The wide range of melting point of milk fat causes it to be unsuitable for many industrial applications. Along with an increase in milk production and health consciousness of the consumers, this has caused a mountain of milk fat surplus in 1992. Therefore, many researchers have explored alternative uses of milk fat as well as altering milk fat properties. Crystallization could produce various milk fat fractions with different physico-chemical properties that are industrially more suitable, techno-economically feasible with increased added value, and which are easy to tailor to a healthier "formulated milk fat" that is socially and legally acceptable.; The goals of this research were: (1) to evaluate use of the population balance model for determining kinetics of continuous crystallization of milk fat; (2) to determine the kinetics of crystallization of milk fat in a continuous, stirred tank crystallizer at the 2-L laboratory scale; (3) to evaluate characteristics of milk fat crystals generated at various conditions. Four factors were studied: temperature of crystallization (30{dollar}spcirc{dollar}C and 20{dollar}spcirc{dollar}C), residence time in the continuous crystallizer (1 and 2 hr), two different lots of milk fat (early and late pasture feeding), and energy input, i.e., agitator speed (75, 100, 150 and 200 RPM).; Among the four factors studied in this research, temperature had the most prominent role in determining the characteristics of milk fat fractions, and affecting kinetic parameters (nucleation and growth rates). Agitator speed significantly affected nucleation and crystal growth rates. Two different mechanisms occurred to different extent depending on temperature of crystallization.; Residence time (one or two hour) did not give a statistically significant effect on kinetic parameters. However, it gave a consistent trend on thermal properties and chemical composition of the fractions, and significantly affected melting point of hard fractions. Different lots of milk fat, although affecting the characteristics of the fractions, did not significantly affect the kinetic parameters in this study. Here again, the slightly different chemical composition did not necessarily affect the physical properties of fractions produced.; Use of population balance model for predicting kinetic parameters had the maximum benefit when all formalism of the model was obeyed. In this study, this occurred when agglomeration was not a predominant mechanism during continuous crystallization. Forcing the population balance model when agglomeration process was important, resulted in underprediction of mean crystal size. Nucleation rate was most affected by the fitting technique of this model. Effective nucleation rate for continuous crystallization of milk fat as found by this model were three orders of magnitude higher than in batch system, whereas growth rates were comparable. (Abstract shortened by UMI.)
展开▼
