Evaluation of Spray Drift of Plant Protection Drone Nozzles Based on Wind Tunnel Test

Abstract

The use of drones in agriculture is expanding at a brisk pace in crop production due to the superiority in precision, efficiency, and safety of their applicators. However, their potential drift risk also raises concern for users and regulatory authorities. The method of wind tunnel research can effectively evaluate the weighted influence of each drift factor, especially the drift characteristics of the nozzle and spray solution. Based on the wind tunnel test results, centrifugal nozzles have a higher drift risk than hydraulic nozzles, even with a similar DV₅₀. The cumulative drift rate of the centrifugal nozzle at 2 m downwind was 90.1% compared to the LU12001 nozzle’s 40.6% under the wind speed of 3.5 m/s. Compared with the same coding as the flat fan hydraulic nozzle, the IDK nozzle can effectively reduce the drift rate. For the tested nozzles, DV₅₀ and wind speed had a linear relationship with drift rate, and the sampling location had an exponential or logarithmic relationship with drift rate. Spray adjuvants, especially modified vegetable oils, had a significant effect on reducing the amount of drift. The results of this experiment provide a reference for the selection of nozzles and the addition of spray adjuvants. Further clarifying the spray drift characteristics of drones until a drift prediction model is available is still the focus of research.

1. Introduction

Spray drift is defined as a physical movement in which pesticide droplets move from the target area to the non-target area through the air during the application process or soon thereafter [1]. It is expressed as the proportion of the target droplets to the total spray volume. Drifting agricultural chemicals can reach and adversely affect nearby people, livestock, neighboring crops, fields, streams or bodies of water, and surrounding areas and cause serious consequences, such as environmental contamination, health risks to people and animals, damage to adjoining crops, lower control efficacy, and pesticide and money waste [2,3,4].

Spray drift is mainly affected by four categories of factors: environmental parameters, application parameters, characteristics of the surrounding application environment, and the properties of the spray solution [5]. The environmental parameters include temperature and humidity, wind speed, etc.; the application parameters include equipment parameters, spraying height, spraying speed, etc.; the characteristics of the surrounding application environment mainly include crops and shelter; the properties of the solution mainly include solution viscosity, surface tension, density, etc. The methods of measuring drift mainly include field tests, wind tunnel tests, and computer simulations.

The wind tunnel is a kind of pipeline experimental equipment that artificially generates and controls airflow and is used to simulate the flow of air around aircraft or other entities, measure the effect of airflow on entities, and observe physical phenomena [6]. The air is created by a powerful fan system that blows over the object under test. The wind tunnel laboratory simulates the actual field conditions, and its advantage lies in that it can easily control the wind speed and wind direction and accurately measure variables such as spray pressure, the moving speed of the nozzle, and the vertical distance between the nozzle and the spray target. The use of this experimental condition can effectively avoid the difficulty in evaluating various factors caused by temporal changes in the field. However, the wind tunnel test also has its defects, such as the inability to simulate the real drift distance of the droplets after spraying by plant protection equipment like the field test.

In the field of agriculture, among the various types of wind tunnels, low-speed wind tunnels are the most developed, including direct current closed, air conditioning backflow closed, etc. [6]. The test section has two configurations: open and closed. The main principle of a backflow-closed wind tunnel is to drive the airflow to flow continuously in the wind tunnel loop through the fan system. The direct-current closed wind tunnel is driven by the fan system to continuously draw air into the wind tunnel from the outside atmosphere through the air inlet and then discharge it to the outside atmosphere through the exhaust port. According to the requirements of ISO standard [7], the wind tunnel should be large enough to generate and maintain the rated airflow velocity for measurement in a uniform airflow environment, and the local change rate of the airflow velocity should not be greater than 5%. The fluidity should not be greater than 8%, and the air velocity used for the graded measurement of relative spray drift is usually 2 m/s. It is still one of the important methods for carrying out research on the measurement of droplet drift in the wind tunnel. Ferguson et al. [8] used a wind tunnel to quantify the droplet size spectrum and spray drift potential of different nozzle types in order to select technologies that reduce spray drift. Hoffmann et al. [9] established a WTDISP (wind tunnel dispersion) model by measuring the size and flow of a series of nozzles in a low-speed wind tunnel and obtained the droplet size distribution rules of different nozzles under wind tunnel conditions. Fritz et al. [10] compared the modeled downwind deposition to that measured in the wind tunnel to evaluate the accuracy of the wind tunnel dispersion (WTDISP) model using a low-speed wind tunnel in USDA-ARS. Torrent et al. [2] used a phase Doppler particle analyzer (PDPA) and two different wind tunnel methods to assess the potential reduction of spray drift for hollow-cone nozzles.

