119 Instrumental detection of peak dough resistance in real-time using a prototype mixer for experimental baking.

H. D. SAPIRSTEIN and R. Roller. Dept. of Food Science, University of Manitoba, Winnipeg, MB R3T 2N2.

Electronic recording dough mixers are excellent tools to compare dough mixing requirements of different flours for experimental baking. In general, mixing time to peak resistance is taken as a practical measure of optimum dough development. Whereas the determination of peak dough resistance (PDR) from a complete dough mixing curve is relatively straightforward, determining PDR in real-time, i.e. while a dough is undergoing development in the mixer is much more challenging. In this study, a 100 g (flour basis) National pin mixer was modified to permit computerized recording of torque-time data supplied by a strain- gauge sensor mounted beneath a water-jacketed moving bowl. The data was acquired using a 12 bit A/D conversion circuit and a sampling rate of 360 Hz. The resulting output was typical for a torque-recording sensor, i.e. broad and noisy. A variety of digital filtering algorithms were evaluated for data smoothing, including triangular averaging and Chebychev low pass filters. Filtered data was subsequently processed to isolate top- and bottom-curve envelopes, and to compute a midline curve. Detection of a candidate mixing peak was based on the midline curve maximum subject to preceding slope criteria and absence of an additional maximum following the candidate peak time by 10%. Accordingly, full formula doughs are mixed to time of PDR + 10% which is a common usage in experimental baking.