Performance Test of Artificial Defoliating Broccoli Conveyor Line and Analysis of Defoliating Broccoli Inflorescences

Abstract

There is a close relationship between stem and leaf biological characteristics of mature broccoli plants and defoliation technology. Morphological parameters such as the spread degree and diameter of cauliflower stem and leaf and the connection performance of cauliflower stem and leaf were studied. These experiments are helpful to the research of defoliation devices and methods for large-scale manual transportation lines. In this paper, according to the damage forms that may be suffered in the separation process of mature broccoli stems and leaves, “sunny” broccoli varieties were selected. Firstly, the mechanical properties of mature broccoli stems were obtained through different loading methods such as stretching, compression, shearing and bending. Secondly, the stress relaxation characteristics of broccoli were analyzed by static compression of broccoli flower balls. Finally, the separation method of broccoli stems and leaves was simulated by ANSYS. The simulation results showed that shear fracture was more suitable for stem and leaf separation of broccoli than tensile fracture. Then, aiming at the separation of stems and leaves of broccoli, an artificial broccoli assembly line was proposed to ensure the efficiency of leaf removal through man–machine cooperation. The dynamic characteristics of the belt of broccoli leaf removal line were studied to ensure the efficient and stable operation of the conveyor system of broccoli artificial leaf removal line.

1. Introduction

As an important cash crop, broccoli is increasingly favored by consumers because it is rich in a variety of bioactive compounds such as vitamins [1]. Broccoli stems and flowers have a high water content. The broccoli plant has only flower balls and a few stalks that can be used, and other parts such as leaves, stalks, and small flower balls are not harvested [2]. Because the existence of broccoli leaves can delay their aging and prolong their storage time [3], broccoli often has leaves after the initial harvest. The last step before packing the broccoli is to remove the stems, leaves and stubble near the broccoli balls. Therefore, it is of great significance to study the mechanical properties of broccoli stem and leaf connection for the work of broccoli leaf removal. The study on the mechanical properties of broccoli stem and leaf connection and the process of stem and leaf separation has a guiding role in determining the suitable methods for broccoli stem and leaf separation. In order to improve the efficiency and standardization of broccoli defoliation, a kind of human–machine cooperation broccoli defoliation transport line was proposed. And through the analysis of the dynamic characteristics of the conveyor belt, the mechanism and structural characteristics of the broccoli artificial defoliation line were revealed.

The mechanical characteristics analysis of stems and leaves of broccoli and other plants is the basis of the study of broccoli leaf removal. The study of the mechanical properties of plants is often the basis for other applied research. Therefore, many scholars have carried out a lot of research on the mechanical properties of each part of the plant. Because of the different biophysical properties of different plants, the test method is suitable for some plants to measure their mechanical properties, and the method of constructing simulation model is suitable for other plants to study their mechanical properties.

Zhao et al. [4] obtained the mechanical property parameters of broccoli stalks through different forms of load tests such as stretching, compression and bending. Kamandar [5] studied the bending behavior of plants with elliptical cross sections and conical stems: loading rate, internode location and water content have significant effects on Young’s modulus and bending strength of straw. Chandio et al. [6] measured vertical and lateral rupture force, deformation, hardness, toughness and other parameters of corn grains by changing water content and compression loading speed, so as to study the effects of water content and compression loading speed on mechanical properties of corn grains in different directions. Liu et al. [7] measured the mechanical strength, elasticity and other related physical properties of cucumber stems, and analyzed the changes of these properties under different treatment conditions to reveal the mechanical properties and physiological and biochemical reactions of cucumber stems under different nitrogen (N) and potassium (K) conditions. Zhu et al. [8] used a texture analyzer to apply different levels of compressive stress to “Shine Muscat” grapes to determine the physical and physiological mass index of fresh grapes. By observing the changes in grape tissue and cell microstructure, it was found that the quality of table grapes gradually decreased with the increase in compression level.

The experimental method is mostly used by the above experts to determine the mechanical properties of the plants and explore the influence of different parameters. However, the method of software modeling is more suitable for some plants to explore the mechanical and physical characteristics of plants. Sadrmanesh et al. [9] used PFC 5.0 3D software to simulate the stretching behavior of hemp fiber, and combined it with the results of physical experiments to calibrate key parameters in the discrete element model. Wu et al. [10] established a segmental flexural stiffness model of tea stems with multiple virtual nodes. The curve of tea stem deflection was predicted by determining the position of the virtual node under load. Singhal et al. [11] conducted compression and bending tests on flax stems. The load state of the stalk was analyzed by finite element modeling (FEM), and the roll shape of the stalk crusher was optimized. Wu et al. [12] in order to explore the influence of different ditcher parameters (such as height) on rice straw displacement, established a rice plant simulation model by measuring the morphological and structural parameters of rice straw and material parameters. The optimal canopy opening position was obtained according to the maximum displacement of rice plants. Jing et al. [13] determined the mechanical characteristic parameters of rice at the heading stage through experiments, and simulated the canopy opening process based on the explicit dynamic finite element method.

