The characteristic morphological features of wheat grain include sections of embryo, endosperm and the layers of seed coat and pericarp. Bran is a mill fraction composed of pericarp, seed coat and pigment strand, hyaline layer and aleurone layer.1
The objective of milling is to break open the grain, scrape off as much as endosperm from the bran as possible, and gradually reduce the practically pure endosperm into the flour. The effectiveness of the milling depends on how well the miller is able to recover endosperm essentially uncontaminated with bran.2
The main methods currently in use for determining bran contamination are ash content and color grade measurement.2-5 Although the ash content — the inorganic residue remaining after incineration of sample at high temperatures — is highest in the bran layers, minerals are distributed in all parts of grain, including the endosperm.
Endosperm ash levels vary by wheat variety, location and crop year.6 Also, ash measurement would be useless if flour is enriched with nutrients such as calcium carbonate and iron.7
Recent advances in technology have introduced tristimulus (L*a*b), near infrared (NIR) and digital image analysis (DIA) methods as rapid, nondestructive bran measurement alternatives. All the methods except DIA are an indirect measure of bran.5
A new image analysis system, the Branscan 1000, developed by Parascan Technologies Ltd., Redditch, U.K., uses reflected visible light for on-line and at-line analysis of bran in flour The algorithms used are designed to be insensitive to variations in color and brightness of bran specks and of endosperm and respond solely to the quantity of darkly contrasted tissues (specks).3, 4, 8
The Turkish Wheat Flour Codex classifies bread flours into three types (Type 550, Type 650, and Type 850), based on their ash content of 0.55%, 0.65%, and 0.85%, respectively.9 The objective of our study was to investigate the relationship between ash content and speck count and Branscan percent values of bread flours by Branscan 1000 in a commercial mill.
MATERIAL AND METHODS. Bread flours were produced from various blends of Turkish, Australian, Russian and Kazakhstanian wheat on a 300 tonne per day commercial mill. One hundred and twenty six flour samples having extraction rates about 70% and 52 samples with approximately 40% extraction rates were collected several times during a shift for 21 days and 10 days, respectively, through a sampling flap located immediately below the measuring station. Moisture contents were determined by holding samples at 130oC for two hours according to method TS-1135 of the Turkish Standards Institute, Ankara, Turkey10. Ash contents were measured according to method TS-151111 of the Turkish Standards Institute by incinerating samples at 900oC until the weight of ash was unchanged. Results were expressed on a dry matter basis.
Data were analyzed using the linear regression procedure in the SPSS statistical software.
RESULTS AND DISCUSSION. Relationships between ash content and Branscan data in all flour samples are shown in the charts on page 36. The regression variance analysis in SPSS suggested that the Branscan % and speck count values were significantly related to the ash content having 0.7334 and 0.7501 coefficients of correlation values, respectively, at P<0.05 significance level.
A recent evaluation of milling protocols for noodle flours using Branscan 2000 indicated that the cumulative Branscan % curves provides similar information about the purity of flour streams to cumulative ash curves, although ash value of flours in the range of 0.28% to 0.65% was insufficient to determine if a specification based on ash had been met on a pilot mill.12
Another on-line measurement of bran in flour by image analysis reported that Branscan 1000 values increased with bran content having a correlation of coefficient of 0.975 for single grist flour streams in a commercial mill.4
Data from our study suggested that Branscan could be used to estimate ash contents of bread flours in the range of 0.45% to 0.65% or to monitor bran contamination in a commercial mill.
This data was first presented at the 1999 ICC Conference in Valencia, Spain. The authors, Kursat Kurt, a miller with Fatoglu Milling Co., Giresun, Turkey, and Dilek Boyacioglu and M. Hikmet Boyacioglu, professors in the Department of Food Engineering at Istanbul Technical University, Maslak, Istanbul, Turkey, wish to express their appreciation to Branscan Ltd., U.K., for providing the Branscan 1000 for the experiments.