Vorticity and convective heat transfer downstream of a vortex generator

Abstract

Vorticity generation has been identified, since the 80's, as an efficient means for enhancing heat transfer; the mean radial velocity component due to the induced flow pattern contributes to the heat removal. In the present work, momentum and heat transfer are studied in a test section designed to mimic the industrial HEV (High-Efficiency Vorticity) mixer. It consists of a basic configuration with a unique vorticity generator inserted on the bottom wall of a heated straight channel. The aim of this work is to analyze to which extend the convective heat transfer is correlated to the vorticity, as it is presumed to cause the intensification. In this case, the driving vorticity is the streamwise vorticity flux Ω, and the heat transfer is characterized by the Nusselt number Nu, both quantities being spanwise averaged. The study is mainly numerical; we have used the previous PIV measurements and DNS data from the open literature to validate the numerical simulations. It is shown that there exists a strong correlation between the vorticity flux and Nusselt number close to the vortex generator. However, the axial variation diverges for these quantities when moving downstream. The Nusselt number presents a sharp peak over the VG and decreases nearly to its basic level just behind the VG, while the vortex persists far downstream from the tab and relaxes very slowly. Heat transfer intensification at the Nusselt peak is about 100%, and reduces to about 6% downstream of the VG, the intensity of the vorticity momentum being decreased only to about 50% of its peak value at the test section outlet.

Flow and heat transfer characteristics in square channel with concave–convex vortex generators based on numerical simulations

“…However, compared with PIV and DNS, the current simulation does not generate a valley at the peak and shows a dropping‐rising velocity fluctuation at x/H = 0. The discrepancy in these results is mainly caused by lack of tab thickness in DNS and/or PIV inaccuracy in the low velocity region 30 …”

“…To verify the prediction, the simulation for validation is performed under the same conditions of the previous PIV, 7 DNS, 40 and numerical simulation 30 studies on the trapezoidal VG, so that the results from this study can be used to validate the results in the present research. However, for the near wall region treatment, the previous RSM study 30 used two‐layer model instead of standard wall treatment. The standard and two‐layer near wall treatment methods were compared here, and it was found that the results using the standard wall treatment are closer to the experimental data from PIV 7 .…”

“…In this paper, two new types of VGs (shown in Figure 3a,b, concave and convex structures) are designed based on the original trapezoidal VG from literature 30 …”

“…Configurations of different vortex generators: (a) concave VG, (b) convex VG, and (c) trapezoidal VG 30 …”

“…Inspired by the punched holes VGs and VGs combined with grooves, two new types of VGs, concave and convex structures, were designed herein. To better compare the pressure loss and heat transfer performance of the two new VGs and the trapezoidal VG, the geometric dimensions are same with the trapezoidal tab structure study 30 . Numerical simulations of the interaction between two counter‐rotating longitudinal vortices were carried out for the plate‐fin heat exchanger using two rows of delta winglet type 31 .…”