The standardized defoliation of mature broccoli is often assisted by artificial defoliation conveyor lines. The performance of a conveyor belt is closely related to the dynamic characteristics of the conveyor belt. The characteristics of conveyor belts have been applied to various fields such as algorithms, mining, vegetables and fruits [14,15,16]. In view of the characteristics of conveyor belts, some scholars have carried out relevant research by constructing models.

Bortnowski et al. [17,18,19] investigated and analyzed the transverse vibration model of a conveyor belt. Subsequently, the vibration frequency of the working conveyor was measured to determine the transverse vibration frequency distribution of the belt along the entire length of the conveyor route when no material was available and the belt was loaded. Finally, a transverse vibration model of the conveyor belt is proposed, which interprets the conveyor belt as a known beam. This method allows the lateral bending stiffness of the grooved belt to be included in the model, which is closely related to the belt geometry. Zhang et al. [20] obtained the theoretical maximum position of additional deformation and additional stress in the transition section of the pipe belt conveyor by analyzing the forming mechanism of the conveyor belt in the transition section. And the three-dimensional simulation model of the transition section of the pipeline belt conveyor is established. Sun et al. [21] established a dynamic model of a belt conveyor and carried out a simulation analysis using RecurDyn 2022 software. Krauze et al. [22] developed an analytical model of a typical belt feeder and determined its stability and supporting force. Hao et al. [23] established a dynamic model of the rail coupling in the turnover system and analyzed the vibration characteristics of the rail, so as to gain an in-depth understanding of the dynamic behavior of the turnover system in the track belt conveyor.

The study of plant stem and leaf connection performance serves certain application scenarios. Post-harvest treatment of plants is considered a special application direction for plant harvest. The research on the defoliation method and artificial defoliation pipeline of mature broccoli is very rare. Compared with other plants, broccoli has the problem of wide leaves and cumbersome removal of leaves. Compared with the non-standardized manual stretching method, more reasonable defoliating methods and the establishment of a standardized manual defoliating pipeline are helpful in improving the efficiency of defoliating and the quality of broccoli after defoliating. Therefore, it is of great guiding significance to study the stem and leaf separation characteristics of mature broccoli and the dynamic characteristics of the defoliating pipeline.

2. Materials and Methods

2.1. Physical Morphological Characteristics of Broccoli

The broccoli variety in this study is “sunny”. The mature broccoli plant structure is shown in Figure 1, the broccoli plant is tall and upright and the main stem tip is a green or purplish-green flower ball. This variety is characterized by its good appearance, large size, and high yield. The buds are evenly granulated, with a vibrant dark green color, and they are fresh and crispy to the taste. The leaves of broccoli are blue-green, and the thicker mid-rib leaves stand upright. The base of the stem is thin and woody, gradually thickening from bottom to top. The outer skin is waxy green. Usually, 1 to 4 lateral branches grow at the base of the stem.

After harvesting, all the outer large leaves of the broccoli are usually removed. The two-dimensional parameter model of mature broccoli is shown in Figure 2, where h₁ is the plant height of broccoli, h₂ is the stem height, l₁ is the leaf expansion degree, and d₁ is the stem diameter.

2.2. Experimental Design of Broccoli Stem and Leaf Mechanics

The mechanical properties of broccoli stems and leaves include the stretching, compression, bending, shearing and friction properties of broccoli stems and leaves, which is also the basis of broccoli assembly line leaf removal, which are also the basis of broccoli assembly line leaf removal. The mechanical test of broccoli stems and leaves was conducted by TA.XTPLUS texture analyzer (British Stable Micro System). It can complete various forms of mechanical experiments by replacing various types of indenters. The parameters of the texture analyzer and its matching indenter are shown in

Table 1. The parameter of Texture analyzer and indenter.

Instrument Name	Parameter Type	Specification
TA.XTPLUS Texture analyzer	Range	50 kg
Craft Knife Adaptor	Serial number	A/CKB 13232
3 Point Bend Rig	Serial number	HDP/3PB 13269
Tensile Grips	Serial number	A/TG 13249

.

The texture analyzer can measure and record the forces under the stretching, shearing, bending and compression of broccoli petioles in real-time. The test items included stem stretching, stem shear, stem bending, stem radial compression, and root shear.

2.3. Mechanical Strength Analysis of Broccoli Stems and Leaves

Based on the size and mechanical tests of broccoli petioles, a number of mechanical properties can be calculated, including compressive strength, tensile strength, shear strength and elastic modulus. The compressive strength σ_max and elastic modulus E₁ of the broccoli petiole can be calculated by the following formulas:

$${\sigma }_{max}={\displaystyle \frac{F_{max}}{lD}}$$

(1)

$$E1={\displaystyle \frac{\sigma }{\epsilon }}$$

(2)

where F_max is the maximum compressive force, N; l is the sample length, m; D is the sample diameter, m; σ is compressive stress, Pa; ε is compressive strain.

The tensile strength of broccoli petiole component sample:

$$\sigma ={\displaystyle \frac{F}{A}}$$

(3)

where F is the maximum load of the stretching process, N; A is the cross product at the fracture, m².

