Evaluation of particle measures relevant for powder bed porosity—A study of spray dried dairy powders.

Abstract: The formation of lumps during large scale wetting of spray dried powders is a common problem in the industry. The result is a large energy demand if an efficient mixing is to be achieved at the industrial scale. If more knowledge regarding the phenomena governing lump formation is obtained, the unit operation could be optimized and made more efficient. As a part of increasing the understanding of lump formation, the wetting of the porous media that is a powder bed needs to be understood. This wetting relates to the structure and void space (porosity) of the powder bed, which is why a relation between powder bed porosity and readily obtainable powder particle morphology parameters is sought. In this study a thorough investigation of the morphology of five different spray dried milk powders, with arbitrary particle shapes, was performed using manual image analysis of scanning electron microscope (SEM) images as well as automated methods based on light scattering and projected image analysis. The results of the automated methods, when utilized on the complexly shaped samples, were validated by a comparison to the SEM image analysis. This enabled the development of a relation between powder morphology and powder bed porosity based on data from automated methods. The results indicate that automated methods based on light scattering underestimate the particle sizes of arbitrary shaped particles. Methods based on projected image analysis, however, show a better agreement with SEM image analysis. The developed relation between powder bed porosity and powder particle morphology is based on defining the bed as having either an “Agglomerates in fines”, “Bicontinuous” or “Fines in agglomerates” structure. The structure of each powder is determined based on calculation of a “BedRatio” which is a function of the ϕagglomerate and ϕfines both shown to be readily obtainable by automated methods. The calculated porosities of the powders investigated in this study show a good agreement with experimental measurements and are considered to correspond well to earlier literature. It is, however, clear from the experiments that care needs to be taken during bulk density measurement of spray dried powders so that no intra particle void space is taken into consideration in the calculations. [Copyright &y& Elsevier]