Flow boiling experimental data for liquid hydrogen (LH2), liquid methane (LCH4), and liquid oxygen (LO2) were compiled. The collected data represent saturated nucleate/forced convective evaporation boiling (pre-CHF), and film/dispersed droplet boiling (post-CHF). For LH2, pre-CHF data were compiled from six sources, and post-CHF data were compiled from 12 sources. Useful pre-CHF boiling data for LCH4 were found only in two sources, and no data for post-CHF boiling of LCH4 could be found. For LO2, pre-CHF (only two data points) and post-CHF data could be extracted only from a single source. The predictions of 15 pre-CHF and nine film/dispersed flow boiling correlations are compared with data for the three cryogens. The post-CHF boiling data are only compared with appropriate models and correlations based on surface geometry and orientation. Predictive methods with best agreement with experimental data are identified.

For pre-CHF boiling the correlations of Steiner and Taborek (1992) and Bennett and Chen (1980) perform best for LH2, and can predict the bulk of the heat flux data within an order of magnitude. They can predict respectively, 81% and 79% of the data within a factor of two. The correlations of Steiner and Taborek (1992) and Liu and Winterton (1991) perform best for pre-CHF boiling of LCH4 and can predict, respectively, 100% and 98% of the data within a factor of two. For LO2, the correlations of Steiner and Taborek (1992) and Kandlikar (1990) both predict the only two available data points within a factor of two. For post-CHF, the correlations of Shiotsu and Hama (2000) and Miropolski (1963) perform best for LH2, and can predict the bulk of the existing heat flux data within an order of magnitude. They can predict respectively, 76% and 73% of the data within a factor of two. The correlations of Shiotsu and Hama (2000) and Groeneveld v5.7 (1973) perform best for LO2, and can predict the bulk of the existing heat flux data within an order of magnitude. They can predict respectively, 96% and 80% of the data within a factor of two.

编制了液态氢（LH2），液态甲烷（LCH4）和液态氧（LO2）的沸腾实验数据。收集的数据表示饱和的成核/强制对流蒸发沸腾（CHF前）和薄膜/分散液滴沸腾（CHF后）。对于LH2，从6个来源汇编了CHF之前的数据，而从12个来源汇编了CHF之后的数据。仅在两个来源中找到了LCH4有用的CHF沸腾前数据，而找不到LCH4的CHF沸腾后数据。对于LO2，只能从单个来源提取CHF之前的数据（仅两个数据点）和CHF之后的数据。将15种预CHF和9种薄膜/分散流沸腾相关性的预测与三种制冷剂的数据进行了比较。CHF后的沸腾数据仅与基于表面几何形状和方向的适当模型和相关性进行比较。

对于CHF之前的沸腾，Steiner和Taborek（1992）以及Bennett和Chen（1980）的相关性对于LH2表现最佳，并且可以在一个数量级内预测大量的热通量数据。他们可以在两倍的范围内分别预测81％和79％的数据。Steiner和Taborek（1992）以及Liu和Winterton（1991）的相关性对CHF沸腾前的LCH4沸腾效果最好，并且可以分别在两个因子内预测100％和98％的数据。对于LO2，Steiner和Taborek（1992）和Kandlikar（1990）的相关性都预测了因子2内仅有的两个可用数据点。对于后CHF，Shiotsu和Hama（2000）以及Miropolski（1963）的相关性对于LH2表现最佳，并且可以在一个数量级内预测大量现有热通量数据。他们可以分别预测 数据的76％和73％在两倍之内。Shiotsu和Hama（2000）和Groeneveld v5.7（1973）的相关性对于LO2表现最佳，并且可以在一个数量级内预测大量现有热通量数据。他们可以在两倍的范围内分别预测96％和80％的数据。