Analysis of the Factors Affecting the Deposition Coverage of Air-Assisted Electrostatic Spray on Tomato Leaves

Abstract

In order to investigate the effects of various factors (charging voltage, spray distance and spray pattern) on the deposition coverage of tomato leaves, the Box–Behnken surface response methodology was used to design an outdoor air-assisted electrostatic spraying experiment with three factors and three levels. The deposition coverage of tomato leaves in the upper, middle and lower layers was collected under different polarity charging voltages (0, +10 kV, −10 kV), spray distances (1, 3, 5 m) and spray patterns (ascending spray, descending spray, fixed height spray). Regression analysis and variance analysis were performed on the experimental data to determine the optimal working parameters. The results showed that (1) spray distance is the most important factor affecting the droplet coverage rate in the process of air-assisted electrostatic spraying; (2) the droplet coverage rate of air-assisted electrostatic spraying is optimal when the charging voltage polarity is negative voltage, the spray distance is 2.75 m, and the spray pattern is descending spray. The following conclusions were obtained. (1) In air-assisted electrostatic spraying, the distribution of air flow had the greatest effect on droplet deposition on tomato leaf surface. (2) Compared with air-assisted non-electrostatic spray, air-assisted electrostatic spray had a better deposition effect.

1. Introduction

A greenhouse environment has the characteristics of high temperature and humidity, lack of natural enemies of pests, high planting density of crops and so on. It is easily affected by pests and diseases in the actual production process [1,2]. Due to the intensive planting of greenhouse crops, there are many problems in the greenhouse application environment, such as narrow crop rows and dense canopy. This makes it difficult for traditional manual sprayers and backpack sprayers to effectively deposit pesticide solution on the surface of the middle and lower leaves of crops during application, and it also increases the health risks for its operators [3,4,5]. In order to cope with the complex working environment, more and more efficient spraying technologies have been put into use, including air-assisted spraying technology, electrostatic spraying technology, air-assisted electrostatic spraying technology, etc. [6,7,8].

Air-assisted spraying technology refers to the plant protection machinery in the process of spraying crops through the fan generated by the airflow field; the droplets are transported to the target position and deposited on the target crop surface. In the airflow field, crop leaves will also be disturbed by the airflow field, which improves the droplet deposition amount, penetration and deposition uniformity on the back of leaves and inside the canopy [6,9,10,11]. In air-assisted spraying, the main factors affecting droplet deposition on crop leaf surface are air velocity, spray distance, moving speed of spraying platform, etc. [12,13].

Electrostatic spraying is a spraying technique that charges droplets. When the charged droplet cloud composed of many charged droplets approaches the plant leaves, an induced electric field will be formed around the target leaves, so that the charged droplets will be affected by the electric field force near the target leaves, and the status quo of “electrostatic deflection and circling” will appear, thus reducing the droplet drift and improving the deposition of charged droplets on the target leaves [14,15]. In electrostatic spraying, the main factors affecting droplet deposition on crop leaf surface are charging voltage, spraying pressure, spraying height, droplet size, etc. [16,17,18].

Air-assisted electrostatic spraying technology combines the advantages of air-assisted spraying technology and electrostatic spraying technology, which can effectively improve the droplet deposition coverage of spray distance. In addition, air-assisted electrostatic spraying technology also inherits many factors that can affect the deposition effect of air-assisted spray and electrostatic spray. These influencing factors mainly include charged voltage, spray distance, spray pattern, air velocity and spray pressure [19,20,21].

At present, there are few studies on the influence of air-assisted electrostatic spraying platform dynamics on droplet deposition coverage, and the related studies are not comprehensive. Dai et al. examined the effects of spray pressure, air pressure and applied voltage on droplet size and charge-to-mass ratio [19]. Neto et al. compared the effects of different spray volumes and charges on spray deposition [20]. Zampiroli et al. evaluated the control effect of different spray rates and different nozzles [21]. These studies mainly compared the different spray pressures, different air pressures, different types of nozzles, charge sizes and other factors in the spray deposition effect. There are few studies on the polarity of charging voltage and the spray pattern of the sprayer.

