15D-7 |
Mixing in a twin screw extruder evaluated by MRI |
Y. J. CHOI, K. L. McCarthy, and M. J. McCarthy. Biological and Agricultural Engineering, University of California, Davis, One Shields Ave., Davis, CA 95616 Mixing is an important process for ensuring process uniformity of food products. There currently exist few process sensors capable of measuring the uniformity of mixing during processing. This work is aimed at developing magnetic resonance imaging (MRI) based sensors for characterizing the degree of mixing during processing. This study used a noninvasive technique to quantify mixing as a function of axial position in an APV 30 mm corotating twin screw extruder. Two model systems were chosen, representing the case of two components mixing under laminar shear conditions. The test materials were glycerin and starch gel chosen based on their high viscosity and structural stability. To improve the signal to noise ratio, each was mixed with paramagnetic doping material (i.e., CuSO4) which changed only NMR properties. The extent of mixing of the doped and undoped material was evaluated statistically at axial cross sections of concentration images obtained by MRI. A multislice imaging sequence was performed to obtain concentration images over 9 cm sections of the extruder. Each slice was 15 mm thick with 160 mm field-of-view. Concentration profiles were obtained by MRI to provide information regarding mixing in consecutive axial slices. The extent of mixing was characterized best in terms of coefficient of variation (CV), which allows comparison from one image to another. For glycerin model system, the CV changed by 60% during the mixing process. In the starch gel, there was a 12% change in mean signal intensity during the mixing process. The MRI technique differentiates between axial mixing and transverse mixing that takes place in the extruder channel.
Session 15D, Food Engineering: Processing Technologies
|