Analysis of Volume Distribution and Evaluation of the Spraying Spectrum in Terms of Spraying Quality

Assessment of the quality of the operation of agricultural nozzles on the basis of transverse volume distribution and spatial methods of analysis for stream spraying spectra is insufficient, and positive result do not guarantee that the intended and effective spraying effects are obtained. Tests were carried out to assess the quality of nozzles on the basis of transverse volume distribution analysis, microstructure characteristics, and detailed analysis of places where an unexpected change in the nature of the transverse volume distribution (increase in volume) was noted. The subjects of the study were RS11003 flat fan nozzles and a measuring stand equipped with a grooved table, which was used to carry out tests. During the tests, the unit flow rate from the nozzles, the transverse volume distribution of liquids from individual table grooves, and the corresponding CV distribution coefficients of variation were recorded. Detailed tests were carried out for the selected nozzle, consisting of spot measurement of droplet characteristics in individual liquid stream bands. The widths of these bands were constant and equal to the width of the measuring table groove. Measurements were made using analyzer 2D-Laser Doppler Anemometry/Phase Doppler Anemometry (2D-LDA/PDA) from Dantec Dynamics. The analysis of the results obtained from the grooved table and the droplet characteristics in individual stream bands showed clear and unexpected changes in the nature of the transverse volume distribution for all tested nozzles. These changes, consisting of a local increases in droplet diameters (with a reduced number of occurrences), can cause a significant reduction in the quality and effectiveness of spraying, despite the positive fulfillment of generalized normative criteria for their assessment.


Introduction
There is no doubt that the droplet size spectra produced by different nozzle types (e.g., air induction or conventional) [1] and also by liquid atomization (e.g., hydraulic or pneumatic) [2] affect the efficiency of spray application and the efficacy of chemical plant protection products (PPP). An efficient spray application is the main strategy to simultaneously increase the benefits of PPP, reduce the risk of environmental and human contamination, and produce high-quality and safe food in a sustainable way [3,4]. Therefore, the efficiency of spray applications depends primarily on the proper functioning of the nozzles and on the work parameters selected [5,6]. Conventional flat fan nozzles, characterized by hydraulic atomization, are the most commonly used in spraying of arable field crops [6][7][8] and for herbicide applications in 3D crops [9,10]. The selection of nozzle type according to the needs of the treatment specifications should aim to achieve the required total application volume rate, taking into account the spray quality [11], to both maximize the efficacy of treatments and minimize the drift risk. The PPP label recommendations should also be taken into account to select the best working

Materials and Methods
The subjects of the research were new standard flat fan nozzles RS11003 from the MMAT Agro Technology, Leszno, Poland. Seven copies were randomly selected from a batch of 75 pieces. According to the manufacturer's data, the nozzle produces a droplet spectrum ranging from fine to medium droplets with regard to ANSI/ASAE S572. 1,2009. The optimum fraction of the droplets produced is 100-250 µm. Due to the reduction of the risk of drift of the finest droplet fraction (<100 µm), it is recommended to use nozzles in the following conditions: wind speed up to 2 m/s, temperature 12-20 • C and relative humidity above 60%, maximum pressure range 1. 5-5 bar [21]. Certification tests for this group of nozzles were carried out for a pressure of 3 bar, as a result of the tests obtaining a unit flow rate of 1.15 dm 3 /min, spray angle α of 103 • , and a coefficient of uniformity of the transverse volume distribution (characterized by CV coefficient) of 7.6%.