In this paper, the effects of different polarity charging voltages (0, +10 kV, −10 kV), spray distances (1, 3, 5 m) and spray patterns (ascending spray, descending spray, fixed height spray) on droplet deposition coverage on the upper, middle and lower tomato leaves were investigated by using the greenhouse single-rail air-assisted sprayer [22] developed by Jiangsu University. Three regression equations were obtained between the polarity charge voltage, spray distance, spray pattern and droplet deposition coverage on tomato leaves in each layer. The purpose of this study was to explore the charging voltage polarity and spray pattern suitable for this air-assisted electrostatic sprayer. And it provided a reference for greenhouse air-assisted electrostatic sprayer application.

2. Materials and Methods

2.1. Equipment

The greenhouse single-rail air-assisted electrostatic sprayer used in this experiment was developed by Jiangsu University. Its main structure consists of a control box, cistern, lifting structure, spraying platform, walking structure and track. There is a rotary atomizer in the middle of the spray platform for spraying the area directly below the machine location. However, since the test does not involve walking spray and does not measure the spray effect in the area directly below the spray platform, the function and effect of the walking mechanism and rotary atomizer are not described too much. The structure of the sprayer is shown in Figure 1.

The lifting structure of the sprayer is mainly composed of a single-phase AC motor, winch wheel, worm reducer and guide rod. Two winch wheels are, respectively, fixed at two ends of the worm reducer, and the lifting movement of the spraying device is realized by winding and retracting the steel rope. The lifting motor configuration has a brake function and can realize the lifting function of power failure that is stopped. In addition, the worm reducer also has a self-locking function to ensure that the rising speed is consistent with the falling speed.

The spray platform consists of two centrifugal fans (YWL 2E-150, Hangda Electric Co., China) and four electrostatic flat-fan nozzles (FVP 110-02, HYPRO Manufacturer Co., USA). The two fans and four electrostatic flat-fan nozzles are symmetrically fixed on two sides of the bottom plate of the spraying device, respectively, so that the spraying device can simultaneously apply pesticide to crops in two directions. The electrostatic nozzle is fixed above the air outlet of the fan, so that the droplets have sufficient charging time. The spray angle of the electrostatic nozzle is 110°, and inductive charging is adopted. Its main operating parameters are shown in

Table 1. Working parameters of the sprayer.

Working Parameters	Number
Spray pressure	0.3 MPa
Outlet diameter	0.1 m
Outlet air velocity	20 m/s
Charged voltage	0, ±10 kV
Spray platform lifting speed	0.3 m/s

.

2.2. Air Velocity Measurement Test

In order to understand the variation in air velocity during spraying, the air velocity along the axis of the centrifugal fan outlet was measured. The test was conducted in a corridor at Jiangsu University with a length of 10 m, a width of 2 m and a height of 3 m. The test was measured by a handheld anemometer, and the wind speed change within the range of 0–7 m from the air outlet in the axis direction of the air outlet was measured. The measurement point interval was 0.5 m (Figure 2a). The test results are shown in Figure 2b.

According to the curve of air velocity change in Figure 2b, the attenuation of air velocity is faster in the range of 0–3 m, and the attenuation of air velocity is gentle in the range of 3–7 m.

2.3. Deposition Test Arrangement

The application scenario of the greenhouse single-rail air-assisted electrostatic sprayer is a solar greenhouse in Shouguang City, Shandong Province. The main crops planted in the solar greenhouse are tomato, eggplant, cucumber and other crops, which are planted in the close planting mode. The ridge distance between two rows of crops is 1.2 m; the canopy distance of mature crops is approximately 0.6 m; and the length of single crops is approximately 12 m. The greenhouse single-rail air-assisted electrostatic sprayer is hung on the greenhouse beam through the rail and suspended above the crop canopy. During spraying, the sprayer moves to the middle of two rows of crops and sprays with the aid of the lifting structure.

The experiment was carried out in Jiangsu University agricultural machinery compound using the greenhouse single-rail air-assisted electrostatic sprayer. The tomato crops used in the experiment were potted tomatoes planted in a self-built small greenhouse. In the experiment, the placement method was intended to simulate the planting method in the Shouguang greenhouse, as shown in Figure 3.

