In many practical situations the erosive wear process until failure may last several months, with the eroded surfaces gradually changing their surface topology. The geometry modifications can then considerably affect the performance of the equipment, such as in wind turbines and rotating machinery. Such geometry modifications are then of interest to predict the life of the equipment as well as the consequences on performance, for instance. Simulations accounting for geometry changes are typically CPU-intensive because they are inherently transient, and involve mesh adaptations and/or remeshing. In this work, we analyze the impact of geometry modifications on the erosive process in three test cases. To reduce the associated computational cost, a simpler approach is experimented: the fluid flow is calculated once, and assumed not to change with the subsequent geometry deformations. Results from the most rigorous approach, $i.e.$, solving for the fluid flow every timestep; from this simplified approach, herein named frozen flowfield; and from the static mesh are compared against experimental results for the test cases. Expectedly, the rigorous approach is the one that best matches the experiments in all cases, whereas the simplified procedure produces geometries in reasonable agreement with experiments. By using the simplified approach, considerable savings in CPU times are obtained, while still reproducing the shapes observed experimentally.

在许多实际情况下，直至失效的侵蚀磨损过程可能会持续数月，侵蚀表面会逐渐改变其表面拓扑结构。几何形状的修改会显着影响设备的性能，例如在风力涡轮机和旋转机械中。例如，这样的几何修改对于预测设备的寿命以及对性能的影响是有意义的。考虑几何变化的模拟通常是 CPU 密集型的，因为它们本质上是瞬态的，并且涉及网格调整和/或重新划分网格。在这项工作中，我们在三个测试案例中分析了几何修改对侵蚀过程的影响。为了减少相关的计算成本，实验了一种更简单的方法：流体流量计算一次，并假设不随随后的几何变形而改变。最严格的方法的结果，$\mathrm{一世}.\mathrm{电子}.$, 求解每个时间步长的流体流动；从这个简化的方法，这里命名为冻结流场；和来自静态网格的数据与测试用例的实验结果进行比较。可以预料，严格的方法是在所有情况下与实验最匹配的方法，而简化程序产生的几何形状与实验合理一致。通过使用简化的方法，可显着节省 CPU 时间，同时仍可再现实验观察到的形状。