MPF-BENCH
收藏资源简介:
MPF-BENCH数据集是由爱荷华州立大学研究人员创建的大型数据集,包含11000个模拟,涵盖100万个时间快照,使用验证良好的Lattice Boltzmann方法框架生成。该数据集记录了多种密度比、粘度比、雷诺数和Bond数的两相流动现象,包括波纹模式、气泡和液滴动力学及破裂过程。该数据集为科学机器学习在多相流问题中的应用提供了丰富的训练和评估资源。
MPF-BENCH dataset was developed by researchers from Iowa State University. It is a large-scale dataset consisting of 11,000 simulations spanning 1 million temporal snapshots, generated using a well-validated Lattice Boltzmann method framework. This dataset documents two-phase flow phenomena across varying density ratios, viscosity ratios, Reynolds numbers, and Bond numbers, including ripple patterns, bubble and droplet dynamics, and breakup processes. It provides a rich training and evaluation resource for scientific machine learning applications targeting multiphase flow problems.