In the first stage of research, the nozzle quality was assessed based on the results of the lateral uniformity of volume distribution, characterized by the CV coefficient. Seven samples of the RS11003 nozzles and the reference TP11003-SS nozzle from TeeJet were used in the tests. The CV coefficient was determined based on the results of measurements of liquid volume distribution on a standardized groove table with a groove width of 25 mm, carried out in accordance with ISO 5682-1-1996 standards Laboratory tests were carried out for the following parameters and accepted conditions:  The working medium was pure water and its temperature was 20 ± 2 °C;  During measurements, a constant value of the spraying boom height above the measuring table surface of 0.4 m and working pressure of 3 bar were maintained;  The duration of each measurement was 120 s;  The accuracy of liquid volume reading in a single measuring vessel was ±1 mL;  The ambient temperature during tests was 20 ± 2 °C, with relative humidity of 40%-80%;  The accuracy of the working pressure reading was ±0.1 bar;  The accuracy of the nozzle's height reading above the measuring table was ±0.005 m;  The accuracy of the nozzle's angle reading in relation to the horizontal was ±1° [15. The automatic monitoring and control system developed by the authors, which was equipped with a laboratory stand, allowed fluctuations in working pressure to be eliminated and the assumed spray dose to be obtained at a constant level during measurements. This system consisted of a precise flow meter, pressure sensor, and valves (including stepless, pressure control, and shut-off valves) that work with a measuring card and computer. During the measurements, the station operating values and the volume of liquid accumulated in individual grooving table measuring cylinders were recorded in the computer memory. The patented computer program equipped with the stand enabled the determination of the CV coefficient value for subsequent samples of tested nozzles, as well as a spraying boom composed of the appropriate number of tested samples. The volumes of liquids coming from adjacent nozzles were aggregated in such a way that the liquid streams overlapped and the axis of the nozzles were spaced every 0.5 m, which gave the same liquid distribution as for the actual field spraying boom [23]. Based on the results obtained in this way, the values of the CV coefficient for the spraying boom were calculated.
In the main part of the research, measurements of the droplet spectrum in the liquid stream were performed using the 2D-LDA/PDA laser analyzer by Dantec Dynamics.
The relevant elements of the Doppler analyzer are shown in Figure 2. Laboratory tests were carried out for the following parameters and accepted conditions: • The working medium was pure water and its temperature was 20 ± 2 • C; • During measurements, a constant value of the spraying boom height above the measuring table surface of 0.4 m and working pressure of 3 bar were maintained; • The duration of each measurement was 120 s; • The accuracy of liquid volume reading in a single measuring vessel was ±1 mL; • The ambient temperature during tests was 20 ± 2 • C, with relative humidity of 40%-80%; • The accuracy of the working pressure reading was ±0.1 bar; • The accuracy of the nozzle's height reading above the measuring table was ±0.005 m; • The accuracy of the nozzle's angle reading in relation to the horizontal was ±1 • [15].
The automatic monitoring and control system developed by the authors, which was equipped with a laboratory stand, allowed fluctuations in working pressure to be eliminated and the assumed spray dose to be obtained at a constant level during measurements. This system consisted of a precise flow meter, pressure sensor, and valves (including stepless, pressure control, and shut-off valves) that work with a measuring card and computer. During the measurements, the station operating values and the volume of liquid accumulated in individual grooving table measuring cylinders were recorded in the computer memory. The patented computer program equipped with the stand enabled the determination of the CV coefficient value for subsequent samples of tested nozzles, as well as a spraying boom composed of the appropriate number of tested samples. The volumes of liquids coming from adjacent nozzles were aggregated in such a way that the liquid streams overlapped and the axis of the nozzles were spaced every 0.5 m, which gave the same liquid distribution as for the actual field spraying boom [23]. Based on the results obtained in this way, the values of the CV coefficient for the spraying boom were calculated.
In the main part of the research, measurements of the droplet spectrum in the liquid stream were performed using the 2D-LDA/PDA laser analyzer by Dantec Dynamics.
The relevant elements of the Doppler analyzer are shown in Figure 2.
streams overlapped and the axis of the nozzles were spaced every 0.5 m, which gave the same liquid distribution as for the actual field spraying boom [23]. Based on the results obtained in this way, the values of the CV coefficient for the spraying boom were calculated. In the main part of the research, measurements of the droplet spectrum in the liquid stream were performed using the 2D-LDA/PDA laser analyzer by Dantec Dynamics.
