Characterisation of activation pressure, flowrate and spray angle for hollow-cone nozzles controlled by pulse width modulation

Biosystems Engineering Characterisation of activation pressure, flowrate and spray angle for hollow-cone nozzles controlled by Pulse Width Modulation --Manuscript Draft-- Manuscript Number: YBENG-D-21-01267R2 Article Type: Research Paper Keywords: Orchard sprayer; pulse width modulation; variable-rate; precision pesticide application; intelligent sprayer Corresponding Author: Heping Zhu UNITED STATES First Author: Ramon Salcedo, PhD Order of Authors: Ramon Salcedo, PhD Heping Zhu, PhD Hongyoung Jeon, PhD Erdal Ozkan, PhD Zhiming Wei Emilio Gil Ramón Salcedo Manuscript Region of Origin: North America Abstract: Pulse width modulation (PWM) solenoid valves are used for controlling flowrates of hollow-cone nozzles on variable-rate air-assisted orchard sprayers. However, little information is available on the spray characteristics of these PWM-controlled nozzles. Laboratory tests were performed to evaluate the influences of duty cycles of two different design PWM solenoid valves along with five operating pressures on spray characteristics of five hollow-cone nozzles with different disc-core combinations. Parameters of the spray characteristics investigated were nozzle flowrates, upstream and downstream pressures of the PWM valves, nozzle activation pressures and times, and spray angles of water discharged from nozzles. Test results illustrated that these parameters except for the spray angle were greatly affected by the operating pressure, nozzle selection (disc orifice size and quantity of holes on core), PWM duty cycle and solenoid valve design. Spray angles were affected by the duty cycle but not by the PWM valve design. In general, nozzles with larger disc orifice and higher operating pressures resulted in higher flowrates as expected, whilst the nozzle activation pressure and the spray angle decreased as the duty cycle decreased. Thus, when designing orchard sprayers for precision variable-rate pesticide applications, variations in spray characteristics of disc-core type hollow-cone nozzles controlled by PWM solenoid valves should be considered and minimised. This prospect could be facilitated by establishing a digital database of these characteristics. Opposed Reviewers: Response to Reviewers: Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation Declaration of interests ☒ The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. ☐The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Declaration of Interest Statement Highlights:  Spray parameters were measured for PWM-controlled disc-core hollow-cone nozzles.  Pressures acting on PWM valves and nozzle orifices were significantly different.  Nozzle activation pressures decreased as duty cycles decreased.  PWM valve design affected nozzle pressures and flowrates but not spray angles.  Digital tables were established for improvement of precision variable-rate sprayers. Highlights 1 Characterisation of activation pressure, flowrate and spray angle for hollow-cone 1 nozzles controlled by Pulse Width Modulation 2 Ramón Salcedoa, b, Heping Zhub,*, Hongyoung Jeonb, Erdal Ozkana, Zhiming Weib, c, Emilio Gild 3 a Department of Food Agricultural and Biological Engineering (FABE), The Ohio State University, 4 Columbus, Ohio, United States; 5 b Application Technology Research Unit (ATRU), Agricultural Research Service of the United States 6 Department of Agriculture (USDA-ARS), Wooster, Ohio, United States; 7 c Shandong Academy of Agricultural Machinery Sciences, Jinan, Shandong, China. 8 d Department of Agri-Food Engineering and Biotechnology (DEAB), Polytechnical University of 9 Catalonia, Castelldefels, Barcelona, Spain 10 11 * Corresponding author: Heping Zhu, USDA-ARS ATRU, 1680 Madison Ave., Wooster, OH 44691; 12 Tel: 330-465-4101; E-mail: [email protected]ov 13 14 Abstract 15 Pulse width modulation (PWM) solenoid valves are used for controlling flowrates of hollow-cone 16 nozzles on variable-rate air-assisted orchard sprayers. However, little information is available on the 17 spray characteristics of these PWM-controlled nozzles. Laboratory tests were performed to evaluate the 18 influences of duty cycles of two different design PWM solenoid valves along with five operating 19 pressures on spray characteristics of five hollow-cone nozzles with different disc-core combinations. 20 Parameters of the spray characteristics investigated were nozzle flowrates, upstream and downstream 21 pressures of the PWM valves, nozzle activation pressures and times, and spray angles of water 22 discharged from nozzles. Test results illustrated that these parameters except for the spray angle were 23 Manuscript File Click here to view linked References 2 greatly affected by the operating pressure, nozzle selection (disc orifice size and quantity of holes on 24 core), PWM duty cycle and solenoid valve design. Spray angles were affected by the duty cycle but not 25 by the PWM valve design. In general, nozzles with larger disc orifice and higher operating pressures 26 resulted in higher flowrates as expected, whilst the nozzle activation pressure and the spray angle 27 decreased as the duty cycle decreased. Thus, when designing orchard sprayers for precision variable28 rate pesticide applications, variations in spray characteristics of disc-core type hollow-cone nozzles 29 controlled by PWM solenoid valves should be considered and minimised. This prospect could be 30 facilitated by establishing a digital database of these characteristics. 31 32 Keywords 33 Orchard sprayer, pulse width modulation, variable-rate, precision pesticide application, intelligent 34 sprayer 35 36 1. Introduction 37 Pesticide applications on tree crops and vineyards are traditionally achieved with sprayers 38 delivering constant spray rates (l ha-1) throughout the growing season without considering crop 39 variations in the field. However, because of both the intrinsic variability of the crop (e.g., foliar 40 density, spacing between trees, and canopy size) and sprayer setup, this conventional constant-rate 41 practice can cause crops to be frequently either overor under-sprayed (Salyani, Farooq, and Sweeb, 42 2007; Zhu, Zondag, Derksen, Reding, and Krause,2008; Khot et al., 2014; Garcerá, Moltó, and 43 Chueca, 2017). In addition, this conventional practice can result in pesticides reaching the ground and 44 dispersed into the atmosphere , thereby posing health risks to human and livestock as well as 45 unnecessary increase in production cost. To address such issues, researchers have developed laser46 3 guided precision intelligent sprayers (Chen, Zhu, and Ozkan, 2012; Shen, Zhu, Liu, Chen, and Ozkan, 47 2017) and other sensor systems (Gil et al., 2013; Esau et al., 2014; Maghsoudi, Minaei, Ghobadian, 48 and Masoudi, 2015; Román et al., 2020) for variable-rate applications of pesticides depending on the 49 crop architecture conditions. 50 The pulse width modulation (PWM) solenoid valves were introduced to control flowrates of 51 hydraulic nozzles over 30 years ago (Giles and Comino, 1989 and 1990). The flowrates are 52 manipulated by these electromechanical valves through modulating the valve opening time within a 53 waveform of PWM signals at a specific frequency. The valve opening time, called as “duty cycle” 54 (DUC), varies from 0% for the valve fully closed to 100% for the valve fully opened. Compared with 55 the other alternatives, the valves consisting of solid-state relays with the PWM technique can provide 56 a relatively constant pressure whilst changing the flowrate, resulting in consistent droplet size 57 spectrum (Giles and Comino, 1990; Giles, 1997). This technology has been widely adpated mainly on 58 horizontal boom sprayers for row crop applications normally operated at low pressures between 200 59 and 400 kPa (Han, Hendrickson, Ni, and Zhang, 2001; Bennur and Taylor, 2010; Mangus et al., 2017; 60 Womac, Melnichenko, Steckel, Montgomery, and Hayes, 2016; Butts et al., 2019). However, research 61 conducted in recent years has indicated that changes in DUCs resulted in variations in spray 62 characteristics, such as final flowrate, spray pressure, and spray angle (Needham, Holtz, Giles, 2012; 63 Butts et al., 2018; Wei, Zhu, Zhang, Salcedo, Duan, 2021; Fabula, Sharda, Kang, Flippo, 2021). 64 Conventional air-assisted orchard sprayers apply pesticides at a constant rate and most of them 65 are equipped with disc-core type hollow-cone nozzles operated at spray pressures much greater than 66 those pressures for operating the flat-fan nozzles commonly used on boom sprayers (Zhu et al., 2006; 67 Garcerá et al., 2017; Grella, Marucco, Manzone, Gallart, Balsari, 2017; Kasner et al., 2020; Miranda 68 et al., 2021; Rathnayake, Chandel, Schrader, Hoheisel, Khot, 2021). Recently, Smart Guided Systems 69 4 LLC (Indianapolis, IN, USA) have manufactured a laser-guided intelligent spray control system based 70 on the prototypes developed for orchard sprayers (Chen et al., 2012; Shen et al., 2017), which allows 71 growers to upgrade their conventional constant rate orchard sprayers to achieve variable-rate spray 72 applications. The spray control system is implemented with PWM valves to control hollow-cone 73 nozzle flowrates. However, because the implementation of this technology is new to the orchard spray 74 technologies, there is a lack of adequate information on spray characteristics (such as flowrate, spray 75 angle and simultaneous pressure acting on nozzle orifice) of these hollow-cone nozzles when coupled 76 with PWM valves. Questions have raised concerning the variation in spray characteristics when the 77 PWM valves are used to modulate nozzle flowrates. It is therefore necessary to evaluate the 78 performance of these PWM-controlled nozzles when operated under varying pressures and DUCs. 79 Furthermore, because PWM solenoid valve manufacturers have their own unique designs of valve, it 80 is beneficial to establish the understanding of how the different valve designs influence spray 81 parameters for PWM-controlled hollow-cone nozzles. It is highly likely that valves with different 82 designs will have different response times and flowrates at a specific DUC (Wei et al., 2020). These 83 variations may have a direct effect on spray characteristics which in turn can effect the quality of 84 foliar coverage and deposition as well as the potential risk of spray drift and ground contamination. 85 Therefore, the objective of this research was to determine and interpolate the spray 86 characteristics of disc-core hollow-cone nozzles coupled with two different commercially available 87 designs of PWM valve operating at various spray pressures and DUCs, in an effort to understand the 88 performance of hollow-cone nozzles manipulated with PWM valves when utilised in new variable89 rate orchard sprayers. Spray characteristics measured in this work included nozzle flowrates, upstream 90 and downstream pressures of the PWM valves, nozzle activation pressures and times, and spray 91 angles. 