Clarification of formative mechanism of uneven grain moisture in silo and the moisture distribution monitoring

本研究的目的是探讨水分分布不均匀导致粮食霉变变质的形成机理，并开发水分分布异常监测系统。即使筒仓中装有水分均匀的谷物，在储存过程中也会形成谷物水分的不均匀。我们假设筒仓壁和粮食之间的温差引起水分运动。本文首先对某粮食筒仓上部空间和仓外壁的温湿度进行了测量。利用实测值和模型试验，对筒仓内水分运动进行了数值模拟。结果表明，造成粮层水分分布不均匀的主要原因是结露和壁面或顶棚的水滴进入粮层，而不是粮层间水分的直接运动。其次，为了监测粮食水分分布的不均匀性，在厚粮层中插入长金属棒等，进行了时域光谱TDR和TDT试验。由于在水分不同的边界中产生反射波，因此根据反射波形检测到水分分布的异常。从传播延迟时间来看，平均湿度至少检测到约2m的粮层。第三，在喇叭天线非接触测量的基础上，进行了时域光谱TDR和时域光谱TDT的相同实验。TDT至少检测到1m左右的粮层平均水分，TDR检测到距粮层20cm左右深度的基于水分不规则性的异常反射。第四，针对粮层表层水分分布，进行了利用970nm附近水分吸收带的多个波长获取光谱图像的实验。作为可视化处理的结果，获得了粮层表面的水分分布图像。

The purpose of this study is to investigate the formative mechanism of uneven moisture distribution that cause mold and degradation of grain in storage silo and develop the monitoring system to detect abnormal moisture distribution. Unevenness of grain moisture forms during storage, even if a silo was charged with uniform moisture grain. We assumed the temperature difference between silo wall and grain causes the moisture movement. Firstly, we measured the temperature and humidity of the upper airspace and outside wall of a grain silo. Using the measured values and model experiment, we simulated the moisture movement in the grain silo. We reached the conclusion that main cause for the uneven moisture distribution is the dew condensation and the water droplet from wall or ceiling into the grain Iayer rather than the direct water movement between grain layers. Secondly, to monitor the irregularity of grain moisture distribution, inserting long metal bar etc.into the thick grain layer, time-domain spectroscopy TDR and TDT were experimented. As the reflected wave is generated in the boundary where the moisture differs, from the reflected wave form abnormality of the moisture distribution was detected. From the propagation delay time, the average moisture was detected at least to grain layer of about 2m. Thirdly, with non-contact measurement using horn antennas, same experiments of time-domain spectroscopy TDR and TDT were experimented. The average moisture was detected at least to grain layer of about 1m by TDT and anomalous reflection based on the moisture irregularity was detected to the depth of about 20cm from surface layer by TDR.Fourthly, on moisture distribution of surface grain layer, experiments which obtain the spectral image using the multiple wavelengths near 970nm of the water absorption band was carried out. As the result of the visualization processing, the moisture distribution image of the grain layer surface was obtained.