Size reduction in feed technology and methods for determining, expressing and predicting particle size: A review

Particle size of diets or ingredients plays an important role in pig growth and gut health. The way the size of particles is measured and expressed, however, is limited in explaining pig growth performance differences. This review explores new possibilities to determine, express and predict particle size. Different grinding methods, including the use of roller mills, hammer mills, multicracker and multi-stage grinding were reviewed. Roller milling tends to produce a more uniform particle size distribution (PSD) and consumes less energy, whilst hammer mills have a greater grinding capacity and a higher reduction ratio compared to roller mill. The multicracker system, a more recently developed technology, can be considered cost-effective and ensures grinding capacity. Since the effects of different grinding methods vary, multi-stage grinding, combining different grinding methods, might be a solution to obtain a defined PSD. Particle size determination techniques, including dry/wet sieving, laser diffraction, microscopy, and static/dynamic image analysis are described and compared. It is concluded that more characteristics of particles (e.g. shape, volume or surface area) should be investigated. Besides geometric mean diameter (GMD), particle size can also be expressed with parameters such as D50, D4,3 and span of PSD. Equivalent particle size (EPS) is introduced as a mean of describing the size of particles related to a functional trait of the particles. A meta-analysis was performed by collecting particle size and pig performance data from scientific studies examining the effect of recalculated EPS on pig performance (feed conversion ratio, FCR). Regression/linear modelling shows that recalculated EPS was not better than GMD in explaining pig performance differences due to the high variation among studies. Different expressions of PSD may result in different conclusions. An introduction of describing the breaking behavior of diet ingredients via mathematical models i