High-Resolution Experiments of Flow Phenomena in Dead-ended Branch Lines for the Validation and Advancement of Computational Fluid Dynamics Modeling

Abstract

Thermal fatigue (TF) is one of the major degradation mechanisms that can lead to material failures affecting the safety relevant components of a nuclear power plant (NPP), such as the emergency core coolant system (ECCS) branch lines of the primary coolant circuit. Regarding ECCS piping, previous TF management programs and predictive computational models have proven insufficient. On the one hand, they appear to be overly conservative, leading to an unnecessarily large number of pipes screened for TF related issues, and on the other they do not capture all relevant physics. Correlations used to calculate the location of TF onset appear to be too case specific to be applied across diverse scenarios and configurations, while more advanced tools like computational fluid dynamics (CFD) models lack data for rigorous validation. The previous body of experimental results was too coarse to meaningfully enhance the results of computational methods for predicting the location of TF onset in branch line piping across various geometries and NPP conditions. In the present dissertation, novel, high-resolution, high-fidelity quantitative measurements of flow fields in isolated branch lines are presented. The aim of building a robust database of such data is to aid in overcoming the deficiencies of past experiments, yielding a greater understanding of the associated flow phenomena and validating predictive CFD models. The present data were acquired utilizing four experimental facilities designed and constructed to systematically investigate the flow phenomena responsible for turbulence-induced TF in isolated branch lines, including scaling effects. High-resolution measurements were obtained from the experimental facilities utilizing advanced measurement techniques such as particle image velocimetry. These measurements have clarified and solidified understanding of the flow phenomena present in dead-ended branch lines that are responsible for thermal fatigue. Comparing the results between multiple facilities has demonstrated that the complexity and demands of the measurement apparatus used to study the penetration of flow swirls in branch lines can be reduced significantly from what has previously been employed – i.e., a comparable penetrating flow swirl in a dead-ended branch line can be driven by a stirring paddle instead of a main line flow crossing the branch line opening. Quantitative measurements of the flow fields in the branch lines have aided the conclusion that Low-Re k-ε CFD models are sufficient for penetration depth predictive modeling, a stance that has been adopted by members of the NPP industry. Portions of the data presented are also being utilized in an international benchmark study in an effort to validate CFD models for penetrating flows in dead-ended branch lines.