92 5 93 2. Materials and Methods 94 2.1 PWM-controlled nozzles 95 Disc-core nozzle assemblies from TeeJet Technologies (Spraying Systems Co., Wheaton, IL, 96 USA) were selected to generate hollow-cone spray patterns. Each nozzle assembly consisted of a 97 ceramic orifice disc (disc number D2, D4, or D5) and appropriate cores for each disc (core number 98 DC25 or DC45). The orifice diameters of D2, D4, and D5 discs were 1.0, 1.6, and 2.0 mm, 99 respectively. The DC25 core contained two holes and each had a diameter of 1.0 mm whilst the DC45 100 core contained four holes and each had a diameter of 1.6 mm. The holes in the core were positioned 101 obliquely to the central axis to provide swirl forming an internal vortex and a hollow-cone spray 102 pattern after discharge. Different combinations of discs and cores in the nozzle assembly operated at 103 the same pressure could achieve different flowrates and spray angles. Five combinations were 104 selected to form five nozzle tip assemblies for tests: D2-DC25, D2-DC45, D4-DC25, D4-DC45, and 105 D5-DC25. These nozzle tip combinations covered the range of nozzle flowrates commonly used in 106 orchard sprayers. According to the nozzle manufacturer specifications, amongst the five nozzle 107 assemblies the D4-DC45 and D2-DC25 nozzles produced the highest and the lowest flowrates whilst 108 the D5-DC25 and D2-DC45 nozzles generated a hollow-cone spray pattern with the widest and 109 narrowest spray angles, respectively. 110 Each nozzle was connected to a PWM solenoid valve during the tests. Two commonly-used 111 10-Hz PWM valves with different internal designs were selected from different manufacturers: the 112 Capstan valve 540012-24T, from Capstan Ag Systems, Inc. (Topeka, KS, USA) and the TeeJet 113 115880-1-12 from Spraying Systems Co. (Wheaton, IL, USA). They were referred to PWM-1 and 114 PWM-2 hereafter. The PWM-1 valve was fabricated with four inlets each with a diameter of 3.75 mm 115 6 concentrically distributed around an outlet of 3.81 mm internal diameter. The PWM-2 valve consisted 116 of five concentric equidistant inlets each with a diameter of 2.30 mm distributed concentrically 117 around an outlet with an internal diameter of 3.10 mm. Each valve was powered with 14.5 V DC to 118 ensure it could be activated at operating pressures up to 827 kPa, the maximum pressure based on the 119 valve specification. A PWM signal generator with a power supply, designed by Liu et al. (2014), was 120 used to generate standard electronic pulses to activate each solenoid valve with different DUCs 121 ranging from 0% (0 ms ) to 100% (100 ms). The nozzle flowrates were manipulated by the PWM 122 valve opening time which was controlled by the pulse width in each DUC. For the 10 Hz valve, each 123 DUC duration was 100 ms. At 0% DUC, the pulse width was 0 ms and the valve was completely 124 closed. At 60% DUC, the pulse width was 60 ms and technically the valve was opened for 60 ms and 125 closed for 40 ms during each cycle. On the other hand, at 100% DUC, the pulse width was 100 ms 126 and the valve was completely opened. 127 128 2.2 Test setup 129 Experiments were conducted under environmentally controlled conditions at USDA-ARS 130 Application Technology Research Unit in Wooster, Ohio, USA. A mean ambient temperature of 23ºC 131 with mean relative humidity of 26% was maintained through an automatic central control system 132 during the tests. The respective nozzle coupled with a specific PWM valve was assembled on a spray 133 line that also included two additional manual valves (Fig. 1): a shut-off valve for opening or closing 134 the liquid passage from an external source to the nozzle, and an air bleed valve for exiting air trapped 135 in the spray line to avoid air bubbles entering the nozzle chamber. The nozzle discharged sprays 136 straight downward, perpendicular to the floor. The PWM valve was 60 mm above the nozzle body. 137 There were also two pressure sensors mounted in the spray line to measure the pressures upstream and 138 13 PWM-2 valve were greater than the calculated flowrates while nozzles with PWM-2 valve discharged 256 flowrates closer to the calculated flowrates than the nozzles with PWM-1 valve. For each nozzle, 257 using either PWM-1 or PWM-2 valve, the differences between the calculated and measured flowrates 258 decreased as the DUCs increased (Fig. 3a) and slightly increased within a range of 5% to 10% as the 259 operating pressure increased (Fig. 3b). The differences were reduced as the nozzle orifice size 260 increased from D2-DC25 to D4-DC45. These differences between the measured and calculated 261 flowrates were also reported for flat-fan nozzles operating at low pressures (Wei et al., 2021; Fabula 262 et al., 2021). 263 264 (a) 689 kPa pressure (b) 50% duty cycle Fig. 3. Measured and calculated flowrates discharged from D2-DC25 and D4-DC45 hollow-cone nozzles coupled with either PWM-1 or PWM-2 valve operated at (a) 689 kPa pressure with duty cycles ranging from 10% to 100%, and (b) 50% duty cycle with operating pressures ranging from 276 to 827 kPa. 265 The measured flowrates were significantly affected by nozzle orifice size (F-ratio = 234988; 266 df = 4; p< 0.05), operating pressure (F-ratio = 70171; df = 4; p< 0.05), DUC (F = 234562; df = 9; p< 267 0.05), and PWM valve design (F = 84651; df = 1; p< 0.05). For all nozzles coupled with either PWM268 14 1 or PWM-2 valves, as might be expected, the mean flowrates generally increased as both DUCs 269 increased from 10% to 100% and operating pressures from 276 to 827 kPa (Table 1). However, the 270 flowrates of all nozzles coupled with the PWM-1 valve remained unchanged between 90% and 100% 271 DUCs, which was similar to the flat-fan nozzles operated at 276 kPa pressure (Wei et al., 2021). This 272 observation suggested that a 10 ms interval time was not sufficient to completely close or open the 273 PWM-1 valve when using the valve at a modulation frequency of 10 Hz. 274 Furthermore, the flowrate at 80% DUC was still within 97% to 99% of the mean flowrate at 275 100% DUC for the smaller nozzles (D2-DC25 and D2-DC45) or the larger nozzle (D4-DC25) 276 operated at pressures below 552 kPa. Thus, the flowrate increased linearly with DUCs in most 277 treatments when DUCs were between 10% and 80%. For example, as shown in Table 1, for D2-DC25 278 at 276 and 827 kPa pressures, increments of the mean flowrate with a DUC increase of 10% within 279 the DUC range between 10% and 80% were 0.05 l min-1 and 0.09 l min-1, respectively. Similar trends 280 were also observed for D4-DC45 as the flowrate increased by 0.13 and 0.24 l min-1 for 276 and 827 281 kPa, respectively. 282 Nozzles coupled with the PWM-2 valve discharged lower flowrates than those with the PWM283 1 valve with DUCs less than 90% whilst flowrates from nozzles with both PWM valves were equal at 284 100% DUC for a given pressure. These differences could be explained by the internal flow chamber 285 design of each valve since the inlet and outlet diameters of the PWM-1 valve were greater than those 286 of the PWM-2 valves. For a given nozzle and pressure, the mean flowrate of the PWM-2 valve at 90% 287 DUC was closer to the flowrate of the PWM-1 valve at 80% DUC, and this trend continued until the 288 flowrates from the PWM-2 valve at 20% DUC matched the flowrates from the PWM-1 valve at 10% 289 DUC. It should be noted that the PWM-2 valve started making changes in flowrate at 80% DUC 290 whilst the PWM-1 valve started making changes at 90% DUC. 291 15 In addition, the standard deviations of the mean flowrates suggested that the nozzles with 292 PWM-2 valve had less variation in the mean flowrate than those with the PWM-1 valve as the spray 293 pressure changed from 276 to 827 kPa, but the maximum coefficient of variations for mean flowrates 294 was 3.1%. Flowrates of all nozzles coupled with both PWM-1 and PWM-2 valves at all assigned 295 pressures were measured for 30%, 40%, 60%, and 70% DUCs but were not reported in Table 1 296 because they followed the same trends. Therefore, different PWM solenoid valve designs could have 297 different DUC ranges where their discharge flowrates varied linearly, indicating the necessity to 298 evaluate nozzle flowrates manipulated with different PWM valve designs to achieve accurate 299 variable-rate applications. Thus, although PWM valves should modulate nozzle flowrates within their 300 specified ranges of DUCs; in practice their modulation accuracy varied with the valve design. Thus, 301 flowrates of nozzles coupled with different PWM solenoid valve designs should be evaluated before 302 their integration into variable rate applications. 303 304 16 Table 1. Mean flowrates of hollow-cone nozzles coupled with PWM-1 and PWM-2 valves operated at duty cycles (DUC) ranging from 10% to 100% and pressures at 276, 414, 552, 689 and 827 kPa Hollowcone nozzle Pressure (kPa) Mean flowrate (l min-1)[a] DUC (%) for PWM-1 DUC (%) for PWM-2 100 90 80 50 20 10 100 90 80 50 20 10 D2-DC25 276 0.56 0.56 0.55 0.41 0.25 0.19 0.57 0.56 0.51 0.35 0.18 0.13 414 0.68 0.68 0.67 0.49 0.29 0.23 0.68 0.67 0.61 0.42 0.21 0.15 552 0.77 0.77 0.75 0.55 0.33 0.25 0.78 0.77 0.70 0.47 0.24 0.17 689 0.86 0.86 0.85 0.62 0.36 0.27 0.87 0.85 0.76 0.52 0.25 0.17 827 0.94 0.94 0.92 0.66 0.39 0.29 0.95 0.91 0.81 0.55 0.26 0.17 D2-DC45 276 0.72 0.72 0.71 0.51 0.31 0.24 0.72 0.70 0.63 0.43 0.22 0.15 414 0.87 0.87 0.86 0.62 0.37 0.28 0.87 0.85 0.77 0.52 0.26 0.17 552 0.99 0.99 0.97 0.70 0.41 0.31 1.00 0.96 0.87 0.59 0.29 0.19 689 1.11 1.11 1.08 0.78 0.45 0.34 1.11 1.07 0.97 0.65 0.32 0.21 827 1.21 1.21 1.18 0.84 0.48 0.35 1.22 1.16 1.05 0.70 0.32 0.21 D4-DC25 276 0.97 0.97 0.96 0.66 0.39 0.29 1.01 0.97 0.86 0.56 0.27 0.18 414 1.18 1.18 1.15 0.80 0.46 0.34 1.22 1.17 1.05 0.68 0.32 0.21 552 1.35 1.35 1.31 0.91 0.52 0.39 1.37 1.32 1.19 0.78 0.37 0.24 689 1.51 1.51 1.44 1.01 0.57 0.41 1.53 1.47 1.30 0.88 0.40 0.26 827 1.64 1.64 1.57 1.09 0.61 0.44 1.66 1.58 1.40 0.92 0.40 0.25 D5-DC25 276 1.30 1.30 1.26 0.86 0.52 0.38 1.30 1.26 1.13 0.71 0.33 0.22 414 1.57 1.56 1.49 1.06 0.61 0.45 1.57 1.51 1.35 0.86 0.40 0.26 552 1.80 1.79 1.72 1.20 0.68 0.49 1.81 1.72 1.52 0.99 0.45 0.28 689 1.97 1.97 1.89 1.31 0.74 0.53 2.01 1.89 1.67 1.09 0.46 0.28 827 2.15 2.14 2.02 1.39 0.77 0.56 2.20 2.04 1.80 1.17 0.48 0.28 D4-DC45 276 1.35 1.34 1.30 0.90 0.53 0.40 1.35 1.33 1.19 0.75 0.36 0.24 414 1.63 1.63 1.58 1.10 0.63 0.46 1.64 1.56 1.39 0.89 0.41 0.26 552 1.88 1.87 1.81 1.24 0.71 0.52 1.90 1.80 1.60 1.03 0.46 0.28 689 2.09 2.09 1.98 1.38 0.77 0.55 2.11 1.95 1.73 1.13 0.49 0.29 827 2.28 2.28 2.10 1.44 0.78 0.55 2.32 2.15 1.89 1.23 0.51 0.30 [a] Standard deviations of all mean flowrates were not greater than 0.06 l min-1 for PWM-1 valve while not greater than 0.02 l min-1 for PWM-2 valve. 305 The measured flowrates in Table 1 could be the basis of a database to assign DUCs for 306 accurate spray volume discharged for variable-rate orchard sprayers equipped with PWM-controlled 307 nozzles. The utilisation of this database could help reduce chemical consumption, and minimise risks 308 associated with off-target losses of pesticides. For example, if a variable-rate application was 309 performed at 827 kPa operating pressure with D4-DC45 nozzles coupled with PWM-1 valve and the 310 desired flowrate was 1.82 l min-1, then the calculated DUC should be 80% (1.82 = 2.28 × 80%). 