Measurement and CFD-DEM Simulation of Suspension Velocity of Peanut and Clay-Heavy Soil at Harvest Time

Abstract

The suspension velocity is the core of the cleaning and sorting mechanisms that utilize a combination of a fan and vibrating sieve. To investigate this, various experimental subjects, such as peanuts with different kernels and clay-heavy clods in different states, were used. The experiment involved simulating the suspension velocity of materials through numerical calculations using fluid dynamics and particle discrete element coupling. The Eularian model was employed to study the coupled gas-solid two-phase flow. The experiment measured the suspension velocities of single and double kernel peanuts, which were found to be 8.34~9.40 m/s and 8.13~9.51 m/s, respectively. Under 20.4% water content and lumpy conditions, the suspension velocities of smaller clods, side by side clods, and larger clods were 12.61~14.30 m/s, 14.16~15.76 m/s and 16.44~18.72 m/s, respectively; under 20.4% water content and smaller clods, the suspension velocities of lumpy and strip of clods were 12.61~14.30 m/s, 11.90~14.13 m/s, respectively; under lumpy and smaller clods, the suspension velocity at 17.6%, 20.4%, and 23.9% water content ranged from 12.38 to 14.20 m/s, 12.61 to 14.30 m/s, and 12.62 to 14.49 m/s, respectively. The simulations showed that the suspension velocity for different types of peanuts, clod sizes, shapes, and water contents was less different from the actual experiments. Specifically, the relative errors in suspension velocity for single-kernel peanuts, double-kernel peanuts, smaller clods, side-by-side clods, larger clods, lumpy clods, strips of clods, and clods with 17.3%, 20.4%, and 23.9% water content were 1.2%, 4.1%, 0.4%, 2.0%, 4.4%, 0.4%, 5.1%, 5.4%, 0.4%, and 1.9%, respectively, compared to actual experiment measurements. The results indicate a significant difference in the suspension velocity between peanuts and clay-heavy clods, which can be distinguished from each other based on this difference. Furthermore, the simulation results have been found to be consistent with the experimental results, thus verifying the feasibility of measuring the material suspension velocity using CFD-DEM gas-solid coupling.

1. Introduction

Currently, both domestic and foreign peanut harvesting machinery utilize cleaning and sorting mechanisms that combine a fan and vibrating sieve. This principle is based on the difference in suspension velocity caused by varying material densities, allowing for the selection of appropriate air speeds to achieve material separation [1,2]. In addition, pneumatic conveying is also widely used in mechanical devices such as sowing, fertilizing, picking up, and harvesting [3,4,5,6]. As can be seen, the suspension velocity of the material is an important parameter in the design of agricultural machinery, such as scavenging and conveying devices. China’s peanut planting area ranks second in the world. Nearly one-third is planted in hilly mountainous areas with mostly clay-heavy soil [7]. Due to its small fields, sloping arable land, complex and diverse planting agronomy, and other multiple factors, peanut production mechanization and industrial development is severely restricted. Current peanut combine harvesting machines are designed for sandy loam soil planting, making them poorly adaptable for direct application to clay-heavy soil planting. This results in low harvesting efficiency, significant impurity rates, and high loss rates. These defects highlight the need for improved peanut harvesting machinery specifically designed for clay soil planting. Therefore, there is an urgent need for the research and development of combined harvesting machinery specifically designed for peanut cultivation on clay-heavy soils. This development is of great significance as it can solve the problems of low harvesting efficiency and increase farmers’ income. Furthermore, cleaning is a crucial step in peanut combined harvesting machinery that directly affects the rate of impurities and losses. Therefore, the separation and clearing of peanuts grown on clay-heavy soils for combined harvesting is of great importance to fill the gaps in the industry, while suspension velocity plays a central role in this process.

So far, many scholars have conducted suspension velocity measurement experiments and numerical simulation analyses related to different crop particles and debris. B.Y. Gorial et al. [8] numerically solved a mathematical model of particle motion to simulate the trajectory of wheat and straw in an air stream, suggesting the feasibility of achieving their separation based on particle suspension velocity. El-Sayed A.S. et al. [9] studied the physical and aerodynamic characteristics of different peanut varieties (American, Chinese, and Egyptian) and derived optimum airflow velocities for peanut shell separation, broken seed kernels, and intact seed kernels by means of suspension velocity experiments. Dai Fei et al. [10] measured the suspension velocities of flax grains, capsules, short stems, and capsule husks at different stages to derive the overall interval distribution and separation characteristics of each component of the flax threshing material, and used the gas-solid coupling method to simulate the suspension velocities of each component of the flax threshing material to verify the feasibility of measuring suspension velocities based on CFD-DEM simulation experiments. Yaquan Liang et al. [11] used a DFPF-25 material suspension velocity measurement device to measure the suspension velocity of each component of rice material, and verified through fluid simulation of the cleaning and sorting chamber that a dual fan cleaning structure consisting of a cross-flow fan and a centrifugal fan can significantly improve the cleaning efficiency. Lianxing Gao et al. [12] conducted aerodynamic experiments on the main components of a peanut hulling machine and impurities, and obtained the floating coefficients of broken peanut pods, unpurified peanut pods requiring secondary shelling, peanut rice, and peanut shells and stones, which provided a reference basis for the design of peanut hulling machine sorting devices and pneumatic conveying devices. Wang Yifei [13] measured the suspension velocity of peanut threshing material, including pods, long stalks, short stalks, stalks, and miscellaneous leaves, and used coupled CFD-DEM to simulate the material transport process and flow field simulation for the cleaning and sorting device of the peanut combine harvester.

