Using the Inverse Method to Investigate Flow Models for Mixed Convection of Annular Finned Tube Heat Exchanger

Abstract

The inverse heat conduction method (IHCM) and three-dimensional (3D) computational fluid dynamic (CFD) in combination with T-k(m) are used to select the appropriate flow model and near-wall treatment for the mixed convection of the annular finned tube heat exchanger. Thus, the numerical results obtained by all k-epsilon models and near-wall treatments are yielded for 4 m/s <= V-a <= 5 m/s. First, IHCM combined with T-k(m) is applied to estimate (h) over bar and Q values. The obtained estimates of (h) over bar and Q are used as reference values for CFD. The results show that the T-k and (h) over bar results obtained by the standard (STD) k-epsilon model with the standard wall function (SWF) are closer to T-k(m) and the estimates of (h) over bar than those by the realizable (REAL) and RNG k-epsilon models with various wall functions for V-a of 4-5 m/s. The (h) over bar value obtained from the STD k-epsilon model with SWF, V-a = 5 m/s and S = 15 mm may be 1.14 times that of the RNG k-epsilon model and 0.48 times that of the REAL k-epsilon model. The (h) over bar value of the RNG k-epsilon model with SWF is about 1.6-1.7 times that with EWT for S = 5 mm and V-a = 4 m/s and 5 m/s. These differences mean that the appropriate flow model needs to be varied with V-a to obtain more accurate numerical results. The y(+) of RNG and STD k-epsilon models does not exceed 3. However, the y(+) of the REAL k-epsilon model exceeds 6. To our best knowledge, no studies have explored this issue and the present investigation tries to bridge this gap.