311 However, the measured flowrate at 80% DUC shown Table 1 was 2.10 l min-1 which is 15.3% greater 312 than the calculated flowrate. By using the database established in Table 1, the flow rate of 1.82 l min313 17 1 could be achieved by using 69% DUC. Thus, the database could assist in the selection of proper 314 values for DUC. Nevertherless, more experiments to establish a more complete database could be 315 necessary for different nozzles operated at different pressures and DUCs. 316 317 3.2 Upstream and downstream pressures 318 Upstream pressures fluctuated around the operating pressures for nozzles with both PWM 319 valves (Fig. 4). The greatest variations occurred when the PWM valve was closed due to the hydraulic 320 shock. The sudden stop of the flow caused water hammer to trigger a momentary sharp increase of the 321 upstream pressure greater than the operating pressure. In a multiple nozzle system, however, this 322 upstream pressure spike in the spray boom might be reduced if individual nozzles were not activated 323 at the same time or increased if all nozzles were operated with synchronized open-close actions. The 324 upstream pressure dropped to levels below the operating pressure when the valve opened and the 325 liquid passed through the valve, and then gradually stabilised when the valve remained open. This 326 hydraulic shock was more severe for nozzles with a larger orifice since the nozzle discharged greater 327 volumes of water during each duty cycle. For example, when PWM valves were operated at 50% 328 DUC and 827 kPa assigned spray pressure, the upstream pressures fluctuated between 700 and 900 329 kPa and between 600 and 1100 kPa for D2-DC25 and D4-DC45 nozzles, respectively. Thus, the 330 larger D4-DC45 nozzle had greater upstream pressure surge than the smaller D2-DC25 nozzle when 331 opening and closing actions occurred. 332 333 18 (a) D2-DC25 (b) D4-DC45 Fig. 4. Instantaneous upstream pressures on 10 Hz PWM-1 and PWM-2 valves to manipulate (a) D2-DC25, and (b) D4-DC45 nozzles at 50% DUC (green waveforms) whilst operated at assigned 276 and 827 kPa pressures. 334 Despite these fluctuations, the mean upstream pressure changed < 2% between DUCs for both 335 valves (Fig. 5). The mean upstream pressure increased along with decreasing DUCs as the duration of 336 the valve opening time decreased; as a result, overpressure occurred when the valve remained closed. 337 This trend was more notable with larger nozzles and higher operating pressures. Other factors, 338 inherent to the hydraulic system such as the pipe size or accumulator elements, could also influence 339 the hydraulic shock. Thus, future investigations should include the influence of hydraulic system 340 design on the pressure fluctuations for the PWM valves. 341 In addition, maximum difference of mean upstream pressures for DUCs between 10% to 342 100% was 64 kPa for PWM-1 and was always below 40 kPa for the PWM-2 valve. That is, the PWM343 1 valve had slightly higher differences in mean upstream pressure than the PWM-2 valve. For 344 example, the mean differences between 100% and 10% DUCs were 24 and 19 kPa for PWM-1 and 345 PWM-2, respectively. This effect was intensified by increasing the pressure or the nozzle orifice size 346 because they could produce a greater volume of liquid flowing through the valve. 347 19 348 (a) At 276 kPa (b) At 827 kPa Fig. 5. Mean upstream pressures when duty cycle (DUC) changed from 100% to 10% for two nozzles (D2-DC25 and D4-DC45) connected to two different PWM valves at operating pressures of: (a) 276 kPa, and (b) 827 kPa. 349 Figure 6 shows instantaneous downstream pressures acting on D2-DC25 and D4-DC45 350 nozzles coupled with PWM-1 and PWM-2 valves with 30% and 90% DUCs operated at 827 kPa. 351 Unlike the upstream pressure profiles, during each DUC process, downstream pressures reached zero 352 when the valve was closed and sharply increased to the operating pressure when the valve was open. 353 It was obvious that the mean downstream pressures decreased as DUCs decreased for all nozzles 354 because the nozzles would have longer closed time with lower DUCs. However, the PWM-1 valve at 355 90% DUC remained fully open regardless of the nozzle size or the operating pressure (Fig. 6c, 6d), 356 whilst the PWM-2 could open and close at the same 90% DUC. 357 20 (a) D2-DC25 at 30% DUC (b) D4-DC45 at 30% DUC (c) D2-DC25 at 90% DUC (d) D4-DC45 at 90% DUC Fig. 6. Instantaneous downstream pressures over three 100-ms modulation cycles (green lines presented ideal pressure curves) for D2-DC25 and D4-DC45 nozzles operated at 827 kPa and 30% and 90% DUCs with the 10 Hz PWM-1 and PWM-2 valves. 358 Table 2 summarises the mean percentage deviations of upstream and downstream pressures, 359 and the mean percentage deviations of the upstream pressures from 100% to 10% DUC. The 360 maximum pressure dropped due to PWM valve operations were less than 10%, which was also 361 observed for flat-fan pattern nozzles operated under 448 kPa (Fabula et al., 2021). Flows through the 362 PWM-2 valve normally had greater pressure drop than the PWM-1 valve although the mean pressure 363 deviations slightly increased for both valves as spray pressures increased. Both the PWM-1 and 364 21 PWM-2 valves had higher pressure deviations for larger orifice nozzles (i.e. D4-DC45 and D5365 DC25). 366 Table 2. Mean percent deviations of upstream and downstream pressures (∆𝑃𝑑 𝑢), and mean percent upstream pressure deviations of upstream pressures (∆𝑃100 10 ) between changes of DUCs from 100% to 10% with assigned operating pressures of 276, 414, 552, 689, and 827 kPa for five different sized hollow-cone nozzles (D2-DC25, D2-DC45, D4-DC25, D5-DC25, D4-DC45) coupled with either PWM-1 or PWM-2 solenoid valve. Valve Operating pressure (kPa) Mean pressure change (%) D2-DC25 D2-DC45 D4-DC25 D5-DC25 D4-DC45 ∆𝑃𝑑 𝑢 ∆𝑃100 10 ∆𝑃𝑑 𝑢 ∆𝑃100 10 ∆𝑃𝑑 𝑢 ∆𝑃100 10 ∆𝑃𝑑 𝑢 ∆𝑃100 10 ∆𝑃𝑑 𝑢 ∆𝑃100 10 PWM-1 276 -2.1 2.5 -1.3 3.9 0.7 3.8 3.6 5.7 3.1 7.5 414 0.2 2.5 1.0 2.4 2.2 4.0 5.1 6.5 5.5 7.3 552 0.6 1.5 1.3 2.8 3.2 3.7 6.1 7.0 6.4 6.8 689 1.0 2.0 2.2 2.5 3.9 3.2 6.2 7.3 7.7 5.0 827 1.5 1.1 2.3 2.0 4.0 2.7 4.4 7.4 7.3 5.3 PWM-2 276 -1.4 3.3 0.0 3.0 2.1 4.3 6.4 5.2 6.0 2.5 414 0.7 2.2 1.8 2.6 4.9 3.4 7.9 5.2 7.2 6.2 552 1.1 2.1 2.8 2.0 5.7 3.1 7.9 4.4 9.1 5.3 689 2.2 1.2 3.2 1.5 5.8 3.7 7.2 4.8 9.9 4.5 827 2.1 1.3 3.4 2.6 6.0 2.5 7.5 3.4 9.5 4.5 367 Under ideal conditions, the upstream pressure should remain constant and equal to the 368 downstream pressure acting on the nozzle when the PWM valve was open during each DUC. 369 However, the flow chamber structure and the intrinsic characteristics of each PWM valve could cause 370 energy loss as indicated by Bernoulli's principle. This energy loss could lead to pressure losses and 371 therefore, for optimal field applications, suitable adjustments to the operating pressure should be 372 made to achieve the desired pressure acting on the nozzle orifice. 373 374 22 3.3 Nozzle activation pressure 375 The activation pressure profiles were also shown in Fig. 6 as the pulse portion of downstream 376 pressures acting on D2-DC25 and D4-DC45 nozzles. In general, activation pressures stabilised more 377 rapidly for smaller nozzles regardless of either the DUC or the operating pressure. Initial activation 378 pressure peaks were observed for the small nozzle (D2-DC25) under high operating pressure (827 379 kPa). The pressure peak might be due to momentary flow resistance caused by passing fluid through 380 small orifices at high pressures. However, the peak diminished as nozzle orifice size was increased. In 381 addition, the PWM-2 valve had more fluctuations in the activation pressure than the PWM-1 valve, 382 which was similar to the observation by Wei et al. (2021). 383 Table 3 illustrates the mean activation pressures on five nozzles coupled with either PWM-1 or 384 PWM-2 valve with DUCs ranging from 10% to 100% when spray operating pressures were 276, 552, 385 and 827 kPa, respectively. The mean activation pressure decreased as DUCs decreased, which was 386 similar to the trends for flowrates shown in Table 1. This reduction was related to the kinetic energy 387 of fluid passing through the nozzle orifice and the PWM valve assembly. Discharging a greater 388 amount of liquid through the nozzle could consume higher kinetic energy, causing an energy 389 reduction in the form of downstream pressure reduction. Activation pressures for 30%, 40%, 60% and 390 70% DUCs followed the same trends as other DUCs (data were not shown in Table 3). PWM-2 valve 391 generally had relatively higher activation pressure reduction than PWM-1 valve. The greatest 392 differences in the activation pressures between the two valves occurred when DUC reached 30% to 393 40%. 394 Differences in activation pressures between two PWM valve designs were more evident with 395 higher operating pressures or flowrates, which was likely the result of different internal chambers and 396 component designs of the two PWM-valves that might restrict liquid movement differently. 397 29 activation pressures, activation times, and spray angles. The PWM valve design also affected these 495 spray characteristics except for the spray angle since there were no significant differences in spray 496 angles between the two PWM valve designs. However, when DUC and pressure were constant, spray 497 angles increased as disc orifice diameters increased for both PWM valves, but decreased as the 498 quantity of exiting holes in the core increased. Increasing the operating pressure increased both 499 activation pressures and activation times thereby increasing the flowrates at all DUCs for all nozzles 500 tested. 501 The calculated flowrates of nozzles with both PWM-1 and PWM-2 valves were greater than 502 the measured flowrates at all DUCs. Also, the measured flowrates at each DUC with the PWM-1 503 valve were greater than those with the PWM-2 valve, indicating that the flowrates with the PWM-2 504 valve were closer to the calculated flowrates than those with the PWM-1 valve. The measured 505 flowrates proportionally decreased as DUCs decreased for both PWM valves while the decrease 506 started at 100% DUC with PWM-2 valve and at 90% with PWM-1 valve. Thus flowrates did not 507 change for all nozzles coupled with the PWM-1 valve when DUC changed from 90% to 100%. 508 Evaluations of different PWM-controlled nozzles should be continued to establish a large 509 digital database of spray characteristics, which will be used for the future optimal design and 510 improvement of precision air-assisted sprayers to efficiently and effectively protect crops and 511 safeguard the environment. 512 513 5 Acknowledgements 514 Mention of company or trade names is for description only and does not imply endorsement by 515 the USDA. The USDA is an equal opportunity provider and employer. 516 The authors would like to express their gratitude to Adam Clark, Barry Nudd, and Andy 517 30 Doklovic for their technical assistance throughout the laboratory tests. We also appreciate the USDA518 NIFA Specialty Crop Initiative (Grant No. 2015-51181-24253) for partially financial support of this 519 research. 