Neptuna benchmark dataset
收藏资源简介:
Neptuna数据集是由慕尼黑工业大学团队创建的高保真可压缩多相流基准数据集,专门针对冲击驱动的气泡破裂和液滴破碎等复杂物理现象。该数据集包含2.4 TB的二维和三维高分辨率模拟数据,涵盖了空气泡在水中破裂、液滴在空气中破碎以及R22气泡在空气中相互作用等多种场景,数据通过ALPACA求解器和RDEMIC方法生成。数据集旨在为机器学习代理模型提供全面的训练和评估基准,解决可压缩多相流中冲击波与物质界面相互作用的建模难题,应用于计算流体动力学、航空航天和能源工程等领域。
The Neptuna dataset is a high-fidelity compressible multiphase flow benchmark dataset created by the research team at the Technical University of Munich, specifically targeting complex physical phenomena such as shock-driven bubble collapse and droplet breakup. This dataset encompasses 2.4 terabytes of 2D and 3D high-resolution simulation data, covering diverse scenarios including air bubble collapse in water, droplet breakup in air, and R22 bubble interactions in air, with all data generated using the ALPACA solver and the RDEMIC method. The dataset aims to provide comprehensive training and evaluation benchmarks for machine learning surrogate models, address the modeling challenges associated with shock wave-matter interface interactions in compressible multiphase flows, and find applications in fields such as computational fluid dynamics, aerospace engineering, and energy engineering.