Existing studies have shown the value of the coupled CFD-DEM-based theory as a guide and practical application for the analysis of the aerodynamic characteristics, study separation, and clearing of peanuts grown in clay-heavy soils and side-by-side clods, which can save costs and improve research efficiency.

2. Materials and Methods

2.1. Structure and Working Principle of the Levitation Wind Tunnel

The system composition of the device for measuring the suspension velocity of peanuts and clay-heavy clods is shown in Figure 1 and mainly consists of a holding net, an inlet, a stable section, a test section, a damping net, a buzzer, a motor, and a frequency converter. The working wind speed of the wind tunnel is adjusted by controlling the speed of the motor through the frequency converter.

The main technical parameters of the peanut clay suspension velocity measuring device are shown in

Table 1. Technical parameters of the suspension velocity measuring device for peanuts and clay-heavy clods.

Technical Parameters	Values
External dimensions (L × W × H)/(mm × mm × mm)	810 × 755 × 3359
Power/kW	2.2
Flow rate/(m3·h−1)	2450
Inner diameter of the flow stabilizer/mm	90
Inner diameter of conical tube tip/mm	190
Inner diameter of stable section/mm	500
Conical tube inclination/(°)	2.6
Holding and damping nets	20-mesh nylon mesh

.

The velocity measurement system consists of a differential pressure transmitter, an industrial control computer, and velocity measurement using the drop method, that is, by measuring the static pressure difference between the test section of the wind tunnel and the upstream stable section according to the fluid dynamics “Bernoulli” theorem to convert it to the test section’s working wind velocity. The differential pressure is first converted by the differential pressure transmitter into a voltage value, then the computer collects the voltage value as a digital quantity, which is calculated into the wind tunnel working wind speed value (unit: m/s) by the program cured in the computer and displayed to the digital tube in real time.

To ensure the accuracy of the measurement results, corrections should be made after the measurements are completed compared to the theoretically calculated values. First, set the material in the conical observation tube at a relatively stable suspension height, read the height range value and inverter frequency, and close the power supply to remove the material. Then, re-open the power supply and allow the fan speed to reach the same value through the high-precision rotating blade type anemometer, and read the airflow speed at the small end of the conical observation tube, according to Formula (1), to calculate the suspension velocity of the material [14].

$$V_h=V_0{\left[\frac{r}{r+2H\tan \theta }\right]}^2$$

(1)

where, $V_h$ is the airflow velocity at any section of the conical observation tube, m/s; $V_0$ is the air velocity at the small end of the conical observation tube, m/s; $r$ is the diameter of the small end of the conical observation tube, m; $H$ is the suspension height of the material in the conical tube, m; $\theta$ is the angle of inclination of the conical observation tube °.

2.2. Experimental Materials and Suspension Velocity Measurement Methods

The experiment material consisted of Wanhua 17 peanuts and clay-heavy clods, all taken from the semi-fed peanut combine harvester in the field after cleaning and sorting into the fruit collection box section, as shown in Figure 2. The water content of the peanuts was measured at 24.7% and the clods at 20.4%. After sorting, the average geometry of the single-kernel peanuts was 25.04 × 16.95 × 16.11 mm, with an average mass of 1.92 g. The average geometry of the double-kernel peanuts was 39.24 × 16.17 × 17.02 mm, with an average mass of 3.06 g. Ten random samples of each were taken.

The clay-heavy clods were divided into three size ranges based on the actual harvest: smaller clods, side-by-side clods, and larger clods. The size of the side-by-side clods is essentially the same as the peanut pods, with a mass of 7.71 ± 0.50 g. The smaller clods are 1/2 the size of the side-by-side clods and are the smallest clods that can remain on the sieve surface under the action of the vibrating sieve, with a mass of 3.80 ± 0.50 g. The larger clods are three times the size of the smaller clods, with a mass of 11.74 ± 0.50 g. The shapes of the clods are divided into two types: lumps and strips. The soil water content is divided into three classes: 17.6%, 20.4%, and 23.9%. The water content of the soils in this sample was 20.4%, which was scaled appropriately to consider factors such as clay-heavy soils and harvesting weather in different areas. Ten random samples were taken from each.

