Modelling of ship behaviour in wind and current: proof‐of‐concept of a method to account for arbitrary wind fields and wind field gradients in real‐time simulations
Van Hoydonck, W.; Van Zwijnsvoorde, T.; Verwilligen, J. (2022). Modelling of ship behaviour in wind and current: proof‐of‐concept of a method to account for arbitrary wind fields and wind field gradients in real‐time simulations. Version 2.0. FHR reports, 21_001_1. Flanders Hydraulics Research: Antwerp. IX, 30 + 9 p. app. pp. https://dx.doi.org/10.48607/135 Part of: FHR reports. Flanders Hydraulics Research: Antwerp, more | |
| Available in | Authors | | Document type: Project report
|
| Keywords | Harbours and waterways > Manoeuvring behaviour > Wind
Numerical modelling
| | Author keywords | CFD, wind coefficients, wind gradients, sheltering, atmospheric profile |
| Authors | | Top | - Van Hoydonck, W., more
- Van Zwijnsvoorde, T., more
- Verwilligen, J., more
| | |
| Abstract | The objectve of this report is to investgate a method to account for arbitrary wind felds and wind feld gradi‐ ents in the ship simulators at Flanders Hydraulics (FH). Instead of requiring specifc sets of coefcients that each correspond to a certain wind velocity profle, a set of coefcients (determined in a uniform velocity feld) is used that contains contributons of diferent parts of the hull: the ship is divided vertcally in a number of layers and longitudinally in a number of sectons. For each of these ship parts, the force and moment contribu‐ tons at the global ship reference point are determined from the pressure distributon. To determine the total forces and moments actng on the hull during a simulaton, a representatve wind velocity and wind directon is required for each hull part. In the current investgaton, vertcal gradients in the wind feld (as caused by the atmospheric boundary layer) and horizontal gradients in the wind feld (due to sheltering behind objects) are treated separately. For these simplifed conditons, the velocity feld is sampled at one or multple points per hull part either on a horizontal or vertcal line and the reference velocity is obtained by taking the squared average. Repeatng this for each hull part, all contributons can be added together to obtain the total wind forces and moments.For a simplifed cruise ship, partal wind coefcients are determined for diferent hull divisions. It is found that for three of the components (Cx, CM and CN), the contributons of the diferent hull parts counteract each other. For the other three force components (CY, CZ and CK), the contributons amplify each other. For example, all hull parts contain a lateral force component CY that points in the same directon, whereas for the yawing moment CN (which itself is caused by the lateral force), the frontal and af hull parts contribute opposing values to the total yawing moment.The infuence is determined on the forces and moments of the number of layers and sectons in which the cruise ship hull is parttoned. Dividing the hull in three equally high layers is sufcient for simulatng the ef‐ fects of widely varying vertcal gradients in the wind feld. For horizontal gradients in the wind feld, at least three sectons are required to simulate with reasonable accuracy non‐linearites caused by the counteracting force components CX, CM and CN. Five sectons are likely sufcient for use in a real‐tme simulaton environment.It is concluded that, qualitatvely speaking, the method works as expected for the cases that were investgated. With this method, both horizontal and vertcal gradients in the wind feld can be combined without requiring modifcatons on the ship side. It is recommended to implement a method like this in the simulator at FH in the near future. Future research could amongst others focus on how the global wind feld in the simulator can be altered locally to account for the presence of large objects close to the simulaton vessel. |
|
