Design and optimization of micro-textured tool for CFRP cutting based on response surface method

Abstract

Carbon fiber reinforced plastics (CFRP) are widely used in aerospace applications due to their high specific strength and design flexibility. However, significant subsurface damage often occurs during machining, which severely impacts the service life of the parts. In this paper, the influence of tool parameters and texture parameters on subsurface damage is comprehensively considered, and the Placket-Burman experimental design is used to screen the significance of these factors. Four key parameters that have great influence on the subsurface damage depth are determined: tool Angle, edge shape, texture width, and texture depth. Then, the response surface method is used to establish the prediction model of subsurface damage, and the optimal machining parameters to minimize subsurface damage are determined by combining the white whale optimization algorithm and the genetic algorithm. The experimental results show that the sub-surface damage is reduced by about 50% when using the optimized micro-braided tool compared with the traditional tool. These research results provide new ideas and methods for improving the machining quality of CFRP and optimizing tool parameters.

Keywords

Carbon fiber reinforced plastics micro-textured tool microscopic model response surface methodology subsurface damage

Introduction

The use of carbon fiber reinforced plastics (CFRP) is prevalent in aerospace¹ and various other advanced manufacturing² sectors due to their exceptional properties, such as high rigidity and low density. The microstructure of CFRP features a complex multiphase blend of fiber-resin interfaces, while their macroscopic form is characterized by laminar, irregular, and anisotropic properties.^3,4 During machining, the thermoset matrix material tends to break and propagate beneath the machined surface, leading to defects such as delamination, burrs, and subsurface tears. These defects compromise the machining precision and affect critical properties such as fatigue resistance and assembly accuracy, thereby limiting the use of CFRP in high-end applications and low-end scale industries.^5,6 As the demand for lightweight performance increases, suppressing uncut fibers to improve surface quality has become crucial for manufacturing complex carbon fiber composite parts.

Many researchers have employed finite element simulation technology to explore methods for reducing CFRP machining damage, thereby saving time and reducing economic costs. For instance, Alessandro et al.⁷ developed a CFRP orthogonal cutting finite element model using the SPH method, demonstrating that the chip formation process in the simulation closely matches experimental results. Qin Xuda⁸ established a two-dimensional macroscopic orthogonal cutting model using the Hashin-Damage failure criterion and validated it against milling experiments.

Boughdiri et al. developed a three-dimensional cutting model with constitutive relationships and damage criteria for each constituent of GLARE® composite. The model’s cutting force simulations closely matched experimental results, validating its accuracy.⁹

Numerous other researchers have also enhanced CFRP machining performance through finite element simulations.^10,11 To suppress subsurface damage in CFRP, Lu Dong et al.¹² established a macroscopic CFRP cutting finite element simulation model based on the Hashin-Damage failure criterion, finding that subsurface damage is influenced by fiber direction, with the most severe damage occurring at a 90° fiber direction angle. In addition, Redouane Zitoune et al. found through an experimental comparative study that the delamination phenomenon of composite materials is significantly affected by the processing parameters and the composite manufacturing process.¹³

Recent studies have demonstrated that micro-textured tools are the optimal choice for machining hard materials. These tools effectively reduce friction, manage chip removal, and enhance the surface quality of machined components.^14–16 Studies have shown that micro-textured tools can reduce cutting forces and improve the surface quality of machined parts, making them ideal for machining hard materials. Fatima A and Mativenga¹⁷ PT investigated the effect of groove micro-texturing on the cutting process of AISI4140 steel using carbide tools, finding a reduction in friction coefficients by 17% and 18% on the front and rear faces, respectively. Similarly, You et al.¹⁸ successfully reduced the main cutting force in titanium alloys using bionic micro-textured carbide tools. The weave morphology facilitates chip interaction, leading to secondary cutting phenomena.¹⁹ In previous studies, Xie et al.²⁰ conducted dry turning experiments on titanium alloys, comparing conventional, orthogonal weaving, and diagonal weaving tools, and found that weaving tools increased shear angles, reduced burr formation, and improved surface quality. Sun et al.²¹ discovered through experiments on pure iron that weave parameters could alter chip flow patterns from segmented flow with redundant deformation to stable, uniform flow. Other scholars have also noted that micro-textured tools accommodate chips and change chip flow direction.^22–24 Furthermore, You et al. fabricated a bionic micro-textured tool to cut CFRP based on clam surface, and the results also showed that the textured tool significantly improved the cutting performance of CFRP.²⁵ Chen et al.²⁶ used linear weave milling cutters with different directions to mill CFRP, showing that micro-weaving milling improved surface quality compared to conventional milling. Cheng et al.²⁷ found in CFRP orthogonal cutting experiments that weaving tools effectively suppressed subsurface damage and optimized tool weave parameters through two-dimensional simulation and response surface methods.²⁸