Positive strain of test sample:

$$\epsilon ={\displaystyle \frac{\Delta l}{l}}$$

(4)

l is the length of the “pre-failure zone” of the sample, and l is 0.05 m in this test; Δl is the elongation of the sample at the stage of tensile elastic deformation, m.

The tensile elastic modulus E₂ of the stalk component sample is:

$$\mathrm{E}2={\displaystyle \frac{\sigma }{\epsilon }}$$

(5)

The bending force is considered to be a physical quantity that can be used to evaluate the bending limit of a sample. At the same time, in the bending test, the deflection of the broccoli petiole under the critical force can also be used as another physical quantity to evaluate the mechanical properties of the sample. The stress condition of the sample during the bending test belongs to a single concentrated load, and the calculation formula of deflection is shown in Formula (6). Where P is the breaking force, N; l is stalk length, mm; E₃ is the elastic modulus of the stalk and I is the moment of inertia of the stalk section.

$$Y_{max}={\displaystyle \frac{8P{l}^3}{384EI}}$$

(6)

$$Y_{max}=vt$$

(7)

$$\mathrm{E}3={\displaystyle \frac{8P{l}^3}{384I·Y_{max}}}$$

(8)

The loading speed of the texture analyzer is constant, and the displacement of the loading position can be obtained according to the breaking time of the petiole. This displacement value is the stem deflection at Fracture. The calculation formula is shown in Formula (7). v is the loading speed, mm/s; t is the loading time, s. According to the calculated value of the deflection of the actual load, the elastic modulus of the stalk can be obtained as shown in the Formula (8). The elastic modulus plays an important role in the study of stalk materials.

2.4. Broccoli Stress Relaxation Model Theory

Investigate the influence of different loading rates on the compressive mechanical properties of broccoli, to identify the strain rate dependence of mature broccoli sample materials, and evaluate the impact of different strain rates, so as to determine the loading rate conditions suitable for subsequent relaxation tests. Since it is not possible to determine the mechanical properties by modifying them to obtain standard specimens with regular geometrical shapes, the complete sample is placed horizontally between two rigid parallel plates, and loaded at a certain rate with a texture analyzer.

According to the parallel plate stress contact theory specified in the ASAE standard (2008), the elastic modulus of broccoli was estimated using Formula (9).

$$E_2=0.338F\left(1-{\mu }_2^2\right){\left[K_U{\left(1/R_U+1/R_U\right)}^{1/3}+K_L{\left(1/R_L+1/R_L\right)}^{1/3}\right]}^{3/2}/D^{3/2}$$

(9)

Formula: E₂ is the elastic modulus of broccoli sample, MPa; F is the compressive force, N; μ₂ is the Poisson’s ratio of broccoli sample, without dimension; D is the amount of deformation, m; R_U and R_L are the minimum curvature radius of the convex surface at the contact point between the sample and the upper and lower plates, m; R_U′ and R_L′ are the maximum curvature radius of the convex surface at the contact point between the sample and the upper and lower plates, m; K_U and K_L are coefficient constants determined by the radius of curvature, determined by cosθ (Figure 3 and Figure 4).

For cosθ, it is calculated from the curvature radius of the upper and lower surface panels, respectively. As the contact surface of the compression device is flat, the radius of curvature R₂ = R₂′, and [1/R₂] = [1/R₂′] = 0, cos θ can be calculated by Formula (10):

$$\cos \theta =\left(1/R_1-1/R_1^{\prime }\right)/\left(1/R_1+1/R_1^{\prime }+1/R_2+1/R_2^{\prime }\right)$$

(10)

The radii of curvature R₁ and R₁′ of the convex surface of mature broccoli can be determined by Formulas (11) and (12):

$$R=R_1=d/2$$

(11)

$$R^{\prime }=R_1^{\prime }=\left(d^2+H^2/4\right)/(2d)$$

(12)

where: d is the cross-sectional diameter of mature broccoli, m; H is the ball height of broccoli, m.

Therefore, after determining cosθ according to the above formula, the approximate values of R_U, R_L, R_U′ and R_L′ are calculated, respectively, according to the formula above. Finally, the above data are substituted into the previous equation to calculate the value of the elastic modulus of broccoli, and the stress values σ for different loading directions are calculated according to Equation (13).

Stress values can be calculated according to the Hertz equation:

$${\sigma }_t={\displaystyle \frac{F_N}{A}}={\displaystyle \frac{1.5F}{\pi dl}}{\displaystyle \frac{1}{\left(1-{\mu }_1^2\right)/E_1+\left(1-{\mu }_2^2\right)/E_2}}$$

(13)

In the equation, σ_t is the lateral load stress, MPa; F_N is the normal contact pressure, N; A is the compression area, m²; F is the compression force, N; d is the diametral sphere of the broccoli, m; L is the height of the broccoli, m; E₁ and E₂ are the elastic moduli of the broccoli sample and the compression plate, MPa; μ₁ and μ₂ are the Poisson’s ratios of the broccoli sample and the compression plate.