The relevant elements of the Doppler analyzer are shown in Figure 2. The analyzer probes were placed on a three-axis traverse system, enabling automated multipoint measurement in the measurement space. The 3D traverse system had a positioning resolution of 10 µm and was controlled by 2D-LDA/PDA analyzer software. During the tests, subsequent droplets passed through a small sample volume in the area of two intersecting laser beams, scattering the light by refraction [24][25][26]. The receiving optics were set at an angle of 43.5 • in relation to the laser beam, which ensured the dominance of the refraction of the light scattering mode. For the receiving probe, a B-type aperture mask (average size up to 644.2 µm) and a 0.2 mm gap were used in the test. Table 1 presents the main parameters of the 2D-LDA/PDA analyzer. The spectrum of the droplets produced by the nozzle was examined at a pressure of 3 bar. Measurements were made by spraying water at a temperature of 20 ± 2 • C. Environmental conditions were kept constant by conducting tests at a temperature of 20 ± 2 • C and relative humidity of 40%-80%. Measurements were made at a measuring distance of hz = 0.4 m from the nozzle ( Figure 3) (ISO/FDIS 25358-2018) [27]. The assumed measuring height is consistent with the height of the nozzle position above the surface of the multigrooved table, which was adopted during tests and for evaluation of the transverse distribution of liquid volume. A multipoint measurement was made in the x-axis of the nozzle, with a measuring step of ∆x = 25 mm. Chapple and Hall (1993) showed the usefulness of this method for measuring the properties of agricultural spraying and found that a single measurement along the long axis gave an appropriate representation of the entire spraying process [28]. During the measurements, the average diameters were recorded for each measuring point, namely arithmetic, surface, volumetric, Sauter (D10, D20, D30, D32), and volumetric diameters, below which smaller drops represent 10%, 50%, 90%, and 98% of the total volume at this point (Dv0.1, Dv0.5, Dv0.9 and Dv0.98, respectively). Relative span factor (RSF) values and percentage of total droplet volume less than 100 μm were also recorded at subsequent measuring points (V100).
The relative span factor (RSF) is a dimensionless parameter indicating the homogeneity of the droplet size distribution, defined as: Table 2 presents the test results for 7 randomly selected RS11003 nozzles and the TP11003-SS reference nozzle, determined on the basis of measurements of the transverse distribution of liquid volume on a standardized grooved table, along with the results of measured unit flow rates. A graphic interpretation of the transverse distribution of the liquid collected in grooves equal to 25 mm in width for the tested specimens is presented in the chart below ( Figure 4). For each sample tested, the values of the CV coefficient determined for a single nozzle, the spraying boom, and nozzle assembly are presented. The results are ordered by measurement order. The determined values of the CV coefficient for individual nozzles fulfill a comparative role, while those determined for the spraying boom become a direct assessment of the spraying quality. During the measurements, the average diameters were recorded for each measuring point, namely arithmetic, surface, volumetric, Sauter (D10, D20, D30, D32), and volumetric diameters, below which smaller drops represent 10%, 50%, 90%, and 98% of the total volume at this point (Dv0.1, Dv0.5, Dv0.9 and Dv0.98, respectively). Relative span factor (RSF) values and percentage of total droplet volume less than 100 µm were also recorded at subsequent measuring points (V100).

Results and Discussion
The relative span factor (RSF) is a dimensionless parameter indicating the homogeneity of the droplet size distribution, defined as: Table 2 presents the test results for 7 randomly selected RS11003 nozzles and the TP11003-SS reference nozzle, determined on the basis of measurements of the transverse distribution of liquid volume on a standardized grooved table, along with the results of measured unit flow rates. A graphic interpretation of the transverse distribution of the liquid collected in grooves equal to 25 mm in width for the tested specimens is presented in the chart below ( Figure 4). For each sample tested, the values of the CV coefficient determined for a single nozzle, the spraying boom, and nozzle assembly are presented. The results are ordered by measurement order. The determined values of the CV coefficient for individual nozzles fulfill a comparative role, while those determined for the spraying boom become a direct assessment of the spraying quality. The analysis of the results (Table 2) indicates that when submitting identical nozzle samples for the spraying boom, the obtained were CV values below the maximum permissible value of 10%. The unit flow rate for the tested samples was within ±5% of the nominal flow rate. It can be concluded that in light of the presented work quality assessments, the tested nozzles meet the criteria, which are additionally supplemented by the manufacturer's information on the optimal fraction of produced droplets. It was noted that despite positive assessments, there are large differences in the distribution of the plant protection product. Observations of a similar nature were observed in studies conducted by other authors [29]. The analysis of the transverse volume distribution, measured on a grooved table (Figure 4), shows a clear change in the nature of the volume distribution for individual spray Appl. Sci. 2020, 10, 2395 7 of 13 assessments. It is natural for this type of nozzle to achieve a decreasing liquid volume distribution as it moves away from the nozzle axis (Figure 4, TeeJet TP11003-SS reference nozzle). A significant and unforeseen increase in volume for the RS11003 nozzles subjected to the test was observed at measuring positions 20 to 22 (462.5-512.5 mm; Figure 4) for the right side of the stream. For nozzle number 2, an increase in volume was also noted in the left part of the stream at measuring positions 18 to 20. For basic research in the field of spray spectrum evaluation in selected bands (measuring ranges) and for analysis of spray stream areas in which the volume increase was noted, nozzle number 2 was selected, the volume distribution histogram for which is presented in Figure 5. The list of measurement results is presented in Table 3.     