
A Comparative Evaluation of The Interface Capturing Schemes in A Bundle Bubbly Flow
Please login to view abstract download link
The volume of fluid (VOF) method is arguably the most prevalent CFD method. Among the numbers of interface capturing schemes (ICSs), the scheme of the piecewise linear interface calculation (PLIC) is regarded as the most faithful to the geometrical theory. However, the PLIC is also recognized for its complexity in exhaustively accounting for all potential geometrical arrangement possibilities. The tangent of hyperbola interface capturing/weighted line interface calculation (THINC/WLIC) is an alternative ICS, in which the advection of the VOF value is calculated with a considerably more simplified algorithm than that used in the PLIC. However, the THINC/WLIC has been demonstrated to deteriorate the quality of the interface by the numerical diffusion of the VOF value. In response, we developed an enhanced version of THINC/WLIC, designated THINC/Advanced WLIC (THINC/AWLIC), to achieve a more optimal balance between implementation cost and interface sharpness. While these distinct ICSs have been rigorously validated through numerical benchmarking, their efficacy in addressing complex engineering challenges remains largely unconfirmed. Accordingly, with the objective of applying these ICSs to the nuclear reactor core flow, the liquid-gas two-phase flow in a nuclear fuel-mocking pin bundles has been analysed. The comparative evaluation revealed that the PLIC and THINC/AWLIC maintains a very sharp interface, which produces realistic result. However, the computational time of the PLIC and THINC/AWLIC is nearly three times that of the THINC/WLIC. Conversely, while the THINC/WLIC simulation is relatively rapid, the resulting data differs qualitatively from that of the experiment. Of particular note are the unphysical coalescence of injected bubbles and the dissipation of bubble volumes. Furthermore, the time-averaged cross-sectional void fraction distribution of THINC/WLIC demonstrates a notable reduction in void fraction in the vicinity of the pin-gap region, in comparison against the results of the PLIC and the THINC/AWLIC.