μ₁ and μ₂ are set to 0.30 and 0.25, respectively; E₂ is set to 103 MPa according to the material of the structural steel plate. The value of E₁ is determined based on the calculation of the modulus of the elastic model in the aforementioned compression characteristic test.

In SolidWorks 2023 software, a 3D model with the same size as the broccoli stem and leaf tensile test sample was built, saved as a file with the suffix “.x-t” format, and imported into the workbench’s Display Dynamics module. According to the physical property test results, broccoli stalk material data files were added to the Engineering Date Sources of workbench. The material properties are assigned to the Model in the model, and then the meshing operation is carried out, and the mesh cell size is 1 mm.

2.5. Broccoli Drop Injury Test

In order to explore the influence of factors such as drop height, contact plate Angle, contact plate quality and initial placement posture on the mechanical damage degree of broccoli, the experimental design was carried out according to the orthogonal experimental design pseudo-horizontal method, and the levels of experimental factors were shown in the Table above. According to the pre-test results, the pseudo-horizontal method was used to design the experiment, and the rubber plate was used as the pseudo-horizontal index. The comprehensive damage index was used as the test index, and orthogonal table L₉ (3⁴) was selected to carry out the test [24]. Finally, the primary and secondary relationship of each factor was determined by ANOVA. A total of 9 tests were carried out, and 5 broccoli samples were randomly selected for repeated tests each time, and a total of 45 samples were used for orthogonal tests.

According to the actual pre-test damage, the damage of broccoli can be roughly divided into three levels: epidermal abrasions (level 1, flower surface abrasions, vegetable body is relatively firm), internal tissue damage (level 2, local and tissue bruising, tighter body), Broken (level 3, obvious cracks in the flower bulb and stalk, and the vegetable body is not firm enough). The Damage composite index (DCI) [25] was introduced to evaluate the damage at all levels (Figure 5).

(1) Surface abrasion area S

$$S=\sum _{i=1}^n\pi w_{1i}{w}_{2i}$$

(14)

where: S is the surface abrasion area, mm²; w_1i and w_2i are, respectively, the diameter of the long and short axis of the elliptic damage area at the i place, mm; n is the number of lesions on a single broccoli.

(2) Internal tissue damage volume V

$$V=\sum _{j=1}^n\pi {\displaystyle \frac{(d_{1j}-d_{0j})}{24}}[3w_{1j}{w}_{2j}+4{(d_{1j}-d_{0j})}^2]$$

(15)

where: V is the internal damage volume, mm³; w_1j and w_2j are, respectively, the diameter of the long and short axis of the elliptic damage area at the J-place, mm; d_0j is the distance between the top outline of the j place of broccoli and the top of the bruise, mm; d_1j is the distance from the top outline of the J-place broccoli to the bottom of the bruise, mm.

(3) Break length L

$$L=\sum _{k=1}^n{l}_k$$

(16)

where: L is the broken length of broccoli, mm; l_k is the rupture length at the k point of Broccoli, mm.

(4) Damage composite index DCI

Since the above three damage indexes of different grades are all reverse indexes with different dimensions, it is necessary to carry out standardized dimensionless processing of different indexes according to the following formula to obtain standardized S′, V′ and L′ values during comprehensive evaluation.

$$x^{\prime }={\displaystyle \frac{|x-\overline{x}|}{\sigma }}$$

(17)

where: $x^{\mathrm{’}}$ is the standardized value of measurement data, x, $\overline{x}$, σ are the original value, average value and standard deviation of measurement data, respectively.

Considering the fluctuation of the above damage index data and the correlation between the indexes, the CRITIC weight method is adopted to evaluate the weight of the comprehensive index DCI.

This method mainly uses the internal difference of each evaluation index, namely standard deviation, to determine the weight difference assigned to each index, and its weight size is mainly obtained by normalizing the product of contrast intensity and conflict indicators. Generally speaking, indicators with higher contrast intensity and higher degree of conflict with the standard are given more weight, which indicates that the indicator can reflect more effective information.

The value of damage comprehensive index DCI is calculated according to the following formula:

$$DCI=W_1{S}^{\prime }+W_2{V}^{\prime }+W_3{L}^{\prime }$$

(18)

where: S′, V′ and L′ are the standardized values of the abrasion area, damage volume and rupture length, respectively; W₁, W₂, and W₃ are the weight coefficients of each indicator coefficient, respectively, determined by the CRITIC weight method [26].

2.6. Working Mechanism of Broccoli Artificial Leaf Removal Line

Considering the need to remove leaves from harvested broccoli, and issues with conventional de-leafing methods being unsatisfactory, improper disposal after removal, and low efficiency, an artificial leaf removal conveyor line suitable for the characteristics of broccoli was designed to ensure the efficiency and standardization of the leaf removal process. The overall structure of the conveyor line is shown in Figure 6. The factors affecting the leaf removal speed and subassembly efficiency of broccoli are analyzed, and the conveyor line is optimized on this basis.

