Toward Industrial Applicability of DNB Predictions in CFD With Improved Wall Boiling Models

The design of Pressurized Water Reactor (PWR) fuel has to rely on costly prototypical experimental campaigns, due to the inherent limitations of legacy lumped-parameter thermal-hydraulics codes. By resolving the fine spatial scales, computational fluid dynamics (CFD) methods offer the opportunity of delivering improved fuel design, taking full advantage of three-dimensional mixing and turbulence control, but must extend their predictive capabilities to multiphase conditions up to DNB, where the maturity and accuracy of the methods is still developing. In this paper, the closures developed within the Eulerian-Eulerian two-fluid model framework are assembled and extended to a general boiling formulation to deliver prediction of DNB at reactor conditions. The classic heat flux partitioning approach is advanced through selection of optimal closures for interaction length scale, nucleation site density, and bubble departure diameter, and further extended though the addition of a consistent DNB detection criterion, to reflect the near wall two phase characteristics. Assessment is performed at 138 bar against the experimental dataset for vertical pipe, where the inlet subcooling ranges from 1 to 150 K and the mass flux varies from 600 kg/(m².s) to 2650 kg/(m².s). The presented closures demonstrate consistent agreement with the experimental data, in particular in light of the calibration free application, while further work is ongoing to close the remaining gap and support the introduction of further modeling improvements.