Predicting the vibration response of an elastic structure excited by a turbulent flow is of interest for the civil and military transportation sector. The models proposed in the literature are generally based on the assumption that the turbulent boundary layer (noted TBL in the following) exciting the structure is spatially homogeneous. However, this assumption is not always fulfilled in practice, in particular when the excited area is close to the starting point of the TBL or with curved structures. To overcome this issue, this work proposes to extend two approaches generally used for dealing with homogeneous TBL, namely the spatial and the wavenumber approaches. The extension of the spatial approach to non-homogeneous excitation is relatively straightforward and gives us a reference, however with costly computation. On the contrary, extending the wavenumber approach requires more developments and assumptions that have led the authors to develop a sub-area decomposition technique (SDT). It consists in partitioning the excited area in several sub-areas, assuming a homogeneous TBL pressure field on each sub-area and neglecting certain interactions between the sub-areas. A criterion on the sub-area size as a function of the minimum wavelength of the wall pressure field is proposed on the basis of the numerical calculations. A test case on a plate with varying thickness and excited by a growing TBL allows us to highlight: (1) the interest of the SDT compared to a classical calculation considering a homogeneous TBL; (2) the efficiency of the SDT in terms of computing time compared to the spatial approach.

预测由湍流激发的弹性结构的振动响应是民用和军用运输部门的兴趣所在。文献中提出的模型通常基于以下假设：激发结构的湍流边界层（以下称为 TBL）在空间上是均匀的。然而，这种假设在实践中并不总是满足，特别是当激发区靠近 TBL 的起点或具有弯曲结构时。为了克服这个问题，这项工作建议扩展通常用于处理均匀 TBL 的两种方法，即空间方法和波数方法。将空间方法扩展到非均匀激发相对简单，并为我们提供了参考，但计算成本很高。相反，扩展波数方法需要更多的发展和假设，这导致作者开发了子区域分解技术 (SDT)。它包括将激发区域划分为几个子区域，假设每个子区域上有均匀的 TBL 压力场，并忽略子区域之间的某些相互作用。在数值计算的基础上，提出了作为壁面压力场最小波长函数的分区大小判据。在具有不同厚度并由不断增长的 TBL 激发的板上的测试案例使我们能够强调：（1）与考虑均匀 TBL 的经典计算相比，SDT 的兴趣；(2) 与空间方法相比，SDT 在计算时间方面的效率。