Inferential framework for two-fluid model of cryogenic chilldown

TL;DR

This paper develops a probabilistic inference framework for cryogenic two-phase flow models, enabling accurate parameter estimation and improved predictions of cryogenic chilldown processes using a fast two-fluid solver.

Contribution

It introduces a novel probabilistic approach with cryogenic correlations for parameter inference in two-fluid models, enhancing predictive accuracy for cryogenic chilldown.

Findings

Accurate predictions of experimental cryogenic flow data.

Successful simultaneous optimization of multiple model parameters.

Framework applicable to both saturated and sub-cooled nitrogen flow.

Abstract

We report a development of probabilistic framework for parameter inference of cryogenic two-phase flow based on fast two-fluid solver. We introduce a concise set of cryogenic correlations and discuss its parameterization. We present results of application of proposed approach to the analysis of cryogenic chilldoown in horizontal transfer line. We demonstrate simultaneous optimization of large number of model parameters obtained using global optimization algorithms. It is shown that the proposed approach allows accurate predictions of experimental data obtained both with saturated and sub-cooled liquid nitrogen flow. We discuss extension of predictive capabilities of the model to practical full scale systems.