Abstract

The effects of different modeling and solving approaches on the simulation of a steam ejector have been investigated with the computational fluid dynamics (CFD) technique. The four most frequently used and recommended turbulence models (standard k-ε, RNG k-ε, realizable k-ε and SST k-ω), two near-wall treatments (standard wall function and enhanced wall treatment), two solvers (pressure- and density-based solvers) and two spatial discretization schemes ( the second-order upwind scheme and the quadratic upstream interpolation for convective kinematics (QUICK) of the convection term have been tested and compared for a supersonic steam ejector under the same conditions as experimental data. In total, more than 185 cases of 17 different modeling and solving approaches have been carried out in this work. The simulation results from the pressure-based solver (PBS) are slightly closer to the experimental data than those from the density-based solver (DBS) and are thus utilized in the subsequent simulations. When a high-density mesh with y+ < 1 is used, the SST k-ω model can obtain the best predictions of the maximum entrainment ratio (ER) and an adequate prediction of the critical back pressure (CBP), while the realizable k-ε model with the enhanced wall treatment can obtain the best prediction of the CBP and an adequate prediction of the ER. When the standard wall function is used with the three k-ε models, the realizable k-ε model can obtain the best predictions of the maximum ER, and the three k-ε models can gain the same CBP value. For a steam ejector with recirculation inside the diffuser, the realizable k-ε model or the enhanced wall treatment is recommended for adoption in the modeling approach. When the spatial discretization scheme of the convection term changes from a second-order upwind scheme to a QUICK scheme, the effect can be ignored for the maximum ER calculation, while only the CBP value from the standard k-ε model with the standard wall function is reduced by 2.13%. The calculation deviation of the ER between the two schemes increases with the back pressure at the unchoked flow region, especially when the standard k-ε model is adopted. The realizable k-ε model with the two wall treatments and the SST k-ω model is recommended, while the standard k-ε is more sensitive to the near-wall treatment and the spatial discretization scheme and is not recommended for an ejector simulation.

Highlights

  • The continuous increase in energy consumption and the decrease in natural resources require a more efficient use of energy

  • The former three meshes were tested under the RKE model with the standard wall function, while the latter three meshes were tested under the RKE model with the enhanced wall treatment

  • We numerically studied the effects of different modeling and solving approaches on the simulation of a supersonic steam ejector

Read more

Summary

Introduction

The continuous increase in energy consumption and the decrease in natural resources require a more efficient use of energy. Steam ejector refrigeration systems (SERSs) utilizing low-grade thermal energy to drive environmentally friendly refrigerants can be an attractive technology for the efficient use of available energy (e.g., solar energy, geothermal energy, industrial waste heat) [2,3]. These systems have many advantages, such as reliability, limited maintenance needs and low initial and operational costs [4]. A well-designed ejector with high efficiency (efficiency here refers to the ratio of the increase of exergy of secondary flow to the decrease of exergy of primary flow) can improve the overall system performance and make SERSs economically more attractive [6]

Objectives
Results
Conclusion
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