Tidsskrift: Journal of Cereal Science, vol. 37, p. 275–284, 2003
Open Access: none
Using the nine different wheat varieties, sixteen different flours were blended to achieve two protein levels 11 and 13%. Mixing characteristics of the flours were analysed by Farinograph and Mixograph. Thermal properties of flour were measured by differential scanning calorimetry (DSC). Starch gelatinisation, pasting viscosity, and gelation properties were assessed by a Rapid Visco-Analyser (RVA). Hearth bread loaves were produced from the flours using a small-scale baking method. The results from the analyses of flour were related to the properties of hearth bread loaves by multivariate statistical methods. The Partial Least Squares Regression (PLS) model (Model 1) obtained using results from grain and flour analysis (no starch analysis included) as X - matrix explained 70% of the total variance in hearth bread characteristic form ratio. If flour-pasting properties were included in the PLS1 model (Model 2) the explained variance increased to 77%. The model obtained including the DSC parameters of wheat flour explained 76% of the total variance in form ratio (Model 3). This emphasises the importance of starch in wheat flour and that it is possible to find wheat flour starch quality parameters in addition to protein quality that is important for hearth bread quality. No good models were obtained for hearth bread characteristics loaf volume and weight. (C) 2003 Elsevier Science Ltd. All rights reserved.