At present, the evaluation of nozzles installed on fixed-wing aircraft or the effect of liquid chemicals using wind tunnels has been relatively complete; a large amount of data has been accumulated, and a certain model has been established [11,12]. However, the evaluation of plant protection drone spraying using wind tunnels is still in its infancy. Recently, plant protection drones have developed rapidly, and the number and application area of drones during 2021 had exceeded more than 150,000 and 1.5 billion mu in China [13]. The nozzles installed on drones by world-renowned drone companies, such as DJI (Shenzhen Da-Jiang Innovations Science and Technology Co., Ltd.) and XAG (XAG Co., Ltd.), have all adopted the method of centrifugal atomization. They fitted below the rotors, which produce 85 μm to 550 μm sized spray droplets. This type of nozzle is more suitable for low-volume and variable-rate spraying. Especially when spraying some special pesticide formulations, such as citrus whitening agents, these nozzles are not easy to block. Although some companies also choose to install hydraulic nozzles on their drones, the quantity sold by these companies is much lower. According to our survey, currently, in the Chinese market, about 30% of crop spraying drones use hydraulic nozzles, while 70% use centrifugal nozzles.

At present, most of the research on plant protection drones is focused on the control and efficacy of low-volume spraying on pests and diseases [13,14,15,16]. Data collection on droplet drift is very rare. Compared with ground-based spraying equipment, plant protection drones spray at a higher distance from the crop canopy and fly faster, which is bound to cause more serious droplet drift and environmental safety problems [5]. Therefore, how to clarify and alleviate the risk of pesticide drift during the spraying process of drones has become a technical bottleneck restricting the healthy and orderly development of the plant protection drone industry.

How to reduce the drift of pesticide droplets, improve the efficiency of pesticide use, and reduce the impact of pesticides on non-target organisms and the pressure on the environment (Effectiveness-Efficiency-Environment, 3E) has always been an important issue in plant protection application technology. In order to further clarify the drift characteristics of plant protection drones, this study explores the drift characteristics of different types of nozzles installed on plant protection drones and explores anti-drift technology for plant protection drones. This experiment was conducted in the wind tunnel of the China National Center for International Collaboration Research on Precision Agricultural Aviation Pesticide Spraying Technology.

2. Materials and Methods

2.1. Wind Tunnel

The low- and high-speed wind tunnels built in the National Center for International Collaboration Research on Precision Agricultural Aviation Pesticide Spraying Technology (NPAAC) at South China Agricultural University, Guangzhou, China, can generate airspeeds of 2–51 m/s (Figure 1a). The wind tunnel includes 8 parts, including a mounting bracket, observation window, test section, droplet size test system, spray system, contractive segment, steady section, and drive section. The test section size is 2.0 m × 1.1 m × 20.0 m (width × height × length) (Figure 1b). Due to the limitations of the closed environment and size of the wind tunnel, we did not simulate rotor wash in our experiment, which is different from field conditions and should be taken into consideration.

To test the droplet size, a droplet size testing system (Figure 1f) and a nozzle controlling system for the hydraulic nozzle (Figure 1e) and centrifugal nozzle (Figure 1d) were designed in the wind tunnel lab. The DP-02 laser diffraction instrument (OMC Instrument Co., Ltd., Zhuhai, China) was used in the test section to measure the droplet size. This laser-lens combination was able to detect droplets in the range of 1 to 1500 μm. The laser particle size analyzer consists of a collimated laser generator, a signal acquisition device, and a data processing system. The distance between the transmitter and receiver is 2.0 m. The specific measurement method for droplet particle size was detailed by Wang et al. [17].