The configuration scheme of the conveying line must meet the requirements of broccoli conveying. Considering the size of the site and whether it is suitable for human–machine cooperation, and referring to relevant parameters of the conveyor line and matching power, the conveying and packaging method of the conveyor line is determined with reference to the mechanical physical characteristics of broccoli. The main design parameters of the broccoli leaf removal and packaging double-layer conveyor line are determined.

The leaf removal and packaging conveyor line is composed of a driving drum, rotating drum, conveyor belt, upper support roller, lower support roller, deflector plate, bracket and other parts. The driving drum is connected one by one with the driven wheel system that drives its rotation. The power transmitted by the motor drives the driven wheel system and the driving drum to rotate through the active wheel system by means of chain transmission. The rotation of the driving drum drives the conveyor belt and the rotating drum to rotate, thus achieving the function of conveying broccoli.

As shown in Figure 7, the three-dimensional model of the leaf removal conveyor line, where the conveying part is the most important part of the leaf removal and packaging conveyor line, mainly undertakes the conveying function. The device mainly includes a conveyor belt, drive roller, driven roller, roller group and support plate. The force situation borne by the roller group and support plate determines the service life of the device, and the rotation speed of the active drum determines the conveying speed of the device, thereby determining the overall leaf removal and packaging efficiency. As shown in Figure 8, in the conveying process, the torque delivered by the motor drives the active drum to rotate at a certain speed, and then drives the conveyor belt and the crops on it together with the driven drum to complete the conveying operation.

As shown in Figure 9, the transmission system consists of two parts: one part is driven by the motor, the active wheel system drives the upper layer passive wheel system to ensure that the upper layer conveyor belt produces a conveying speed v1 to the right; the other part is that the active wheel system drives the lower layer conveyor belt to produce a conveying speed v2 to the right. Workers stand on both sides of the inclined plate manually, cooperating with the conveyor belt to achieve the upper conveyor belt transporting broccoli, and the lower conveyor belt transporting the removed broccoli leaves.

3. Results and Discussion

3.1. Physical Experimental Analysis of Broccoli Stems and Leaves

Plant height h1, stem height h2, leaf spread l1 and petiole diameter d1 of mature broccoli were measured with vernier calipers and tape measures, and morphological characteristics such as the number of large leaves and the number of small leaves were counted, respectively, as shown in

Table 2. Morphological parameters of mature broccoli.

Serial Number	Plant Height h1/cm	Stem Length h2/cm	Stem Diameter d1/mm	Leaf Spread l1/cm	Number of Small Leaves /Piece	Number of Large Leaves/Piece
1	65	20	12	81	12	5
2	70	17	11	69	14	4
3	64	16	13	78	15	5
4	63	18	12	73	10	5
5	60	24	13	69	16	4
6	76	20	14	70	10	6
7	68	21	14	74	16	3
8	62	21	15	80	13	4
9	60	18	17	70	14	4
10	70	16	9	72	14	3
11	58	23	15	80	11	5
12	56	22	10	66	13	5
13	59	22	13	80	10	5
14	68	18	11	75	15	3
15	62	20	11	78	12	7
16	67	21	13	65	15	5
17	59	20	14	63	9	6
18	61	17	14	73	13	6
19	64	18	15	79	13	4
20	70	23	14	70	13	5
21	62	23	17	75	12	6
22	71	21	10	81	13	7
23	66	24	11	77	15	4
24	74	22	8	75	16	4
25	70	21	12	69	13	5
Average value	65.00	20.24	12.72	73.68	13.08	4.8
Standard deviation	5.25	2.42	2.28	5.27	2.00	1.12

.

The plant height at maturity of this type of broccoli follows a normal distribution, as shown in Figure 10, with an average plant height of 65 cm. In addition, the leaf spread is 73.68 cm. The larger leaf spread indicates the occurrence of overlapping and interlacing of leaves between broccoli plants. The stem length is 20.24 cm, the petiole diameter is 12.72 mm, the average number of small leaves is 13.08, and the average number of large leaves is 4.8. Compared with the average value and standard deviation of each parameter, it is found that the growth difference of broccoli plants of this variety is small, leaves distribution is shown in Figure 11.

3.2. Experimental Analysis of Broccoli Stem and Leaf Mechanics

3.2.1. Broccoli Stem and Leaf Tensile Test

As shown in Figure 12, the tensile test was carried out on the TA.XTPLUS physical property tester (texture meter), since the loading speed and maximum loading force of the texture analyzer will have a significant impact on the test results. When conducting tensile testing of stems, the loading speed of the texture analyzer probe is set to 0.5mm/s. The tensile test is relatively complex and requires the use of weights to zero the probe force before the test can be carried out.

Before conducting the tensile test, the separated stalks are first cut along the root to prepare 50 mm long test samples, and eight stalks are prepared to perform eight repeated tests. To prevent slippage and clamp cracking of the test sample, wrap gauze around the clamp part of the sample.

