Thomas Danner

Product Engineering Using the Example of Extruded Instant Powders

Starch-based instant powders are usually produced by cooking/gelatinizing in an excess of water with subsequent drying and milling. Cooking extrusion is an alternative energy-saving process. This high-temperature/short-time process is determined by various process which influence the properties of the finished product, e.g. the specific weight and the solubility of the powder, as well as the viscosity and the mouthfeel of the reconstituted gel. This work aims at demonstrating the functions between product characteristics and process parameters by statistical trial planning so that a target finished product can be designed by choosing the required process parameters.

Dielectric Spectroscopy and Principal Component Analysis as a Method for Oil Fraction Determination in oil-in Water-Emulsions with Varying Salt Content

Important parameters, that define the physical properties of an emulsion, are the concentrations of the constituents. In the case of a continuous production of emulsions an exact and fast control of the fractions is essential to ensure a constant product quality. Whereas the method of sample drying, weighing and chemical analysis is too slow for an online product control, the dispersed phase fraction determinition by density only works for two-component-systems. For the purpose of more practical emulsions with possible variations in both, the dispersed phase fraction and the salt content, in this work dielectric spectroscopy in combination with principal component analysis is studied. Introduction The properties of an emulsion are strongly dependent on the concentratations of constituents. As an example, the rheological and the sensory properties depend on the fraction of the dispersed phase (here the oil phase, of an oil-in-water emulsion). Emulsions in practice often consist of more than two chemical species. It is therefore not sufficient to chracterise the emulsions by the measurement of its dielectric properties at one frequency. For a sufficient characterisation the measurement of the dielectric constant over a frequency range is necessary, yielding a large data set, difficult to analyse. One possible method for its analysis is the principal factor analysis, which is a mathematical-statistical technique [1]. It was first used in the behavioral sciences