Analysis of free surface anti-roll tank using URANS. Verification and validation

To prevent excessive roll motion of ships operating in seas, damping systems are often required. Exterior systems can be used like, bilge keels or active stabilizer fins, or interior systems like anti-roll tanks (ARTs). There are mainly two sorts of ARTs, i.e. free surface tanks and Utype tanks. Both types have been studied extensively in the past, e.g. by Watts (1883), Frahm (1911) and Verhagen and Wijngaarden (1965), but also more recently, e.g. by Lee and Vassalos (1996), Kerkvliet et al. (2014) and Carette (2015). The content of this paper is restricted to the free surface type ART, which is nowadays often used to increase the roll damping of ships passively. The main advantages of such an ART are the large damping moment at small roll amplitudes and the ease to adapt the response by changing the water level. The response of the tank is highly frequency and amplitude dependent with a strong non-linear character (Carette et al., 2016). Also the shape of the interior geometry, e.g. additional struts, plates or other flow obstructions, will have an effect on the response, which makes it difficult to predict the response by analytical models. Therefore, systematic oscillation tests are often performed by modelscale experiments or by use of Computational Fluid Dynamics (CFD). This paper shows the response of a twodimensional (2D) and three-dimensional (3D) modelscale free surface ART using the CFD code ReFRESCO (www.refresco.org). The objective of this paper is to show which issues are important when CFD is used as a research and design tool. A verification and validation study is performed to determine numerical settings to obtain a good trade-off between accuracy and computational costs. The CFD results are validated against model-scale experimental results, based on the work of Carette (2015). The results show that CFD can be used as a simulation driven design tool to accurately predict the response of an ART.