In the greenhouse spraying operation in Shouguang, the crops in the working scene of the greenhouse single-rail air-assisted electrostatic sprayer are located at both ends of the machine horizontal direction (Y axis direction) and have certain symmetry; therefore, the tomato leaves at one end are selected to arrange the sampling points. In order to reflect the law of droplet deposition on the surface of each layer of tomato leaves, the sampling points are divided into upper, middle and lower layers from the height direction (Z axis direction); each layer is 0.5 m apart, and the upper tomato leaves are approximately 1.5 m above the ground (Z axis direction). And from the horizontal distance direction (Y axis direction), they are divided into near-end, mid-end and far-end. The distance is 2 m, and the distance between the near-end and the air outlet in the horizontal direction is 1 m. The sampling point layout is shown in Figure 4.

During the test, the outdoor temperature was 27 °C, the humidity was 44%, and there was no wind.

2.4. Experimental Factors

There are many factors affecting the droplet deposition coverage of air-assisted electrostatic spray on tomato leaf surface, including but not limited to the charging voltage polarity, spray distance, spray pattern, spray height, spray angle size and target characteristics. The three factors selected in this experiment are the charging voltage polarity, spray distance and spray pattern, and the other conditions remain the same. In order to facilitate the subsequent data processing and calculation, X₁ represents the charging voltage polarity; X₂ represents the spray distance; X₃ represents the spray pattern; Y₁ represents the droplet deposition coverage on the upper tomato leaf surface; Y₂ represents the droplet deposition coverage on the middle tomato leaf surface; and Y₃ represents the droplet deposition coverage on the lower tomato leaf surface. X₁, X₂ and X₃ are the independent variables; Y₁, Y₂ and Y₃ are the dependent variables; and droplet deposition coverage is used as an evaluation index. The factor levels tested are shown in

Table 2. Test factor level table.

Factors	Levels
	−1	0	1
X1 (kV)	−10	0	+10
X2 (m)	1	3	5
X3 (m/s)	+0.3	0	−0.3

.

Among them, the spray patterns are divided into ascending spray, fixed height spray and descending spray, and the ascending velocity direction of the ascending spray is taken as the positive direction; therefore, the ascending spray velocity is +0.3 m/s, the fixed height spray velocity is 0, and the descending spray velocity is −0.3 m/s (Figure 5).

Ascending spraying: spray for 3 s from 0.5 m above the ground (Z axis direction), then rise at +0.3 m/s and spray at the same time. When the spraying platform rises to 2 m above the ground, it stops rising and continues spraying for 2 s before stopping the application operation. The time required for rising is 5 s.

Descending spraying: spray for 3 s from 2 m above the ground (Z axis direction), then descend at −0.3 m/s and spray at the same time. When the spraying platform descends to a height of 0.5 m above the ground, it stops descending and continues spraying for 2 s before stopping the application operation. The descending time is 5 s.

Fixed height spray: spray continuously for 10 s at a position 2 m away from the ground (Z axis direction).

In order to determine the optimum spray parameter combination of air-assisted electrostatic spray, response surface methodology was used to design the experiments. The Box–Behnken orthogonal test method was used to select the test groups. And the Develve software (Version 4.16) was used to generate the test schemes. A total of 17 groups of tests were obtained, as shown in

Table 3. Experimental design scheme.

Test Number	X1 (kV)	X2 (m)	X3 (m/s)
1	−10	1	0
2	0	3	0
3	0	3	0
4	−10	3	−0.3
5	10	5	0
6	10	1	0
7	0	3	0
8	−10	5	0
9	0	3	0
10	0	1	0.3
11	−10	3	0.3
12	0	5	0.3
13	10	3	−0.3
14	0	1	−0.3
15	10	3	0.3
16	0	3	0
17	0	5	−0.3

.

2.5. Data Processing and Analysis

In this test, droplets were collected using a 15 mm × 20 mm white paper card. Before each experiment, white paper cards were fixed on the target leaves of tomato plants with paper clips as droplet collection devices. In order to observe the effect of droplet deposition, carmine solution with a concentration of 5 g/L was used for spraying. In order to ensure that the test data are not contaminated by other factors, at the end of each test, the white paper card needs to be dried and stored separately in a sound self-sealing plastic bag.

At the end of the test, the white paper card used to collect droplets was brought back to the laboratory. For the collected white paper card, the following steps need to be performed: (1) Scan one by one with a scanner into digitized .jpg format files; (2) Adjust image digits; (3) Binarization; (4) Adjust threshold; (5) The droplet deposition coverage is obtained by the software Depositscan (Version 1.2) [23] (Figure 6).

