Steam-air reflux condensation with realistic cooling channel boundary conditions: A parametric study using a fully coupled ALE-IT method

This work presents a high-fidelity numerical study of reflux film condensation of a saturated steam-air mixture in a vertical channel that is thermally coupled to an explicit cooling-water channel. A fully coupled, pressure-based ALE interface-tracking solver is extended to a three-phase (gas–liquid–solid) formulation. A systematic parametric study spans mixture-coolant temperature difference, inlet air mass fraction, inlet gas velocity, coolant mass flow rate, cooling-flow direction (co-/counter-current), and channel length. The results show that replacing the idealized isothermal wall with a conjugate water-cooled wall is essential for realistic prediction since the isothermal assumption overestimates condensation up to 17% at low temperature differences. Co-current cooling yields 8% to 12% increase in total condensate at cases with inlet air mass fractions of 10% and higher. More importantly, co-current cooling shifts film starting point downstream, shortening the required active length. Non-condensable gases build a dominant diffusion barrier at the interface; even small air fractions can prevent complete condensation unless sufficient thermal driving is provided. Finally, for the studied operating window the channel length can be reduced by half to two-thirds with negligible loss of total condensate flow, thereby limiting sensible heat duty and improving compactness. These insights provide actionable guidance for designing compact reflux condensers under practical boundary conditions. • Extended fully-coupled three-phase model for realistic condensation. • Isothermal wall models overpredict condensation mass flow rate up. • Non-condensable gases are the dominant performance limiter. • Co-current cooling optimizes surface use and shortens channel length. • Shorter channels maintain condensation but reduce sensible heat.