|Trabelsi, Samir - OICD|
Submitted to: Journal of Agricultural Engineering Research
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: April 8, 1998
Publication Date: N/A
Interpretive Summary: Reliable knowledge of the moisture content of grain is very important, because wet grain will spoil in storage or during transit. The grain selling price depends upon the moisture content, because if grain must be dried before storage, the cost of drying must be taken into account. Also, for processing, grain moisture content must be adjusted to certain levels, and the accuracy of the adjustments is reflected in the final quality of products. Standard methods for moisture content determination require many hours of drying in an oven, and they are much too slow for application in trade or in grain processing. Therefore electronic instruments have been developed that sense moisture in grain through the electrical properties of the grain. This method requires the correlation between grain properties and moisture content to be corrected for grain temperature and for variation in the bulk density, or packing, of the grain. Our studies on microwave measurements of moisture in grain have shown that both effects can be corrected electronically by using certain mathematical expressions to determine the moisture content independent of density. Various expressions have been claimed to provide density-independent determination. In the work reported, an objective comparison of the existing expressions is performed and their effectiveness is checked on the same set of experimental data for wheat obtained in this laboratory. Conclusions drawn from the comparison can be used by scientists and equipment manufacturers.
Technical Abstract: The accuracy of moisture content determination in grain from measurements of its electromagnetic parameters is known to be dependent on the bulk density of the sample used for the measurement. Various means of limiting this effect were considered in the past. This paper presents a review of density-independent expressions developed for this purpose and compares their effectiveness in minimizing the density dependence of the predicted moisture content. Because of different numerical coefficients involved in various expressions, the same procedure and the same data set of over 180 measurements on hard red winter wheat, Triticum aestivum L., at frequencies of 11.3 and 16.8 GHz on samples in free space were used to provide a fair comparison. Several density-independent expressions were identified that provide prediction of wheat moisture content with a standard error of calibration of about 0.2 percent moisture, for moisture contents ranging from 10.6% to 18.2%. Although these results were obtained for static samples of grain, it is most likely that they can be extended to grain flowing through a tubular conduit, or moving on a conveyor, and measurements can be taken continuously.