In this paper, the droplet deposition coverage rate on the surface of each layer leaf was used as an evaluation index. Generally speaking, the larger the droplet deposition coverage rate on the tomato leaf surface, the better the spray effect.

$$\overline{X}=\frac{{\displaystyle \sum _{j=1}^m{X}_j}}{n-1}$$

(1)

where $\overline{X}$ is the sample mean; $X_j$ is the droplet deposition coverage rate on each white paper card; n is the total number of droplet collection cards at a certain position.

3. Results

The test was designed using response surface methodology, and the final test results are shown in

Table 4. Experimental results.

Test Number	X1 (kV)	X2 (m)	X3 (m/s)	Y1 (%)	Y2 (%)	Y3 (%)
1	−10	1	0	16.32	7.45	2.95
2	0	3	0	14.64	7.41	1.8
3	0	3	0	15.96	6.9	1.57
4	−10	3	−0.3	12.35	17.51	18.63
5	10	5	0	6.61	1.98	0.35
6	10	1	0	13.76	6.71	2.03
7	0	3	0	13.45	10.35	1.73
8	−10	5	0	5.74	1.79	1.5
9	0	3	0	14.33	7.61	2.96
10	0	1	0.3	10.5	9.11	5.34
11	−10	3	0.3	11.25	9.78	3.56
12	0	5	0.3	0.89	0.66	1.27
13	10	3	−0.3	10.18	12.23	4.38
14	0	1	−0.3	10.54	8.82	3.55
15	10	3	0.3	12.45	10.57	15.62
16	0	3	0	10.18	5.74	1.19
17	0	5	−0.3	0.93	0.48	0.44

. X₁ represents the charging voltage polarity; X₂ represents the spray distance; X₃ represents the spray pattern; Y₁ represents the droplet deposition coverage on the upper tomato leaf surface; Y₂ represents the droplet deposition coverage on the middle tomato leaf surface; and Y₃ represents the droplet deposition coverage on the lower tomato leaf surface.

The test data in Table 4 were imported into the Develve software (Version 4.16), and the test data were further processed by the Develve software. The response relationship between air-assisted electrostatic spray parameters and droplet deposition coverage on the surface of upper, middle and lower tomato leaves was obtained using regression analysis, and the droplet deposition coverage on the surface of upper, middle and lower tomato leaves was established based on the charging voltage polarity, spray distance and spray pattern. A quadratic polynomial regression model for the dynamic response relationship of spray platforms was constructed, the regression equations of which are shown in Equations (2)–(4):

$$\begin{array}{l}{Y}_1=10.57-0.16 X_1+4.40 X_2+0.45 X_3+0.0429 X_1{X}_2+0.2808 X_1{X}_3\\ - 3.6575\times {10}^{-16} X_2{X}_3+1.37 {X_1}^2-4.47 {X_2}^2-3.52 {X_3}^2\end{array}$$

(2)

$$\begin{array}{l}{Y}_2=0.46-0.10 X_1+6.46 X_2-3.58 X_3+0.0116 X_1{X}_2+0.5058 X_1{X}_3\\ - 0.0458 X_2{X}_3+0.02 {X_1}^2-1.36 {X_2}^2+28.92 {X_3}^2\end{array}$$

(3)

$$\begin{array}{l}{Y}_3=-5.26-0.04 X_1+5.39 X_2+0.70 X_3-0.0029 X_1{X}_2+2.1925 X_1{X}_3\\ - 0.4000 X_2{X}_3+0.04 {X_1}^2-1.01 {X_2}^2+53.56 {X_3}^2\end{array}$$

(4)

According to the test results recorded in Table 4, the regression model variance analysis tables of droplet deposition coverage on the upper, middle and lower tomato leaf surfaces can be obtained through further calculation, and the results are shown in

Table 5. Variance analysis of upper tomato leaves’ droplet deposition coverage model.

