CFD Verification and Validation Study for a Captive Bullet Entry in Calm Water
Author Maximiano, A., Vaz, G. and Scharnke, J.
Title CFD Verification and Validation Study for a Captive Bullet Entry in Calm Water
Conference/Journal OMAE ASME 36th International Conference on Ocean, Offshore and Arctic Engineering, Trondheim, Norway
Paper no. OMAE2017-61666
Month June
Year 2017


Abstract
As a step towards complex impact loads cases, e.g. lifeboat drop tests or ship/platform slamming in waves, a verification and validation (V&V) study is carried out with an open-usage community based CFD code ReFRESCO for a simple impact load test case: a captive axisymmetric generic lifeboat shape (bullet) that penetrates the water surface at a constant velocity and angle of attack. The quantities of interest are the body fixed longitudinal force FX, vertical force FZ, and pitch moment MYY.
The influence of the iterative convergence level, domain size and free surface modelling are investigated. Seven different grids and four time steps were used to assess the grid and time step sensitivity, in a total of 28 calculations. For the tested grids and time steps it was found that the results are more sensitive to the grid resolution than to the time step. The pressure distribution on the hull is correlated with the trends observed in the loads, and the relation between between relative and static pressure is found to be important for the calculated loads. An experimental test campaign was previously carried out by MARIN, and its results are used to validate the simulations performed. A very good match between experiments and simulations is found.
A V&V study is performed for the quantities of interest at nine different time instants covering the impact phase. The numerical uncertainties are obtained from a solution verification procedure [1]. The experimental uncertainties are estimated, and a validation exercise carried out according to the ASME standards [2]. The outcome of the validation exercise is an estimated 95% confidence interval for the modelling error, M. For FX the modelling error is below 15 N, for 8 out of 9 time instants. For FZ the modelling error is below 14 N, except at the time instants where, due to vibrations in the experimental setup, a larger value (up to 23 N) is found. For MYY the modelling error is under 5Nm. These results provide confidence in ReFRESCO for the simulation of free surface impact flows.

   Subject to copyright regulations
Disclaimer | print