The velocity of suspension of each classified material was tested 10 times, with each test repeated 3 times in a single-factor test. In the test, to measure the suspension velocity of different masses of clods, the shape of the clods was taken to be lumpy based on the number and mass share of the two shapes in the fruiting soil mixture, and a water content of 20.4% was taken. When conducting experiments to measure the suspension velocity of different shapes of clods, the smaller clods were taken based on the number and mass share of clods in the three size zones and the ease of performing clod removal. Then, a water content of 20.4% was taken. When carrying out experiments to measure the suspension velocity of clods with different water contents, smaller and lumpy clods were selected.

The experimental tools include height measuring scales, the HHF142 high-precision rotary vane anemometer (measuring range 0.2 to 40 MPS), vernier calipers, and electronic measuring scales, etc.

First, turn on the wind tunnel power supply, warm up for half an hour, and then place the material to be tested on the holding grid from the inlet, adjust the wind tunnel wind speed by inverter, and observe the suspension of the material in the conical tube (Figure 3). As the materials to be tested are irregularly shaped and unevenly distributed in mass, a fixed and precise suspension height value will not be obtained when the suspension velocity is measured. When the material suspension height is stabilized in a certain height range in the conical observation tube, stop pressing the wind speed adjustment button on the inverter panel and record the suspension height range [H1, H2] and the frequency of the inverter. Calculate the average suspension height range for the three replicate tests [h1, h2], and calculate the material suspension velocity range according to Equation (1). After recording, turn off the power to the wind tunnel, remove the material, turn the power back on, and adjust the fan speed to the recorded value. The air velocity at the small end of the conical tube is measured with the “OMEGA” HHF142 high-precision rotary vane anemometer and the arithmetic average of six measurement points is taken. The six measuring points are determined in accordance with the equal area method for circular duct sections, i.e., the small end of the conical observation tube section is divided into three concentric circles of equal area, and then each circle is divided into two parts of equal area, and the measuring points are chosen on the dividing line of these two parts [15].

3. Numerical Simulation Modelling

3.1. Controlling Equations for the Fluid and Particle Phases

In the CFD-DEM coupling process, the fluid phase is controlled by the Navier–Stokes equations for incompressible fluids:

$$\frac{\partial \left(\alpha \rho \right)}{\partial t}+\Delta ·\left(\alpha \rho u\right)=0$$

(2)

where, $\rho$ is the is the density of the gas, kg/m³; $\alpha$ is the volume fraction of the gas; $u$ is the rate of gas flow, m/s; $t$ is the time, s; $\Delta$ is the Hamiltonian differential operator.

The equation of conservation of momentum is

$$\frac{\partial \left(\alpha \rho u\right)}{\partial t}+\Delta ·\left(\alpha \rho u^2\right)=\alpha \Delta \tau +\alpha \rho g-\alpha \Delta p-\frac{{\sum }_{i=1}^N{F}_i}{V_c}$$

(3)

where, $p$ is the is pressure on the micro-element, Pa; $g$ is the acceleration of gravity, m/s²; $\tau$ is the gas kinetic viscosity, Pa$·$s; $F_i$ is the resistance of particle $i$ to the gas, N; $V_c$ is the volume of the cell grid, m³.

The particle phase control equation is as follows:

$$m\frac{d{v}_p}{dt}=mg+F_p$$

(4)

$$\frac{d{\omega }_p}{dt}=\frac{M_p}{I_p}$$

(5)

where, $m$ is the mass of the particle, kg; $v_p$ is the velocity of the particle, m/s; $F_p$ is the combined force on the particle other than gravity, such as buoyancy, drag and additional mass force, N; ${\omega }_P$ is the angular velocity of the particle, rad/s; $M_p$ is the combined moment of rotation of the particles, $I_p$ is the rotational inertia of the particle.

3.2. Material Model and Parameter Setting

Based on the actual distribution of fruit and debris in the fruit collection box after completion of harvesting by the peanut combine in hilly mountainous areas, the simulation experiments mainly analyzed the suspension state of different kernels of peanuts and clay-heavy clods with different masses and shapes as well as different water contents. The experiments were based on the discrete element simulation software EDEM 2021.2 with the multi-sphere filling method to create parametric models of differently shaped particle entities (shown in Figure 4) [16], analyze particle movement of the peanut pods and clay-heavy clods, and analyze the fluid dynamic properties based on Ansys 2022 R1 [17].

The physical and relevant action parameters of the particle model, derived from measurements, and a review of the relevant literature, are shown in

Table 2. Basic material parameters.

Materials	Parameters	Values
Peanut pods	Poisson’s ratio	0.4
	Density/kg·m−3	465.08
	Shear modulus/Pa	6.5 × 106
Clay heavy soil	Poisson’s ratio	0.46
	Density/kg·m−3	2054
	Shear modulus/Pa	1 × 106
Rigid PVC	Poisson’s ratio	0.32~0.47
	Density/kg·m−3	1418
	Shear modulus/Pa	1.50 × 109~3.92 × 109
Peanut pods–rigid PVC	Coefficient of restitution	0.275
	Coefficient of static friction	0.583
	Coefficient of rolling friction	0.104
Clay heavy soil–rigid PVC	Coefficient of restitution	0.31
	Coefficient of static friction	0.522
	Coefficient of rolling friction	0.43

[18,19].