Source	Sum of Square	df	Mean Square	F-Value	p-Value
Model	325.02	9	36.11	13.05	0.0013
X1	0.8845	1	0.8845	0.3195	0.5896
X2	170.66	1	170.66	61.65	0.0001
X3	0.1485	1	0.1485	0.0536	0.8235
X1X2	2.94	1	2.94	1.06	0.3369
X1X3	2.84	1	2.84	1.03	0.3449
X2X3	0.0000	1	0.0000	0.0000	1.0000
X12	7.89	1	7.89	2.85	0.1352
X22	84.26	1	84.26	30.44	0.0009
X32	52.27	1	52.27	18.88	0.0034
Residual	19.38	7	2.77
Lack of fit	0.5376	3	0.1792	0.0380	0.9886
Pure error	18.84	4	4.71
Cor total	344.40	16

,

Table 6. Variance analysis of middle tomato leaves’ droplet deposition coverage model.

Source	Sum of Square	df	Mean Square	F-Value	p-Value
Model	281.80	9	31.31	7.13	0.0084
X1	3.18	1	3.18	0.7225	0.4234
X2	92.34	1	92.34	21.01	0.0025
X3	9.95	1	9.95	2.26	0.1762
X1X2	0.2162	1	0.2162	0.0492	0.8308
X1X3	9.21	1	9.21	2.10	0.1909
X2X3	0.0030	1	0.0030	0.0007	0.9798
X12	22.62	1	22.62	5.15	0.0576
X22	124.48	1	124.48	28.33	0.0011
X32	28.52	1	28.52	6.49	0.0382
Residual	30.76	7	4.39
Lack of fit	19.21	3	6.40	2.22	0.2284
Pure error	11.55	4	2.89
Cor total	312.56	16

and

Table 7. Variance analysis of lower tomato leaves’ droplet deposition coverage model.

Source	Sum of Square	df	Mean Square	F-Value	p-Value
Model	411.90	9	45.77	35.53	<0.0001
X1	2.27	1	2.27	1.76	0.2261
X2	13.29	1	13.29	10.32	0.0148
X3	0.1830	1	0.1830	0.1421	0.7174
X1X2	0.0132	1	0.0132	0.0103	0.9221
X1X3	173.05	1	173.05	134.37	<0.0001
X2X3	0.2304	1	0.2304	0.1789	0.6850
X12	63.31	1	63.31	49.15	0.0002
X22	68.04	1	68.04	52.83	0.0002
X32	97.82	1	97.82	75.95	<0.0001
Residual	9.02	7	1.29
Lack of fit	7.25	3	2.42	5.48	0.0668
Pure error	1.76	4	0.4408
Cor total	420.91	16

.

Based on the observation of Table 5, Table 6 and Table 7, it can be seen that the p-values of the three regression models are all less than 0.01, and the values of the respective lack of fit terms are all greater than 0.05, which indicates that the three models are suitable and significant. At the same time, for the determination coefficient R² of the three regression equations, the closer the calculation result of R² is to 1, the better the fitting degree and the greater the correlation. After optimizing the three models Y₁, Y₂ and Y₃, the coefficient of determination R² of Y₁ is 94.37%; the coefficient of determination R² of Y₂ is 90.16%; and the coefficient of determination R² of Y₃ is 97.86%. All three coefficients of determination R² are greater than 90%, indicating that the model can explain more than 90% of the variation in response values. To sum up, the three regression models established by this test data can predict and analyze the test indicators.

4. Discussion

4.1. Analysis of Droplet Deposition Coverage

By comparing the F-value (the larger the F-value, the greater the influence of this factor on the response variable) and the p-value (the smaller the p-value, the more significant the influence of this factor on the response variable) in Table 5, it can be seen that among the three factors X₁, X₂ and X₃, the order of their influence on Y₁ is X₂ > X₁ > X₃.

By comparing the F-value and p-value in Table 6, it can be seen that among the three factors X₁, X₂ and X₃, the order of influence on Y₂ is X₂ > X₃ > X₁.

By comparing the F-value and p-value in Table 7, it can be seen that among the three factors X₁, X₂ and X₃, the order of influence on Y₃ is X₂ > X₁ > X₃.

Figure 7a is a curve plot of the interaction of X₁ and X₂ on Y₁ when the spray platform sprays at a fixed height. From the point of view of X₁, electrostatic spray is better than non-electrostatic spray in droplet deposition. This is because of the electrostatic force, which means that charged droplets in the vicinity of tomato leaves occur as an “electrostatic circle or deflection” phenomenon, and then, charged droplets become attached to the surface of tomato leaves.

