28.02.2013 Views

Practical Ship Hydrodynamics

Practical Ship Hydrodynamics

Practical Ship Hydrodynamics

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

166 <strong>Practical</strong> <strong>Ship</strong> <strong>Hydrodynamics</strong><br />

of the ship. Installed power P, cross-section area of the pipe A, andflow<br />

velocity v in the jet thruster are related by:<br />

T D Ð A Ð v 2<br />

P D 1 Ð 1<br />

2 Av2<br />

is here the efficiency of the thruster propeller. These equations yield:<br />

P v<br />

D<br />

T 2<br />

T<br />

D v2<br />

A<br />

D 0.8andv D 11 m/s yield typical relations: approximately 120 kN/m 2 thrust<br />

per thruster cross-section area and 7 kW power per kN thrust.<br />

With increasing speed, jet thrusters become less effective and rudders<br />

become more effective. The reason is that the jet is bent backwards and may<br />

reattach to the ship hull. The thrust is then partially compensated by an opposite<br />

suction force. This effect may be reduced by installing a second (passive) pipe<br />

without a propeller downstream of the thruster (Brix (1993)).<br />

5.2.8 CFD for ship manoeuvring<br />

For most ships, the linear system of equations determining the drift and yaw<br />

velocity in steady turning motion is nearly singular. This produces large relative<br />

errors in the predicted steady turning rate especially for small rudder angles<br />

and turning rates. For large rudder angle and turning rate, non-linear forces<br />

alleviate these problems somewhat. But non-linear hull forces depend crucially<br />

on the cross-flow resistance or the direction of the longitudinal vortices, i.e.<br />

on quantities which are determined empirically and which vary widely. In<br />

addition, extreme rudder forces depend strongly on the rudder stall angle<br />

which – for a rudder behind the hull and propeller – requires at least twodimensional<br />

RANSE simulations. Thus large errors are frequently made in<br />

predicting both the ship’s path in hard manoeuvres and the course-keeping<br />

qualities. (The prediction of the full ship is fortunately easier as at the higher<br />

Reynolds numbers stall rarely occurs). In spite of that, published comparisons<br />

between predictions and measurements almost always indicate excellent accuracy.<br />

A notable exception is Söding (1993a). The difference is that Söding<br />

avoids all information which would not be available had the respective model<br />

not been tested previously. The typical very good agreement published by<br />

others is then suspected to be either chosen as best results from a larger set of<br />

predictions or due to empirical corrections of the calculation method based on<br />

experiments which include the ship used for demonstrating the attained accuracy.<br />

Naturally, these tricks are not possible for a practical prediction where<br />

no previous test results for the ship design can be used. Thus the accuracy of<br />

manoeuvring predictions is still unsatisfactory, but differences between alternative<br />

designs and totally unacceptable designs may be easily detected using<br />

the available methods for manoeuvring prediction. With appropriate validation,<br />

it may also be possible to predict full-scale ship motions with sufficient

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!