3.3. CFD-DEM Numerical Simulation Experiments Based on DDPM Model

The structural model of the suspension velocity experimental device is created using the 3D solid modeling software SpaceClaim 2022R1. To improve the quality of the mesh and reduce the solution time (the computer’s solution time increases exponentially with the number of mesh cells), the model should be simplified as much as possible. The wind tunnel model is extracted from the fluid domain, grouped, and named boundaries are created and saved in .scdoc format before being imported into Fluent Meshing for meshing. The minimum sized cell is 1.4 mm, and the maximum sized cell is 10 mm. The hexahedron dominant method is used to fill the body mesh to improve the calculation accuracy. A total of 89,999 cells are generated after the division is completed, with a maximum skewness of 0.48 and a mesh quality of 0.2, achieving the desired effect. The 3D model and the divided body mesh are shown in Figure 5. Save as a .msh file and import into EDEM and Fluent Solution, respectively, to prepare the coupling simulation.

After setting the parameters and importing the model in EDEM, a virtual plane of the dynamic particle plant was created with a diameter of 80 mm, slightly smaller than the diameter of the small end of the conical tube. The plane is located 30 mm from the bottom of the small end of the conical tube to ensure that the particles are completely in the fluid region when they are generated. Setting the particle generation after 0.01 s allows the wind field to be in a stable phase when the particles appear, limiting the number of generated particles to 1. Set the Rayleigh time step to 3 × 10⁻⁵ s [20]. Turn on the coupling switch after setting.

Import the mesh file in Fluent Solution with the solver as pressure basis and time as transient. After setting up the gravity field, import the coupling interface file to connect Fluent and EDEM. Activate the dense discrete phase model (DDPM) in the Eulerian model, which is solved using a multiphase flow framework with the fluid equations incorporating a volume fraction term [21]. Set up the discrete phase, define the jet source and phase, and define the inlet air flow rate in the boundary conditions and the outlet as a pressure outlet. After initialization, enter the solution interface and set the time step to 1.5 × 10⁻³ s, which should be no less than the Rayleigh time step in EDEM and should be an integer multiple. Then, set the number of time steps to 10,000 and the simulation time to 15 s.

In the calculation process, the setting of the air flow rate has a significant effect on the suspension of the material. A low air flow rate will prevent the material from overcoming gravity and will thus rest on the holding grid, while a high flow rate will cause the material to fly out of the outlet. It is therefore necessary to carry out pre-calculations with reference to the theoretically calculated values of the suspension velocity of each material to find the boundary conditions that allow the material to achieve the desired suspension effect.

The simulations show that when the inlet wind velocity is set at 17.86 m/s, 24.38 m/s, 26.76 m/s, and 30.78 m/s for the peanut pods, smaller clods, side by side clods, and larger clods, each material can reach a stable suspension and is at one third of the conical observation tube. In the experiment, it was found that each material was relatively close to the wall of the conical observation tube when it reached a stable suspension state under the action of the vertical wind field, and the smaller the mass of the material, the closer it was to the wall of the tube, as shown in Figure 6. The reason for this phenomenon is that the velocity of the airflow in any cross-section of the conical observation tube decreases from the center of the tube to the wall, and the pressure difference on the surface of the material particles causes the particles to be pushed upwards by the airflow while also moving in the direction of the tube wall. When the velocity of the airflow near the pipe wall is less than the velocity required to overcome the gravitational work of the particles, the particles begin to fall, creating a horizontal acceleration from the pipe wall towards the center of the pipe, moving towards the center of the pipe. When the particle falls to a position where the velocity of the airflow is greater than the velocity required to overcome the work of gravity, it rises again, and so on, finally reaching a relatively stable state of suspension near the wall of the tube. The smaller the mass of the material found in the experiments, the closer it is to the wall of the pipe because the suspension velocity is significantly influenced by the mass factor; the smaller the mass of the material suspension velocity, the more likely it will reach a stable suspension state in the cross-section of the pipe where the airflow is smaller, that is, near the wall of the pipe. If the mass of the experimental material is further reduced to a certain level, it will cling to the wall of the tube to achieve a stable suspension. The study by Xiangyu Wen et al. [22] also verifies the conjecture of this study.

4. Results and Analysis

4.1. Experimental Results and Analysis

4.1.1. Suspension Velocity of Peanuts of Different Kernels

The suspension velocity measurements of peanuts are shown in

Table 3. Suspension velocity of peanuts of different kernels.

