Всероссийский научно-исследовательский институт физиологии, биохимии и питания животных – филиал Федерального государственного бюджетного научного учреждения «Федеральный научный центр животноводства – ВИЖ имени академика Л.К. Эрнста»
ABSTRACT. The aim of the work is to give a comparative assessment of the procedures for analyzing the dispersion of milled forage for cattle by analytical sieving with the subsequent calculation of statistical parameters. Grinding of individual feeds to a certain particle size has a significant effect on the processes of their digestion in the gastrointestinal tract; when the grinding is insufficient, the unevenness of mixing does not allow to exclude selective eating of certain components of the feed, and considerable regrinding causes digestive upset. The necessity of selecting an adequate statistical analysis procedure is caused by possible differences in the distribution of size fractions in the material under study, using different measurement scales (the scale of intervals, the scale of ratios, etc.) and different mean values (arithmetic mean, geometric mean, etc.). In the computational experiment, a discrete set of hole sizes in two sieve systems (GOST R51568-99 and ASABE, 2007) and several variants of the theoretical particle size distribution in the material under study were used as input data. The Gaussian curve was approximated by a column diagram in which the base of each column has a unit size, and the height expresses the relative fraction of particles falling in a single interval. A series of calculations were carried out for variants with different combinations of mean and variance values. A number of particle sizes were formed in 1 mm steps and the relative fraction of particles for each column of the histogram was calculated. The resulting series of sizes were divided into groups according to standard sizes of standard sieve sets. The sum of the relative particle fractions for each group (in the range between two sizes of the apertures in adjacent sieves) was taken as an estimate of the yield of the fraction that had lingered on the smaller sieve number. It is established that the estimation of the geometric dimensions of feed particles by sieving and the subsequent calculation of the geometric weighted mean is significantly inferior to the reliability of the estimate of the general average of the mean arithmetic weighted method. The choice of the method of statistical data processing for determining the fractional composition of feeds and their mixtures by the sieve method significantly affects the reliability of the average estimate in the analyzed samples. The average particle size, calculated as the arithmetic average, weighted according to the sieve analysis, is an adequate estimate of the structure of the feed and their mixtures. The evaluation of the shape of the distribution and the average particle size of the feed mix does not depend on the series of standard sets of screens used. Increasing the reliability of estimating the average particle size in the studied samples is possible by increasing the number of determined fractions (the number of sieves).
REFERENCES