Numerical re-creation of multi-directional waves in a circular basin using a particle based method
Numerical re-creation of multi-directional waves in a circular basin using a particle based method
复制标题
DOI：
10.1016/j.oceaneng.2020.107446
复制
发表时间：
2020-08-01
影响因子：
5
通讯作者：
Ingram, David
中科院分区：
文献类型：
--
作者：
Kanehira, Taiga;Mutsuda, Hidemi;Ingram, David
数值波浪水池是海岸和海洋工程领域研究流固耦合问题的有力工具。一旦得到很好的验证，数值波浪盆地可以提供显着的优势，实验测试方面的数量和类型的可用数据流，以及相关的成本。然而，传统的数值波浪水槽往往局限于长波峰条件的生成，并且难以正确地模拟大的流体变形沿着的波浪诱导浮体运动和结构上的压力分布。在这里，我们第一次创建，并验证，短峰波场在圆形波盆，使用基于粒子的方法来解决完整的三维Navier-Stokes方程。该模型基于爱丁堡大学的FloWave设施：几何形状可以生成多向波，基于粒子的方法可以自动模拟大的流体变形。多方向波产生不同的值的陡度和方向传播，模拟的表面高程进行了比较，实验获得的data.Acceptable协议之间的测量和模拟的表面高程值。较低陡度波的情况下，更好的性能得到证实;其中r（2）值大于0.7的所有情况下，和有效波高的误差范围在-8%和+1%之间。在更高的陡度条件下，发现生产不足更大，有效波高误差为-10%至-16%。得出的结论是，需要更大数量的粒子具有较小的半径，以准确地捕获较小的高频干扰。然而，有人建议，作为较低的频率分量以及生产的所有情况下，该模型已经适合于大多数的FSI问题，在单向和多向波场。
Numerical wave basins are a powerful tool for the study of fluid-structure interaction (FSI) problems in coastal and ocean engineering fields. Once well validated, numerical wave basins can offer significant advantages over experimental testing in terms of the number and type of available data-streams, and the associated cost. However, conventional numerical wave tanks tend to be limited to the generation of long-crested conditions, and struggle to properly model large fluid deformations along with wave-induced floating body motions and pressure distribution on structures. Here, for the first time we create, and validate, the short-crested wave fields in a circular wave basin, using a particle-based method to solve the full 3D Navier-Stokes equation. The model is based on the FloWave facility at the University of Edinburgh: the geometry enables the generation of multidirectional waves, and the particle-based-approach enables large fluid deformations to be automatically modelled. Multi-directional waves are generated with differing values of steepness and directional spreading, and simulated surface elevations are compared to experimentally-obtained data.Acceptable agreement is found between measured and modelled surface elevation values. Better performance is confirmed for lower-steepness wave cases; where r(2) values of greater than 0.7 are found for all cases, and errors in significant wave height range between -8% and +1%. Greater under-production is found for higher-steepness conditions with a -10% to -16% error in significant wave height. It is concluded that a greater number of particles with smaller radius is required to accurately capture the smaller higher frequency disturbances. It is suggested, however, that as the lower-frequency components are well produced for all cases that the model is already suitable for the majority of FSI problems in both uni- and multi-directional wave fields.