Three approaches (experimental, simulation and ANN) based investigation of thermal characteristic and correlation for frustrum roughened solar air heater

This investigation is an attempt to use the concept of artificial roughness (AR) and enhance the heating capacity of solar air heater (SAH). Frustrum shaped roughness geometry has been fabricated using different sizes punch and die. Frustum roughness having more contact surface area with flowing fluid leading to more heat transfer. The work reported experimental, simulation and artificial neural network (ANN)-based investigation. Enhancement in thermal characteristic of frustrum shaped roughened SAH has been discussed with including fluid flow physics. Main roughness and flow parameter was relative frustum pitch (p/e = 8–14), relative frustum height (e/D_h = 0.013–0.054), Relative frustum height to base diameter (e/d₁ = 0.37–0.75), relative frustum diametral ratio (d₁/d₂ = 1.33–3). Maximum augmentation in Nusselt number (Nu) for varying e/D_h, p/e, d₁/d₂ and e/d₁ was 222, 226, 241 and 237 that of friction factor (f) was 0.019, 0.021, 0.022, and 0.021. Collected experimental data was also used to develop mathematical correlation between thermal characteristic (Nu and f) with roughness and flow parameter. Derived mathematical also validate by comparing correlation predicted value with ANN predicted values. ‘Nu’ and ‘f’ values from statistical correlation was in close agreement to the ANN predicted values with mean deviation being ±9.5 % for Nu and 12 % for f.