Comparisons of In Situ Ship Air Wakes with Wind Tunnel Measurements and Computational Fluid Dynamics Simulations

Numerical simulations using computational fluid dynamics are frequently applied to analyze complex flow fields. However, they have to be validated by matching simulation results to those from canonical flows or experimental measurements. The objective of the present research is to compare results from numerical simulations and wind tunnel measurements for air wakes generated behind ships' superstructures to those from direct in situ measurements. Numerical simulations are performed using COBALT on an unstructured grid system, wind tunnel data are collected from a 4%-scale model, and in situ measurement data are sampled using ultrasonic anemometers mounted above an aft flight deck on a 32.9-m (108 ft)-long research vessel. Reynolds numbers are closely matched for all three approaches concurrently. Two different incoming velocity conditions are compared: a head wind condition and wind 15 degrees off the starboard bow (beta = 0 degrees,-15 degrees, respectively, where beta is the wind yaw angle). Differences in velocity and boundary layers between the three approaches are resolved using unique velocity normalization. The flow structures between beta = 0 degrees and beta = -15 degrees are quite different, i.e., there appears to be strong asymmetric vortical structures over the flight deck for beta = -15 degrees. In general, in situ, computational, and wind tunnel data all show large-scale recirculation motion behind the ship's hangar. However, there are nonnegligible differences between the simulations and wind tunnel measurements compared to the in situ measurements. Differences in velocity angles increase with the yaw angle of the incoming flow.