2.2. Test Nozzles

Two types of nozzles including hydraulic nozzle (Lechler Nozzle Systems Co., Ltd, Changzhou, China) and centrifugal nozzle (XAG Technology Co., Ltd, Guangzhou, China) were selected for this drift test, which are commonly used nozzles installed on plant protection drones. Hydraulic nozzles include the Lechler LU120 series (LU12001, LU12002, LU12003) and the air-injector anti-drift hydraulic nozzles: IDK120 series (IDK12001, IDK12002, IDK12003). The spray pressure of hydraulic nozzles is shown in

Table 1. Droplet size for the centrifugal nozzle and hydraulic nozzle.

Treatment	Nozzle Type	Nozzle Coding	Rotation Speed (Input Voltage)/Spray Pressure	Flow/L·min−1	DV50/μm *	Droplet Size Category
1	Centrifugal Nozzle	-	4000 rpm (10.5 V)	0.5 (30.1 V)	277.3 ± 9.4	M
2		-	6000 rpm (16.3)		194.9 ± 10.2	F
3		-	8000 rpm (21.1)		153.5 ± 6.6	F
4		-	12,000 rpm (31.1)		111.6 ± 0.9	F
5	Universal flat fan hydraulic nozzle	LU12001	0.3 MPa	0.39	114.4 ± 0.4	F
6		LU12002	0.3 MPa	0.80	130.2 ± 0.4	F
7		LU12003	0.2 MPa	1.19	150.6 ± 0.2	F
8	Anti-drift hydraulic nozzle	IDK12001	0.3 MPa	0.39	266.3 ± 0.7	M
9		IDK12002	0.3 MPa	0.80	346.3 ± 1.3	C
10		IDK12003	0.2 MPa	1.19	385.2 ± 0.6	C

* data in the table are means ± standard deviation.

. The setting of the rotational speed of the centrifugal nozzle was based on the measurement results of the hydraulic nozzle. The rotation speed of the centrifugal nozzle is to ensure that the DV₅₀ of the two types of nozzles is similar, so as to facilitate the comparison of subsequent drift results.

In the drift test, nozzles were installed at a height of 0.8 m above the wind tunnel floor, with the spray fan oriented perpendicular to the airflow direction. Each nozzle was tested at three wind speeds: 1.5 m/s, 2.5 m/s, and 3.5 m/s. A total of 10 treatments under 3 wind speeds were designed in the experiment; each test condition was repeated 3 times.

2.3. Drift Test in the Wind Tunnel

2.3.1. Layout of Sampling Lines

The ground and airborne deposition in the wind tunnel were collected by samplers of the monofilament line (φ = 2.0 mm). The arrangement of sampling lines in the wind tunnel is shown in Figure 2. The ground deposition was measured at 3~12 m downwind with 1 m intervals using a monofilament line placed on holders held 10 cm above the floor. The airborne deposition was measured 2 m downwind from the spray nozzle at different heights. Due to the different atomization methods of the two kinds of nozzles, the density of the arrangement of monofilament lines varied. Because the initial velocity of the droplets created by the centrifugal nozzle was horizontal and the sampling location at 2 m was close to the edge of the spray swath, the monofilament lines for airborne deposit collection were suspended from 0.1 m above the wind tunnel floor and 0.05 m apart up to a height of 1 m. The purpose of this arrangement was to avoid the spray plume passing mainly through a certain sampling interval. For the hydraulic nozzle, the sampling line was also suspended from 0.1 m above the wind tunnel floor and 0.1 m apart, up to a height of 1 m. The highest sampling line was 0.2 m above the nozzle.

2.3.2. Spray Control

During the test, the spray pressure was monitored through the pressure display dial, and the flow rate was monitored by the flow sensor. Each test parameter was changed, and the flow rate was recorded in real-time and compared with the calibration value provided by the manufacturer to ensure the normal working state of the nozzle. Spray time was controlled by a solenoid valve with a timing function. In order to avoid over-saturation of spray deposition, the spray was emitted for 10 s for all replications.

