Hot air drying is one of the primary processing steps after paddy harvest. There are complex heat and moisture transfers inside the paddy during drying. The distribution of moisture can also dominate the quality of dried food in industrial production, such as the nutritional content, safety, and weight. The water transfer can greatly contribute to the product quality and the efficiency of drying. Low-field nuclear magnetic resonance (LF-NMR) technology can be expected to rapidly monitor and observe the internal moisture distribution of the sample in real time, due to its easy operation, wide applicability, high accuracy, and stability. Furthermore, the LF-NMR has been widely applied to monitor the moisture status and distribution during food drying, storage, and rehydration, whether the sample morphology, color, and size or not. The transverse relaxation time T₂ without damage or invasion can be used to analyze the migration of sample moisture from a microscopic perspective. This study aims to clarify the internal water migration during drying paddy using LF-NMR and imaging techniques. The drying characteristics were also determined using the transverse relaxation time T₂ inversion spectrum. A systematic analysis was implemented to explore the proportion of water in each state of paddy and magnetic resonance imaging (MRI) images under different drying conditions. The results showed that the higher the tempering temperature was, the shorter the drying time and the higher the drying rate of paddy was. The tempering temperature dominated the waist-bursting rate of paddy. Specifically, the lowest and highest waist bursting rates of paddy were achieved in the tempering temperatures of 40 and 20 ℃, respectively. Three peaks were detected to represent three types of water in paddy grains. Each peak precisely indicated the T₂ value for the respective type of water, namely: bound, semi-bound, and free water, representing successively from left to right as T₂₁, T₂₂, and T₂₃. There was the largest proportion (about 90.5%) of bound water in paddy. While the proportions of free water and semi-bound water were relatively small, only 9.5%. There was a significant linear relationship between the moisture content of the paddy and the amplitude of the NMR signal. The peak area of the T₂ inversion spectrum better reflected the moisture content of the paddy. The lateral relaxation time T₂ of the paddy gradually shifted to the left with increasing drying time, indicating the tighter binding of water and substrate in each state. Drying also changed the distribution of water and the moisture content in the paddy. There was a transformation between moisture in different states when the external environmental conditions changed. There was a variation among different states of moisture inside the paddy during drying, indicating the mutual transformation between bound and semi-bound water, as well as between semi-bound and free water. Tempering reduced the rate of moisture transformation. MRI images showed that the first thing to be lost was the free water attached to the surface of the paddy during drying. Among them, the moisture in the embryo spread outwards, as the moisture decreased on the surface of the paddy. The moisture in the embryo spread out again until the paddy reached a safe storage water content after the surface moisture was dried out again. Therefore, LF-NMR technology can serve as an effective way for the rapid detection of water migration during paddy drying. The finding can provide a strong reference to determine the optimal paddy drying.

This study aims to improve the storage efficiency of longitudinal ventilation in vertical silos during mechanical ventilation of grain storage. A circulation ventilation silo model was established to combine the longitudinal and transverse ventilation. EDEM Fluent software was used to simulate the velocity and temperature fields of air duct structures in a grain storage silo under mechanical ventilation. The results show that the velocity uniformity indexes of the circulating and longitudinal ventilation were 0.92 and 0.88, respectively. The average velocities of circulating and longitudinal ventilation airflow were 2.51 and 1.91m/s, respectively. The circulation ventilation airflow was more evenly distributed in the grain silo. In the simulation of the temperature field, the temperature of the grain pile near the inlet of the longitudinal ventilation initially decreased in a gradient, which was significantly lower than the initial temperature. There was no outstanding heat transfer caused by longitudinal ventilation in the upper layer of the grain layer. But the temperature of the grain surface was lower than that inside the grain pile, due mainly to the environmental temperature at the top of the warehouse. Specifically, the upper layer of grain started the heat exchange in the early stage of circulation ventilation, thus generating a temperature gradient. Furthermore, the temperature of the longitudinal ventilation chamber dropped to 27.35℃ at H=0.5 m and 28.85 ℃ at H=0.8 m after ventilating for 20 s. There was no significant variation in the temperature of the grain pile near the warehouse wall. Correspondingly, the temperatures of the circulating ventilation warehouse were 26.85 ℃ and 28.35 ℃, respectively. When ventilated for 30 s, the temperature of the longitudinal ventilation chamber dropped to 26.85 ℃ at H=0.5 m, and to 28.35 ℃ at H=0.8 m. The temperature of the circulating ventilation chamber were 25.35 ℃ and 27.35 ℃, respectively. After the ventilation was completed, the longitudinal ventilation was maintained at a temperature of 26.85 ℃ at H =0.5 m, and the temperature dropped to 27.85 ℃ at H =0.8 m. The temperature values of the circulating ventilation chamber were 24.85 ℃ and 25.85 ℃, respectively; The overall average temperature dropped to 24.77 ℃ and 23.43 ℃, respectively, at the end of ventilation in the longitudinal and circulating ventilation. The cooling performance of circulating ventilation was significantly improved with the high and more uniform cooling rate, compared with the longitudinal one. In the simulation of the particle temperature field, the temperature of the lower half of the particles in the silo was relatively low after longitudinal ventilation, with an average particle temperature of 21.20 ℃ and an average particle temperature of 22.85 ℃ after circulating ventilation. There was relatively low cooling in the upper part of the particles after longitudinal ventilation. There was no significant change in the temperature of the particles near the warehouse wall. The average temperature of particles was 27.34 ℃ after ventilation. There was more significant cooling of the particles in the upper part of the warehouse after circulating ventilation, where the average temperature of particles was 23.07 ℃ after ventilation. Compared with longitudinal ventilation, the temperature difference of particles between the upper and lower parts of the circulating ventilation warehouse was relatively small, indicating uniform cooling. In the temperature changes in different layers of grain silos under the same ventilation conditions with different heights of grain piles, the temperature of circulating ventilation was lower than that of longitudinal ventilation grain piles. After ventilation, the average temperature of the circulating and longitudinal ventilation chamber dropped to 23.43 ℃, and 24.77 ℃, respectively. The temperature of circulating ventilation decreased significantly, where the average temperature of each layer was lower than that of longitudinal ventilation, resulting in better cooling. The grain pile in the vertical silo with a circulating air duct was better cooling than that of the longitudinal ventilation. This finding can provide an important basis to optimize the air duct in the vertical silo for better ventilation in grain storage.