In the broccoli petiole tensile test, the fracture location of the petiole was distributed in the middle of the sample, and the probe was set to automatically reset after the end of one test, which was convenient for the next clamp and test. The results of the stems and leaves tensile test are shown in Figure 13. In the initial stage of the tensile test, the pulling force increases linearly with displacement, and the pulling force drops sharply after reaching a peak. The decline of some curves is divided into two stages, that is, after dropping to a certain extent, they rise again, and eventually the pulling force drops to 0. The peak of the curve is between 170–270 N, and the time to reach the peak is different, ranging from 0.5 s–2 s.

When the petiole is stretched, the force direction is normal to the growth direction of the petiole, so the load required for fracture is larger. The peak point of the curve represents the critical value at which the petiole reaches fracture when stretched. The value of the peak point of each group of data is taken out and used as the tensile limit to represent the fracture force of the petiole. As can be seen from Figure 13, the petiole stretching process is mainly divided into the elastic stage, the linear elastic stage, and the breaking failure stage. The change law of the curve in the first half of the curve conforms to the stretching curve of plastic materials, and the change law of the curve after breaking conforms to the stretching curve of brittle materials. In addition, petiole samples with smaller outer diameters not only have smaller breaking forces, but also break earlier.

3.2.2. Broccoli Stem and Leaf Shear Test

The shear test is carried out from the axial direction of the petiole or stem. The mature broccoli petiole shear samples are prepared from those which are 50 mm in length, complete, straight, and free of pests and mechanical damage. The stem shear samples are made from the semi-stems that are 40 mm long, straight, and free from pests and mechanical damage. The test loading speed is set at 0.2 mm/s. Shear test as shown in Figure 14 and Figure 15, A/CKB 13232 shear fixture is used to shear the stalk and petiole.

The petiole shear curve is shown in Figure 16, which is divided into a continuous loading stage and a stable shear stage, and the shear resistance mainly comes from the epidermis and xylem. In the stable shear stage, the shear force rises stably without large fluctuations, and the shear force drops sharply after reaching the peak, at which point the petiole sample is cut open. During the continuous loading stage and stable shearing stage, the shear force curve has small fluctuations due to the uneven circumferential distribution of petiole components.

The stem shear curve is shown in Figure 17, and its change trend is similar to that of the petiole shear curve, but there are fluctuations in the curve at the initial shear stage. The reason for the fluctuation is that the stem samples are thicker and the components are unevenly distributed. The peak value of stem shearing is between 550 N–600 N. In addition, the thickness of the stem sample is much larger than that of the petiole sample, thus the difference in the thickness of the stem sample has a great impact on the peak value. Some stem samples with smaller thicknesses have lower peak values, ranging from 350 N–400 N.

3.2.3. Broccoli Stem and Leaf Bending Test

The bending test is different from the shear test and the tensile test in that there is no significant damage to the stalk during the test. Broccoli petioles with a length of 80 mm were taken as bending samples, as shown in Figure 18. HDP/3PB 13269 fixtures were used to perform bending tests on the stems, and the loading speed was 2 mm/s. The resulting curve of the stem bending test is shown in Figure 19. In the beginning, the force gradually increases with time, and after reaching the peak, it slowly decreases. The peak range of bending force is between 15 N–20 N. The rising stage of bending force corresponds to the bending deformation of the stem, and the amount of deformation can be measured by deflection. As the loading force increases, the petiole of broccoli eventually breaks.

3.2.4. Broccoli Stem and Leaf Compression Test

Studying the parameters and mechanical properties of mature broccoli stems and leaves is the first step in the study of broccoli stems and leaves separation. The stress of petiole under radial extrusion load was investigated experimentally. Firstly, the loading speed of the compression test project is set to 0.2 mm/s. The experimental sample is a mature broccoli petiole that is 50 mm long, complete, straight, and free of pests and mechanical damage. The experimental process is shown in Figure 20. The compression samples are placed directly underneath the press head, repeated for eight groups of experiments, and the data are recorded.

The required indenter for petiole compression test is the P/36R cylindrical indenter. Figure 21 shows the pressure–time characteristic curves of eight groups of mature broccoli petiole samples. Before the first peak of the curve appears, the pressure increases approximatively linearly, at which time the petiole undergoes elastic deformation. The pressure drop after the first peak is due to the pressure exceeding the compression strength of the pith, and rupture occurs in the bottom xylem and epidermis. From the curves, it can be observed that the peak value range of the compression load is 350 N–400 N.

3.3. Primary and Secondary Analysis of Mechanical Injury Characteristic Factors of Broccoli

In order to determine the weight of mechanical injury characteristic factors of broccoli, the pre-test of broccoli fall injury was first carried out. Under the same condition, the drop test of five groups of broccoli was carried out, and the damage evaluation indexes of broccoli were calculated statistically. Statistical results are shown in the following table (

Table 3. Statistical results of pre-test of broccoli fall injury.

Sample	Epidermal Abrasion Area S/mm2	Internal Tissue Damage Volume V/mm3	Breaking Length L/mm
1	201.06	27.88	52
2	197.92	17.67	78
3	172.79	4.46	44
4	235.62	53.93	57
5	289.03	42.67	50
Mean value	219.28	29.32	56.2
Standard deviation	44.95	19.62	13.05

).