Classification	Test Number	External Dimensions			Mass/g	Average Suspension Height Range/mm	Suspension Velocity/m·s−1
		Length/mm	Width/mm	Height/mm
Single-Kernel Peanuts	1	23.57	16.18	15.34	1.8	386~456	8.37~9.24
	2	24.53	16.06	15.54	1.74	359~457	8.35~9.61
	3	24.29	16.37	15.54	2.09	344~413	8.89~9.83
	4	24.91	16.3	15.5	1.48	416~586	7.04~8.85
	5	26.9	18.62	18.61	2.17	402~469	8.22~9.03
	6	25.37	16.88	16.91	2	376~424	8.75~9.38
	7	26.74	17.52	15.36	2.12	359~451	8.42~9.61
	8	24.19	16.2	15.88	1.84	356~434	8.63~9.66
	9	23.67	17.47	15.53	1.68	372~443	8.52~9.43
	10	26.23	18.22	16.91	2.28	376~472	8.18~9.38
$$\overline{x}$$		25.04	16.95	16.11	1.92	374.6~460.5	8.34~9.40
Double-Kernel Peanuts	1	39.94	15.85	16.03	3.24	318~457	8.35~10.23
	2	37.18	16.51	17.37	3.06	354~504	7.84~9.69
	3	44.01	17.08	18.05	3.44	392~471	8.20~9.16
	4	36.5	15.42	16.05	2.63	364~467	8.24~9.54
	5	44.7	17.41	18.17	3.26	418~536	7.51~8.82
	6	36.39	16.39	17	2.86	357~456	8.37~9.64
	7	41.03	16.45	17.84	3.74	349~448	8.46~9.76
	8	33.8	15.28	15.85	2.35	352~461	8.31~9.71
	9	36.81	15.17	16.27	2.48	389~513	7.74~9.20
	10	41.99	16.17	17.15	3.52	375~464	8.27~9.40
$$\overline{x}$$		39.24	16.17	17.02	3.06	366.8~477.7	8.13~9.51

, where the peanuts have a water content of 25% and the frequency converter is 28 Hz, and $\overline{x}$ is the arithmetic mean.

As can be seen from the results of the experiment in Table 2 and the comparative graphs of suspension velocities of peanuts of different kernels in Figure 7a, the suspension velocities of single-kernel peanuts ranged from 8.34 to 9.40 m/s and double-kernel peanuts from 8.13 to 9.51 m/s, with little difference between the two. The overall data show that the suspension velocities of the double-kernel peanut are even slightly lower than that of the single-kernel peanut, which is related to the difference in the windward area of the wind tunnel due to the difference in shape and size between the two. The single-kernel peanut pods have a small difference in length, width, and height dimensions, with a shape close to an ellipsoidal shape, relatively uniform forces, and a relatively small floating range in the wind field when they reach a steady state of suspension. On the other hand, the double-kernel peanut pods have a large area of force in the vertical airflow despite their large mass and shape dimensions, which results in the suspension velocities of the double-kernel peanut pods being slightly less than that of the single-kernel peanut pods. In the practical experiments, it was evident that, due to the heterogeneous shape and size of the double-kernel peanut pods, although the stability of the suspended state was better than that of the single-kernel peanut pods, their up and down range in the wind tunnel was significantly greater. The independent sample t-test yielded a significance level of 0.768, which is greater than 0.05, showing that the effect of single and double kernel peanuts on their suspension velocity did not reach a significant state. Thus, overall, it can be concluded that the number of kernels in the peanut pods has a small effect on their suspension velocity.

4.1.2. Suspension Velocity of Different Masses of Clods

The clods were taken in lump form and had a water content of 20.4%. The values of suspension velocity for different masses of clay-heavy clods are shown in

Table 4. Suspension velocity of peanuts of different masses of clods.

Test Number	Smaller Clods		Side-by-Side Clods		Larger Clods
	Average Suspension Height Range/mm	Suspension Velocity/m·s−1	Average Suspension Height Range/mm	Suspension Velocity/m·s−1	Average Suspension Height Range/mm	Suspension Velocity/m·s−1
1	278~390	12.54~14.86	288~371	14.15~16.05	319~392	15.79~17.60
2	277~366	12.99~14.88	296~362	14.34~15.85	261~329	17.33~19.27
3	313~412	12.15~14.07	277~336	14.91~16.33	192~288	17.46~21.59
4	254~318	13.96~15.43	305~375	14.07~15.63	244~342	16.99~19.80
5	282~346	13.38~14.76	296~374	14.09~15.85	271~341	17.02~18.96
6	276~335	13.60~14.90	324~386	13.84~15.18	351~422	15.12~16.77
7	356~450	11.52~13.18	323~382	13.93~15.21	284~369	16.33~18.58
8	384~436	11.74~12.65	322~396	13.65~15.23	326~383	15.99~17.41
9	314~396	12.43~14.05	281~364	14.30~16.23	319~394	15.74~17.60
10	288~410	12.18~14.62	287~358	14.43~16.08	294~382	16.02~18.29
$\overline{x}$	302~386	12.61~14.30	300~370	14.16~15.76	286~364	16.44~18.52

. During the experiment, the wind velocity was adjusted to different levels to achieve a relatively stable condition for all three size ranges of clay-heavy clods and to be located at 1/3 of the conical observation tube. The frequency converter is 38 Hz for smaller clods, 42 Hz for side-by-side clods, and 48 Hz for larger clods.

