Abstract The first part of the paper presents a partitioned fluid–structure interaction (FSI) coupling for the non-uniform flow hydro-elastic analysis of highly flexible propellers in cavitating and non-cavitating conditions. The chosen fluid model is a potential flow solved with a boundary element method (BEM). The structural sub-problem has been modelled with a finite element method (FEM). In the present method, the fully partitioned framework allows one to use another flow or structural solver. An important feature of the present method is the time periodic way of solving the FSI problem. In a time periodic coupling, the coupling iterations are not performed per time step but on a periodic level, which is necessary for the present BEM–FEM coupling, but can also offer an improved convergence rate compared to a time step coupled method. Thus, it allows to solve the structural problem in the frequency domain, meaning that any transients, which slow down the convergence process, are not computed. As proposed in the method, the structural equations of motion can be solved in modal space, which allows for a model reduction by involving only a limited number of mode shapes. The second part of the paper includes a validation study on full-scale. For the full-scale validation study a purposely designed composite propeller with a diameter of 1 m has been manufactured. Also an underwater measurement set-up including a stereo camera system, remote control of the optics and illumination system has been developed. The propeller design and the underwater measurement set-up are described in the paper. During sea trials blade deflections have been measured in three different positions. A comparison between measured and calculated torque shows that the measured torque is much larger than computed. This is attributed to the differences between effective and nominal wakefields, where the latter one has been used for the calculations. To correct for the differences between measured and computed torque the calculated pressures have been amplified accordingly. In that way the deformations which have been computed with the BEM–FEM coupling for non-uniform flows became very similar to the measured results.

中文翻译：

---

BEM-FEM 耦合用于分析非均匀流动中的柔性螺旋桨并通过全尺寸测量进行验证

摘要 本文的第一部分介绍了一种分区流固耦合 (FSI) 耦合，用于在空化和非空化条件下对高柔性螺旋桨进行非均匀流动水弹性分析。所选流体模型是使用边界元法 (BEM) 求解的势流。结构子问题已使用有限元方法 (FEM) 建模。在本方法中，完全分区的框架允许使用另一种流或结构求解器。本方法的一个重要特征是求解 FSI 问题的时间周期方式。在时间周期耦合中，耦合迭代不是按时间步长执行，而是在周期级别上执行，这对于当前 BEM-FEM 耦合是必要的，但与时间步长耦合方法相比，它也可以提供更高的收敛速度。因此，它允许解决频域中的结构问题，这意味着不会计算任何减慢收敛过程的瞬态。正如该方法所提出的，运动的结构方程可以在模态空间中求解，这允许通过仅涉及有限数量的模态振型来简化模型。论文的第二部分包括全面的验证研究。为了进行全面的验证研究，专门设计的直径为 1 m 的复合螺旋桨已经制造完成。还开发了一种水下测量装置，包括立体相机系统、光学元件的远程控制和照明系统。论文中描述了螺旋桨设计和水下测量装置。在海上试验期间，在三个不同位置测量了叶片偏转。测量扭矩和计算扭矩之间的比较表明，测量扭矩比计算扭矩大得多。这归因于有效尾流场和标称尾流场之间的差异，后者已用于计算。为了校正测量扭矩和计算扭矩之间的差异，计算压力已相应放大。通过这种方式，非均匀流动的边界元-有限元耦合计算的变形变得与测量结果非常相似。为了校正测量扭矩和计算扭矩之间的差异，计算压力已相应放大。通过这种方式，非均匀流动的边界元-有限元耦合计算的变形变得与测量结果非常相似。为了校正测量扭矩和计算扭矩之间的差异，计算压力已相应放大。通过这种方式，非均匀流动的边界元-有限元耦合计算的变形变得与测量结果非常相似。