CFDBench
收藏arXiv2024-02-24 更新2024-07-30 收录
下载链接：
http://re-training.In
下载链接
链接失效反馈官方服务：
资源简介：
CFDBench是一个专为评估计算流体动力学(CFD)问题中神经网络训练后泛化能力而设计的大型基准。它包含四个经典的CFD问题：盖驱动腔流、圆管中的层流边界层、通过台阶的坝流和周期性卡门涡街。数据集总共包含302,000帧的速度和压力场，涉及739个不同操作条件参数的案例，这些数据是通过数值方法生成的。
CFDBench is a large-scale benchmark specifically designed to evaluate the post-training generalization capability of neural networks for computational fluid dynamics (CFD) problems. It covers four classic CFD test cases: lid-driven cavity flow, laminar boundary layer flow in a circular pipe, dam-break flow over a step, and periodic Karman vortex street. The dataset contains a total of 302,000 frames of velocity and pressure fields, involving 739 distinct cases with varying operating condition parameters, and all data were generated via numerical methods.
创建时间：
2023-09-13