As can be seen from Table 3, the suspension velocity of smaller clods ranged from 12.61 to 14.30 m/s, side-by-side clods from 14.16 to 15.76 m/s, and larger clods from 16.44 to 18.52 m/s. It is clear that the magnitude of suspension velocity is directly related to the mass of the clods. The one-sample K-S test yielded an asymptotic significance level p = 0.93 > 0.05, meaning that the sample data obeyed a normal distribution. The chi-square test yielded a significance level based on the mean p = 0.29 > 0.05, i.e., the variance is chi-square. A one-way ANOVA test showed that the significance level of the effect of the clods on the suspension velocity was less than 0.001 for all three size intervals, reaching a highly significant level. The level of significant difference in suspension velocity between the smaller clods and the side-by-side clods was 0.001, and the level of significant difference between the smaller clods and the larger clods, and the side-by-side clods and the larger clods was less than 0.001, with significant differences in suspension velocity between the three sizes. From the box line comparison plot in Figure 7b, it can be seen that the up and down range of the clods in the wind field for the three size ranges tends to be essentially the same. Therefore, it is considered that the mass of the clods is positively correlated with the suspension velocity values.

4.1.3. Suspension Velocity of Different Shapes of Clods

Smaller clods were taken for the experiment, with a water content of 20.4%. The values of suspension velocity for different shapes of clods are shown in

Table 5. Suspension velocity of peanuts of different shapes of clods.

Test Number	Lumpy Clods		Strip of Clods
	Average Suspension Height Range/mm	Suspension Velocity/m·s−1	Average Suspension Height Range/mm	Suspension Velocity/m·s−1
1	221~329	13.73~16.29	265~350	13.30~15.16
2	223~339	13.52~16.23	282~385	12.63~14.76
3	274~341	13.48~14.95	275~390	12.54~14.93
4	268~352	13.26~15.09	285~355	13.20~14.69
5	282~374	12.84~14.76	280~395	12.45~14.81
6	286~381	12.71~14.67	294~410	12.18~14.49
7	308~389	12.56~14.18	274~386	12.61~14.95
8	281~378	12.76~14.78	292~385	12.63~12.53
9	289~396	12.43~14.60	276~389	12.67~14.76
10	299~403	12.31~14.37	296~374	12.84~14.44
$\overline{x}$	273~368	12.95~14.97	282~383	12.71~14.75

. The frequency converter is 38 Hz.

The data show that the suspension velocities of 12.95~14.97 m/s for lumpy clods and 12.71~14.75 m/s for strip of clods are not very different. As can be seen from the box line comparison plot in Figure 7c, the median line of suspension velocity for the two shapes of clods is essentially at the same level. In practical trials, it was found that the suspension height range of the clods was significantly lower than that of the strips, similar to the suspension of single and double kernel peanuts. The reason for this is that the windward area of the strips in the vertical airflow is large and stable (this phenomenon is also illustrated by the fact that the value of the suspension velocity of the strips in the box line diagram is smaller than that of the lumpy), but the windward area decreases abruptly when the airflow force is encountered at both ends during the turning process, resulting in a greater height difference between the up and down floating of the strips in the conical observation tube than that of the lumpy. The independent samples T-test yielded a significance level of 0.165, which is greater than 0.05. It is therefore concluded that the shape of the clods has less influence on their suspension velocity values, given the same water content and mass size.

4.1.4. Suspension Velocity of Clods at Different Water Contents

Smaller, lumpy clods were taken for the experiment. The suspension velocity values of clods with different water contents are shown in

Table 6. Suspension velocity of clods at different water contents.

Test Number	17.6%		20.4%		23.9%
	Average Suspension Height Range/mm	Suspension Velocity/m·s−1	Average Suspension Height Range/mm	Suspension Velocity/m·s−1	Average Suspension Height Range/mm	Suspension Velocity/m·s−1
1	278~336	13.58~14.86	277~366	12.99~14.88	302~396	12.43~14.31
2	326~411	12.17~13.79	313~412	12.15~14.07	288~369	12.93~14.62
3	307~389	12.56~14.20	254~318	13.96~15.43	297~394	12.47~14.42
4	317~375	12.82~13.98	282~346	13.38~14.76	287~376	12.80~14.65
5	307~406	12.25~14.20	276~335	13.60~14.90	290~364	13.03~14.58
6	298~372	12.87~14.40	356~450	11.52~13.18	312~391	12.52~14.09
7	310~394	12.47~14.13	384~436	11.74~12.65	294~406	12.25~14.49
8	322~460	11.36~13.87	314~396	12.43~14.05	291~374	12.84~14.55
9	297~415	12.10~14.92	288~410	12.18~14.62	286~367	12.97~14.67
10	307~399	12.38~14.20	302~386	12.61~14.30	294~386	12.61~14.49
$\overline{x}$	306~429	11.86~14.22	278~390	12.54~14.86	293~422	11.98~14.51

, where the frequency converter frequency is 38 Hz.