The analysis of the results ( Table 2) indicates that when submitting identical nozzle samples for the spraying boom, the obtained were CV values below the maximum permissible value of 10%. The unit flow rate for the tested samples was within ±5% of the nominal flow rate. It can be concluded that in light of the presented work quality assessments, the tested nozzles meet the criteria, which are additionally supplemented by the manufacturer's information on the optimal fraction of produced droplets. It was noted that despite positive assessments, there are large differences in the distribution of the plant protection product. Observations of a similar nature were observed in studies conducted by other authors [29]. The analysis of the transverse volume distribution, measured on a grooved table (Figure 4), shows a clear change in the nature of the volume distribution for individual spray assessments. It is natural for this type of nozzle to achieve a decreasing liquid volume distribution as it moves away from the nozzle axis ( Figure 4, TeeJet TP11003-SS reference nozzle). A significant and unforeseen increase in volume for the RS11003 nozzles subjected to the test was observed at measuring positions 20 to 22 (462.5-512.5 mm; Figure 4) for the right side of the stream. For nozzle number 2, an increase in volume was also noted in the left part of the stream at measuring positions 18 to 20. For basic research in the field of spray spectrum evaluation in selected bands (measuring ranges) and for analysis of spray stream areas in which the volume increase was noted, nozzle number 2 was selected, the volume distribution histogram for which is presented in Figure 5. The list of measurement results is presented in Table 3.   Note: D10, arithmetic average diameter; D20, surface average diameter; D30, volumetric average diameter; D32, Sauter diameter; Dv0.1, volumetric diameter, below which smaller drops represent 10% of the total volume; Dv0.5, volumetric diameter, below which smaller drops represent 50% of the total volume; Dv0.9, volumetric diameter, below which smaller drops represent 90% of the total volume; Dv0.98, volumetric diameter, below which smaller drops represent 98% of the total volume; RSF, relative span factor; V100, percentage of total droplet volume less than 100 µm.

Results and Discussion
Based on the analysis of the test results, Dv0.5 (VMD) variability was found in the range of 241-262 µm as it moved away from the nozzle axis. Such a small change in volume median diameter (VMD) was observed up to scanning position x = 337.5 mm for the left and right parts of the tested stream (Table 3, Figures 6 and 7). Based on the analysis of the test results, Dv0.5 (VMD) variability was found in the range of 241-262 µ m as it moved away from the nozzle axis. Such a small change in volume median diameter (VMD) was observed up to scanning position x = 337.5 mm for the left and right parts of the tested stream (Table 3, Figures 6 and 7).  Past research [30] has shown that VMD grows in the stream as it moves away from the axis, which is natural for flat fan agricultural nozzles. It was also found that despite a decrease in the number of measured droplets in subsequent samples as they move away from the spray stream axis, the measurement accuracy was reduced, however the obtained volume median diameter (VMD) value was reliable. Above x = 337.5 mm, an increase of Dv0.5 by 20-40 µ m on average was noted at subsequent measuring points, which is natural for these types of nozzles. The number of discrete measurement points was adjusted to obtain a maximum deviation of Dv0.5 between subsequent measurements of ± 10% (ISO/FDIS 25358-2018). An increase in the value of the diameter of the Past research [30] has shown that VMD grows in the stream as it moves away from the axis, which is natural for flat fan agricultural nozzles. It was also found that despite a decrease in the number of measured droplets in subsequent samples as they move away from the spray stream axis, the measurement accuracy was reduced, however the obtained volume median diameter (VMD) value was reliable. Above x = 337.5 mm, an increase of Dv0.5 by 20-40 µm on average was noted at subsequent measuring points, which is natural for these types of nozzles. The number of discrete measurement points was adjusted to obtain a maximum deviation of Dv0.5 between subsequent measurements of ± 10% (ISO/FDIS 25358-2018). An increase in the value of the diameter of the droplet spectrum and a decreasing number of occurrences in this area of the stream results in the correct distribution of liquid volume, which was also noted in studies that assessed the transverse distribution of liquid volume. A significant increase in the value of VMD was found in measuring positions 20 to 22 (462.5-512.5 mm; Figure 7). On the right side, streams averaged 170 µm, while a slightly lower value of 95 µm on average was noted in the same positions for the left. A significant decrease in the value of the relative span factor (RSF) was also noted in these measuring positions, which clearly indicates a decrease in the uniformity of the droplet size distribution. A decrease in the value (V100) from a level of about 5% in the area of the nozzle axis to values of 0.1%-0.2% at these measuring points indicates the occurrence of droplet spectra with diameters greater than 100 µm.