520 521 6 References 522 Bennur, P. J., & Taylor, R. K. (2010). Evaluating the response time of a rate controller used with a 523 sensor-based, variable rate application system. Applied Engineering in Agriculture, 26(6), 524 1069-1075. https://doi.org/10.13031/2013.35903 525 Butts, T. R., Samples, C. A., Franca, L. X., Dodds, D. M., Reynolds, D. B., Adams, J. W., Zollinger, 526 R. K., Howatt, K. A., Fritz, B. K., Hoffmann, W. C., & Kruger, G. R. (2018). Spray droplet 527 size and carrier volume effect on dicamba and glufosinate efficacy. Pest Management Science, 528 74(9), 2020-2029. https://doi.org/10.1002/ps.4913 529 Butts, T. R., Butts, L. E., Luck, J. D., Fritz, B. K., Hoffmann, W. C., & Kruger, G. R. (2019). Droplet 530 size and nozzle tip pressure from a pulse-width modulation sprayer. Biosystems engineering, 531 178, 52-69. https://doi.org/10.1016/j.biosystemseng.2018.11.004 532 Chen, Y., Zhu, H., & Ozkan, H. E. (2012). Development of a variable-rate sprayer with laser scanning 533 sensor to synchronize spray outputs to tree structures. Transactions of the ASABE, 55(3), 773534 781. https://doi.org/10.13031/2013.41509 535 Esau, T. J., Zaman, Q. U., Chang, Y. K., Schumann, A. W., Percival, D. C., & Farooque, A. A. (2014). 536 Spot-application of fungicide for wild blueberry using an automated prototype variable rate 537 sprayer. Precision agriculture, 15(2), 147-161. https://doi.org/10.1007/s11119-013-9319-4 538 Fabula, J., Sharda, A., Kang, Q., & Flippo, D. (2021). Nozzle flowrate, pressure drop, and response 539 time of pulse width modulation (PWM) nozzle control systems. Transactions of the ASABE, 540 31 64(5): 1519-1532. https://doi.org/10.13031/trans.14360 541 Garcerá, C., Moltó, E., & Chueca, P. (2017). Spray pesticide applications in Mediterranean citrus 542 orchards: Canopy deposition and off-target losses. Science of the Total Environment, 599, 543 1344-1362. https://doi.org/10.1016/j.scitotenv.2017.05.029 544 Gil, E., Llorens, J., Llop, J., Fàbregas, X., Escolà, A., & Rosell-Polo, J. R. (2013). Variable rate 545 sprayer. Part 2-Vineyard prototype: Design, implementation, and validation. Computers and 546 Electronics in Agriculture, 95, 136-150. https://doi.org/10.1016/j.compag.2013.02.010 547 Giles, D. K., & Comino, J. A. (1989). Variable flow control for pressure atomization nozzles. 548 Transactions of SAE, Journal of Commercial Vehicles, 98(2), 237-249. 549 Giles, D. K., & Comino, J. A. (1990). Droplet size and spray pattern characteristics of an electronic 550 flow controller for spray nozzles. Journal of Agricultural Engineering Research, 47, 249-267. 551 https://doi.org/10.1016/0021-8634(90)80045-V 552 Giles, D. K. (1997). Independent control of liquid flowrate and spray droplet size from hydraulic 553 atomizers. Atomization and Sprays, 7(2). https://doi.org/10.1615/AtomizSpr.v7.i2.40 554 Grella, M., Marucco, P., Manzone, M., Gallart, M., & Balsari, P. (2017). Effect of sprayer settings on 555 spray drift during pesticide application in poplar plantations (Populus spp.). Science of the 556 Total Environment, 578, 427-439. https://doi.org/10.1016/j.scitotenv.2016.10.205 557 Grella, M., Gioelli, F., Marucco, P., Zwertvaegher, I., Mozzanini, E., Mylonas, N., Nuyttens, D. & 558 Balsari, P. (2021). Field assessment of a pulse width modulation (PWM) spray system 559 applying different spray volumes: duty cycle and forward speed effects on vines spray 560 coverage. Precision Agriculture, 1-34. https://doi.org/10.1007/s11119-021-09835-6 561 Han, S., Hendrickson, L. L., Ni, B., & Zhang, Q. (2001). Modification and testing of a commercial 562 sprayer with PWM solenoids for precision spraying. Applied Engineering in Agriculture, 563 32 17(5), 591. https://doi.org/10.13031/2013.6906 564 Kasner, E.J., Fenske, R.A., Hoheisel, G.A., Galvin, K., Blanco, M.N., Seto, E.Y., & Yost, M.G. 565 (2020). Spray drift from three airblast sprayer technologies in a modern orchard work 566 environment. Annals of Work Exposures and Health, 64, 25-37. 567 https://doi.org/10.1093/annweh/wxz080 568 Khot, L. R., Ehsani, R., Maja, J. M. J., Campoy, J. M., Wellington, C., & Al-Jumaili, A. (2014). 569 Evaluation of deposition and coverage by an air-assisted sprayer and two air-blast sprayers in a 570 citrus orchard. Transactions of the ASABE, 57(4), 1007-1013. 571 https://doi.org/10.13031/trans.57.10419 572 Liu, H., Zhu, H., Shen, Y., Chen, Y., & Ozkan, H. E. (2014). Development of digital flow control 573 system for multi-channel variable-rate sprayers. Transactions of the ASABE, 57(1), 273-281. 574 https://doi.org/10.13031/trans.57.10216 575 Maghsoudi, H., Minaei, S., Ghobadian, B., & Masoudi, H. (2015). Ultrasonic sensing of pistachio 576 canopy for low-volume precision spraying. Computers and Electronics in Agriculture, 112, 577 149-160. https://doi.org/10.1016/j.compag.2014.12.015 578 Mangus, D. L., Sharda, A., Engelhardt, A., Flippo, D., Strasser, R., Luck, J. D., & Griffin, T. (2017). 579 Analyzing the nozzle spray fan pattern of an agricultural sprayer using pulse width modulation 580 technology to generate an on-ground coverage map. Transactions of the ASABE, 60(2), 315581 325. https://doi.org/10.13031/trans.11835 582 Miranda, M. P., Scapin, M. S., Vizoni, M. C., Zanardi, O. Z., Eduardo, W. I., & Volpe, H. X .L. 583 (2021). Spray volumes and frequencies of insecticide applications for suppressing Diaphorina 584 citri populations in orchards. Crop Protection, 140, 105406. 585 https://doi.org/10.1016/j.cropro.2020.105406 586 33 Needham, D. L., Holtz, A. J., & Giles, D. K. (2012). Actuator system for individual nozzle control of 587 flowrate and spray droplet size. Transactions of the ASABE, 55(2), 379-386. 588 https://doi.org/10.13031/2013.4137 589 Rathnayake, A. P., Chandel, A. K., Schrader, M. J., Hoheisel, G. A., & Khot, L. R. (2021). Spray 590 patterns and perceptive canopy interaction assessment of commercial airblast sprayers used in 591 Pacific Northwest perennial specialty crop production. Computers and Electronics in 592 Agriculture, 184, 106097. https://doi.org/10.1016/j.cropro.2022.105921 593 Román, C., Llorens, J., Uribeetxebarria, A., Sanz, R., Planas, S., & Arnó, J. (2020). Spatially variable 594 pesticide application in vineyards: Part II, field comparison of uniform and map-based variable 595 dose treatments. Biosystems Engineering, 195, 42-53. 596 https://doi.org/10.1016/j.biosystemseng.2020.04.013 597 Saha, A., Lee, J. D., Basu, S., & Kumar, R. (2012). Breakup and coalescence characteristics of a 598 hollow cone swirling spray. Physics of fluids, 24(12), 124103. 599 https://doi.org/10.1063/1.4773065 600 Salcedo, R., Zhu, H., Zhang, Z., Wei, Z., Chen, L., Ozkan, E., & Falchieri, D. (2020). Foliar 601 deposition and coverage on young apple trees with PWM-controlled spray systems. 602 Computers and Electronics in Agriculture, 178, 105794. 603 https://doi.org/10.1016/j.compag.2020.105794 604 Salcedo, R., Zhu, H., Ozkan, E., Falchieri, D., Zhang, Z., & Wei, Z. (2021). Reducing ground and 605 airborne drift losses in young apple orchards with PWM-controlled spray systems. Computers 606 and Electronics in Agriculture, 189, 106389. https://doi.org/10.1016/j.compag.2021.106389 607 Salcedo, R., Zhu, H., Jeon, H., Ozkan, E., Wei, Z., Gil, E., Campos, J., and Román, C. (2022). Droplet 608 size distributions from hollow-cone nozzles coupled with PWM valves. Submitted to Journal 609 34 of the ASABE for publication. (In review) 610 Salyani, M., Farooq, M., & Sweeb, R. D. (2007). Spray deposition and mass balance in citrus orchard 611 applications. Transactions of the ASABE, 50(6), 1963-1969. 612 https://doi.org/10.13031/2013.24092 613 Shen, Y., Zhu, H., Liu, H., Chen, Y., & Ozkan, H. E. (2017). Development of a laser-guided, 614 embedded-computer-controlled, air-assisted precision sprayer. Transactions of the ASABE, 615 60(6), 1827-1838. https://doi.org/10.13031/trans.1245 616 Wei, Z., Zhu, H., Zhang, Z., Salcedo, R., & Duan, D. (2021). Droplet size spectrum, activation 617 pressure, and flowrate discharged from pwm flat-fan nozzles. Transactions of the ASABE, 618 64(1), 313-325. https://doi.org/10.13031/trans.14100 619 Womac, A. R., Melnichenko, G., Steckel, L., Montgomery, G., & Hayes, R. M. (2016). Spray tip 620 effect on glufosinate canopy deposits in Palmer amaranth (Amaranthus palmeri) for pulse621 width modulation versus air-induction technologies. Transactions of the ASABE, 59(6), 1597622 1608. https://doi.org/10.13031/trans.59.11642 623 Zhu, H., Derksen, R. C., Guler, H., Krause, C. R., and Ozkan, H.E. (2006). Foliar deposition and off624 target loss with different spray techniques in nursery applications. Transactions of the ASABE, 625 49(2): 325-334. 626 Zhu, H., Zondag, R. H., Derksen, R. C., Reding, M., & Krause, C. R. (2008). Influence of spray 627 volume on spray deposition and coverage within nursery trees. Journal of Environmental 628 Horticulture, 26(1), 51-57. https://doi.org/10.24266/0738-2898-26.1.51 629 1 Characterization Characterisation of activation pressure, flow rateflowrate and 1 spray angle for hollow-cone nozzles controlled by PPulse Width Modulation M2 controlled hollow-cone nozzles 3 Ramón Salcedoa, b, Heping Zhub,*, Hongyoung Jeonb, Erdal Ozkana, Zhiming Weib, c, Emilio Gild 4 a Department of Food Agricultural and Biological Engineering (FABE), The Ohio State University, 5 Columbus, Ohio, United States; 6 b Application Technology Research Unit (ATRU), Agricultural Research Service of the United States 7 Department of Agriculture (USDA-ARS), Wooster, Ohio, United States; 8 c Shandong Academy of Agricultural Machinery Sciences, Jinan, Shandong, China. 9 d Department of Agri-Food Engineering and Biotechnology (DEAB), Polytechnical University of 10 Catalonia, Castelldefels, Barcelona, Spain 11 12 * Corresponding author: Heping Zhu, USDA-ARS ATRU, 1680 Madison Ave., Wooster, OH 44691; 13 Tel: 330-465-4101; E-mail: hepi[email protected] 14 15 Abstract 16 Pulse width modulation (PWM) solenoid valves have beenare used for controlling flow rateflowrates 17 of hollow-cone nozzles on variable-rate air-assisted orchard sprayers. However, little information is 18 available on the spray characteristics of these PWM-controlled nozzles. Laboratory tests were 19 performed to evaluate the influences of duty cycles of two different fabricated design PWM solenoid 20 valves along with five operating pressures on spray characteristics of five hollow-cone nozzles with 21 different disc-core combinations. Parameters of the spray characteristics investigated were nozzle flow 22 rateflowrates, upstream and downstream pressures of the PWM valves, nozzle activation pressures and 23 Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Editor-Annotated Manuscript 2 times, and spray angles of water discharged from nozzles. Test results illustrated that these parameters 24 except for the spray angle were greatly affected by the operating pressure, nozzle selection (disc orifice 25 size and quantity of holes on core), PWM duty cycle and solenoid valve fabricationdesign. Spray angles 26 were affected by the duty cycle but not the by the PWM valve fabricationdesign. In general, nozzles 27 with larger disc orifice and higher operating pressures resulted in higher flow rateflowrates as expected, 28 while whilst the nozzle activation pressure and the spray angle decreased as the duty cycle decreased. 