2.3.3. Sample Processing

For quantitative characterization of deposition, rhodamine-B, a water-soluble fluorescent tracer (Surround WP, Engelhard Corp., Iselin, NJ, USA), was used in each treatment at a concentration of 1 g/L. After each spray replication, the monofilament lines at each position were collected and put into labeled zip-top bags, which were then transported to the lab for analysis. For full wash-off, 100 mL of 10% ethanol was pipetted into each bag to wash off the tracer on the samplers, and then the bags were shaken for almost 30 s.

After sufficient agitation, part of the effluent was poured into a cuvette. The cuvette containing the effluent was measured using a spectrofluorophotometer with an excitation wavelength of 550 nm and an emission wavelength of 575 nm. The spray deposition on monofilament lines was calculated and converted to μg/mL from a calibration curve determined through sampling known concentrations of the tracer. The relationship between the fluorimeter reading and the tracer concentration (mg/L) was determined by the calibration factor F_cal. The calibration factor F_cal is determined based on the relationship curve between the fluorescence value and the known concentration value under the same test conditions as rhodamine-B.

2.3.4. Data Analysis

All the zip-lock bags with monofilament lines were stored in closed ice chests immediately to avoid decomposition. Each sample bag was labeled with unique identifiers that included the treatment and sampling location.

The cumulative drift rate of the vertical sampling lines at 2 m downwind can be calculated by the following formula:

$$Q_v(\%)=\left[\left[{\displaystyle \sum }_{i=1}^n{A}_i\right]\cdot \frac{s}{d}\right]/N\times 100$$

(1)

where Q_v: cumulative drift rate (%); A_i: The amount of deposition measured on one sampling line (μg); n: Number of sampling lines; s: Sampling interval (cm); d: Sampling line diameter (cm); N: Tracer spray amount (μg).

The drift rate (%) of the ground deposition at different distances from the nozzle can be calculated by the following formula:

$$Q_h(\%)=\left[A·s/d\right]/N\times 100$$

(2)

where Q_h: Drift rate (%); A: The amount of deposition measured on one sampling line (μg); s: Sampling interval (cm); d: Sampling line diameter (cm); N: Tracer spray amount (μg).

2.4. Spray Adjuvant

To evaluate the effect of spray adjuvants on reducing drift, four kinds of commonly used spray adjuvants were selected for the test, including Maifei (MF, BeijingGrand AgroChem Co., Ltd., Beijing, China), Beidatong and Mingshun-1 (BDT and MS, Hebei Mingshun Agricultural Technology Co., Ltd., Shijiazhuang, China), and Starguar4A (Solvay S.A. Co., Ltd., Shanghai, China).

Two typical nozzles were selected for the test: the hydraulic nozzle LU12001 under 0.3 MPa and the centrifugal turntable nozzle at a rotational speed of 12,000 rpm were tested to determine the effect of spray adjuvants on droplet drift. During the test, the wind speed in the wind tunnel was 3.5 m/s. The additional amount of spray adjuvant was 1%, and water was used as a control.

3. Results

3.1. Droplet Size

Droplet size data (means ± standard deviation) for the centrifugal nozzle and the hydraulic nozzle are shown in Table 1. For the hydraulic nozzle, the spray pressure was chosen at 0.3 MPa, which is commonly used in the field. Due to the fact that the droplet size of a hydraulic nozzle cannot vary over a wide range like a centrifugal nozzle, in order to achieve a significant difference in droplet size between 03 nozzles and 02 nozzles, as well as for a favorable comparison with the centrifugal nozzle at 8000 rpm, we have reduced the droplet size sprayed by the 03 nozzle with set pressure to 0.2 MPa. The rotational speed of the centrifugal nozzle was chosen to keep the DV₅₀ similar to that of the hydraulic nozzle. Through the test, when the rotational speed of the centrifugal nozzle is 4000 rpm, 8000 rpm, and 12,000 rpm, the droplet size distribution is 277.3 μm, 153.5 μm, and 111.6 μm, which is similar to that of the hydraulic nozzles of LU12001 with a spray pressure of 0.3 MPa, LU12003 with a spray pressure of 0.2 MPa, and IDK12001 with a spray pressure of 0.3 MPa. Based on the test results, the droplet size category of the nozzles was determined according to ASABE S641 [18]. The droplet size category of the centrifugal nozzle was changed from medium (M) to fine (F), the flat fan hydraulic nozzle was fine (F), and the anti-drift hydraulic nozzle was changed from coarse (C) to medium (M).