Nguyen-Dinh found that the four serrated straight flutes micro-textured cutting tool generated the least number of harmful particles when machining CFRP compared to conventional tools.²⁹ Samsudeensadham et al. found that micro-textured cutting tools outperform standard tools when machining Al-7075/CFRP/Ti-6Al-4V composites. Their unique surface design reduces friction and chip adhesion, improving CFRP surface integrity and extending tool life.³⁰

Although it has been demonstrated that micro-structured tools can reduce subsurface damage, minimize harmful particle dispersion, and improve machining quality when cutting carbon fiber composites, there is a lack of research on the interaction between tool parameters and texture parameters. Establishing the relationship between subsurface damage depth and tool structure parameters, and optimizing these parameters to minimize damage, remain a challenge. Therefore, this paper employs a Plackett-Burman experimental design to identify significant weave and tool parameters, constructs a mathematical model linking key structural parameters to subsurface damage depth, and uses optimization algorithms to determine the optimal micro-weave tool parameters. This research has theoretical significance and practical value for reducing subsurface damage and designing new micro-weave tools.

Simulation and experiment

Since CFRP materials exhibit macroscopic anisotropy and a complex microstructure composed of matrix, fiber, and interface phases, the cutting process is a continuous and dynamic one, making underplane damage difficult to observe. This results in challenges such as poor universality, low efficiency, and high costs. To address these issues, a three-dimensional micro-CFRP orthogonal cutting finite element model has been developed. The accuracy of this model is validated by comparing the predicted cutting forces and subsurface damage with experimental results. The influence of micro-texture structure parameters on subsurface damage is then investigated using finite element simulations. Based on these findings, valuable design recommendations are provided to enhance the performance of micro-textured tools, promoting their use as a sustainable and efficient alternative in the machining industry.

Finite element modeling

Using Abaqus software, a three-dimensional microscopic finite element model of orthogonal cutting of CFRP is established, ignoring the effects of tool deformation and temperature. The fibers were represented as cylinders with a diameter of 5 μm, uniformly distributed within the matrix, with the centerlines of adjacent fibers spaced 8 μm apart. To reduce the computation time, the model dimensions were set to 150 μm × 150 μm × 16 μm, with the left and bottom sides of the workpiece fully fixed. The 90° CFRP model was cut using a micro-texture tool, as illustrated in Figure 1. The cutting edge radius of the tool was 10 μm, with a rake angle (α) of 5° and a clearance angle (β) of 15°. The cutting depth was set to 50 μm, and the cutting speed (v) was 600 mm/min.

In the model, the fiber and matrix are modeled separately, with the tool set as a rigid body. The interfacial phase is simulated using zero-thickness cohesive elements. The material properties for each phase are detailed in Table 1³⁰

Table 1.

Three-dimensional microscopic CFRP material parameters.

Materials	Attribute	Numerical value
Carbon fiber	Tensile strength	$E_{1} = 231 G P a, E_{2} = E_{3} = 15 G P a$
		$υ_{12} = υ_{13} = 0.2, υ_{23} = 0.25$
		$G_{12} = G_{13} = 15 G P a, G_{23} = 7 G P a$
	Drawing strength	$X_{t} = 4620 M P a, Y_{t} = 1500 M P a$
	Compressive strength	$X_{c} = 3960 M P a, Y_{c} = 3340 M P a$
	Shear strength	$S = 1500 M P a$
Matrix	Shear strength	$E_{m} = 3.35 G P a, υ_{m} = 0.35$
	Yield strength	$σ_{y 0}^{m} = 120 M P a, {\bar{ε}}_{f}^{p l} = 0.05$
	J-C parameter	$A = 120 M P a, B = 654.18 M P a,$
	J-C parameter	$n = 0.772, c = 0.124, m = 0.304$
	Fracture energy	$G_{c}^{m} = 0.1 N / m m$
Interphase	Regular tenacity	$t_{n} = 50 M P a$
	Shear strength	$t_{s} = 75 M P a$
	Elastic rigidity	$K = 100000 N / m m^{3}$
	Fracture energy	$G_{I}^{C} = \frac{0.002 N}{m m}$
	Fracture energy	$G_{I I}^{C} = G_{I I I}^{C} = 0.006 N / m m$
	Mixture model parameters	$η = 1.45$

Carbon fiber is defined as a linearly elastic, transversely isotropic material, with failure governed by the maximum principal stress criterion. If the stress in any direction exceeds the strength limit, the element is removed. The damage parameters are detailed in Table 2.³¹

Table 2.

Failure criterion of fiber damage.

Failure mode	Determining failure formula
Longitudinal tensile failure (σ₁₁ ≥ 0)	$d_{t 1}^{f} = \frac{σ_{11}}{X_{t}}$
Longitudinal compression failure (σ₁₁ < 0)	$d_{c 1}^{f} = \| \frac{σ_{11}}{X_{c}} \|$
Transverse tensile failure (σ₂₂ ≥ 0,σ₃₃ ≥ 0 )	$d_{t 2}^{f} = \frac{σ_{22}}{Y_{t}}, d_{t 3}^{f} = \frac{σ_{33}}{Y_{t}}$
Lateral compression failure (σ₂₂ < 0,σ₃₃ < 0)	$d_{c 2}^{f} = \| \frac{σ_{22}}{Y_{c}} \|, d_{c 3}^{f} = \| \frac{σ_{33}}{Y_{c}} \|$
Shear failure (in-plane)	$d_{s 12 \pm}^{f} = \| \frac{τ_{12}}{S} \| (τ_{12} \neq 0)$
Shear failure (out of plane)	$d_{s 23}^{f} = \max {\| \frac{τ_{13}}{S} \|, \| \frac{τ_{23}}{S} \|}$

The CFRP exhibits continuous but anisotropic material properties at the macroscopic level. In the finite element analysis, the fiber orientations are clearly defined, and the bottom and left sides of the workpiece are fully constrained. The workpiece in the model has dimensions of 1 mm × 0.5 mm, and both the tool and the workpiece are meshed using 4-node plane stress elements with reduced integration (CPS4R). Different fiber orientations correspond to different friction coefficients.³² In this study, the Coulomb friction model was adopted, with the friction coefficients for fibers at 0°, 45°, 90°, and 135° set at 0.3, 0.6, 0.8, and 0.6, respectively.

Experimental verification

To validate the precision of the microscopic finite element model, a unidirectional thin plate T800SC manufactured by Toray was chosen for orthogonal cutting experiments encompassing four distinct fiber angles: 0°, 45°, 90°, and 135°. The dimensions of the prepared specimen are 100 mm × 50 mm × 5 mm. In accordance with the product specifications provided by the company, Table 3 presents the material properties while Table 4 displays the mechanical properties.

Table 3.

Material properties of T800S CFRP.

Reinforced material	Carbon fiber volume content	Fiber bundle	Form of laying	Number of plies
Toray T800S	60%	5 μm	One way	36
Toray T800S	60%	24,000	One way	36

Table 4.

Mechanical properties of T800S CFRP.

Mechanical properties	Drawing strength (MPa)	Tensile modulus (GPa)	Breaking elongation (%)	Density (g/cm3)
Toray T800S	5880	294	2.0	1.8

CK6150 CNC lathe was used as the experimental platform. First, the workpiece was fixed on the three-jaw chuck using a special fixture, and the level of the workpiece was adjusted using a spirit level. The three-jaw chuck was then locked to prevent the rotation of the workpiece, allowing the tool to move horizontally to complete the orthogonal cutting. A KISTLER 9257BH three-direction dynamometer was used to collect cutting force data during the cutting process, ensuring that the cutting parameters were consistent with the simulation model. The tool material was YG6 carbide, with a rake angle of 5° and a clearance angle of 15°. A femtosecond laser micromachining platform was used to prepare the micro-textured tool. The experimental setup and the profile of the micro-textured tool are shown in Figure 2.

The experimental cutting parameters are consistent with the simulation cutting parameters. A grooved micro-texturing tool with a microtexture parallel to the main cutting direction was used for orthogonal cutting of CFRP at different fiber angles. Each experiment was repeated three times, and the cutting force data were recorded for each trial. The cutting force measurement results are shown in Figure 3.

As shown in Figure 4, the experimental results and the 3D simulation results (average cutting force) demonstrate a strong consistency. The differences between the simulated and experimental values for the same fiber angles were minimal, thereby verifying the accuracy of the 3D finite element model. Notably, CFRP unidirectional plates with a 90° fiber orientation angle produced the largest cutting force. Consequently, this study focused on orthogonal cutting experiments with CFRP at a 90° fiber orientation angle.

Results and discussion

Plackett-Burman test

Different from conventional tools, micro-textured tools are tools with specific shapes machined on the tool surface by laser. In this paper, the surface profile of the micro-textured tool is shown in Figure 5, where T, W, s, and d are texture parameters representing the distance from the first micro-texture to the cutting edge, texture width, texture spacing and texture depth, respectively. Tool parameters include the fillet radius(R), rake angle(α), and tool clearance (β).

The influence of different structural parameters on the subsurface damage produced by cutting CFRP is different in significance. In this paper, the subsurface damage is defined as the maximum distance (h) from the machining upper surface to the fiber tear. In the stress cloud diagram, 1 indicates complete damage and 0 indicates complete undamaged, as shown in Figure 6.The design of screening experiments (Plackett-Burman, PB) is an effective method to screen out the factors that have a significant impact on the response index from multiple factors.

The Plackett-Burman (PB) experimental design is an effective method for identifying significant factors affecting the response index from among multiple variables.³³ Using the PB experimental design method and Abaqus simulation experiments, the results of the test group and the subplane damage measurements are presented in Table 5.

Table 5.

Table of the Plackett-Burman design.

Test serial number	R (μm)	α (°)	β (°)	T (μm)	W (μm)	s (μm)	d (μm)	Subsurface damage depth (μm)
1	0.020	−5	30	20	30	25	5	45.383
2	0.005	25	30	10	50	15	5	15.080
3	0.020	25	10	20	30	15	5	22.740
4	0.005	−5	30	20	50	15	15	32.000
5	0.020	−5	10	10	50	25	15	43.740
6	0.020	25	10	20	50	15	15	20.740
7	0.005	25	10	10	30	25	15	13.000
8	0.020	25	30	10	50	25	5	36.320
9	0.005	25	30	20	30	25	15	15.080
10	0.020	−5	30	10	30	15	15	40.000
11	0.005	−5	10	20	50	25	5	37.300
12	0.005	−5	10	10	30	15	5	31.700

The significance analysis of the experimental results, conducted using Minitab software, is presented in Table 6. A significance level of p < .005 is considered statistically significant. The analysis reveals the order of influence for each parameter on the surface damage depth as follows: α > R > s > d > β > W > T, with β, W, and T being deemed insignificant.

Table 6.

ANOVA table for the depth of damage under the surface.

Source	Degree freedom (DF)	Sum squares (SS)	Mean square error (MSE)	F	p	Significant
Model	7	1468.72	209.817	32.90	0.002	Prominent
Linear	7	1468.72	209.817	32.90	0.002	Prominent
R	1	349.52	349.521	54.81	0.002	Prominent
α	1	956.99	956.992	150.08	0.000	Prominent
β	1	17.87	17.868	2.80	0.169	Inapparent
T	1	3.63	3.627	0.57	0.493	Prominent
W	1	24.87	24.875	3.90	0.119	Inapparent
s	1	67.99	67.987	10.66	0.031	Prominent
d	1	47.85	47.852	7.50	0.050	Prominent

Response surface experiment (BBD)

Design and results of response surface experiment

The response surface method (RSM) combines experimental design with statistical analysis to construct either polynomial or non-polynomial models that approximate implicit functional relationships. Using RSM, the functional relationship between response objectives and design variables can be determined, allowing for the identification of the optimal combination of design variables. As the number of influencing factors increases, the complexity of the RSM model also increases, which can compromise its accuracy. Therefore, RSM is most suitable for models with four factors and three levels.

In this study, the four most significant influencing factors were selected: rake angle (γ), cutting edge radius (R), texture width (W), and texture depth (d). These factors were used to design a Box-Behnken Design (BBD), supplemented with additional experiments to obtain more comprehensive data. The experimental results are shown in Table 8. Based on the design and practical application of the micro-texturing tool, as well as references from the literature, the high and low levels for each factor were determined and are presented in Table 7.

Table 8.

Box-Behnken experimental design and results.

