holder

The following is an abstract for the selected article. A PDF download of the full text of this article is available here. Members may download full texts at no charge. Non-members may be charged a small fee for certain articles.


InSb Focal Plane Array Chemical Imaging Enables Assessment of Unit Process Efficiency for Milling Operation

Volume 64, Number 12 (Dec. 2010) Page 1320-1324

DAVID L. WETZEL,* ELIESER S. POSNER, and HULYA DOGAN


In the dry milling of wheat flour, each unit process (roller mill, purifier, sifter, etc.) produces a mixture with varying amounts of wheat endosperm and non-endosperm byproducts. Chemical images with 82 000 pixels of each intermediate product stream issuing from an individual processing machine are readily analyzed in terms of the relative amount of endosperm and non-endosperm. Approximately three minutes is required to produce an image of each intermediate product stream. Applying partial least squares (PLS) chemometric software to identify individual pixels, which enables calculation of the relative amount of endosperm and non-endosperm, is not a time-limiting factor. When relative flow rates are known for each stream, mass balance can be calculated from each intermediate stream in terms of the product (endosperm content) and the lower value non-endosperm byproduct. Data is presented from a purifier in a commercial flour mill. Intermediate streams collected from a run with optimized operational parameters were compared to those of another run before adjustment. The endosperm (product) mass balance profile for each run enabled assessment of operational efficiency. The devised chemical imaging analysis system would be particularly useful in commissioning of a new mill or to optimize existing wheat milling systems. Also, when raw material differs from that for which previous optimization was established, a new optimization may be in order. The ability to acquire a large number of spectra from a specimen and apply multivariate statistics to identify each pixel and subsequently count pixels accommodates heterogeneity and reports the results from averaging a very large number of individual spectra. A second illustration of the utility of the imaging method is presented centering on streams from the first and second break unit operations at the beginning of the roller mill process.

Index Headings: Chemical imaging; InSb focal plane array; Focal plane array; Unit process control; Hyperspectral imaging; Near-infrared chemical imaging; Wheat.