As can be seen from Table 6, the suspension velocity ranged from 12.38 to 14.20 m/s at 17.6% water content, 12.61 to 14.30 m/s at 20.4%, and 12.61 to 14.49 m/s at 23.9%. As can be seen from the box line comparison plot in Figure 7d, the suspension velocity values and median line for the three water content clods show a slight increasing trend. The one-sample K-S test yielded an asymptotic significance p = 0.165 > 0.05, meaning that the sample data obeyed a normal distribution. The chi-square test yielded a significance based on the mean p = 0.519 > 0.05, i.e., the variance is chi-square. A one-way ANOVA test yielded a significance level of 0.005, which is less than 0.05, suggesting that water content has a significant effect on suspension velocity. It is therefore believed that the soil water content affects its suspension velocity value, i.e., the suspension velocity value increases slowly with increasing water content.

In the experiments, it was found that when the water content reached 23.9%, the soil sample was very easy to adhere to the inner wall of the conical observation tube, which greatly increased the difficulty of the measurement. This is because too much water content in the soil particles will form a water film, which has a hydrogel effect and will make the adhesion between the soil and the tube wall increase [23]. Therefore, harvesting should be carried out as much as possible when the soil water content is below 20%. This not only reduces the difficulty of the operation but also ensures the effectiveness of soil removal and reduces the rate of contamination.

4.2. Numerical Simulation Results and Analysis

4.2.1. Analysis of Simulation Results

Once the calculations were completed, the flow field data calculated by Fluent 2022R1 software and the particle data calculated by EDEM 2021.2 software were imported into the post-processing software EnSight 2022R1 for coupled gas-solid flow analysis when each material was in steady suspension in the conical observation tube [24]. The suspension position and velocity of each material in stable suspension in the conical observation tube and the corresponding velocity distribution of the flow field are shown in Figure 8.

As can be seen from Figure 8, the particles within the flow field are blue in color, which indicates that they are all in a stable suspension. The corresponding airflow velocities were obtained from the coordinates of the particles’ positions within the flow field. Suspension velocities were 8.52~9.43 m/s for single-kernel peanuts and 7.74~9.20 m/s for double-kernel peanuts. Under 20.4% water content and lumpy conditions, suspension velocity was 12.18~14.62 m/s for smaller clods, 14.30~16.23 m/s for side-by-side clods, and 16.99~19.80 m/s for larger clods; under 20.4% water content and smaller clods conditions, suspension velocity was 12.18~14.62 m/s for lumpy clods and 12.67~14.76 m/s for strips; under the conditions of lumpy and smaller clods, the suspension velocity was 11.36~13.87 m/s at 17.6% water content, 12.18~14.62 m/s at 20.4%, and 12.52~14.09 m/s at 23.9%.

A comparison of the simulation experiment with the actual experimental results is shown in Figure 9. As the suspension velocity of each material is a range value, the relative error is calculated by taking its central value. The calculation formula is:

$$ϵ=\frac{v_s-v_e}{v_e}\times 100\%$$

(6)

where, $v_s$ is the center value of the simulated experiment levitation velocity, m/s; $v_e$ is the center of the actual experiments’ levitation velocity, and m/s; $ϵ$ is the relative error.

The relative errors between the simulation and the actual experimental results were calculated to be 1.2%, 4.1%, 0.4%, 2.0%, 4.4%, 0.4%, 5.1%, 5.4%, 0.4%, and 1.9% for single-kernel peanuts, double-kernel peanuts, smaller clods, side-by-side clods, larger clods, lumpy clods, strip of clods, and a water content of 17.6%, 20.4%, and 23.9% respectively. The simulation results are less inaccurate than the actual experimental results, and it can be concluded that the simulation based on the coupled gas-solid two-phase flow method is highly accurate in calculating the suspension velocity of peanuts and clay-heavy clods.

4.2.2. Validation of the Model

In order to further determine the accuracy of the coupled model simulation to calculate the suspension velocity of peanut pods and clay-heavy clods, the material suspension velocity value was used as the evaluation index by increasing the air flow rate at the entrance of the flow field so that the peanut pods and clay-heavy clods reached a stable suspension state at different height positions in the conical observation tube and comparing whether the suspension velocity at this time is consistent with the actual experiment and the previously simulated suspension velocity value. The validation results are shown in Figure 10.

As can be seen from Figure 10, the suspension height of single and double kernel peanuts and clods of different mass size, shape, and water content in the conical observation tube increased with increasing inlet air flow rate, but their suspension velocity values remained stable within the range of suspension velocities measured in the actual experiment. It can therefore be concluded that the peanuts and clay-heavy colds suspension velocity measurement process calculated based on the coupled CFD-DEM model simulations is reliable and the gas-solid two-phase flow model has been validated.

