Effect of milling raw soybean on the compositional uniformity of soybean cake during extrusion-aided screw press oil extraction

Abstract

The effect of particle size of roller-milled raw soybeans on the compositional uniformity of soybean cake during extrusion-aided screw press oil extraction (ESPOE) was studied. In order to obtain reproducible milled soybean samples, a roller mill feeler gauge was constructed and used to calibrate the roller mill nip so that four milling sizes; coarse, intermediate-coarse, intermediate-fine and fine which corresponded to milled soybean particle mass median diameters of 2.90, 1.89, 1.72 and 1.57 mm respectively; were obtained. Whole raw soybeans and raw soybeans roller-milled to represent the four milling sizes were continuously processed using the extrusion-aided screw press oil extraction (ESPOE) technique with an lnsta-Pro single screw autogenous extruder and a Hander screw press oil extractor. The soybean cake moisture content and protein, oil and crude fiber contents in the dry matter were monitored over twenty-five minutes of processing and were the compositiortal variables. The bulk density (unpacked and packed), the mass throughput of the Insta-Pro model 8200 volumetric feeder/blender at constant auger speed setting, the crude oil yield and the foots to oil mass ratio in the crude oil produced were also measured for whole raw soybean and the four milling sizes. The effect of milling on the overall cost of the ESPOE process was estimated. Soybean cake moisture content fluotuated significantly when either whole or milled raw soybeans were processed. Soybean crude fiber content showed significant differences between treatments with time. Protein and oil contents did not show significant differences between treatments with time but trends indicated somewhat more fluctuations when milled soybeans were processed than when whole raw soybeans were used. There was least soybean cake protein content fluctuation when whole raw soybeans were processed. Average soybean cake protein content increased slightly with processing time when soybeans were coarsely milled. Intermediate milling sizes produced the most fluctuation in cake protein content with time, and finely milled beans the least fluctuation with a net increase less than that of coarsely milled soybeans. Soybean cake oil content fluctuated the least with whole raw soybeans and among the milled samples it fluctuated the least with coarsely milled beans. Intermediate-coarse and intermediate-fine milling sizes fluctuated the most. The intermediate-fine and finely milled beans were characterized by periodic peaks in oil content. Soybean cake crude fiber content fluctuated the least when whole raw soybeans were processed. Among the milled soybeans, crude fiber content fluctuated least when beans were coarsely or finely milled and fluctuated most for intermediate milling sizes. Whole raw soybeans had the 'highest bulk density. In the milled samples, the unpacked bulk density decreased as the particle size decreased. Packed density was lowest for the largest particle size and increased as particle size decreased up to intermediate and increased when particle size approached fine. The feeder/blender produced the highest mass throughput with whole raw soybeans and in the milled soybean samples the mass throughput decreased with decrease in particle size up to intermediate-fine and increased with further size decrease. The oil production at constant feeder/blender auger speed rate was highest when whole soybean was processed. For milled soybean, the rate of oil production was lowest for coarse milled and increased with decrease in particle size. The foots to oil mass ratio was lowest with whole beans. In the milled samples, the foots to oil mass ratio increased with decrease in particle size from coarse to fine. The apparent total ESPOE process cost was lowest when whole soybeans were processed. Extrusion and oil extraction costs decreased somewhat with decrease in particlesize. There was most fluctuation of the extruder motor electrical current demand when extrudingwhole soybean and among the milled soybean samples, fluctuation decreased with decrease in particle size. This study showed that milling raw soybean prior to ESPOE was not required.