set into the classic equations of the friction velocity and of the drag coefficient. So, the results of the code correct directly the data collected on board of the ship with respect to the relative wind.