Abstract

To get reliable computational results, the RNG k-e turbulence model with enhanced wall treatment was validated to solve the heat transfer of supercritical RP-3 in a helically coiled tube, and models of the thermo-physical properties of RP-3 were optimally chosen. Most significantly, the grid independence was validated by two-step procedure, and the effect of boundary grids of the supercritical-temperature wall on the computational accuracy was well studied. Through adjusting boundary-layer girds’ size, four regions (increased, pseudo-convergence, decreased and convergence) of the outlet temperature Tout were obtained and analyzed. The results showed that the maximum computation errors of Tout and the pressure differential between the inlet and outlet ΔP reached 20.65% and 98.15%, respectively, indicating that boundary grids have a significant influence on computation of flow and heat transfer. Based on this, a dimensionless distance from the wall-adjacent cell to the wall y+ = Prw−1/1.78 (Prw denotes Wall Prandtl number) was recommended as a convergence point. The variation laws of viscous length scale y were discussed under different structural parameters, operation parameters, and helical lengths. An explicit model of y* was proposed to calculate the height (y) of the first boundary layer grids and refine boundary grids efficiently. A modified model for coefficient of friction factor Cf was proposed based on Rogers’s, and Nusselt number Nu was proposed based on an analogy of momentum and heat transfer. The above models about y*, Cf and Nu could apply to both the entrance region and the whole tube length, and showed good performance when Reynolds number was extended to above 70 000, or whenever the outlet temperature is below or above the critical point.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call