Submitted to: Cereal Chemistry
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: November 5, 1999
Publication Date: N/A
Interpretive Summary: Small scale, micro-test milling of wheat is done to predict the milling quality that will be observed on a large commercial mill. Of necessity, micro-milling uses a small mill and small sample size. The resulting data is greatly influenced by wheat moisture content and kernel texture. Normally, water is added to wheat before milling to achieve a constant moisture level for milling. However, that adds at least four hours to the entire processing time. During commercial milling, the influence of kernel texture is very small, but that influence increases greatly for smaller mills. A mathematic adjustment was developed that corrected for differences in wheat moisture content, precluding the need to temper wheat before micro-milling. Another mathematic adjustment was developed that corrected for the influence of kernel texture on milling data. Compared to unadjusted results, both new adjustments together produced much better agreement between a very small and a larger mill as they assessed milling quality of wheat. Applying the adjustments to micro- milling data saves valuable time and produces more commercially realistic data for predicting commercial milling quality. This procedure is a valuable advantage to everyone who uses small micro-mills to predict milling quality of wheat, including wheat breeders, government quality evaluators, and wheat elevators and handlers, grain traders, and commercial millers.
Technical Abstract: Variations in soft wheat moisture content and kernel texture greatly affected the flour yield produced by a small, Quadrumat Jr. micro-test mill. An algorithm was developed that adjusted Quadrumat flour yield to 15% wheat moisture content, precluding the need to temper the wheat before milling. Another algorithm was developed to adjust flour yield relative to wheat kernel texture obtained during the micro-milling procedure itself. The adjusted micro-milling data had much better agreement (R2=0.90 across three confirmation data sets) with milling data from a larger, longer flow Allis-Chalmers mill than did unadjusted Quadrumat Jr. data (R2=0.55). Representative micro-milling flour yield data could be produced milling as little as ten grams of wheat, totaling less than five minutes of operator time.