Application of experimental, numerical, and machine learning methods to improve drying performance and decrease energy consumption of tunnel-type food dryer

Abstract

In this study, to distribute the drying air uniformly on the product surface, straight and trapeze air barriers were designed in the drying chamber of the existing tunnel dryer. The effects of air barriers on product surface temperature changes were investigated by computational fluid dynamics analysis (CFD). Drying time in the experiment without an air barrier decreased by 45% with the trapeze barrier and 20% with the straight barrier. Likewise, the trapeze barrier provided 53.9% energy savings, and the straight barrier 37.4% energy saving compared to the drying process carried out in the current system. Also, using the experimental data, mathematical equations that can calculate activation energy (E-a) in the drying process were produced with the help of regression-based artificial intelligence methods (Pace and Elastic.Net). With the help of these equations, the E-a values of the drying process performed under different experimental conditions were determined, and a 1.03% error value was calculated between the obtained E-a values and the experimental values.