3.2. Drift Characteristics of Different Nozzles

3.2.1. Cumulative Drift Rate at 2 m Downwind

Wind speeds, nozzle types, and nozzle coding have a significant effect on the cumulative drift rate (Figure 3). From the test result, the cumulative drift rate of the IDK series nozzle was 2.4–8.4 times lower than that of the LU series nozzle at 2 m downwind under the three different wind speeds (1.5 m/s, 2.5 m/s, and 3.5 m/s). Compared with the hydraulic nozzle, the centrifugal nozzle has a higher cumulative drift rate at the same wind speed. For example, the DV₅₀ of the centrifugal atomizing nozzle under 12,000 rpm is similar to the LU12001 nozzle under a spray pressure of 0.3 MPa, but the drift rate of the centrifugal nozzle was 90.1% compared to the LU12001 nozzle’s 40.6% under the wind speed of 3.5 m/s. This may be due to the fact that the droplets produced by centrifugal nozzles have horizontal velocity but no vertical downward velocity, which is more likely to cause droplet drift compared with hydraulic nozzles.

Figure 4 was obtained by fitting the relationship between the DV₅₀ and the cumulative drift rate at 2 m downwind. Due to the big difference in structure and drift characteristics between centrifugal and hydraulic nozzles, they are fitted separately. Through fitting analysis, there is a linear relationship between the DV₅₀ and the cumulative drift rate at 2 m downwind, and the R² > 0.71.

3.2.2. Vertical Airborne Drift at 2 m Downwind

Vertical airborne drift for conventional hydraulic nozzles and centrifugal nozzles under different wind speeds is shown in Figure 5. The airborne drift basically increased as a result of the increased wind speed and decreased DV₅₀. The airborne drift on the monofilament lines for hydraulic nozzles generally increased as the sampling height decreased.

The drift risk of centrifugal nozzles in the airborne was higher. When the wind speed was 3.5 m/s, the maximum drift rate was at 0.55 m height, with the rotation speed of the centrifugal nozzles at 12,000 rpm. That means that the droplets were mainly passing through a height of 0.55 m, which will lead to a further drift distance. Compared with LU nozzles, IDK nozzles have an obvious anti-drift function, and the drift rate at each height was significantly lower than that of LU nozzles.

3.2.3. Ground Deposition

Ground deposition decreases with the increase in the downwind sampling position (Figure 6). The relationship between the sampling position and the drift rate was fitted, and it was found that the two basically conform to the exponential function relationship with R² > 0.84.

Q = a + b × c^D_,

(3)

where, Q is the drift rate (%), a, b, and c are constants, and D is the drift position (m).

The drift rates of centrifugal nozzles and hydraulic nozzles with similar DV₅₀ are compared in Figure 7. Similar to the results of vertical airborne drift, the drift of centrifugal nozzles is significantly higher than that of hydraulic nozzles at each sampling position.

3.2.4. Wind Tunnel Drift Model

Nozzle type (mainly droplet size), wind speed, and sampling position all have a significant impact on the droplet drift rate. By fitting the above three factors, it is found that when the following equations are used to fit the DV₅₀, wind speed, sampling position, and drift rate, it has the best coefficient of determination:

Q = 2.28 − 1.20ln(D_l) – 0.0057D_s + 0.89W_s,

(4)

where D_l is the sampling position, m; D_s is DV₅₀, μm; W_s is wind speed, m/s; Q is drift rate, %. R² is equal to 0.61. The nozzle type (DV₅₀) and wind speed had a linear relationship with the drift rate, and the sampling location has an exponential or logarithmic relationship with the drift rate.

3.3. Effects of Spray Adjuvants on Drift

Compared with water, the spray adjuvant had a significant effect on reducing the amount of drift at each sampling location. For hydraulic nozzles, adding a spray adjuvant can reduce the drift by 24.1–66.4%. Among them, BDT^®, which is mainly composed of modified vegetable oil, has the best effect. When the wind speed is 3.5 m/s, the addition of BDT^® reduces the drift rate by 66.4% compared with water (Figure 8).

