Integrated Study of Teff Grain Characteristics in Positive Pressure Dilute-Phase Pneumatic Conveyor: Experimental, CFD-DPM, and Artificial Neural Network Approaches

Abstract

Teff is a gluten-free grain native to Ethiopia that is gaining global popularity due to its nutritional benefits. Despite its advantages, teff faces challenges in farming and handling because of its tiny grain size, weak stems, high lodging tendency, and reliance on traditional production and postharvest methods. These factors contribute to low yields, poor grain quality, and substantial losses during harvesting, handling, and processing. This study investigates the use of pneumatic conveying systems to improve teff grain handling, with a focus on pressure drop as a key system parameter. The research involved determining engineering properties, designing and constructing a laboratory-scale conveyor system, and analyzing pressure drops across critical components, including the feeder, horizontal and vertical pipes, 90° bends, and a cyclone separator. Laboratory experiments, CFD-DPM simulations, and artificial neural network modeling were employed to comprehensively study the system. Teff grain was found to have a mean particle diameter of 0.68-0.72 mm, particle density of 1120-1381 kg/m³, sphericity of 0.67- 0.69, terminal velocity around 3.6 m/s, and an angle of repose between 20.2° and 23.8°. Compared to other cereal grains, teff’s smaller particle size and lower terminal velocity significantly influence its pneumatic conveying characteristics. Fluidized bed tests revealed that teff behaves as a bubbly, free-flowing material, with pressure drops of 100- 107 mbar/m and a minimum fluidization velocity of 0.56-0.575 m/s, indicating that dilute- phase pneumatic conveying is more suitable than dense-phase systems. Minimum conveying velocities ranged from 6.0-20.7 m/s, depending on solid loading ratios 0.3-5.0 and pipe diameters 50-110 mm, suggesting that teff requires relatively lower airflow rates for full particle suspension. However, it’s very small particle size leads to higher pressure drops and non-uniform air velocity distribution in straight pipes. CFD-DPM simulations closely matched laboratory results, with R² values of 0.990-0.997. ANN models trained on both experimental and simulation data also performed excellently, predicting pressure drops with R² values of 0.971-0.981. Statistical validation demonstrated the reliability of these models, with MSE of 27-7984, RMSE of 5.2-89, MAE of 4.1-65, and MAPE of 6-25%, showing strong agreement with experimental results and reliable predictive capability. The findings emphasize that the design of pneumatic conveyors for teff must consider its small particle size, moderate density, and low terminal velocity, rendering dilute-phase conveying the most suitable option. It is essential to optimize airflow and pressure differentials to ensure a blockage-free and energy-efficient operation. In summary, the integration of experimental analysis, CFD-DPM simulations, and ANN modeling offers a comprehensive approach to understanding teff grain flow, thereby facilitating the design, development, and optimization of efficient pneumatic conveying systems for production, postharvest handling, and processing applications.