5. Discussion

The suspension speed is the central basis of the combination of a fan and vibrating sieve and is essential for the effective operation of the cleaning and sorting device on peanut combine harvesters. Zhichao Hu et al. [25] presented the working principle and structural design of a 4LH2-type, half-feed, and self-propelled peanut combine harvester scavenging system. Lianxing Gao et al. [26] proposed the principle of combined double suction air outlet and vibrating sieve type of cleaning and studied the effect of vibrating frequency, vibrating sieve, height of suction air outlet, and speed of the fan on the rate of cleaning loss and trash content of dialogue peanut through cleaning performance tests. Yifei Wang [13] measured the suspension velocity of peanut threshing material, including pods, long stalks, short stalks, stalks, and miscellaneous leaves, and used coupled CFD-DEM to simulate the material transport process and flow field simulation for the cleaning and sorting device of the peanut combine harvester. Previous research has primarily focused on mechanism design, structure optimization, and material testing for sandy loam soil peanut cleaning machinery, with little attention given to peanut cleaning machinery designed for hilly, mountainous clay-heavy land. Further research is needed in this area to develop effective and efficient cleaning machinery for this type of soil. Therefore, the research in this paper has certain significance in filling the gap in the industry.

In measurements and simulation experiments with different kernel peanut pods and different-shaped clods of clay-heavy soil, we found that there were significant differences in the suspension state of the differently shaped materials in the suspension wind tunnel. The floating range of double- kernel peanut pods and strips of clods within the wind field is significantly greater than that of single-kernel peanut pods and lump clods, which is directly related to the windward area of the different shapes of material within the wind field. However, there was no significant difference in the suspension velocity values for the different shapes of material when a steady-state suspension was reached. As the proportion of double-kernel peanut pods is much greater than that of single-kernel peanut pods at harvest, this suspension characteristic may be more favorable for cleaning and sorting peanuts grown on clay-heavy soils at harvest.

In the measurement experiments with clods of different water content, we found that the higher the water content, the more significant the adhesion effect of the clods in the conical observation tube. During the measurement experiments with a water content of 23.9%, the difficulty of the experiments increased considerably and had to be interrupted several times because the clods adhered to the wall of the tube. During harvesting, it is common for the soil water content to be close to or even exceed the experimental value, resulting in clods sticking to the machine. To avoid operational difficulties and improve sorting efficiency, it is recommended to harvest when the soil water content is below 20%.

This study is based on existing peanut half-feed combine harvesting machinery to propose cleaning and sorting ideas applicable to peanuts grown on clay-heavy soils and is limited to full-feed combine harvesting machinery. In addition, the final shape of the cleaning and sorting device will include the design and optimization of the vibrating sieve and the grouping of the vibrating sieve with the negative pressure soil separation mechanism, which will be the main issue addressed in our subsequent research.

6. Conclusions

• The suspension velocity of the large number of clay-heavy clods present in the fruit collection box during the combined harvest and Wanhua 17 peanuts was measured by means of a suspension velocity measurement device. The suspension velocities of single and double kernel peanuts were 8.34~9.40 m/s and 8.13~9.51 m/s, respectively. Under 20.4% water content and lumpy conditions, the suspension velocities of smaller clods, side by side clods, and larger clods were 12.61~14.30 m/s, 14.16~15.76 m/s, and 16.44~18.72 m/s, respectively; under 20.4% water content and smaller clods, the suspension velocities of lumpy and strips of clods were 12.61~14.30 m/s, 11.90~14.13 m/s, respectively; under lumpy and smaller clods, the suspension velocity at 17.6%, 20.4%, and 23.9% water content ranged from 12.38 to 14.20 m/s, 12.61 to 14.30 m/s, and 12.62 to 14.49 m/s, respectively. Suspension velocity values and suspension characteristics were obtained for different kernelled peanut pods and clay-heavy clods under different conditions.
• Numerical simulations of the experimental process of peanuts and clay-heavy clods in a suspension velocity measurement device based on CFD-DEM theory were carried out, and the model was validated. The coupling calculations gave suspension velocities of 8.52~9.43 m/s for single-kernel peanuts and 7.74~9.20 m/s for double-kernel peanuts; under 20.4% water content and lumpy conditions, suspension velocity was 12.18~14.62 m/s for smaller clods, 14.30~16.23 m/s for side-by-side clods, and 16.99 ~19.80 m/s for larger clods; under 20.4% water content and smaller clod conditions, suspension velocity was 12.18~14.62 m/s for lumpy clods and 12.67~14.76 m/s for strips; under the conditions of lumpy and smaller clods, the suspension velocity was 11.36~13.87 m/s at 17.6% water content, 12.18~14.62 m/s at 20.4%, and 12.52~14.09 m/s at 23.9%; The simulation results for the suspension velocity values of each material are within the ideal range of error from the experimental results.
• The feasibility of the idea of fruit-soil separation through suspension velocity was verified by comparing and analyzing the difference in suspension velocity between peanuts and clay-heavy clods of various sizes, providing a basis for subsequent research.