On the distortion of propeller inflow wakes during RANS calculations on rotating grids

Authors
  1. Hally, D.
Corporate Authors
Defence R&D Canada - Atlantic, Dartmouth NS (CAN)
Abstract
When calculating the flow around a propeller, a measured ship wake is often used to include the effects of the flow near the stern of the ship. Typically the wake is measured at the location of the propeller; however, in a RANS calculation, the measured wake is applied as a boundary condition upstream of the propeller. This report describes a study to determine how much the wake is distorted as it is advected from the upstream boundary to the propeller disk. The RANS solver ANSYS CFX was used on unstructured grids of different node densities. An initial study of the advection of mathematically defined axial wakes was used to develop a simple model of the attenuation and dispersion of the harmonics of the wake field. Three real wakes were then studied: one from an unappended twin screw ship, one from an appended twin screw ship, and one from a single screw ship. Each wake was treated both as an axial wake and as a full three-component wake. Moderate grid density is sufficient to preserve the hull boundary layer and shaft wakes for the axial twin screw wakes; however, grids of the order of tens of millions of nodes (without consideration of the propeller itself) would be necessary for accurate preservation of the wakes of shaft brackets or the deep wakes of single screw ships. Three-component wakes are naturally self-distorting due to advection by the tangential components, so accuracy in all three components cannot normally be achieved. A good compromise is to replace the wake

Il y a un résumé en français ici.

Report Number
DRDC-ATLANTIC-TM-2010-228 — Technical Memorandum
Date of publication
01 Nov 2010
Number of Pages
68
DSTKIM No
CA036100
CANDIS No
535598
Format(s):
Electronic Document(PDF)

Permanent link

Document 1 of 1

Date modified: