Design and Testing of an Impurity Removal Device in a Stripper-and-Stick Cleaner for Machine-Harvested Long-Staple Cotton

Abstract

This study aims to address the low cleaning efficiency (73.12%) and high cotton loss rate (12.50%) of existing stripper-and-stick cleaners when processing machine-harvested long-staple cotton. Based on the material characteristics of Xinjiang long-staple cotton, a novel impurity removal device was designed specifically for its cleaning. A mechanical model of cotton–impurity interaction was established to guide the design of U-shaped saw cylinder parameters. Static analyses conducted via Ansys Workbench revealed that both straight and arc-shaped sawteeth exhibited negligible deformations (≤4.73 × 10⁻⁴ mm) and von Mises stresses (≤9.52 MPa), meeting the practical operational requirements. Through orthogonal experiments (sawtooth type, linear velocity, and feed speed), the optimal operating parameters were determined as arc-shaped sawteeth, 7.5 m/s linear velocity, and 11 t/h feed speed. Prototype testing demonstrated that under these parameters, the impurity removal device achieved an average cleaning efficiency of 82.42% (a 9.30% improvement over conventional devices) and an average cotton loss rate of 8.34% (a 4.16% reduction compared to conventional devices). This study provides a valuable reference for the design and optimization of stripper-and-stick cleaners specifically tailored for machine-harvested long-staple cotton.

1. Introduction

Long-staple cotton (Gossypium barbadense), renowned for fibers exceeding 33 mm in length, serves as a premium raw material for luxury textiles owing to its superior softness and tensile strength [1,2,3]. However, compared to machine-harvested upland cotton, underdeveloped harvesting equipment and limitations in cotton cultivars result in a 5–10% higher impurity content in machine-harvested long-staple cotton [4,5,6]. During the roller ginning process, impurities such as boll husks can contaminate cotton fibers, leading to uneven coloration and reduced quality of lint [7]. Therefore, effective impurity removal prior to ginning is critical to ensuring high-quality lint production. Currently, stripper-and-stick cleaners encounter challenges such as low efficiency in removing impurities and excessive cotton loss when processing machine-harvested long-staple cotton.

Impurity removal devices serve as the core component of stripper-and-stick cleaners, with their structural design directly impacting the overall cleaning quality of the machinery. Extensive research has been conducted globally on optimizing these cleaners. To address the improper cleaning of seed cotton, Shi Y. et al. [8] installed adjustable flaps and baffles on the machine’s upper casing, enabling the adaptive regulation of cleaning modes based on the impurity content of the seed cotton. Chai X. et al. [9] developed an integrated stripper-and-stick cleaner which combines opening, extraction, and recycling functions to improve impurity removal efficiency. Miao H. et al. [10] designed a comb-type solid steel grid bar to prevent seed cotton from being improperly cleaned. Egamberdiev F et al. [11] proposed a dual-cylinder stripper-and-stick cleaner to enhance cleaning efficiency. Armijo C.B. [12] modified a three-roller stripper-and-stick cleaner by adding wedge-shaped components behind the brushing rollers and installing slat grids at the recycling rollers to mitigate low cleaning efficiency and high cotton loss. However, this modification inadvertently increased new defects such as neps and slivers. Current domestic research primarily focuses on stripper-and-stick cleaner for machine-harvested land cotton, while studies on impurity removal devices optimized for long-staple cotton remain limited.

To address the aforementioned challenges, an impurity removal device for a stripper-and-stick cleaner was designed for machine-harvested long-staple cotton cleaning. Based on a physical-mechanical property analysis of long-staple cotton and the dynamic process modeling of seed cotton cleaning, critical components of the device were designed. Key factors influencing cleaning performance were identified through theoretical analysis. The quasi-horizontal method was used to design the three-factor three-level orthogonal test, and the optimal parameter combination of the device was obtained. Finally, a prototype verification was conducted to provide theoretical support for the stripper-and-stick cleaner designed for machine-harvested long-staple seed cotton.

2. Structure and Working Principle

2.1. Structure

The stripper-and-stick cleaner for machine-harvested long-staple cotton comprises three primary subsystems: an impurity removal device, a cotton brushing device, and a cotton recovery device, as shown in Figure 1. Among these subsystems, the impurity removal device determines the quality of cleaning. It primarily consists of a U-shaped saw cylinder, a stationary brush, and cleaning bars, as shown in Figure 2. The structure and performance parameters of the stripper-and-stick cleaner are presented in

Table 1. Structure and performance parameters of stripper-and-stick cleaner for machine-harvested long-staple cotton.

Parameters	Values
Structure size (length × width × height) (mm × mm × mm)	3200 × 1640 × 2100
Motor output power (kW)	15
Feed rate (t/h)	11~13

.

2.2. Working Principle

Under the gravitational force, seed cotton uniformly falls onto the U-shaped saw cylinder. The sawteeth on the cylinder hook the seed cotton and drive it into high-speed rotation. When the cotton reaches the stationary brush, the brush evenly distributes the cotton across the sawtooth surface, forming a thin cotton layer that ensures firm engagement with the teeth. During this brushing process, the curved stationary brush penetrates the cotton, blocking and stripping off a few boll husks. Boll husks attached to the cotton layer surface separate from the seed cotton under centrifugal force, while tightly adhered boll husks are dislodged through impact with the cleaning bars.

3. Critical Component Design

3.1. Material Characterization of Machine-Harvested Long-Staple Cotton

The material characteristics of machine-harvested long-staple cotton are a critical factor in determining the parameters of the U-shaped saw cylinder. Optimal structural and dynamic configurations enhance cleaning efficiency and minimize cotton loss. This study uses Xinhai 62 long-staple cotton from Xinjiang as the test material, with systematic measurements of weight, cotton layer thickness, and coefficients of friction as detailed in

Table 2. Physical characteristics of machine-picked long-staple cotton and boll husk.

Material	Parameters	Mean Value	Maximum	Minimum
Single-grain long-staple cotton	Weight (g)	0.24	0.26	0.14
	Diameter (mm)	19.0	21.3	17.8
	Cotton layer thickness (mm)	5.7	6.0	5.2
	Coefficient of friction	0.34	0.35	0.33
	Moisture regain (%)	5.08	5.37	4.82
Boll husk	Weight (g)	0.26	0.70	0.24

.

Seed cotton–boll husk separation experiments were conducted using a TA.XT Plus Texture Analyzer (manufactured by Stable Micro Systems, Godalming, UK). Preliminary observations indicated that some boll husks attached to the fiber surface exhibited extremely weak adhesion forces and could be easily dislodged. Therefore, these boll husks were removed by shaking prior to formal testing. Fifty samples of long-staple cotton with residual boll husks were randomly selected as experimental specimens. During testing, a boll husk was secured with a clamp on the SMS Texture Analyzer, while the free end of the fiber was fixed to a movable grip. The machine was activated, and data acquisition performed automatically by the computer system. The loading rate was maintained at 2 mm/s throughout the experiment.

The experimental results showed that the separation force between seed cotton and boll husk was between 0.12 N and 1.25 N, with a mean of 0.35 N. As shown in Figure 3, the load–displacement curve for their separation has three distinct phases:

• Elastic-like phase

Before displacement occurred, most fibers were bent, with only a few under tension. As displacement began, the pre-tensioned fibers detached first, while the bent fibers gradually straightened with increasing displacement, causing a linear increase in load.

2.

Yield-like phase

More fibers straightened and detached, with the number of detached fibers approaching that of newly detaching fibers. The load fluctuated within a narrow range during this transition.

3.

Separation phase

The fibers were essentially in a straightened state. With the continued increase in displacement, fewer fibers underwent relative sliding, resulting in a continuous decrease in load until complete separation, at which point it dropped to zero.

This triphasic pattern demonstrated a notable similarity to the tensile mechanical characteristics of plant materials [13,14].

3.2. U-Shaped Saw Cylinder Design

3.2.1. Saw Cylinder Structure

The U-shaped saw cylinder, as the core working component of the impurity removal device, consists of a U-shaped saw cylinder shaft, sawteeth, flanges, and a cladding surface, as shown in Figure 4. The flanges are sleeved onto the U-shaped saw cylinder shaft with key fixation. A cladding surface, fabricated from thin steel plates, is circumferentially mounted on the outer periphery of the flanges, forming the cylinder. The sawteeth are secured to the cylinder via screw fasteners.

The U-shaped saw cylinder has a diameter 1.5–2 times greater than that of conventional seed cotton cleaning machines. The adoption of this enlarged diameter confers dual engineering advantages [15]:

• The reduced rotational speed extends the service life of the U-shaped saw cylinder shaft and bearings while improving system stability;
• The increased quantity of cleaning bars expands the impurity discharge zone area, elevating the collision probability between boll husks and cleaning bars, thereby boosting boll husk removal efficiency.

Therefore, the U-shaped saw cylinder designed in this study has a radius of 0.35 m.

3.2.2. Saw Cylinder Linear Velocity

The separation efficiency between seed cotton and boll husks is fundamentally determined by the linear velocity of the U-shaped saw cylinder. A force analysis of the boll husks provides a theoretical basis for determining the optimal rotational speed of the cylinder. The separation process can be categorized into two distinct scenarios: (1) scenarios without collision between boll husks and cleaning bars, and (2) scenarios with collision.

In non-collision scenarios, boll husks are subjected to four mechanical loads during U-shaped saw cylinder operation: centrifugal force generated by rotational motion, fiber tensile force, aerodynamic drag, and gravitational force. The volume of the boll husks is relatively small, and the centrifugal force, which dominates at 25–35 times the boll husks’ self-weight [16], renders both aerodynamic drag and gravitational force negligible. The force distribution is shown in Figure 5a. To achieve the separation of seed cotton and boll husks, the following conditions must be satisfied:

$$F_1>T_1$$

(1)

$$F_1=m_1{\displaystyle \frac{v^2}{r_1}}$$

(2)

where F₁ is the centrifugal force generated by the rotational motion, N; T₁ is the fiber tensile force, N; m₁ is the boll husks’ weight, kg; v is the saw cylinder’s linear velocity, m/s; and r₁ is the saw cylinder’s radius, m.

The experimental measurements indicated an average detachment force of 0.35 N between seed cotton and boll husks. Based on calculations from Equations (1) and (2), if centrifugal force alone is used to remove the boll husks, the linear velocity of the U-shaped saw cylinder must be at least 21.7 m/s. However, excessive rotational speed causes two critical issues: (1) seed cotton not firmly gripped by the sawteeth detaches, increasing fiber loss; and (2) accelerated wear on the U-shaped saw cylinder shaft and bearings reduces their service life. To solve these problems, a series of cleaning bars are arranged around the U-shaped saw cylinder. By means of collision with these bars, the separation efficiency between seed cotton and boll husks is enhanced, as shown in Figure 5b.

To achieve effective cotton–impurity separation, the following critical conditions must be satisfied:

$$F_1+I_1>T_1$$

(3)

$$I_1=m_1{\displaystyle \frac{\mathsf{\Delta }v}{\mathsf{\Delta }t}}$$

(4)

where I₁ is the collision force between the boll husks and cleaning bars, N; Δv is the velocity variation amplitude of the boll husks, m/s; and Δt is the collision duration, s.

The linear velocity of the U-shaped saw cylinder can be reduced to 5–10 m/s following the circumferential integration of cleaning bars [17].

3.2.3. Sawtooth Design

The sawteeth are classified into two types: straight and arc-shaped. Figure 6 illustrates the structural schematic of the straight sawteeth. The working angle (α) directly influences the cotton-grabbing capacity: a larger α enhances the gripping force but may affect the brushing effect. The tip angle (φ) governs sawtooth strength, with higher φ values improving durability; empirical data from prior studies established φ as 27°. Tooth spacing (l) exhibits a trade-off because increasing l increases the fiber load capacity while reducing the holding force. Similarly, tip depth (h) must be matched to the cotton fiber thickness to balance load capacity and structural integrity [18,19,20].

To determine the specific structural parameters of the sawteeth, a force analysis must be conducted on the seed cotton during motion. During sawtooth-driven rotation, the seed cotton is subjected to four principal forces: a centrifugal force induced by high-speed rotation, a surface friction force from serration contact, an aerodynamic drag force, and a gravitational force. Since the centrifugal force during operation is 25–35 times the seed cotton’s weight, the gravitational force can be neglected. The force distribution acting on the seed cotton is shown in Figure 7. To prevent the premature detachment of the seed cotton, the following conditions must be satisfied:

$$F_2=F_k+f$$

(5)

Including:

$$\left\{\begin{array}{c}{F}_2=m_2{\displaystyle \frac{v^2}{r_1}}\\ F_k={\displaystyle \frac{1}{2}}C\rho S{v}^2\\ f=\mu (F_2\sin \alpha +F_k\cos \alpha )\end{array}\right.$$

(6)

where F₂ is the centrifugal force acting on seed cotton, N; F_k is the aerodynamic drag force, N; f is the frictional force; m₂ is the weight of seed cotton, m₂ = 0.24 g; C is the drag coefficient, C = 0.44 [21]; ρ is the air density, ρ = 1.205 kg/m³; S is the windward area of seed cotton, S = 1.45 × 10⁻³ m²; μ is the friction coefficient, μ = 0.34; and α is the sawtooth working angle, (°).

Based on Equation (5) and in conjunction with the force analysis of seed cotton motion, the following is derived:

$$F_2\mathrm{c}\mathrm{o}\mathrm{s}\alpha =F_k\mathrm{s}\mathrm{i}\mathrm{n}\alpha +\mu (F_2\mathrm{s}\mathrm{i}\mathrm{n}\alpha +F_k\mathrm{c}\mathrm{o}\mathrm{s}\alpha )$$

(7)

Substituting Equation (6) into Equation (7) yields Equation (8).

$$\mathrm{t}\mathrm{a}\mathrm{n}\alpha ={\displaystyle \frac{\frac{m_2{v}^2}{r_1}-\frac{1}{2}\mu C\rho S{v}^2}{\frac{1}{2}C\rho S{v}^2+\mu \frac{m_2{v}^2}{r_1}}}$$

(8)

Let $K={\displaystyle \frac{1}{2}}C\rho S$, simplify Equation (8), and determine the mathematical expression for the working angle α:

$$\alpha =\mathrm{a}\mathrm{r}\mathrm{c}\mathrm{t}\mathrm{a}\mathrm{n}{\displaystyle \frac{m_2-K\mu r_1}{K{r}_1+\mu m_2}}$$

(9)

After calculation and rounding, the working angle α is determined to be 42°.

In the ABC triangle of Figure 6, the mathematical expression for the sawtooth pitch l is determined by combining the area formula of a triangle with the laws of sines and cosines:

$$l={\displaystyle \frac{h\mathrm{s}\mathrm{i}\mathrm{n}\phi }{\mathrm{c}\mathrm{o}\mathrm{s}(\phi +\alpha )\mathrm{c}\mathrm{o}\mathrm{s}\alpha }}$$

(10)

where h is the tooth tip depth, designed to match the cotton layer thickness with h = 5.7 mm; φ denotes the tooth tip angle, which is designed to meet strength requirements and is preliminarily set to 27°; and the calculated sawtooth pitch l is 9.8 mm.

In summary, the specific structural parameters of the sawteeth are listed in

Table 3. Parameters of straight sawtooth structure.

Radius of Sawteeth (m)	Tooth Tip Depth (mm)	Operating Angle (°)	Sharpness Angle (°)	Pitch (mm)
0.35	5.7	42	27	9.8

.

In Figure 6, l_bc denotes the contact length between the straight sawteeth and the fibers. The longer l_bc is, the greater the number of fibers in contact with the sawteeth. This increase leads to stronger friction forces between the fibers and the sawteeth, thereby making the cotton less prone to premature detachment. In the ABC triangle, the mathematical expression for l_bc is derived as follows:

$$l_{bc}={\displaystyle \frac{l\mathrm{s}\mathrm{i}\mathrm{n}\delta }{\mathrm{c}\mathrm{o}\mathrm{s}(\alpha +\delta )}}$$

(11)

where l_bc is the contact length between the straight sawteeth and the fibers, mm, and δ is the angle between the relative velocity of seed cotton and the linear velocity of the sawteeth. Theoretically, the sum of α, φ, and δ equals 90°; however, due to the fillet at the tooth root, δ is set to 16°.

The schematic diagram of the arc-shaped sawtooth structure is shown in Figure 8. Based on the arc length formula, the contact length $l_{\widehat{bd}}$ between the arc-shaped sawteeth and the fibers is determined as follows:

$$l_{\widehat{bd}}={\displaystyle \frac{\pi R}{{180}^{°}}}\theta$$

(12)

where $l_{\widehat{bd}}$ is the contact length between the arc-sawteeth and the fibers, mm; R is the arc radius, which is designed to be 10 mm in this study; and θ is the central angle of the arc, (°).

In the ABD triangle, the following relationship is derived using the law of sines:

$$l_{bd}=l{\displaystyle \frac{\mathrm{s}\mathrm{i}\mathrm{n}\delta }{\mathrm{c}\mathrm{o}\mathrm{s}(\epsilon +\alpha +\delta )}}$$

(13)

where l_bd is the distance between b and d, and ε is the newly introduced working angle, °.

In the OBD triangle, the following relationship is derived using the law of cosines:

$$\mathrm{c}\mathrm{o}\mathrm{s}\theta ={\displaystyle \frac{2R^2{\mathrm{c}\mathrm{o}\mathrm{s}}^2(\epsilon +\alpha +\delta )-l^2{\mathrm{s}\mathrm{i}\mathrm{n}}^2\delta }{2R^2{\mathrm{c}\mathrm{o}\mathrm{s}}^2(\epsilon +\alpha +\delta )}}$$

(14)

Therefore, the mathematical expression for the contact length $l_{\widehat{bd}}$ of the arc-shaped sawteeth is:

$$l_{\widehat{bd}}={\displaystyle \frac{\pi R}{{180}^{°}}}\mathrm{a}\mathrm{r}\mathrm{c}\mathrm{c}\mathrm{o}\mathrm{s}{\displaystyle \frac{2R^2{\mathrm{c}\mathrm{o}\mathrm{s}}^2(\epsilon +\alpha +\delta )-l^2{\mathrm{s}\mathrm{i}\mathrm{n}}^2\delta }{2R^2{\mathrm{c}\mathrm{o}\mathrm{s}}^2(\epsilon +\alpha +\delta )}}$$

(15)

Compared to straight sawteeth, an arc-shaped sawtooth introduces an additional contact length l′:

$$l^{\prime }=l_{\widehat{bd}}-l_{bc}$$

(16)

Figure 9 illustrates the relationship between the newly introduced working angle ε and the additional contact length l′ of the arc-shaped sawteeth under the following conditions: working angle α = 42°; arc radius R = 10 mm; teeth pitch l = 9.8 mm; and δ = 16°. The curve demonstrates that l′ increases with ε. However, due to constraints imposed by the teeth geometry and strength limitations, ε cannot be infinitely increased. In this study, ε is determined to be 16°, resulting in l′ = 5.19 mm. The structural parameters of the arc-shaped sawteeth are listed in

Table 4. Parameters of arc-shaped sawtooth structure.

Radius of Sawteeth (m)	Tooth Tip Depth (mm)	Operating Angle (°)	Radius of the Arc (mm)	Pitch (mm)
0.35	5.7	58	10	9.8

.

3.3. Static Analysis

3.3.1. Static Analysis of Sawtooth Structure

A structural static analysis of the sawtooth was conducted to verify whether its strength design meets operational requirements and to provide a theoretical basis for design optimization. The structural parameters of the straight and arc-shaped sawteeth are detailed in Table 3 and Table 4, respectively. The sawtooth material is Grade 55 steel, with its material parameters listed in

Table 5. Sawtooth and cotton fiber parameters.

Material	Modulus of Elasticity E (MPa)	Poisson’s Ratio μ	Density (kg/m3)
Grade 55 steel	2.17 × 105	0.27	7.83 × 103
Cotton fiber	2400	0.4	400

.

Meshing

The sawtooth structures were meshed using a tetrahedral method with a 0.5 mm grid size. The straight sawtooth model consisted of 296,272 elements and 449,497 nodes, achieving a maximum mesh quality of 1.00 and an average of 0.81. For the arc-shaped sawtooth variant, the mesh contained 398,802 elements and 597,779 nodes, with a maximum mesh quality of 0.99 and an average of 0.82. These metrics confirm satisfactory mesh performance for both configurations.

To ensure the accuracy of static analysis results, a convergence test was performed. Mesh sizes were set to 1 mm, 0.5 mm, and 0.25 mm, with maximum stress and deformation values recorded at each step. The results showed that the fluctuation amplitude of stress and deformation values was less than 1%, as presented in

Table 6. Convergence test results of finite element analysis of mesh generation.

Sawtooth Type	Mesh Size (mm)	Max Stress (Mpa)	Max Deformation (4mm)
Straight	1	9.504	4.71 × 10−4
	0.5	9.519	4.73 × 10−4
	0.25	9.525	4.74 × 10−4
Arc-shaped	1	8.000	4.16 × 10−4
	0.5	8.018	4.19 × 10−4
	0.25	8.018	4.20 × 10−4

. Therefore, the mesh converges.

Boundary Conditions

The loads acting on the sawteeth primarily arise from the impact forces and pressure exerted by seed cotton and impurities. Therefore, an ultimate load of 30 N is applied to the sawtooth surface [22]. The base of the sawtooth rack is configured with fully fixed constraints. The load and constraint configurations are illustrated in Figure 10a,b.

Results and Analysis

As shown in Figure 10b,c, the maximum deformation for both the straight and arc-shaped sawteeth occurred at the tooth tip, measuring 4.73 × 10⁻⁴ mm and 4.19 × 10⁻⁴ mm, respectively. These values indicate extremely small deformations.

From the analysis presented in Figure 10d,e, the maximum stress for the straight and arc-shaped sawteeth was observed at the tooth root, reaching 9.52 MPa and 8.02 MPa, respectively. These stress values are significantly lower than the yield strength of the structural material (380 MPa), confirming that the design satisfies operational requirements under practical working conditions.

3.3.2. Frictional Simulation Analysis of Fibers

During sawtooth rotation, friction stabilizes the seed cotton, a complex process. To enhance simulation efficiency and accuracy, the model was simplified. Three-dimensional models of the sawteeth and cotton fibers were created using SolidWorks 2022 and saved as “.x_t” or “.x_b” files. The models were then imported into Ansys Workbench for friction analysis. The material parameters for the sawteeth and fibers are provided in Table 5 [23].

Meshing

To improve computational precision and reduce simulation time, the contact regions were refined with a mesh size of 0.1 mm, compared to the global mesh size of 0.5 mm. For the straight sawteeth and fibers, the total number of elements generated was 296,832, with 453,750 nodes. The maximum mesh quality was 1, and the average was 0.93. For the arc-shaped sawteeth and fibers, the total number of elements was 399,362, with 602,032 nodes. The maximum mesh quality was 1, and the average was 0.90. The mesh quality results were close to 1, indicating high-quality meshes.

To ensure the accuracy of static analysis results, a convergence test was conducted. Mesh sizes were set to 1 mm, 0.5 mm, and 0.25 mm, with frictional force variations recorded at each step. The results indicated that the frictional force variations in the fibers in both straight and curved teeth remained consistent despite mesh size refinement. Therefore, the mesh converges.

Boundary Conditions

Given the significant difference in material stiffness between the fibers and sawteeth, the fibers were modeled as flexible bodies, and the sawteeth as rigid bodies. The geometric interaction was defined as frictional contact, with static and dynamic friction coefficients set to 0.35 and 0.34, respectively. The boundary conditions were configured as follows: a 3N compressive load was applied to the fiber end face, and the sawteeth were assigned a fully fixed constraint. The fiber moved along the sawtooth surface with a maximum displacement of 5 mm. The fiber movement process is illustrated in Figure 11.

Results and Analysis

The friction force variation curve is illustrated in Figure 12. The results reveal that the friction force on the fibers increased linearly as they moved across the tooth surface. Under identical conditions, the friction force experienced by the fibers on the curved sawteeth was greater than that on the straight sawteeth due to the larger contact area between the arc-shaped sawteeth and the fibers. This finding aligns with theoretical analyses.

4. Experiments and Results

4.1. Materials and Equipment

The experimental samples were selected from machine-harvested long-staple cotton Xinhai 62 in the Aksu region of Xinjiang. Prior to the experiments, the samples underwent pretreatment using equipment such as a foreign fiber separator for seed cotton and an inclined seed cotton cleaner. The moisture content of all test samples, controlled within the range of 4.5–5.5%, conformed to the National Standard GB/T 41228-2021 General Technical Requirements for Cotton Processing and Conditioning [24]. The test equipment is illustrated in Figure 13. The sawteeth were mounted on a U-shaped saw cylinder of the cleaning machine.

4.2. Experimental Design and Results

An orthogonal experimental design method was adopted to formulate the experimental plan. This study involved three factors. Since these factors differed in the number of levels, the quasi-level method was applied to modify the orthogonal experimental design. The L⁹(3⁴) orthogonal array was selected [25,26]. The experiments were conducted in accordance with the National Standard GB/T 19818-2005 Seed Cotton Cleaner [27]. The cleaning efficiency η_z and cotton loss rate η_L were measured and calculated as follows:

$${\eta }_z={\displaystyle \frac{Q_{before}-Q_{after}}{Q_{before}}}\times 100\%$$

(17)

$${\eta }_L={\displaystyle \frac{W}{G}}\times 100\%$$

(18)

where Q_before is the impurity mass in seed cotton before cleaning, g; Q_after is the impurity mass after cleaning, g; W is the mass of seed cotton mixed within impurities, kg; and G is the total mass of processed seed cotton, kg.

The levels of the experimental factors are listed in

Table 7. The levels of the experimental factors.

Level	Sawtooth Type A	Linear Velocity B (m/s)	Feed Speed C (t/h)
1	Straight	5	11
2	Arc-shaped	7.5	12
3		10	13

.

The experimental design and corresponding results are presented in

Table 8. Experimental design and corresponding results.

NO.	Factors and Levels				Evaluation Indicators
	A	B	C	Empty Columns 1	Cleaning Efficiency y1 (%)	Cotton Loss Rate y2 (%)
1	1	1	1	1	77.60	10.04
2	1	2	3	2	80.65	10.57
3	1	3	2	3	80.48	12.93
4	2	1	3	3	80.14	7.79
5	2	2	2	1	82.36	8.12
6	2	3	1	2	85.91	13.84
7	2	1	2	2	80.14	6.96
8	2	2	1	3	82.69	11.48
9	2	3	3	1	85.42	12.09

¹ To reduce experimental random errors, an additional blank column was incorporated into the orthogonal design table.

.

4.3. Analysis of Results

An analysis of variance (ANOVA) was performed on the experimental results using SPSS 27, and the results are presented in

Table 9. Variance analysis results.

Source	Cleaning Efficiency					Cotton Loss Rate
	Sum of Squares	Freedom	Mean Square	F Values	p Values	Sum of Squares	Freedom	Mean Square	F Values	p Values
Model	55.295a	5	11.059	17.006	0.021 *	45.514b	5	9.103	36.912	0.007 **
A	20.48	1	20.48	31.494	0.011 *	2.569	1	2.569	10.417	0.048 *
B	32.503	2	16.252	24.991	0.013 *	33.603	2	16.801	68.13	0.003 **
C	2.311	2	1.156	1.777	0.31	9.343	2	4.671	18.942	0.02 *
Error	1.951	3	0.65			0.74	3	0.247
Total	60,145.962	9				1024.276	9

** is particularly significant at the 0.01 level; * is significant at the 0.05 level.

. The sawtooth type exhibited a significant effect on both cleaning efficiency and cotton loss rate. The linear velocity of the sawteeth significantly influenced cleaning efficiency, while its impact on cotton loss rate was highly significant. In contrast, the cotton feed speed showed no significant effect on cleaning efficiency but demonstrated a significant effect on cotton loss rate.

Due to conflicting optimization results from different indicators, the comprehensive scoring method was employed to evaluate the metrics, with the highest composite score determining the final optimized solution [28,29]. As shown in

Table 10. Comprehensive performance scores based on a composite scoring method.

NO.	Membership		Comprehensive Score Cs	Rank
	Cleaning Efficiency	Cotton Loss Rate
1	0.00	0.58	0.301	9
2	0.62	0.50	0.557	7
3	0.59	0.14	0.354	8
4	0.52	0.92	0.727	5
5	0.97	0.87	0.918	1
6	1.69	0.00	0.813	3
7	0.52	1.05	0.793	4
8	1.04	0.36	0.685	6
9	1.59	0.27	0.904	2

, the combination A₂B₂C₂ achieved the highest comprehensive score of 0.918, representing the optimal level.

4.4. Prototype Testing and Validation

Based on the optimization test results, namely, arc-shaped sawteeth, linear velocity of 7.5 m/s, and feed speed of 11 t/h, prototype testing and validation were conducted, considering practical operability. The tests were repeated five times, yielding a cleaning efficiency of 82.42% and a cotton loss rate of 8.34% for the impurity removal device with the arc-shaped sawteeth. The cleaning performance comparison is illustrated in Figure 14. When the arc-shaped sawteeth were replaced with straight sawteeth under identical conditions, the impurity removal device exhibited a cleaning efficiency of 79.95% and a cotton loss rate of 10.35%. Compared with the straight sawteeth, the arc-shaped sawteeth achieved a 2.47% improvement in cleaning efficiency and a 2.01% reduction in cotton loss rate, as shown in

Table 11. Test verification results.

No.	Impurity Removal Device with Arc-Shaped Sawteeth		Impurity Removal Device with Straight Sawteeth
	Cleaning Efficiency (%)	Cotton Loss Rate (%)	Cleaning Efficiency (%)	Cotton Loss Rate (%)
1	83.24	8.83	79.62	10.33
2	81.95	7.92	79.56	10.65
3	82.32	8.14	80.23	10.46
4	81.73	8.56	79.88	9.97
5	82.86	8.27	80.47	10.38
Average value	82.42	8.34	79.95	10.35

.

5. Discussion

The research findings have significant implications for enhancing the machine-harvested long-staple cotton processing industry. The novel impurity removal device not only improves cleaning efficiency but also reduces cotton loss during processing, addressing the limitations of existing stripper-and-stick cleaner for machine-harvested long-staple cotton.

One innovation of this study is the design of a large-diameter U-shaped saw cylinder, which enhances cleaning efficiency while improving the service life of shafts and bearings. Another innovation lies in the development of arc-shaped sawteeth. Compared with straight sawteeth, these arc-shaped sawteeth increase contact area with cotton fibers, thereby enhancing friction forces and reducing the likelihood of seed cotton dropping during cleaning. Additionally, the puncture capacity of the arc-shaped sawteeth is smaller, which can reduce the grasping of impurities and improve the efficiency of impurity removal. Experimental results demonstrate that compared with straight sawteeth, the cleaning efficiency of arc-shaped sawteeth increases by 2.47%, while the cotton loss rate decreases by 2.01%.

Static analysis reveals that deformation for both tooth types remains minimal, with stress variations within acceptable ranges. These results satisfy operational requirements and can reduce maintenance costs.

Orthogonal experiments identified optimal operating parameters for the impurity removal device: arc-shaped sawteeth, a linear velocity of 7.5 m/s, and a feed speed of 11 t/h. These parameters balance maximum cleaning efficiency with minimal cotton loss, ensuring high-quality production and cost-effectiveness.

The findings of this study provide theoretical support for designing stripper-and-stick cleaners for machine-harvested long-staple cotton. Future research could explore intelligent methods to achieve more precise and adaptive impurity removal during real-time processing.

6. Conclusions

In conclusion, this study successfully designed and tested a novel impurity removal device for a stripper-and-stick cleaner used in machine-harvested long-staple cotton cleaning. The key research findings are as follows:

• The newly designed impurity removal device with arc-shaped sawteeth achieved an average cleaning efficiency of 82.42% and a cotton loss rate of 8.34% under optimal operating parameters (arc-shaped sawteeth, 7.5 m/s linear velocity, and 11 t/h feed speed). This represents a 9.30% improvement in cleaning efficiency and a 4.16% reduction in cotton loss rate compared to conventional devices.
• Under identical operating parameters, the arc-shaped sawteeth outperformed the straight teeth, achieving a 2.47% higher cleaning efficiency and a 2.01% lower cotton loss rate. This exceptional performance is due to the increased contact length and enhanced friction forces between the teeth and cotton fibers, ensuring more stable seed cotton rotation and effective impurity removal.
• Static analysis revealed that both straight and arc-shaped sawteeth satisfied design requirements in terms of strain and stress, ensuring the reliability of the sawteeth during operation.
• This research comprehensively understands the material characteristics of machine-harvested long-staple cotton and their impact on cleaning equipment design. The established cotton–impurity interaction mechanical model and orthogonal experiments provide valuable theoretical and practical references for optimizing stripper-and-stick cleaners.