Numerical simulation of spray ejection from a nozzle for herbicide application: Comparison of drag coefficient expressions

Abstract

This paper compares the different expressions of drag coefficients using a Computational Fluid Dynamics (CFD) code. This code is a Large-eddy Simulation-Lagrangian Stochastic Model (LES-STO) to simulate liquid particle ejection from a HARDI™ ISO F110-03 nozzle. The results are compared to laboratory measurements of the vertical component of the liquid particle velocity. To do this, the physical process was segmented from the trajectory of the particles into the following two stages: a Transitional Stage and a Sedimentation Stage. At each of these stages, speeds, Reynolds numbers and drag coefficients were calculated independently for each particle during each time step. A model was used that allows for a 1:1 simulation with respect to the applied volume, using a time step of 2 · 10−4 s, evaluating a total of 15,500,000 particles in 40 s of simulation. For the aerodynamic drag coefficient, six different expressions were compared with the laboratory tests; the one proposed by Turton and Levenspiel (1986) achieved the best agreement of the velocity values as shown in the Chi-square goodness of Fit and F-test variance. With these data, it was possible to validate what the Environmental Protection Agency (EPA) described in its pesticide specifications. Potential drift is significantly reduced when the nozzle is placed at a height of 0.75 m because particles larger than 300 µm do not reach sedimentation velocity before impacting the ground. For these drops, the inertia forces do not reach equilibrium with the traction forces, so the particles maintain the ejection inertia. Unfortunately, there is not presently a technology that guarantees the exclusive application of the diameters that are required on the product label.