Figure 7b is a curve plot of the interaction of X₂ and X₃ on Y₂ in the case of air-assisted non-electrostatic spray. Analyzed from the angle of X₃, Y₂ is minimum when fixed height spray is employed. However, Y₂ changes little under three spray patterns. This indicates that the effect of X₃ on Y₂ is not significant for the air-assisted non-electrostatic spray. Comparing the Y₂ of ascending spray with the Y₂ of descending spray, it can be seen that the Y₂ of ascending spray is lower than the Y₂ of descending spray. Therefore, for this type of greenhouse single-rail air-assisted electrostatic sprayer, the droplet deposition coverage on the surface of tomato leaves in the middle layer is descending spray > ascending spray > fixed height spray.

Figure 7c shows the interaction of X₁ and X₂ on Y₃ when the spray platform sprays at a fixed height. With the increase in X₁, Y₃ first decreases and then increases, and the variation range is obvious. The main reason is that, at a fixed spray height, the lower leaf is less affected by the air flow, and electrostatic force becomes the main force. Therefore, the Y₃ of air-assisted electrostatic spray is larger than the Y₃ of non-electrostatic spray.

By observing the response curves of Y₁, Y₂ and Y₃ with respect to spray distance X₂ in Figure 7a–c, it can be seen that Y₁, Y₂ and Y₃ all first increase and then decrease as X2 increases. This indicated that the droplet deposition coverage on the upper, middle and lower tomato leaves in the mid-end position was higher than that in the near-end and far-end positions of the air-assisted electrostatic spray. This phenomenon is observed because (1) the wind speed at the near-end (1 m) is relatively high, and a large number of droplets are carried by the air flow, and it is difficult to deposit them on the surface of the proximal leaf; (2) the wind speed at the far-end (5 m) is relatively low, and it is difficult to effectively transport droplets to the far-end, and a large number of droplets are deposited on the surface of the mid-end leaf in the process of transportation, resulting in the generally low droplet deposition coverage on the surface of the far-end leaf.

4.2. Model Validation

In order to determine the optimal parameters for X₁, X₂ and X₃, Y₁, Y₂ and Y₃ were taken as the maximum optimization targets, and parameter optimization was carried out with the help of the Develve software. A model of droplet deposition coverage on the upper, middle and lower leaves of the tomato plant was established with the parameter ranges of X₁, X₂ and X₃ as constraints. The model is

$$\max Y_1=f(X_1,X_2,X_3)$$

(5)

$$\max Y_2=f(X_1,X_2,X_3)$$

(6)

$$\max Y_3=f(X_1,X_2,X_3)$$

(7)

where

$$X_1\in [-10,+10],X_2\in [1,5],X_3\in [-0.3,+0.3]$$

According to the above model and requirements, when X₁ is negative voltage, X₂ is 2.23 m, and X₃ is descending spray, Y₁, Y₂ and Y₃ reach the maximum. At this time, the predicted value of Y₁ is 14.0386%; the predicted value of Y₂ is 16.3669%; and the predicted value of Y₃ is 17.5975%. In order to verify the validity of the model, relevant experiments were carried out according to the optimized spray parameters, and the experiments were repeated three times. Considering that it is difficult to set the decimal place of equipment parameters during the test, the optimized parameters are slightly corrected: X₁ = −10 kV, X₂ = 2.75 m, X₃ = −0.3 m/s. Under these parameters, descending air-assisted electrostatic spraying operation was carried out, and the test results are shown in

Table 8. Comparison of model predicted values with experimental values.

Response Variable	Predicted Value (%)	Test Value (%)	Prediction Error (%)
Y1	14.0386	12.28	12.53
		14.33	2.08
		14.56	3.71
Y2	16.3669	15.07	7.92
		17.35	6.01
		14.46	11.65
Y3	17.5975	16.37	6.98
		15.82	10.10
		16.84	4.30

.

Based on the data in Table 8, it can be seen that the relative error between the experimental value and the predicted value of the model is less than 13% for the deposition coverage on the upper, middle and lower tomato leaf surfaces. This indicates that the model is valid.

5. Conclusions

Based on the analysis of the test results, the following conclusions can be drawn:

(1)

In air-assisted electrostatic spraying, the distribution of air flow had the greatest effect on droplet deposition on tomato leaf surface.

(2)

Compared with air-assisted non-electrostatic spray, air-assisted electrostatic spray had a better deposition effect.