Development and Validation of Pattern Recognition Algorithm and Suppression Pool Flow Modeling

A comprehensive understanding of the dynamics of vapor bubbles and jets discharged from vent pipes and spargers into subcooled liquid has a significant role in assessing the capability of nuclear safety systems to perform their function correctly. However, high-resolution local measurements and numerical calculations of these complex and rapidly condensing gas volumes are challenging because of rapid pressure oscillations, microscopic length, and turbulent two-phase flow timescales. Visual observation of the interior of the test section can make modern measurement techniques a viable alternative in such cases.

This paper presents work on modeling vertical (vent) and horizontal (sparger) steam injection in a water pool by applying the Eulerian-Eulerian two-fluid approach. In this work, the formation and collapse of the steam bubbles in chugging condensation mode and bubbling condensation oscillation mode are evaluated using the pattern recognition (PR) algorithm. The PR algorithm is based on video material recorded during direct contact condensation (DCC) experiments of the PPOOLEX and SEF-POOL test facilities of Lappeenranta-Lahti University of Technology. The velocity of collapsing bubbles is estimated using the PR algorithm. The accuracy of the PR algorithm is cross-checked with computational fluid dynamics simulation results. Results indicate that the presented PR algorithms provide essential information on the dynamics of phase interface in all directions. These details are advantageous for comprehensive DCC and steam-water surface instability model development. Results show that including interfacial instability modeling, i.e. the Rayleigh-Taylor interfacial area model in Eulerian two-fluid simulation, inevitably improved interface roughness and, thereby, heat transfer, which controls bubble growth and collapses.