Test serial number	R (μm)	α (°)	W (μm)	d (μm)	Subsurface damage depth (μm)
1	0.0050	5	40	10	15.4334
2	0.0200	5	40	10	20.3358
3	0.0050	25	40	10	12.0843
4	0.0200	25	40	10	21.8233
5	0.0125	15	30	5	24.4332
6	0.0125	15	50	5	22.4320
7	0.0125	15	30	15	23.0000
8	0.0125	15	50	15	18.4320
9	0.0050	15	40	5	15.0090
10	0.0200	15	40	5	24.8202
11	0.0050	15	40	15	16.5000
12	0.0200	15	40	15	21.3202
13	0.0125	5	30	10	21.3729
14	0.0125	25	30	10	22.3896
15	0.0125	5	50	10	21.3729
16	0.0125	25	50	10	19.8896
17	0.0050	15	30	10	13.5009
18	0.0200	15	30	10	25.3202
19	0.0050	15	50	10	16.5002
20	0.0200	15	50	10	24.3200
21	0.0125	5	40	5	22.3729
22	0.0125	25	40	5	22.3896
23	0.0125	5	40	15	22.0792
24	0.0125	25	40	15	21.3896
25	0.0125	15	40	10	23.4432
26	0.0125	15	40	10	22.9320
27	0.0125	15	40	10	22.9320
28	0.0050	25	30	15	15.5843
29	0.0125	15	40	10	23.4432
30	0.0050	5	50	15	13.4334
31	0.0200	5	30	5	22.3729

Table 7.

Box-Behnken experimental factor level design.

Factor	Symbol	Level
Factor	Symbol	Central point	Low (−1)	High (1)
R (μm)	A	0.0125	0.0050	0.0200
α (°)	B	15	4	25
W (μm)	C	40	30	50
d (μm)	D	10	5	15

Establish and test the prediction model

The response surface analysis of the experimental data was performed using Minitab software. A multivariate regression model was established to describe the relationship between the blade radius (A, μm), tool angle (B, degrees), texture width (C, μm), texture depth (D, μm), and the depth of underplane damage. This was achieved using a polynomial nonlinear fitting method. The resulting regression models for the depth of underplane damage are expressed as regression equations in uncoded units, as follows.

Depth = - 15.8 + 2493 A + 0.598 B + 0.607 C + 1.140 D - 63502 A^{2} - 0.01576 B^{2} - 0.00323 C^{2} + 12.97 AB - 6.32 AC - 29.4 AD - 0.00768 BC - 0.0233 CD

(1)

ANOVA was used to assess the applicability and significance of the model, with the results presented in Table 9. The F-score and p-value were used to evaluate the validity of the regression model and the influence of the combination of factors on the model. A smaller p-value (or a larger F-score) indicates a more significant model, with terms having a p-value less than 0.05 considered significant. The F-score of the model was found to be 26.68, and a p-value close to 0 indicates that the model is highly significant.

Table 9.

ANOVA table for Box-Behnken experimental design.

Source	Symbol	Adj SS	Adj MS	F	p
Model	12	492.697	41.058	26.68	0.000
Linear	4	313.588	78.397	50.95	0.000
A	1	240.240	240.240	156.14	0.000
B	1	0.562	0.562	0.37	0.553
C	1	8.964	8.964	5.83	0.027
D	1	9.001	9.001	5.85	0.026
Square	3	95.743	31.914	20.74	0.000
A*A	1	87.183	87.183	56.66	0.000
B*B	1	16.978	16.978	11.03	0.004
C*C	1	0.711	0.711	0.46	0.505
Two-factor interaction	5	26.508	5.302	3.45	0.023
A*B	1	5.879	5.879	3.82	0.066
A*C	1	1.224	1.224	0.80	0.384
A*D	1	6.585	6.585	4.28	0.053
B*C	1	3.699	3.699	2.40	0.138
C*D	1	7.411	7.411	4.82	0.042
Error	18	27.696	1.539
Lack of fit	16	27.522	1.720	19.75	0.051
Pure error	2	0.174	0.087
Total	30	520.392

In terms of model fitness, three key metrics were considered: R², R²_adj, and R²_pred. These metrics range from 0 to 1, with values closer to 1 indicating higher fitness. The R², R²_adj, and R²_pred values were 0.9468, 0.9113, and 0.8035, respectively, demonstrating the model’s effectiveness in terms of fitness and signal-to-noise ratio (SNR). Therefore, the regression model can reliably be used to predict the depth of subplane damage.

To further evaluate the accuracy of the model in predicting underplane damage under various structural parameters, 10 groups of different structural parameters were randomly selected within the parameter range outlined in Table 7, as shown in Table 10. Simulation experiments were conducted to obtain experimental values of underplane damage for the additional test group, which were then compared with the model-predicted values. The comparison results are depicted in Figure 7.

Table 10.

Additional test group structure parameter table.

