Numerical modeling and simulation of convective water activity in low temperature batch drying of cobbed maize

Abstract

Drying of maize grain is considered an important process in reducing aflatoxins effects on Maize Mechanical drying is recommended over natural drying during wet weather but is costly and energy intensive. Comprehensive understanding of convective water activity during grain drying is critical to the development of more efficient drying process particularly for cobbed maize. The main objective of this study was to analyse batch dehydration of cobbed maize to help understand the dynamics of low temperature processing to facilitate the development of appropriate on-farm drying technologies. The specific objectives were (i) Experimental measurements of drying characteristics and process conditions during batch drying of cobbed maize at low temperatures,(ii) Analysis of the temperature distribution and velocity field of drying air in the batch drier using CFD software (ANSYS Fluent) and (iii) Validation of drying model using experimental results from batch drying at an industrial seed processing facility. Empirical data was obtained using temperature, relative humidity, velocity and moisture content sensors. The sensors were applied within the actual cobbed maize drying operations at Kenya Seed Company Ltd. - Kitale in November 2018.The results showed drying chamber temperatures in the range of 30 - 35 o C while the relative humidity varied from 15 – 35%. The initial moisture content of the Kernels and cob were 19%wb and 38%wb, respectively, reducing to 12%wb for both components after 70hours of dehydration. Numerical modeling was performed using CFD software (ANSYS Fluent 19R1). For ease of analysis, a 2D geometrical model was developed in accordance with the actual parameters of the drying chamber that was used in the experimental analysis. ANSYS CFD-Post was applied to visualize contours, streamline and vector plots inside the drying chamber at superficial air flows of 1m/s, 5m/s and 10m/s, for scenarios with and without the product. The simulated pressure and velocity contours showed lower airflow along the walls and at the corners, consistent with the findings of other studies on non-circular geometries. Similar trends were observed at the increased airflow settings. Non-uniform airflows cause uneven dehydration which presents a challenge for optimal termination of batch operations where over-drying is avoided. The Midili model best fitted the drying kinetics of cobbed maize with R 2 and RMSE values of 0.946 and 0.0127, respectively showing good agreement between the mathematical model and the experimental data. The study concluded that the geometry significantly affects the distribution of air velocity and temperature during drying process and that Midili Model best fits on the drying curves of cobbed maize. Further research is recommended to assess the impact of moisture content, cob size and fines / shell out on batch drying kinetics.