29 Thus, when designing orchard sprayers for precision variable-rate pesticide applications, variations in 30 spray characteristics of disc-core type hollow-cone nozzles coupled withcontrolled by PWM solenoid 31 valves should be considered and minimized minimised. This prospect could be facilitated through by 32 established establishing a digital database of these characteristics in future designs of orchard sprayers 33 to perform precision variable-rate pesticide applications. 34 35 Keywords 36 Orchard sprayer, pulse width modulation, variable-rate, precision pesticide application, intelligent 37 sprayer 38 39 1. Introduction 40 Pesticide applications on tree crops and vineyards are traditionally achieved with sprayers 41 delivering constant spray rates (L l ha-1) throughout the growing season without considering crop 42 variations in the field. However, because of both the intrinsic variability of the crop (e.g., foliar 43 density, spacing between trees, and canopy size) and sprayer setup, this conventional constant-rate 44 practice can cause crops to be frequently either overor under-sprayed (Salyani, Farooq, and Sweeb, 45 2007; Zhu, Zondag, Derksen, Reding, and Krause,2008; Khot et al., 2014; Garcerá, Moltó, and 46 Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Commented [A1]: I think this is the point you are making. Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) 3 Chueca, 2017). In addition, this conventional practice can also result in pesticides reaching the ground 47 and the dispersed into the atmosphere , thereby posing health risks to human and livestock as well as 48 unnecessary increase in production cost. To address such issues, researchers have developed laser49 guided precision intelligent sprayers (Chen, Zhu, and Ozkan, 2012; Shen, Zhu, Liu, Chen, and Ozkan, 50 2017) and other sensor systems (Gil et al., 2013; Esau et al., 2014; Maghsoudi, Minaei, Ghobadian, 51 and Masoudi, 2015; Román et al., 2020) for variable-rate applications of pesticides depending on the 52 crop architecture conditions. 53 The pulse width modulation (PWM) solenoid valves were introduced to control flow rates of 54 hydraulic nozzles over 30 years ago (Giles and Comino, 1989 and 1990). The flow rates are 55 manipulated by these electromechanical valves through modulating the valve opening time within a 56 waveform of PWM signals at a specific frequency. The valve opening time, called as “duty cycle” 57 (DUC), varied varies from 0% for the valve fully closed to 100% for the valve fully opened. 58 Compared with the other alternatives, the valves consisting of solid-state relays with the PWM 59 technique can provide a relatively constant pressure while whilst changing the flow rate, resulting in 60 consistent droplet size spectrum (Giles and Comino, 1990; Giles, 1997). This technology has been 61 widely adpated mainly on horizontal boom sprayers for row crop applications normally operated at 62 low pressures between 200 and 400 kPa (Han, Hendrickson, Ni, and Zhang, 2001; Bennur and Taylor, 63 2010; Mangus et al., 2017; Womac, Melnichenko, Steckel, Montgomery, and Hayes, 2016; Butts et 64 al., 2019). However, research conducted in recent years has indicated that changes in DUCs resulted 65 in variations in spray characteristics, such as final flow rateflowrate, spray pressure, and spray angle 66 (Needham, Holtz, Giles, 2012; Butts et al., 2018; Wei, Zhu, Zhang, Salcedo, Duan, 2021; Fabula, 67 Sharda, Kang, Flippo, 2021). 68 Conventional air-assisted orchard sprayers apply pesticides at a constant rate and most of them 69 Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) 4 are equipped with disc-core type hollow-cone nozzles operated at high spray pressures much greater 70 than those pressures for operating the flat-fan nozzles commonly used on boom sprayers (Zhu et al., 71 2006; Garcerá et al., 2017; Grella, Marucco, Manzone, Gallart, Balsari, 2017; Kasner et al., 2020; 72 Miranda et al., 2021; Rathnayake, Chandel, Schrader, Hoheisel, Khot, 2021). Recently, Smart Guided 73 Systems LLC (Indianapolis, IN, USA) have manufactures manufactured a laser-guided intelligent 74 spray control system based on the prototypes developed for orchard sprayers (Chen et al., 2012; Shen 75 et al., 2017), which allows growers to upgrade their conventional constant rate orchard sprayers to 76 achieve variable-rate spray applications. The spray control system is implemented with PWM valves 77 to control hollow-cone nozzle flow rateflowrates. However, bBecause the implementation of this 78 technology is new to the orchard spray technologies, there is a lack of adequate information on spray 79 characteristics (such as flow rateflowrate, spray angle and simultaneous pressure acting on nozzle 80 orifice) of these hollow-cone nozzles when coupled with PWM valves. Questions have raised if 81 theseconcerning the variation in spray characteristics are significantly changed when the PWM valves 82 are used to modulate the nozzle flow rates. It is therefore necessary to evaluate the performance of 83 these PWM-controlled nozzles when operated under varying pressures and DUCs when improved the 84 variable-rate orchard sprayers are needed. Furthermore, because PWM solenoid valve manufacturers 85 have their own unique fabrication designs of the valve, it will is be beneficial to establish the 86 understanding if of how the different valve fabrications designs influence the spray parameters of 87 thefor PWM-controlled hollow-cone nozzles. It is highly possible likely that these valves with 88 different designs will have different response times and flow rateflowrates at a specific DUC (Wei et 89 al., 2020). These variations may have a direct effect on spray characteristics which, in turn, on the 90 can effect the quality of foliar coverage and deposition as well as the potential risk of airborne spray 91 drift and ground driftcontamination. 92 Commented [A2]: High pressure is not a precise term since it depends on context. Please give a range of values (in kPA) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) 11 10% DUCs (%). 218 219 2.3.3 Nozzle activation pressure and activation time 220 Because PWM valves regulated the flow rateflowrates with by DUCs, the downstream 221 pressure that acted on the nozzle orifice performed ashasd a pulse waveform when the PWM valve 222 was activating. Nozzle activation pressure, as described by Wei et al. (2021), was introduced to 223 define the pulsed part of the downstream pressure profile within the time duration when the PWM 224 valve was open at during each duty cycle. The activation pressure was the actual the pressure to 225 pushthat causesd fluid liquid to pass through the nozzle to produce spray droplets. Similarly, nozzle 226 activation time, as defined by Wei et al. (2021), was also employed to express the actual time 227 durationperiod that the nozzle was discharging discharged sprays at within each duty cycle. The 228 activation time was determined from the activation pressure profile in three phases: 1) an initial 229 period during which the valve was openingopensed (rising time), 2) the period that the valve was 230 fully open; and 3) the last period when the valve was closing closesd (falling time). 231 232 2.3.4 Spray angles 233 Spray pattern profiles discharged from hollow-cone nozzles were determined from images 234 taken with a high-speed camera (Phantom v310; Vision Research Inc., AEMTEK, Wayne, NJ, USA) 235 along withusing an image capturing program (Phantom Camera Control 2.8; Vision Research Inc., 236 AEMTEK, Wayne, NJ, USA) at the resolution of 1280 × 800 pixels (figFig. 2a). LED lights 237 (Multiled LT-V8-11; GS Vitec; Bad Soden-Salmünster, Germany) illuminated the spray pattern 238 profiles, and an opaque black background was placed behind the nozzle ensured enoughto provide 239 image contrast. 240 Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) 12 241 (a) High-speed camera setup (b) Image analysis Figure Fig. 2. Test setup for measurement of spray pattern angles produced from hollow-cone nozzles using: (a) a high-speed camera and LED lights, and (b) an imaging analysis program. 242 Spray angles were measured from the spray pattern images similar to the one shown in figure 243 Fig. 2b using 2D image analysis software (Image Pro Plus 6.1; Media Cybernetics, Inc., Rockville, 244 MD, USA). Considering the general behavior of hollow-cone nozzles (Saha et al., 2012), the spray 245 angle was only measured only nearclose to the nozzle orifice before the liquid breakup occurred in the 246 spray pattern. Since Because the distance between the nozzle and the liquid breakup was not constant, 247 the distance was always observed visually before and determined before starting the spray angle 248 measurements were started. 249 250 2.3.5. Statistical analysis 251 The mean value of each parameter (flow rateflowrate, upstream pressure, downstream 252 pressure, activation pressure, activation time, spray angle) for each nozzle and PWM valve 253 High-speed camera LED light Spray angle Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom), Check spelling and grammar Formatted: English (United Kingdom) Formatted: English (United Kingdom) 13 combination was calculated from using three replications. A four-way analysis of variance (ANOVA) 254 was performed separately for the flow rateflowrate, activation pressure, activation time, and spray 255 angle. Each parameter was considered as a main factor. The influence of four variables (PWM valve, 256 nozzle size, spray pressure, and DUCs) as subfactors on the corresponding parameter was determined. 257 A confidence level of 95% was used in all analyses. Statistical assessments were processed using 258 ProStat version 3.81 (Poly Software International, Pearl River, NY, USA). Before the 259 calculationsdescriptive statistics were calculated, data were evaluated for their normality distribution 260 using the Kolmogorov-Smirnov test by visual observation inspection of the residues. A previous data 261 transformation (∛x) was used to obtain the normality distribution except for the spray angle. 262 263 3. Results and Discussion 264 3.1 Flow rateFlowrates 265 Figure 3a shows the measured and calculated flow rateflowrates at different DUCs for D2266 DC25 and D4-DC45 nozzles operated with PWM-1 and PWM-2 solenoid valves operated at 689 kPa 267 pressure while figure 3b shows these flow rateflowrates at different operating pressures for the two 268 PWM-controlled nozzles at the 50% DUC. The trends shown in figure Fig. 3 were also true for all 269 other nozzles coupled with either PWM-1 or PWM-2 valve. The mean flow rateflowrates for all 270 nozzles coupled with either PWM-1 or PWM-2 valve were greater than the calculated flow 271 rateflowrates while nozzles with PWM-2 valve discharged flow rateflowrates closer to the calculated 272 flow rateflowrates than the nozzles with PWM-1 valve. For the sameeach nozzle, with using either 273 PWM-1 or PWM-2 valve, the differences between the calculated and measured flow rateflowrates 274 decreased as the DUCs increased (figFig. 3a) and slightly increased within a range of 5% to 10% as 275 the operating pressure increased (figFig. 