The weight of CRITIC is calculated as follows (

Table 4. CRITIC weight calculation result.

Item	Index Variability	Index Conflict	Amount of Information	Weight
$S^{\prime }$	0.387	1.319	0.51	28.26%
$V^{\prime }$	0.397	1.224	0.485	26.89%
$L^{\prime }$	0.384	2.11	0.81	44.85%

):

The calculation results of CRITIC weight are shown in the table above. The rupture length L contains the most information and allocates the most weight.

According to Formula (18), the expression of damage comprehensive index DCI can be obtained as follows:

$$DCI=0.28S^{\prime }+0.27V^{\prime }+0.44L^{\prime }$$

(19)

The tests were arranged according to the pseudo-horizontal orthogonal test scheme, and the statistical DCI test results were shown in the following table according to the above formula. After preliminary analysis of variance, it is concluded that the influence of the Angle of the contact plate B on the DCI of the damage composite index is not significant enough, and the mean square error value is small. Therefore, factor B is regarded as the source of error, and the variance test table obtained is shown as follows (

Table 5. Orthogonal test scheme and results of drop damage.

Serial Number	A. Falling Height/mm	B. Contact Plate Angle/°	C. Contact Material	D. Initial Attitude	Damage Composite Index DCI
1	1 (100)	1 (0)	1 (Steel plate)	1 (Broccoli front down)	0.17
2	1 (100)	2 (15)	2 (Rubber sheet)	2 (Broccoli side down)	0.10
3	1 (100)	3 (30)	3 (Rubber sheet)	3 (Broccoli front up)	0.03
4	2 (300)	1 (0)	2 (Rubber sheet)	3 (Broccoli front up)	0.16
5	2 (300)	2 (15)	3 (Rubber sheet)	1 (Broccoli front down)	0.24
6	2 (300)	3 (30)	1 (Steel plate)	2 (Broccoli side down)	0.33
7	3 (500)	1 (0)	3 (Rubber sheet)	2 (Broccoli side down)	0.56
8	3 (500)	2 (15)	1 (Steel plate)	3 (Broccoli front up)	0.72
9	3 (500) W	3 (30)	2 (Rubber sheet)	1 (Broccoli front down)	0.69

):

According to the variance analysis, the factors that have A significant impact on the DCI index of broccoli fall impact process are falling height A, while the contact plate Angle B, contact plate quality C and initial attitude D have no significant impact on the test indexes (p > 0.10). This may be due to the limited fall height of broccoli. Although the center of gravity of broccoli is lower, the posture change is not obvious in a short time, and the contact rebound is stopped manually at the moment, and some injuries are not fully included and included in the calculation of DCI value, resulting in the insignificant influence of the above two factors. In addition, the experimental observation confirmed that the damaged parts of the broccoli falling upside down were concentrated in the petiole and short stem of broccoli, while the damaged parts of broccoli under the side were concentrated in the edge of the broccoli flower ball, but the damage to flower ball was not obvious enough. This is because at the moment of the collision, the contact area between broccoli and the contact plate material is large, and the mechanical damage is small under the impact condition of the same contact energy (

Table 6. DCI Analysis of variance.

Source of Variance	Degree of Freedom	Adj SS	Adj MS	F	P
A	2	0.501	0.251	82.626	0.012
C	2	0.027	0.013	4.385	0.186
D	2	0.006	0.003	1.000	0.500
error (B)	2	0.006	0.003
total	9	1.540

).

3.4. Broccoli Stem and Leaf Separation Model

The main stem and petiole of broccoli were equivalent to a connected cylindrical structure, and a certain tension was applied to the petiole of broccoli, and the Ansys simulation experiment was carried out. The experimental results were shown as follows:

The above is the equivalent stress situation, as shown in Figure 22, in which the equivalent stress is concentrated at the end of the root tip of the petiole and mainly distributed at the broken petiole. As shown in Figure 23, the maximum equivalent stress is concentrated at the lower end of the root tip. The maximum equivalent stress of broccoli petiole was 731.88 MPa when the sample was stretched and broken.

The above is the equivalent elastic strain, as shown in Figure 24. The equivalent elastic strain is concentrated at the root tip of the sample petiole. The maximum equivalent elastic strain is concentrated in the lower part of the root tip, as shown in Figure 25. The maximum equivalent elastic strain of broccoli petiole was 0.0036 mm at the time of tensile fracture of broccoli petiole sample.

From the above virtual experiment, it can be observed that the radial tensile resistance of broccoli petiole is strong, and the shape of the fracture site is irregular. In addition, the fracture stubble affects the appearance of the broccoli. Therefore, compared with the method of stretching and removing leaves, the cutting rules of petioles after cutting leaves are established, and the loading force required by this method is small. Therefore, the cutting and removing method is preferred in the work of removing leaves of mature broccoli.