For centrifugal nozzles, adding a spray adjuvant can reduce drift by 0.68–50.8%. Similar to the results of the hydraulic nozzles, BDT^® also has the best effect, with a 50.8% improvement in the anti-drift performance (Figure 8).

4. Discussion

In just a few years, there’s been a huge rise in the use of drones in agriculture. The popularity of drones in agriculture is set to soar globally as countries grant operators permission to also apply crop protection products. These countries mainly include east Asian and southeast Asian countries such as China, Japan, South Korea, Thailand, and India. Drones used in agriculture can operate at very low altitudes over sodden fields and tall crops where the ground machines could normally move, fly quickly to exact locations to treat target areas precisely, as well as be pre-programmed to navigate their own way around. The rapid development of plant protection drones has solved the problem of a lack of tools for spraying in east Asian and southeast Asian countries, but it is also a double-edged sword. Indeed, like China and India, they have essentially enabled farmers to leap from hand-held applicators, skipping vehicle-mounted boom machines, and going straight to drones. At the same time, drones improve application timeliness, reduce the need for skilled labor, and cut hand-held sprayer operators’ exposure to harmful pesticides [19]. On the other side of the double-edged sword, their drift risk cannot be ignored. The lack of relevant drift data and rotor wake caused by blades increases the potential complexity of droplet drift prediction, which is why many countries still prohibit drone spraying. As in the whole of the EU, due to rules not keeping up with technology, all aerial applications are prohibited.

In order to clarify the drift characteristics of plant protection drones, it is essential to carry out the measurement of droplet size. Based on our research results, the droplet size range of the centrifugal nozzle is narrow, which will lead to greater drift risk in the case of small droplets spraying. Even worse, the principle of the centrifugal nozzle is that the droplets are loaded on a high-speed optional turntable [20], which leads to the droplets sprayed by this type of nozzle having a fast horizontal speed and no vertical downward speed, which further increases the risk of drift.

The wind tunnel is a common equipment tool for nozzle evaluation, spray adjuvant optimization, and drift research. Guler et al. [21] evaluated the spray characteristics and drift reduction potential of air induction nozzles with air-intake holes that were sealed or open; Huang et al. thought the wind tunnel test proved that spraying height, along with wind direction and relative humidity, all had a significant effect on spray drift. Nuyttens et al. [22] compared the drift potential of standard flat-fan nozzles and air-inclusion nozzles and studied different wind tunnel evaluation methods. Wang et al. [23] conducted a study on the drift characteristics of a crop protection unmanned aerial vehicle unit under wind tunnel conditions, and evaluated the effects of nozzle type, meteorological parameters, and spray adjuvants on the drift. From his study, an anti-drift nozzle with a lower flying speed and appropriate spray adjuvants can significantly reduce drift. Liu et al. [24] and Grant et al. [25] also collected the drift of crop protection drones in the wind tunnel or at the wind tunnel entrance and obtained similar conclusions, that is, the effect of wind speed is greater than the effect of fog droplet size or payload. However, considering that the wind tunnel is a closed environment and differs significantly from the field, the effect of adding a downwash on the drift in a closed environment may be quite different from that in the field. Therefore, we did not include factors such as downwash in our experiments.

Since centrifugal nozzles are rarely used in ground equipment, there are relatively few studies on the drift characteristics of centrifugal nozzles in wind tunnels. Bayat et al. [26] studied the effect of air-assist velocities, application volumes, and two droplet sizes by changing disc speed on drift potential in a wind tunnel. There are even fewer studies on the drift characteristics of centrifugal nozzles installed on plant protection drones. Our research complements the current lack of data in this part. Both the data from the cumulative drift rate at 2 m downwind and ground deposition proved that the drift potential of the centrifugal nozzle is several times greater than that of the hydraulic type of the similar DV₅₀, which needs special attention. Fortunately, adding an adjuvant with anti-drift properties can still reduce drift for centrifugal nozzles.