3b). The differences were reduced as the nozzle orifice size 276 Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) 14 increased from D2-DC25 to D4-DC45. These differences between the measured and calculated flow 277 rateflowrates were also reported for flat-fan nozzles at operating at low pressures (Wei et al., 2021; 278 Fabula et al., 2021). 279 280 (a) 689 kPa pressure (b) 50% duty cycle Figure Fig. 3. Measured and calculated flow rateflowrates discharged from D2-DC25 and D4-DC45 hollow-cone nozzles coupled with either PWM-1 or PWM-2 valve operated at (a) 689 kPa pressure with duty cycles ranging from 10% to 100%, and (b) 50% duty cycle with operating pressures ranging from 276 to 827 kPa. 281 The measured flow rateflowrates were significantly affected by nozzle orifice size (F-ratio = 282 Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) 15 234988; df = 4; p< 0.05), operating pressure (F-ratio = 70171; df = 4; p< 0.05), DUC (F = 234562; df 283 = 9; p< 0.05), and PWM valve fabricationdesign (F = 84651; df = 1; p< 0.05). For all nozzles coupled 284 with either PWM-1 or PWM-2 valves, as might be expected, the mean flow rateflowrates generally 285 increased as both DUCs increased from 10% to 100% and operating pressures from 276 to 827 kPa 286 (Table 1). However, the flow rateflowrates of all nozzles coupled with the PWM-1 valve were 287 remained unchanged between 90% and 100% DUCs, which was similar to the flat-fan nozzles 288 operated at 276 kPa pressure (Wei et al., 2021). This observation suggested that a 10 ms interval time 289 was not enough sufficient to completely close or open the PWM-1 valve when using the the valve at a 290 at the modulation frequency of 10 Hz. 291 Furthermore, the flow rateflowrate at 80% DUC was still within 97% to 99% of the mean flow 292 rateflowrate at 100% DUC for the smaller nozzles (D2-DC25 and D2-DC45) or the larger nozzle (D4293 DC25) operated at pressures below 552 kPa. Thus, tThe flow rateflowrate increased linearly with 294 DUCs in most treatments when DUCs were between 10% and 80%. For example, as shown in Table 295 1, for D2-DC25 at 276 and 827 kPa pressures, increments of the mean flow rateflowrate with a DUC 296 increase of 10% within the DUC range between 10% and 80% were 0.05 L l min-1 and 0.09 L l min-1, 297 respectively. Similar trends were also observed for D4-DC45 as the flow rateflowrate increased by 298 0.13 and 0.24 L l min-1 for 276 and 827 kPa, respectively. 299 Nozzles coupled with the PWM-2 valve discharged lower flow rateflowrates than those with 300 the PWM-1 valve with DUCs less than from 90% and lower DUCs whilest flow rateflowrates from 301 nozzles with both PWM valves were equal at 100% DUC for a given pressure. These differences 302 could be explained by the internal flow chamber design of each valve that since the inlet and outlet 303 diameters of the PWM-1 valve were greater than those of the PWM-2 valves. For a given nozzle and 304 pressure, the mean flow rateflowrate of the PWM-2 valve at 90% DUC was closer to the flow 305 Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) 16 rateflowrate of the PWM-1 valve at 80% DUC, and this trend continued until the flow rateflowrates 306 from the PWM-2 valve at 20% DUC matched to the flow rateflowrates from the PWM-1 valve at 307 10% DUC. Itg should be noted that theThe PWM-2 valve was designed to operate with 308 electromagnetic forces to started making changes in flow rateflowrates at 80% DUC while whilst the 309 PWM-1 valve started making changes at 90% DUC. 310 In addition, the standard deviations of the mean flow rateflowrates suggested that the nozzles 311 with PWM-2 valve had less variations in the mean flow rateflowrates than those with the PWM-1 312 valve as the spray pressure changed from 276 to 827 kPa, but the maximum coefficient of variations 313 for mean flow rateflowrates was 3.1%. Flow rateFlowrates of all nozzles coupled with both PWM-1 314 and PWM-2 valves at all assigned pressures were also measured for 30%, 40%, 60%, and 70% DUCs 315 but were not reported in table Table 1 because they followed the same trends as shown in the table. 316 Therefore, different PWM solenoid valve fabrications designs would canould have different DUC 317 ranges to where their linearly discharge flow rateflowrates varyied linearly, which indicatedindicating 318 the necessity to evaluate nozzle flow rateflowrates manipulated with different PWM valve designs 319 fabrications to achieve accurate variable-rate applications. That isThus, although technically all PWM 320 valves should modulate nozzle flow rateflowrates to comply with pulse widths ofwithin their 321 specified specified ranges of DUCs; however, practicallyin practice their modulation accuracy varied 322 with the valve designs. Thus, flow rateflowrates of nozzles coupled with different PWM solenoid 323 valve fabricationdesigns should be evaluated before they were areir integrated integration into 324 variable rate applications. 325 326 Formatted ... Formatted ... Formatted ... Formatted ... Formatted ... Formatted ... Formatted ... Formatted ... Formatted ... Formatted ... Formatted ... Formatted ... Formatted ... Formatted ... Formatted ... Formatted ... Formatted ... Formatted ... Formatted ... Formatted ... Formatted ... Formatted ... Formatted ... Formatted ... Formatted ... Formatted ... Formatted ... Formatted ... Formatted ... Formatted ... Formatted ... Formatted ... Formatted ... Formatted ... Formatted ... Formatted ... Formatted ... Formatted ... Formatted ... 17 327 Table 1. Mean flow rateflowrates of hollow-cone nozzles coupled with PWM-1 and PWM-2 valves operated at duty cycles (DUC) ranging from 10% to 100% and pressures at 276, 414, 552, 689 and 827 kPa Hollowcone nozzle Pressure (kPa) Mean flow rateflowrate (L l min-1)[a] DUC (%) for PWM-1 DUC (%) for PWM-2 100 90 80 50 20 10 100 90 80 50 20 10 D2-DC25 276 0.56 0.56 0.55 0.41 0.25 0.19 0.57 0.56 0.51 0.35 0.18 0.13 414 0.68 0.68 0.67 0.49 0.29 0.23 0.68 0.67 0.61 0.42 0.21 0.15 552 0.77 0.77 0.75 0.55 0.33 0.25 0.78 0.77 0.70 0.47 0.24 0.17 689 0.86 0.86 0.85 0.62 0.36 0.27 0.87 0.85 0.76 0.52 0.25 0.17 827 0.94 0.94 0.92 0.66 0.39 0.29 0.95 0.91 0.81 0.55 0.26 0.17 D2-DC45 276 0.72 0.72 0.71 0.51 0.31 0.24 0.72 0.70 0.63 0.43 0.22 0.15 414 0.87 0.87 0.86 0.62 0.37 0.28 0.87 0.85 0.77 0.52 0.26 0.17 552 0.99 0.99 0.97 0.70 0.41 0.31 1.00 0.96 0.87 0.59 0.29 0.19 689 1.11 1.11 1.08 0.78 0.45 0.34 1.11 1.07 0.97 0.65 0.32 0.21 827 1.21 1.21 1.18 0.84 0.48 0.35 1.22 1.16 1.05 0.70 0.32 0.21 D4-DC25 276 0.97 0.97 0.96 0.66 0.39 0.29 1.01 0.97 0.86 0.56 0.27 0.18 414 1.18 1.18 1.15 0.80 0.46 0.34 1.22 1.17 1.05 0.68 0.32 0.21 552 1.35 1.35 1.31 0.91 0.52 0.39 1.37 1.32 1.19 0.78 0.37 0.24 689 1.51 1.51 1.44 1.01 0.57 0.41 1.53 1.47 1.30 0.88 0.40 0.26 827 1.64 1.64 1.57 1.09 0.61 0.44 1.66 1.58 1.40 0.92 0.40 0.25 D5-DC25 276 1.30 1.30 1.26 0.86 0.52 0.38 1.30 1.26 1.13 0.71 0.33 0.22 414 1.57 1.56 1.49 1.06 0.61 0.45 1.57 1.51 1.35 0.86 0.40 0.26 552 1.80 1.79 1.72 1.20 0.68 0.49 1.81 1.72 1.52 0.99 0.45 0.28 689 1.97 1.97 1.89 1.31 0.74 0.53 2.01 1.89 1.67 1.09 0.46 0.28 827 2.15 2.14 2.02 1.39 0.77 0.56 2.20 2.04 1.80 1.17 0.48 0.28 D4-DC45 276 1.35 1.34 1.30 0.90 0.53 0.40 1.35 1.33 1.19 0.75 0.36 0.24 414 1.63 1.63 1.58 1.10 0.63 0.46 1.64 1.56 1.39 0.89 0.41 0.26 552 1.88 1.87 1.81 1.24 0.71 0.52 1.90 1.80 1.60 1.03 0.46 0.28 689 2.09 2.09 1.98 1.38 0.77 0.55 2.11 1.95 1.73 1.13 0.49 0.29 827 2.28 2.28 2.10 1.44 0.78 0.55 2.32 2.15 1.89 1.23 0.51 0.30 [a] Standard deviations of all mean flow rateflowrates were not greater than 0.06 lL min-1 for PWM-1 valve while not greater than 0.02 L l min-1 for PWM-2 valve. 328 The measured flow rateflowrates in table Table 1 could be the basis of a database for future 329 computer programs to assign DUCs for achieving accurate spray volume discharged from for 330 variable-rate orchard sprayers equipped with PWM-controlled nozzles. The utilization utilisation of 331 the this database could help reduce chemical consumption, and minimize minimise risks associated 332 with off-target losses of pesticides. For example, if a variable-rate application was performed at 827 333 kPa operating pressure with D4-DC45 nozzles coupled with PWM-1 valve and the desired flow 334 rateflowrate was 1.82 L l min-1, then the calculated DUC should be 80% (1.82 = 2.28 × 80%). 335 Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) 18 However, the measured flow rateflowrate at 80% DUC shown Table 1 was 2.10 L l min-1 which was is 336 15.3% greater than the calculated flow rateflowrate. By using the database established in table Table 337 1, the flow rate of 1.82 L l min-1 could be achieved at by using 69% DUC. Thus, the database could 338 assist in the selection of proper values for DUCs to save chemical products and spray volume and 339 further improve the accuracy of variable-rate applications. Nevertherless, more experiments to 340 establish greater a more complete database would could be necessary for different nozzles operated at 341 different pressures and DUCs in the future. 342 343 3.2 Upstream and downstream pressures 344 The uUpstream pressures fluctuated around the operating pressures for nozzles with both 345 PWM valves (figFig. 4). The greatest variations occurred at the moment when the PWM valve was 346 closed due to the hydraulic shock. The sudden stop of the flow caused water hammer to trigger a 347 momentary sharp increase of the upstream pressure greater than the operating pressure. In a multiple 348 nozzle system, however, this upstream pressure spike in the spray boom might be reduced if 349 individual nozzles were not activated at the same time or increased if all nozzles were performed 350 operated with synchronized open-close actions. The upstream pressure dropped to levels below the 351 operating pressure when the valve opened and the fluid liquid passed through the valve, and then 352 gradually stabilized stabilisesd when the valve remained open. This hydraulic shock was more severe 353 for nozzles with a larger orifice as since the nozzle dischargesd greater volumes of water in during 354 each duty cycle. For example, when PWM valves were operated at 50% DUC and 827 kPa assigned 355 spray pressure, the upstream pressures fluctuated between 700 and 900 kPa and between 600 and 356 1100 kPa for D2-DC25 and D4-DC45 nozzles, respectively. That isThus, the larger D4-DC45 nozzle 357 had greater upstream pressure surge than the smaller D2-DC25 nozzle due towhen opening and close 358 Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) 19 closing actions occurred. 359 360 (a) D2-DC25 (b) D4-DC45 Figure Fig. 4. Instantaneous upstream pressures on 10 Hz PWM-1 and PWM-2 valves to manipulate (a) D2-DC25, and (b) D4-DC45 nozzles at 50% DUC (green waveforms) while whilst operated at assigned 276 and 827 kPa pressures. 361 Despite these fluctuations, the mean upstream pressure changed less than< 2% between DUCs 362 for both valves (figFig. 