Validation of the VECTIS steady-state solver

Abstract

Wind speed measurements are obtained from anemometers located on research ships. Even though the anemometers are usually positioned in well-exposed locations the presence of the ship’s hull and superstructure distorts the airflow to the anemometer and biases the wind speed measurements. Previous studies have shown biases of up to 10 % for bow-on flows, and that the biases generally increase for other wind directions. Corrections for the effects of the flow distortion are vital, as these data are used for satellite validation and in climate related studies. Therefore, the computational fluid dynamics (CFD) package VECTIS is used to numerically simulate the airflow over ships and derive corrections for this effect.

A VECTIS simulation of one ship at one wind direction currently takes approximately one month to perform on a typical UNIX workstation. Therefore, it would be impractical to study the airflow over a large number of research ships and/or a large number of wind directions. A faster method (the “steady-state solver”) for VECTIS simulations has been available for some time, but requires significant increases in computational speed and memory which have only recently become widely available. This report presents a comparison of VECTIS simulations using the steady-state solver with both previous VECTIS studies and in situ wind speed measurements.

Use of the steady-state solver requires a higher mesh density but also cuts model convergence times from weeks to days, allowing fine-resolution models to be run without impractical time constraints. The results of this study show that in regions where the flow distortion is high, the increased mesh density results in significant improvement in the comparison between modelled and in-situ wind speeds.

More information

Identifiers

Catalogue record

Export record