Analysis of U-Type Anti-Roll Tank Using URANS: Sensitivity and Validation

Abstract

The roll motion of ships operating in a seaway is often limiting operations. These limits could be due to, e.g. maximum acceleration, green water, capsize risk or just comfort. Therefore additional roll damping is desired to prevent uncontrolled roll motion. Different means are available to decrease the roll motion of a ship, amongst other these include bilge keels, active fin stabilizers (either for forward or zero speed) and U-shape or free surface anti-roll tanks (ART). The amplitude and phase of the roll opposing moment resulting from the water that moves inside the ART are a function of the geometry of the tank and especially its internal damping. Due to the complex and non-linear nature of this flow, the use of Computational Fluid Dynamics (CFD) was chosen to analyse the details of the flow inside the tank and its anti-roll performance. The present paper focuses on the sensitivity and validation of the anti-roll performances of passive U-type ART using CFD. For this, the incompressible Unsteady Reynolds Averaged Navier-Stokes (URANS) code ReFRESCO was used. The sensitivity on the results for the U-tank is analysed by varying the grid resolution and the numerical time step. The two-dimensional (2D) full-scale and Froude based model-scale ReFRESCO results are compared to 2D and 3D full-scale CFD results of Delaunay (2012) [1] and Thanyamanta and Molyneux (2012) [2] and validated with model-scale experimental results of Field and Martin (1975) [3] and MARIN experimental results by Gunsing et al. (2014) [4]. This paper shows the influence of the convective scheme for capturing the free-surface interface and provides recommendations for a time step and grid resolution to effectively calculate the roll damping of an ART.

Numerical and Experimental Investigation of the Dynamics of a U-Shaped Sloshing Tank to Increase the Performance of Wave Energy Converters

“…In this chapter, the lumped parameter model in Equation ( 5) is enhanced via experimental free-decay tests of the U-tank. The tuned model is described the following Equation ( 14), in which the calibration is made on the inertial term via the mass coefficient K in Equation (15), and the damping term b * ττ in Equation (16), by means of the damping term q in Equation ( 17):…”

“…Bhushnan et al [14] proposed a method to estimate the damping coefficient from curve regression of a free decay test in 2D and 3D numerical simulation and stated that the damping is quadratic with the velocity and most of the head losses are due to the bends of the U-tube. Kerkvliet et al [15] tested and validated a new CFD code ReFRESCO with the experimental results and demonstrated the CFD code's effectiveness in calculating the roll damping of a U-shaped ART.…”

Kinematics and force analysis of a novel offshore crane combined compensation system

“…The three-post platform only compensates ship's roll and pitch which means u 2x = u 1x , u 2y = u 1y , h 2 = 0; 2. The active heave compensation system compensates crane tip's vertical motion, which means 200,000 kg m 4 30,000 kg m 5 23,000 kg m p 10,000 kg L r 6 ¼ constant, and the trajectory tracking is expressed in equation (39).…”

“…Nowadays, it is increasingly common for some ships to use dynamic positioning system to control ship’s yaw, surge and sway, thus keeping the ship in the desired position and course. 2–4 Moreover, there are several different devices that aim at reducing and controlling ship’s roll including water tanks, 5–7 gyrostabilizers 8,9 and fins. 10,11 Nonetheless, there is no mature technology or equipment to control ship’s roll, pitch and heave simultaneously until now, and this poses great challenges to offshore crane operations.…”

Numerical Simulation of Nonlinear Free Surface Water Waves Using a Coupled Potential Flow-Urans/Vof Approach

“…In a free-surface capturing method, such as the one here used, the discretization of the convective terms, both for momentum and volume-fraction transport equations, have an important effect in the robustness and in the accuracy of the solutions [15]. For the momentum a standard 2nd order unstructured-grids QUICK scheme is used.…”