Performance enhancement of Solar Air Heater using Artificially Roughness

Abstract

The major issue in the solar air heater is low convective heat transfer coefficient which leads to lower thermal efficiency. The use of artificial roughness on flow facing surface of the absorber plate is an effective way to improve the thermo-hydraulic performance of the solar air heater. In this study, NACA 0040 profile ribs are implemented in the reverse position on the absorber plate with a relative roughness pitch (P/e) of 5 and relative roughness height (e/D) of 0.065. In the first phase of investigations, 2D and 3D numerical analysis of the solar air heater duct subjected to continuous ribs and broken ribs (ribs with grooves) was carried out using ANSYS Fluent.The objective of the present study is to optimize the groove parameters such as number of grooves, relative width (g/e), inclination and the location of grooves. The numerical analysis of solar air heater duct was carried out considering NACA 0040 profile ribs in the reverse position with different number of grooves (2, 3, 4, 5), relative widths (0.667, 1, 1.333, 1.667) and inclinations (15°, 30°, 45°, 60°, 75°, 90°) in the Reynolds number range of 6000-18000. From the numerical analysis, it was found that among the various groove parameters 5 grooves of 1.333 relative width provided at 30° angle (from ribs) results in optimum performance. In the next phase of study, experimental investigations were carried out with the bsorber plates subjected to optimized parameters of the broken ribs. In experimental investigation, the Nusselt number for broken ribs decreased by 0.38-2.26% and friction factor reduced by 6.25-10% in comparison with the continuous ribs for various Reynolds number range. This results in an increase in the THPP in the range of 0.30 to 4.80 %. The maximum value of THPP in the experimental investigation was found to be 2.58 at the Reynolds number of 6000 for broken ribs. Based on the numerical study conducted, an empirical correlation was developed for the prediction of Nusselt number and friction factor in terms of various influencing parameters, which was found in good agreement (within 8%) with the experimental results.