Development of a predictor tool for the estimation of minimum free fat in pasting for achieving desired particle size (Dv90) in chocolate on a given refiner

Abstract

The research in this paper was aimed to develop a predictor tool for the estimation of minimum free fat in pasting in an attempt to increase the refining efficiency to achieve the targeted Dv90 on a given 600mm refiner. In this study the corresponding minimum fat coating thickness on each particle and the total number of particles present per kilogram of refined milk chocolate (RMC) was also calculated. This study also aimed to give a clear in depth understanding of particle size distribution as this project based on PSD. The effect of Dv90 on fat coating thickness and number of particles was evaluated in this paper through the performance of the refiner at different refining condition by using excel calculator. The theory of excel calculator and the process of development of excel calculator was also described in this paper through explaining the derived mathematical equation with the help of chocolate science involved in it. Later this excel calculator was upgraded to predictor tool with the help of regression model. A data pool was created with the obtained data after taking multiple trials in the pilot plant. This data pool was further used as data bank for the predictor tool. This is a statistical tool made over excel based on the real time practical data. Total 7 trials were conducted in this project and particle size was measured. For example for a particular recipe that means at a constant fat percentage coating thickness was varied with particle size. At Dv90 40 micron coating thickness on each particle was calculated 0.25 micron while at Dv90 48 micron coating thickness was found to be 0.27 micron when number of particles are in 1013. This is the first statistical tool based on real time practical data which is able to predict the total number of particles present per kg of mass and the fat coating thickness on each particle in micron based on the user input recipe within just a few a second along with the required minimum fat to obtain desired user input Dv90 from a given 600mm refiner.