Although a drop in the number of droplet occurrences was noted in these areas (Figure 8), such a significant increase in diameter in the spraying spectra characterized by diameters Dv0.1, Dv0.5, and Dv0.9 resulted in disruption of the liquid volume distribution in these spray bands (Figure 9). This state was confirmed by the results obtained in the assessment of the volume distribution, which was determined for the nozzle on a multigrooved table, as well as by the results of the volume distribution determined on the basis of average volume diameter D30 and the number of drops at subsequent measuring points ( Figure 10).     .9) at measuring points in the liquid stream for nozzle number 2. Note: Dv0.1, volumetric diameter, below which smaller drops represent 10% of the total volume; Dv0.5, volumetric diameter, below which smaller drops represent 50% of the total volume; Dv0.9, volumetric diameter, below which smaller drops represent 90% of the total volume.   .9) at measuring points in the liquid stream for nozzle number 2. Note: Dv0.1, volumetric diameter, below which smaller drops represent 10% of the total volume; Dv0.5, volumetric diameter, below which smaller drops represent 50% of the total volume; Dv0.9, volumetric diameter, below which smaller drops represent 90% of the total volume. Figure 8. Distribution of volume diameters (Dv0.1, Dv0.5, Dv0.9) at measuring points in the liquid stream for nozzle number 2. Note: Dv0.1, volumetric diameter, below which smaller drops represent 10% of the total volume; Dv0.5, volumetric diameter, below which smaller drops represent 50% of the total volume; Dv0.9, volumetric diameter, below which smaller drops represent 90% of the total volume.   According to the ASABE S572.1 reference, the diameter of the volume median diameter (Dv0.5 = 280.5 µm) was obtained for the spray stream, which standardizes the nozzle to the medium droplet class (236-340 µm). A decrease in the volume median diameter value was observed to the level of Dv0.5 = 266.5 µm for the stream in which measuring areas were omitted, in which the disturbance of the liquid volume distribution was noted. In the areas of disturbance on the right side, an average Dv0.5 V 417.6 µm was noted, resulting in very coarse droplets (404-502 µm), while for the left side an average volume median diameter Dv0.5 value of 347.9 µm was noted, which corresponds to the coarse droplet class (341-403 µm). It can be concluded that according to the spray spectrum, the droplet class within the fine-medium range dedicated to this group of nozzles (RS11003) was obtained for the entire stream. However, in areas where special attention was paid, an increase in liquid volume was demonstrated as a result of testing the transverse distribution, which results from a clear increase in droplet diameter in the spectrum, despite the reduced occurrence. Studies have shown a significant change in droplet size, ranging from the medium level used to assess the entire stream, to coarse and very coarse droplets in disturbance areas. This affects the spray structure. Significant differences in the droplet size in individual spraying bands can affect the biological effectiveness of the plant protection product used. In such cases, the threat to the natural environment is very significant as a result of concentration (overlapping) of spraying bands of increased volume, especially in the assembly of nozzles on the spraying boom. The variability of the median volume in this range affects the liquid retention on plants, dripping, and penetration of droplets in the crop.

Conclusions
The presented research results and their analysis led to the following conclusions: • The tested nozzles sufficiently meet the generalized criteria regarding macroparameters (unit flow rate, CV coefficient) and microparameters, including qualification in the field of optimal fraction for produced droplets at the medium level (236-340 µm). This is also confirmed by the results of numerous studies carried out in other scientific centers and research institutions.

•
Pronounced changes in the nature of the volume distribution (volume increase) observed in specific bands of the sprayed stream during the nozzle tests caused a significant change in droplet size, ranging from the medium level for the assessment of the entire stream to coarse or even very coarse droplets. The assembly of such streams emitted by nozzles located on the spraying boom can lead to the synergistic effect of the described disturbances in the considered areas and can contribute to a significant deterioration in the quality and effectiveness of spraying. The effects of poor spraying quality include reduced treatment effectiveness (reduced pest or disease control) or crop damage (overdose of a plant protection product). These effects can also lead to environmental hazards and economic losses. The Doppler laser anemometry method can be used to assess the lateral distribution of liquids on sprayed surfaces and to determine the corresponding value of the CV coefficient as an alternative to analogous measurements on a grooved table. This is confirmed by the results presented in Figure 10.