Enhancing solar air heater efficiency with 3D cylinder shaped roughness elements

Abstract

Solar air heaters (SAHs) harness solar radiation to warm the air. This study investigates the efficacy of 3D cylindrical roughness features in enhancing heat transfer and reducing friction losses within SAHs. The configuration involves inline and staggered placement of 3D cylindrical roughness elements beneath the absorber surface. Through ASHRAE standard 93–77 testing, heat transfer and friction factors are examined across various cylinder roughness setups, with parameters such as relative roughness pitch (p/h) and gap (w/h) considered. The channel aspect ratio (W/B) is set at 5, and a relative roughness height (h/D) of 0.06 is maintained and flow Reynolds number (Re) is varied from 3000 to 8000. Regression analysis reveals a significant relationship between heat transfer and friction. Notably, at Re = 7910, p/h = 20, and w/h = 8 for staggered cylinder roughness, average friction factor (fav) increases by 171.3 %, and average nusselt number (Nuav) increases by 69.7 %. Similarly, the efficiency index (η) peaks at 0.8192 with inline roughness at Re = 2969, p/h = 20, and w/h = 8. The introduced 3D cylinder roughness elements disrupt the laminar sub-layer, amplifying convective heat transfer while curbing friction losses. This approach optimizes SAH performance within the studied parameter range, showcasing potential improvements for utilizing solar energy. Experimental results, along with expressions for average friction factor (fav) and average Nusselt number (Nuav), underscore the benefits of cylinder-shaped roughness geometries while maintaining h/D = 0.06. These findings introduce an innovative convective heat transfer method, enhancing SAH efficiency. Furthermore, the anticipated average friction factor (fav) and average Nusselt number (Nuav) values align within error limits of ±6.63 % and ±6.62 %, respectively, with improved regression coefficients indicating the stronger relationship and the influence of independent variables on dependent variables.