5). The mean upstream pressure increased along with decreasing DUCs as the 363 duration of the valve opening time decreased; as a result, there was an increased in overpressure 364 occurred when the valve remained closed. This trend was more notable with larger nozzles and higher 365 operating pressures. Other factors, inherent to the hydraulic system such as the pipe size or 366 accumulator elements, could also influence the hydraulic shock. Thus, future investigations should 367 include the influence of hydraulic system design on the pressure fluctuations for the PWM valves. 368 In addition, maximum difference of mean upstream pressures for DUCs between 10% to 369 100% was 64 kPa for PWM-1 and was always below 40 kPa for the PWM-2 valve. That is, the PWM370 1 valve had slightly higher differences of in mean upstream pressures than the PWM-2 valve. For 371 example, the mean differences between 100% and 10% DUCs were 24 and 19 kPa for PWM-1 and 372 Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) 20 PWM-2, respectively. This effect was intensified by increasing the pressure or the nozzle orifice size 373 because they could produce a greater amount volume of liquid mass flowing through the valve. 374 375 (a) At 276 kPa (b) At 827 kPa Figure Fig. 5. Mean upstream pressures when duty cycle (DUC) changed from 100% to 10% for two nozzles (D2-DC25 and D4-DC45) connected to two different PWM valves at operating pressures of: (a) 276 kPa, and (b) 827 kPa. 376 Figure 6 shows instantaneous downstream pressures acting on D2-DC25 and D4-DC45 377 nozzles coupled with PWM-1 and PWM-2 valves with 30% and 90% DUCs operated at 827 kPa. 378 Differently fromUnlike the upstream pressure profiles, during each DUC process, downstream 379 pressures reached zero when the valve was closed and sharply increased to the operating pressure 380 when the valve was open. It was obvious that the mean downstream pressures decreased as DUCs 381 decreased for all nozzles because the nozzles would have longer closed time with lower DUCs. 382 However, the PWM-1 valve at 90% DUC remained fully open regardless of the nozzle size or the 383 operating pressure (figFig. 6c, 6d), while whilst the PWM-2 valve enabled couldto open and close the 384 valve at the same 90% DUC. 385 Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) 27 474 Table 4. Mean activation times of hollow-cone nozzles coupled with PWM-1 and PWM-2 valves operated at duty cycles (DUC) ranging from 10% to 100% and operating pressures at 276, 414, 552, 689 and 827 kPa Hollowcone nozzle Pressure (kPa) Mean activation time (s)[a] DUC (%) for PWM-1 DUC (%) for PWM-2 100 90 80 50 20 10 100 90 80 50 20 10 D2-DC25 276 100 100 93 63 34 24 100 98 87 58 28 18 414 100 100 94 65 35 24 100 98 89 59 29 20 552 100 100 95 65 36 25 100 98 90 60 30 20 689 100 100 96 67 37 26 100 98 91 62 32 21 827 100 100 97 68 37 27 100 99 93 63 32 22 D2-DC45 276 100 100 92 63 33 23 100 97 87 57 27 17 414 100 100 93 64 34 24 100 98 88 58 28 18 552 100 100 94 64 35 24 100 98 88 59 29 19 689 100 100 95 65 35 24 100 98 90 60 29 19 827 100 100 95 66 36 25 100 98 91 62 31 20 D4-DC25 276 100 100 92 63 33 23 100 97 87 56 27 16 414 100 100 93 64 34 24 100 97 87 58 27 17 552 100 100 94 64 34 24 100 98 88 58 28 18 689 100 100 95 65 35 25 100 98 89 59 29 19 827 100 100 95 66 36 25 100 98 91 62 31 20 D5-DC25 276 100 100 91 62 32 22 100 96 86 55 25 16 414 100 100 92 62 32 22 100 97 87 56 26 16 552 100 100 93 63 33 22 100 98 88 57 27 17 689 100 100 94 64 34 23 100 98 88 58 28 18 827 100 100 94 65 34 24 100 97 87 58 28 17 D4-DC45 276 100 100 91 61 32 21 100 95 85 55 25 14 414 100 100 92 62 32 22 100 95 85 55 25 15 552 100 100 92 62 32 22 100 96 86 56 26 16 689 100 100 93 63 33 22 100 97 87 57 27 17 827 100 100 94 63 34 23 100 97 87 58 28 17 [a] Standard deviations of all mean times were lower than 1 ms for both PWM-1 and PWM-2 valves. 475 3.5 Spray angle 476 Table 5 shows the mean spray angles of hollow-cone nozzles coupled with either PWM-1 or 477 PWM-2 valve operated at pressures ranging from 276 to 827 kPa and DUCs from 10% to 100%. 478 Overall, spray angles increased as the operating pressure increased for all nozzles at all DUCs. Also, 479 there were was less variations in spray angles at higher operating pressures with smaller nozzles or at 480 lower pressures with larger nozzles. However, there was a tendency of reducedfor spray angles to 481 reduce with reduced reducing DUCs for nozzles with both PWM valves. The reason was that the 482 pressure and flow rateflowrate decreased as DUC decreased, resulting in the momentum of fluid 483 Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) 28 exiting the nozzle the nozzle decreasedto decrease, thus; consequently, reducing the spray angle also 484 decreased. 485 A nozzle coupled with either PWM-1 or PWM-2 valve showed similar spray angles under the 486 same operating pressure and DUC conditions, although the mean angle varied slightly more with 487 PWM-2 than that with the PWM-1 valve. For example, the differences in the mean spray angles for 488 the D4-DC45 nozzle between 100% and 10% DUCs were -7.2º and -11.1º with PWM-1 valve and - 489 7.5º and -12.4º with PWM-2 valve when the operating pressures were 276 and 827 kPa, respectively. 490 Unlike other nozzle-PWM combinations, a the D2-DC25 nozzle showed spray angles 491 increased slightly as DUCs decreased at 276 kPa pressure regardless which PWM valve was used. 492 However, this increase diminished as the pressure increased to 827 kPa. For example, the mean 493 difference in spray angles from the D2-DC25 nozzle between 100% and 10% DUCs at 276 kPa 494 pressure was 7.4º with PWM-1 valve and 8.6º with PWM-2 valve. Similar trend was also observed for 495 D2-DC25 and D4 -DC25 nozzles operated at low pressures. 496 497 Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) 29 498 499 Table 5. Mean spray angles of hollow-cone nozzles coupled with PWM-1 and PWM-2 valves operated at duty cycles (DUC) ranging from 10% to 100% and pressures at 276, 414, 552, 689 and 827 kPa Hollowcone nozzle Pressure (kPa) Mean spray angle (º)[a] DUC (%) for PWM-1 DUC (%) for PWM-2 100 90 80 50 20 10 100 90 80 50 20 10 D2-DC25 276 51 50 51 56 55 58 50 52 53 54 58 58 414 55 55 56 59 59 60 55 56 57 57 60 61 552 57 57 57 59 61 61 57 57 57 58 59 62 689 58 58 59 59 59 61 58 58 59 58 61 63 827 60 60 59 60 60 60 60 58 59 60 61 64 D2-DC45 276 38 38 39 41 42 43 39 42 41 42 41 41 414 44 44 45 43 44 45 45 43 42 42 43 44 552 48 48 46 43 44 44 48 45 44 43 41 44 689 51 51 48 45 43 43 51 48 46 46 43 43 827 54 54 53 48 43 43 54 50 49 46 44 43 D4-DC25 276 70 69 70 69 67 70 71 70 69 67 70 75 414 74 74 72 70 68 70 73 73 71 68 70 75 552 76 75 73 72 70 71 75 75 74 73 70 72 689 78 77 74 72 71 71 78 77 73 74 71 74 827 80 79 77 74 71 73 79 78 75 74 73 74 D5-DC25 276 75 75 74 71 71 74 76 75 74 74 75 77 414 78 77 75 74 73 75 78 78 76 77 74 77 552 79 79 78 76 74 75 80 79 79 78 75 78 689 81 80 80 78 76 76 81 80 81 78 76 78 827 82 81 80 79 77 77 83 82 80 80 78 79 D4-DC45 276 61 61 59 55 55 53 60 59 59 57 54 53 414 62 62 59 57 56 54 63 61 59 57 56 53 552 65 64 61 59 57 56 64 62 61 60 57 55 689 68 67 64 61 58 57 68 65 65 60 57 56 827 69 69 64 62 59 58 70 68 65 62 59 57 [a] Standard deviations of all mean spray angles were lower than 4º for both PWM-1 and PWM-2 valves. 500 As mentioned aboveearlier, the orifice diameters of disc D2, D4 and D5 were 1.0, 1.6, and 2.0 501 mm, and numbers of holes on cores DC25 and DC45 were two and four, respectively. Data in table 502 Table 5 illustrates that with the same core in the nozzle assembly operated at the same pressure and 503 same DUC, spray angles increased as disc orifice diameters increased for both PWM valves. For 504 example, for nozzles with the DC25 core operated at 552 kPa pressure and 50% DUC, the spray angle 505 from nozzles coupled with the PWM-1 valve increased from 59º to 72º and then 76º when the disc 506 increased from D2 to D4 and then D5. With the same conditions for nozzles with the PWM-2 valve, 507 Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) 30 the spray angle increased from 58º to 73º and then 78º. Under the same operating pressure and same 508 DUC, larger disc orifice diameters allowed more fluids liquid to be discharged with greater tangential 509 momentum into the atmosphere, resulting in wider spray angles. 510 On the other handHowever, with the same disc in the nozzle assembly operated at the same 511 pressure and same DUC, spray angles decreased considerably as the number of holes on the core 512 increased (table Table 5). For example, for nozzles with the D4 disc operated at 827 kPa pressure and 513 50% DUC, the spray angle from nozzles coupled with the PWM-1 valve decreased from 77º to 64º 514 when the hole number in the core increased from 2 (DC25) to 4 (DC45). Similarly, under the same 515 conditions, the spray angle from nozzles coupled with the PWM-2 valve decreased from 74º to 62º. 516 More holes on in the core might cause weaker tangential momentum of fluids the liquid when as 517 exiting it is emitted from the nozzle disc orifice, causing narrower spray angles. 518 Statistic analyses revealed that the nozzle disc-core assembly (F-ratio = 17486; df = 4; p< 519 0.05), spray pressure (F-ratio = 622; df = 4; p< 0.05), and DUC (F-ratio = 566; df = 4; p< 0.05) had 520 significant effects on the spray angle, but not the PWM valve used (F-ratio = 1.4; df = 1; p = 0.235). 521 For sprayers without air assistance, the variations of spray angles might provide fluctuations in spray 522 deposition distributions across target areas. However, the initial spray clouds plume from air-assisted 523 sprayers could be spread dispersed liquid evenly since air from fans iswould be highly turbulent and 524 likely to cause significant mixing particularly as the distance from the sprayer increased 525 increasesdbecause the discharged air disturbed and agitated the spray clouds consistently. This 526 tendence was also reported for PWM-vineyard sprayers tested under field conditions (Grella et al., 527 2021). Thus, the influence of DUCs and other parameters on spray angles would should have a 528 minimal impact on applicationss of with air-assisted orchard sprayers. 529 530 Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) 31 4 Conclusions 531 For disc-core hollow-cone pattern nozzles coupled with PWM solenoid valves, operating 532 pressures and DUCs significantly affected the nozzle flow rateflowrates, upstream and downstream 533 pressures, activation pressures, activation times, and spray angles. The PWM valve fabrication design 534 also affected these spray characteristics except for the spray angle since . Tthere were no significant 535 differences in spray angles between the two PWM valve fabricationdesigns. However, ; when DUC 536 and pressure were constant, . Hhowever, spray angles increased as disc orifice diameters increased for 537 both PWM valves, but decreased as the quantity of exiting holes in the core increased. Increasing the 538 operating pressure increased both activation pressures and activation times and thereby increasing the 539 flow rateflowrates at all DUCs for all nozzles tested. 