Avaliação do desempenho de dosadores de fertilizantes de semeadoras adubadoras

Abstract

The increase in the world population and the professionalization of productive agriculture, which requires better productivity in the same area of cultivation. Among several alternatives, there is the precise dosage of the fertilizer at the time of sowin g, which guarantees an increase in production and cost reduction, being in charge of the fertilizer dispenser installed in the fertilizer seeder. This work presents the comparison between the main fertilizer feeders used in fertilizer sowing machines as to the uniformity in the longitudinal distribution and a variation in the dosage in longitudinal slopes. Worm feeder with side discharge, double worm feeder, toothed rotor feeder and transverse overflow feeder using a solid worm were taken. The experiment was carried out in a laboratory of an agricultural products company, using a bench developed especially for this type of study. To make the fertilizer function, the use of polyamide pellets has been validated, which has characteristics very similar to the commercial fertilizer 02-20-20, besides having good mechanical resistance, not being hygroscopic and having homogeneous granulometry, which guarantees the repeatability of the experiment. In the evaluation of the longitudinal distribution, the doser that shows the best performance was the toothed rotor type, followed by the double-screw, solid screw dosers in a transverse overflow and screw-type doser with lateral discharge, which are associated values of variation coefficient 6.1%, 7.7%, 9.4% and 13.3% respectively. The experiment of variation of dosage in different longitudinal inclinations indicated that the best performance is observed in any doser of the endless type with lateral discharge, with CV equal to 4.4%, followed respectively by the dosers of the toothed rotor type (CV 7.6 %), double screw (CV 8.2%) and massive screw in a transverse overflow dispenser (CV 11.5%). As for the number of collection units to evaluate the coefficient of variation of the longitudinal distribution of fertilizers, using this bench and other conditions mentioned in this work, the results indicated that the number of 24 collection units is the smallest that does not differ statistically from the other quantity of collection units, level of 5% probability of error.