In terms of drift assessment, field experiments are also an important means in addition to wind tunnel tests. Wang et al. [27] studied the drift test using gasoline-powered single-rotor unmanned helicopters in pineapple fields under different meteorological conditions. The experimental results showed that the position of cumulative spray drift that accounted for 90% of the total spray drift was from 3.70 m to 46.5 m. Zhang et al. [28] used computational fluid dynamics (CFD) methods to simulate and analyze the distribution of the downwash of an unmanned agricultural helicopter (UAH) N-3 and found that crosswind has a greater effect on droplet drift than flight height. Wang et al. [17] tested the drift of a commercial quadcopter equipped with centrifugal nozzles and found that the deposition at 12 m downwind decreased by an order of magnitude compared with the in-swath zone. However, when the wind speed is high, there are still droplets at the position of 50 m downwind. Teske et al. [29] summarized the ability of the Comprehensive Hierarchical Aeromechanics Rotorcraft Model, CHARM, and AGricultural DISPersal, AGDISP to predict the deposition and drift of sprays released from drones. Until now, field and computer simulation methods have been used to assess the drift of plant protection drones, but relatively few wind tunnel studies have been conducted and employed. Despite the numerous field trials and studies conducted, there are still some issues that exist. These tests have significant differences in aircraft types, nozzle types, and testing environments, making it difficult to compare them with each other. Additionally, field experiments mostly use commercial drones, making it difficult to compare the situation of different types of nozzles on the same type of drone. This issue also limits the reference value of field experiment results for our wind tunnel tests.

The downwash airflow is unique to agricultural drones, especially multi-rotor drones. The combination of the downwash airflow and natural wind has a significant impact on the deposition and drift of droplets. Guo et al. [30] used CFD simulation and experimental verification of the spatial and temporal distributions of the downwash airflow of a quad-rotor UAV in hover. Through his test, he found that locating the nozzle position at 0.5–1.2 m below the rotor can better utilize the rotor’s airflow and increase the speed of droplet sedimentation.

Wen et al. [31] performed numerical simulation and validation analysis on spray distributions disturbed by quadrotor drone wake at different flight speeds. The accuracy of numerical simulations has been verified through wind tunnel tests simultaneously. Ruiz et al. [32] have also conducted similar studies, simulating the rotor wind field of drones and conducting wind tunnel tests to verify them. The purpose of these types of experiments is mostly to verify the accuracy of the model simulation. Chen et al. [33] measured the wind field distribution by using a wireless wind speed sensor network measurement system and found the wind field in the Z direction had an extremely significant effect on droplet deposition in the effective spray area.

Past data in the literature has shown that the nozzle type, spray liquid, and spray pressure all determine spray performance. Spray liquid, including adding an adjuvant, further affects the spray performance and anti-drift effect by changing the properties of the solution, including viscosity, surface tension, and inhomogeneities [34,35]. Among the multi-type adjuvants in this study, modified vegetable oil has the best effect, which is also the most commonly used type of anti-drift adjuvant in drone field spraying.

5. Conclusions

The use of drone applications has many potential benefits, such as feasible applications in difficult-to-access scenarios (e.g., rice fields, hilly fields, and sloped vineyards), reduction of operator exposure, and precise spot spraying. However, these benefits can be confirmed and realized only with an adequately evaluated drift risk assessment and risk management perspective. Based on our experimental results, centrifugal nozzles with a similar DV₅₀ have a higher drift risk than hydraulic nozzles. The cumulative drift rate of the centrifugal nozzle at 2 m downwind was 90.1% compared to the LU12001 nozzle’s 40.6% under the wind speed of 3.5 m/s. Compared with the same coding as the flat fan hydraulic nozzle, the air-induction IDK nozzle can effectively reduce the drift rate, which includes airborne drift and ground deposition. For the tested nozzles, DV₅₀, the wind speed had a linear relationship with drift rate, and the sampling location had an exponential or logarithmic relationship with drift rate. The spray adjuvant had a significant effect on reducing the amount of drift. By comparing various adjuvants, modified vegetable oils have the best anti-drift effect. How to obtain more drift assessment data and reduce the drift will be the key issue of continuous concern for plant protection drone applications in the future.