Two-group interfacial drag force model for dispersed gas-liquid flows in subchannels of a vertical rod bundle

Subchannels are the unit geometric divisions in rod bundles, enabling detailed thermal-hydraulic analysis in bundle structures. Subchannel analysis codes (SACs) widely adopt the two-fluid model for two-phase flow simulations at subchannel-averaged spatial scales. The interfacial drag force (IDF) is a major component of the interfacial momentum transfer terms in the two-fluid momentum formulation. It affects the predictive accuracy of the key flow parameters. The drag feature depends on the bubble shapes and sizes, which vary significantly in dispersed flows. However, the existing IDF models in SACs treat all bubbles as a single population and rely on empirical flow regime transition criteria. This study aimed to develop a flow-regime-independent and physically robust IDF model for subchannels by classifying bubbles into two groups based on drag features (2G approach). Spherical and distorted bubbles formed group one; cap, slug, and churn-turbulent bubbles formed group two. This study developed the 2G IDF model based on the drag coefficient approach, which formulated the IDF as the product of the interfacial area concentration (IAC), the drag coefficient, the relative velocity, and the shape factor. These component parameters were investigated and modeled for each group across different subchannel types: interior, edge, and corner. Local data from air-water dispersed flows in an 8 × 8 rod bundle were post-processed to obtain subchannel-averaged data for 2G IDF and component parameters. The post-processed data were used for model development and evaluation of the 2G IDF in dispersed flows. For interior subchannels, the G1, G2, and total IDFs are predicted with mean absolute relative errors (MAREs) of 20.2%, 23.8%, and 15.0%, respectively. For edge subchannels, the G1, G2, and total IDFs are predicted with MAREs of 42.3%, 46.9%, and 10.4%, respectively. For corner subchannels, the G1, G2, and total IDFs are predicted with MAREs of 35.9%, 64.5%, and 17.3%, respectively. The applicability of the developed 2G IDF model to steam-water boiling flows was also examined.