Numerical investigation of scale effects of Schneekluth's duct

Abstract The power-saving effect of Schneekluth's duct (wake-equalizing duct) on the required ship propulsion power is studied by means of numerical flow computations using the viscous flow code (RANSE solver) Comet. Especially the effect of the Reynolds number (the so-called scale effect) is investigated. Numerical results are compared for a ship hull with and without duct. The „E3 tanker“ is used as example. 6 duct variants are compared with each other and with the case of no duct. Computations simulating physical resistance and pro pulsion tests are performed for Reynolds numbers between 7.98 . 106 and 1.36 . 108. For the propulsion tests the propeller is not modeled geometrically, but substituted by forces acting on the fluid cells in way of the propeller disco Only axial force components are applied. The fluid forces on the hull and the wake distribution in the propeller disc are evaluated. The dependance of the computed results on the numerical scheme (central or upwind) is investigated in order to eliminate discretization errors as far as possible by extrapolating results to 100% central differences. A linear extrapolation of hull force depending on propeller thrust is used to obtain the thrust under self-propulsion conditions. This approach is accurate and efficient; it requires 2 flow computations per case. The numerical results show an acceptable agreement with experiments. In the numerical resistance tests, all duct variants cause a moderate resistance increase compared to the case of the ship without duct. In the pro pulsion test simulation apower saving effect is found only for one duct. For higher Reynolds numbers, which corresponding to full scale conditions, all investigated duct variants produced an increase of required propulsion power. In case of low Reynolds number, the amount of power saving is insignificant. Inappropriate ducts cause a substantial resistance and power requirement. Contrary to what was expected by Schneekluth and other authors it is shown that the ducts do not accelerate the fluid passing through the duct, but they decelerate it, whereas fluid passing outside but near to the duct is accelerated slightly.