540 The calculated flow rateflowrates of nozzles with both PWM-1 and PWM-2 valves were 541 greater than the measured flow rateflowrates at all DUCs. Also, the measured flow rateflowrates at 542 each DUC with the PWM-1 valve were greater than those with the PWM-2 valve, indicating that the 543 flow rateflowrates with the PWM-2 valve were closer to the calculated flow rateflowrates than those 544 with the PWM-1 valve. The measured flow rateflowrates proportionally decreased as DUCs decreased 545 for both PWM valves while the decrease started at 100% DUC with PWM-2 valve and at 90% with 546 PWM-1 valve. Thus The measured flow rateflowrates did not change for all nozzles coupled with the 547 PWM-1 valve when DUC changed from 90% to 100%. 548 The spray droplet size spectrum, another key spray characteristics, produced from these 549 PWM-controlled nozzles would could be discussed and reported in future publications. Evaluations of 550 different PWM-controlled hollow-cone nozzles would should be continued to establish a large digital 551 database of spray characteristics, which will be used for the future optimal design and improvement of 552 precision air-assisted sprayers to efficiently and effectively protect crops and safeguard the 553 Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) 32 environment. 554 555 5 Acknowledgements 556 Mention of company or trade names is for description only and does not imply endorsement by 557 the USDA. The USDA is an equal opportunity provider and employer. 558 The authors would like to express their gratitude to Adam Clark, Barry Nudd, and Andy 559 Doklovic for their technical assistance throughout the laboratory tests. We also appreciate the USDA560 NIFA Specialty Crop Initiative (Grant No. 2015-51181-24253) for partially financial support of this 561 research. 562 563 6 References 564 Bennur, P. J., & Taylor, R. K. (2010). Evaluating the response time of a rate controller used with a 565 sensor-based, variable rate application system. Applied Engineering in Agriculture, 26(6), 566 1069-1075. https://doi.org/10.13031/2013.35903 567 Butts, T. R., Samples, C. A., Franca, L. X., Dodds, D. M., Reynolds, D. B., Adams, J. W., Zollinger, 568 R. K., Howatt, K. A., Fritz, B. K., Hoffmann, W. C., & Kruger, G. R. (2018). Spray droplet 569 size and carrier volume effect on dicamba and glufosinate efficacy. Pest Management Science, 570 74(9), 2020-2029. https://doi.org/10.1002/ps.4913 571 Butts, T. R., Butts, L. E., Luck, J. D., Fritz, B. K., Hoffmann, W. C., & Kruger, G. R. (2019). Droplet 572 size and nozzle tip pressure from a pulse-width modulation sprayer. Biosystems engineering, 573 178, 52-69. https://doi.org/10.1016/j.biosystemseng.2018.11.004 574 Chen, Y., Zhu, H., & Ozkan, H. E. (2012). Development of a variable-rate sprayer with laser scanning 575 sensor to synchronize spray outputs to tree structures. Transactions of the ASABE, 55(3), 773576 Formatted: English (United Kingdom) Commented [A3]: Doesn’t that conflict with - Salcedo, R., Zhu, H., Jeon, H., Ozkan, E., Wei, Z., Gil, E., Campos, J., and Román, C. (2022). Droplet size distributions from hollow-cone nozzles coupled with PWM valves. Submitted to Journal of the ASABE for publication. (In review). Surely your database will be a combination of both sets of results. It is possible that combining the two sets of data could provide “added value” but the problem in terms of publiocation would be the reduced novelty of this with all the data involved being already published. I think this is something for you to ponder. The work as it stands is publishable without this discussion on droplet size.. Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) 33 781. https://doi.org/10.13031/2013.41509 577 Esau, T. J., Zaman, Q. U., Chang, Y. K., Schumann, A. W., Percival, D. C., & Farooque, A. A. (2014). 578 Spot-application of fungicide for wild blueberry using an automated prototype variable rate 579 sprayer. Precision agriculture, 15(2), 147-161. https://doi.org/10.1007/s11119-013-9319-4 580 Fabula, J., Sharda, A., Kang, Q., & Flippo, D. (2021). Nozzle flow rateflowrate, pressure drop, and 581 response time of pulse width modulation (PWM) nozzle control systems. Transactions of the 582 ASABE, 64(5): 1519-1532. https://doi.org/10.13031/trans.14360 583 Garcerá, C., Moltó, E., & Chueca, P. (2017). Spray pesticide applications in Mediterranean citrus 584 orchards: Canopy deposition and off-target losses. Science of the Total Environment, 599, 585 1344-1362. https://doi.org/10.1016/j.scitotenv.2017.05.029 586 Gil, E., Llorens, J., Llop, J., Fàbregas, X., Escolà, A., & Rosell-Polo, J. R. (2013). Variable rate 587 sprayer. Part 2-Vineyard prototype: Design, implementation, and validation. Computers and 588 Electronics in Agriculture, 95, 136-150. https://doi.org/10.1016/j.compag.2013.02.010 589 Giles, D. K., & Comino, J. A. (1989). Variable flow control for pressure atomization nozzles. 590 Transactions of SAE, Journal of Commercial Vehicles, 98(2), 237-249. 591 Giles, D. K., & Comino, J. A. (1990). Droplet size and spray pattern characteristics of an electronic 592 flow controller for spray nozzles. Journal of Agricultural Engineering Research, 47, 249-267. 593 https://doi.org/10.1016/0021-8634(90)80045-V 594 Giles, D. K. (1997). Independent control of liquid flow rateflowrate and spray droplet size from 595 hydraulic atomizers. Atomization and Sprays, 7(2). https://doi.org/10.1615/AtomizSpr.v7.i2.40 596 Grella, M., Marucco, P., Manzone, M., Gallart, M., & Balsari, P. (2017). Effect of sprayer settings on 597 spray drift during pesticide application in poplar plantations (Populus spp.). Science of the 598 Total Environment, 578, 427-439. https://doi.org/10.1016/j.scitotenv.2016.10.205 599 Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) 34 Grella, M., Gioelli, F., Marucco, P., Zwertvaegher, I., Mozzanini, E., Mylonas, N., Nuyttens, D. & 600 Balsari, P. (2021). Field assessment of a pulse width modulation (PWM) spray system 601 applying different spray volumes: duty cycle and forward speed effects on vines spray 602 coverage. Precision Agriculture, 1-34. https://doi.org/10.1007/s11119-021-09835-6 603 Han, S., Hendrickson, L. L., Ni, B., & Zhang, Q. (2001). Modification and testing of a commercial 604 sprayer with PWM solenoids for precision spraying. Applied Engineering in Agriculture, 605 17(5), 591. https://doi.org/10.13031/2013.6906 606 Kasner, E.J., Fenske, R.A., Hoheisel, G.A., Galvin, K., Blanco, M.N., Seto, E.Y., & Yost, M.G. 607 (2020). Spray drift from three airblast sprayer technologies in a modern orchard work 608 environment. Annals of Work Exposures and Health, 64, 25-37. 609 https://doi.org/10.1093/annweh/wxz080 610 Khot, L. R., Ehsani, R., Maja, J. M. J., Campoy, J. M., Wellington, C., & Al-Jumaili, A. (2014). 611 Evaluation of deposition and coverage by an air-assisted sprayer and two air-blast sprayers in a 612 citrus orchard. Transactions of the ASABE, 57(4), 1007-1013. 613 https://doi.org/10.13031/trans.57.10419 614 Liu, H., Zhu, H., Shen, Y., Chen, Y., & Ozkan, H. E. (2014). Development of digital flow control 615 system for multi-channel variable-rate sprayers. Transactions of the ASABE, 57(1), 273-281. 616 https://doi.org/10.13031/trans.57.10216 617 Maghsoudi, H., Minaei, S., Ghobadian, B., & Masoudi, H. (2015). Ultrasonic sensing of pistachio 618 canopy for low-volume precision spraying. Computers and Electronics in Agriculture, 112, 619 149-160. https://doi.org/10.1016/j.compag.2014.12.015 620 Mangus, D. L., Sharda, A., Engelhardt, A., Flippo, D., Strasser, R., Luck, J. D., & Griffin, T. (2017). 621 Analyzing the nozzle spray fan pattern of an agricultural sprayer using pulse width modulation 622 Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) 35 technology to generate an on-ground coverage map. Transactions of the ASABE, 60(2), 315623 325. https://doi.org/10.13031/trans.11835 624 Miranda, M. P., Scapin, M. S., Vizoni, M. C., Zanardi, O. Z., Eduardo, W. I., & Volpe, H. X .L. 625 (2021). Spray volumes and frequencies of insecticide applications for suppressing Diaphorina 626 citri populations in orchards. Crop Protection, 140, 105406. 627 https://doi.org/10.1016/j.cropro.2020.105406 628 Needham, D. L., Holtz, A. J., & Giles, D. K. (2012). Actuator system for individual nozzle control of 629 flow rateflowrate and spray droplet size. Transactions of the ASABE, 55(2), 379-386. 630 https://doi.org/10.13031/2013.4137 631 Rathnayake, A. P., Chandel, A. K., Schrader, M. J., Hoheisel, G. A., & Khot, L. R. (2021). Spray 632 patterns and perceptive canopy interaction assessment of commercial airblast sprayers used in 633 Pacific Northwest perennial specialty crop production. Computers and Electronics in 634 Agriculture, 184, 106097. https://doi.org/10.1016/j.cropro.2022.105921 635 Román, C., Llorens, J., Uribeetxebarria, A., Sanz, R., Planas, S., & Arnó, J. (2020). Spatially variable 636 pesticide application in vineyards: Part II, field comparison of uniform and map-based variable 637 dose treatments. Biosystems Engineering, 195, 42-53. 638 https://doi.org/10.1016/j.biosystemseng.2020.04.013 639 Saha, A., Lee, J. D., Basu, S., & Kumar, R. (2012). Breakup and coalescence characteristics of a 640 hollow cone swirling spray. Physics of fluids, 24(12), 124103. 641 https://doi.org/10.1063/1.4773065 642 Salcedo, R., Zhu, H., Zhang, Z., Wei, Z., Chen, L., Ozkan, E., & Falchieri, D. (2020). Foliar 643 deposition and coverage on young apple trees with PWM-controlled spray systems. 644 Computers and Electronics in Agriculture, 178, 105794. 645 Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) 36 https://doi.org/10.1016/j.compag.2020.105794 646 Salcedo, R., Zhu, H., Ozkan, E., Falchieri, D., Zhang, Z., & Wei, Z. (2021). Reducing ground and 647 airborne drift losses in young apple orchards with PWM-controlled spray systems. Computers 648 and Electronics in Agriculture, 189, 106389. https://doi.org/10.1016/j.compag.2021.106389 649 Salcedo, R., Zhu, H., Jeon, H., Ozkan, E., Wei, Z., Gil, E., Campos, J., and Román, C. (2022). Droplet 650 size distributions from hollow-cone nozzles coupled with PWM valves. Submitted to Journal 651 of the ASABE for publication. (In review) 652 Salyani, M., Farooq, M., & Sweeb, R. D. (2007). Spray deposition and mass balance in citrus orchard 653 applications. Transactions of the ASABE, 50(6), 1963-1969. 654 https://doi.org/10.13031/2013.24092 655 Shen, Y., Zhu, H., Liu, H., Chen, Y., & Ozkan, H. E. (2017). Development of a laser-guided, 656 embedded-computer-controlled, air-assisted precision sprayer. Transactions of the ASABE, 657 60(6), 1827-1838. https://doi.org/10.13031/trans.1245 658 Wei, Z., Zhu, H., Zhang, Z., Salcedo, R., & Duan, D. (2021). Droplet size spectrum, activation 659 pressure, and flow rateflowrate discharged from pwm flat-fan nozzles. Transactions of the 660 ASABE, 64(1), 313-325. https://doi.org/10.13031/trans.14100 661 Womac, A. R., Melnichenko, G., Steckel, L., Montgomery, G., & Hayes, R. M. (2016). Spray tip 662 effect on glufosinate canopy deposits in Palmer amaranth (Amaranthus palmeri) for pulse663 width modulation versus air-induction technologies. Transactions of the ASABE, 59(6), 1597664 1608. https://doi.org/10.13031/trans.59.11642 665 Zhu, H., Derksen, R. C., Guler, H., Krause, C. R., and Ozkan, H.E. (2006). Foliar deposition and off666 target loss with different spray techniques in nursery applications. Transactions of the ASABE, 667 49(2): 325-334. 668 Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom) Formatted: English (United Kingdom)