3.5. Three-Dimensional Model and Dynamic Conveyor Model of Broccoli Leaf Removal and Packaging Conveyor Line

When the whole machine is working, the operator at the starting end will manually remove the leaves of the initially harvested broccoli and put them on the conveyor belt. Due to the difference in quality and shape of the broccoli, the state of dispersed transportation is finally formed, as shown in Figure 26. The total height of the conveyor line is 110 cm, and the width is 70 cm, which can accommodate two broccoli plants at the same time. The length of the conveyor line is set to 8–10 m according to the site conditions. Operators stand at every 0.5 m at both ends of the conveyor belt to manually remove the leaves of the broccoli and let the removed leaves enter the lower conveyor belt via the slanting board, forming a work flow where the upper conveyor belt transports the broccoli and the lower conveyor belt transports the removed leaves.

The dynamic characteristics of the conveyor belt refer to the complex relationship between the strain of the conveyor belt and the size and change frequency of the tension under the action of external tension. It is of great significance to study the dynamic characteristics of conveyor belts to ensure the efficient, stable and safe operation of the conveyor system. It can guide conveyor belt design and material selection, predict and reduce wear, avoid failures, reduce maintenance costs, improve system reliability, optimize energy use, reduce energy consumption, and improve production efficiency by precisely controlling conveyor belt speed and tension. In addition, a deep understanding of the dynamics will help develop intelligent monitoring systems that enable early fault diagnosis and preventive maintenance, thereby reducing downtime and ensuring continuity in production.

The dynamic characteristics of the conveyor belt mainly refer to the hysteresis, creep, and relaxation properties of the belt. As shown in Figure 27a, hysteresis is the phenomenon where, during the process of unloading the pulling force, the strain of the conveyor belt does not immediately return to its initial state. This causes a delay, i.e., the stress–strain curves of loading and unloading do not coincide. This indicates that the conveyor belt has obvious viscoelasticity, especially for canvas and woven core conveyor belts, while the steel wire rope core conveyor belts are relatively less obvious. As shown in Figure 27b, creep refers to the nonlinear relationship between the elongation of the conveyor belt and time under a constant pull force. As shown in Figure 27c, relaxation refers to the phenomenon where, after extending to a certain length and over a period of time, the stress on the conveyor belt decays to a stable value.

Where σ is the stress of the conveyor belt and ε is the strain of the conveyor belt. Studying the dynamic model of the conveyor belt of broccoli leaf removal conveyor line can better understand and predict the behavior of the conveyor belt system in practical work, and has a guiding role for the practical application of broccoli conveyor line.

Mature broccoli has large leaves and stems. Workers take out mature broccoli from the vegetable basket, use a cutter to remove the large leaves and petioles, then place the decorticated broccoli on the first layer of the conveyor belt, and put the cut leaves and petioles on the second layer of the conveyor belt. The conveyor belt runs at a constant speed. After the broccoli is transported to the end of the conveyor belt, workers load it into foam boxes, and complete the packaging after adding a suitable amount of ice cubes (Figure 28). A second conveyor belt carries the leaves and stems to the waste basket for other uses. In the whole working process, workers stand on both sides of the conveyor belt, and the distance between workers is about 0.3 m. In order to ensure man–machine coordination, the conveying speed of the conveyor belt is 0.3 m/s–0.5 m/s, so the overall average working efficiency is more than 40 plants/min–50 plants/min.

4. Conclusions

• The mechanical and physical characteristics of mature broccoli have guiding significance for the separation of stems and leaves. In this paper, the relative physical property parameters of “sunny” broccoli were measured. The height of mature broccoli plants presents as a normal distribution in the field, with an average height of about 65 cm, an average petiole diameter of 12.72 mm, and an average leaf number of 17.56 leaves. The mechanical properties of broccoli petiole were tested by a texture analyzer. The mean stretch limit force of broccoli petiole is 193.41 N, the mean shear limit force is 393.53, the mean bending limit force is 19.60 N, and the mean compressive limit force is 391.39 N. The model of stem and leaf separation also showed that the shear method was more suitable for stem and leaf separation than the stretching method.
• The dynamic characteristics of a conveyor belt mainly refer to the hysteresis characteristics, creep characteristics and relaxation characteristics of the belt. The conveyor belt in the manual leaf-removing conveyor line often uses a rubber belt, which has certain creep properties. Therefore, the spacing of broccoli on the conveyor belt is not entirely a fixed value, but shows a distribution situation from small to large spacing. Using this, a combination of man and machine can be used to match the manual packing speed with the conveyor belt transportation speed.
• The optimized design of the manual leaf-removing and dispensing conveyor line improves the stability and efficiency of the leaf removal. The conveyor uses a motor and chain wheel to control the simultaneous operation of two layers. To ensure the cooperation of humans and machines, the conveying speed of the conveyor belt is 0.3 m/s–0.5 m/s. When the conveyor belt is working, the workers are evenly distributed on both sides of the conveyor belt, with a spacing of about 0.3 m between workers. The distance between the broccoli after leaf removal on the conveyor belt is also 0.3 m, therefore the average working efficiency per person is 40 plants/min.