Test serial number	R (μm)	α (°)	W (μm)	d (μm)
1	0.0100	25	35	8
2	0.0100	20	35	7
3	0.0050	18	45	6
4	0.0200	18	30	10
5	0.0125	15	45	5
6	0.0125	20	50	9
7	0.0125	15	45	12
8	0.0125	20	45	15
9	0.0050	20	40	8
10	0.0200	15	50	6

Using equation (2), the average relative error Δ was calculated to be 7.58%, which is relatively small and demonstrates good consistency. This verifies the feasibility of the model and the accuracy of its predictions. In the equation, “e” represents the predicted value, “t” denotes the actual value from simulation, “n” represents the number of additional test groups, and “i” indicates the sequence number of the additional test.

Δ = \sum_{i = 1}^{n} \frac{e i - t i}{n t i} \times 100 %

(2)

Optimization design

Genetic algorithms are computational methods inspired by biological systems, known for their high efficiency, parallel processing, and global search capabilities. They possess the ability to automatically acquire and accumulate knowledge during the search process, while adaptively controlling the search to obtain optimal solutions. Despite these advantages, genetic algorithms may encounter challenges such as getting trapped in local optimal solutions, even with increased solution space diversity through mutation operations.

To address this issue, we introduce the Beluga optimization algorithm, which maintains a population of beluga whales and continuously adjusts their positions to search for optimal solutions. This algorithm consists of two main phases: exploration and exploitation. During the exploration phase, beluga whales move randomly to explore the solution space. In the exploitation phase, they move towards areas with better objective function values to approach the optimal solution more efficiently.

By combining the genetic algorithm with the Beluga whale optimization algorithm, we can leverage their respective strengths to achieve a balance between global search and local optimization. This enhances the convergence speed and search ability of the algorithm, enabling more effective solutions to complex optimization problems.

To minimize subplane damage, this paper adopts the Beluga genetic algorithm, integrating the Beluga optimization algorithm and genetic algorithm. The objective is to minimize subplane damage depth by optimizing the tool’s front angle, edge angle, texture width (W), and texture depth (d) in micro-texture tool design. The optimization design process is illustrated in Figure 8.

Since the model is built based on the response surface method, the optimal value is obtained in the range of response parameters. Figure 9 shows the comparison between the optimization of the common genetic algorithm and the Beluga genetic algorithm. It can be found that the genetic algorithm is easy to fall into local solutions, and the optimal solution combination after optimization is as follows: tool front Angle 25°, blade 0.005 μm, texture width 32.5 μm and texture depth 5.7 μm, and the predicted value of underplane damage depth is 12.9 μm. After 1000 iterations of population evaluation, it is found that when the population tends to be stable, four optimal structural parameters are obtained: tool front Angle 25°, blade 0.005 μm, texture width 50 μm, and texture depth 15 μm. Under these parameters, the cutting performance is the best and the underplane damage is the least. The predicted value is 10.3143 μm, which is significantly better than the genetic algorithm. Under this parameter, the simulation experimental value is 9.5843 μm, and the error rate is 7.61%.

Under the same conditions, the subplane damage for a common textured tool and a common micro-texture tool is 24.175 μm and 22.8961 μm, respectively. Compared to the subplane damage caused by the optimized micro-texture tool, the subplane damage is reduced by 60.35% and 58.14%, respectively. Additionally, the cutting force of the optimized micro-texture tool is also reduced, as detailed in Table 11. These results demonstrate that the optimized micro-texture tool exhibits superior performance in cutting CFRP. Figure 10 illustrates the comparison of CFRP results before and after optimization with the same cutting depth.

Table 11.

Comparison table of optimized parameter results.

Tool	R (μm)	α (°)	β (°)	T (μm)	W (μm)	s (μm)	d (μm)	Subsurface damage depth (μm)
Ordinary cutting tool	0.01	5	15					24.175
Micro texture tool	0.01	5	15	10	30	20	10	22.896
Optimized tool	0.005	25	30	20	20	50	15	9.5843

Conclusion

The simulation model of CFRP cutting with 3D micro-texture cutter is established by Abaqus, and the main factors affecting the underplane damage are analyzed. The study aims to perform the best tool parameter optimization within the selected parameter range. Here are the key findings:

(1) By considering both tool and texture parameters, a significance analysis of structural parameters affecting subsurface damage during CFRP cutting was conducted through PB test design. The significance order of structural parameters, from highest to lowest impact, was determined as follows: tool front angle, cutting edge, texture width, texture depth, tool clearance angle, texture spacing, and distance between texture and cutting edge.

(2) Utilizing the response surface method, we developed a prediction model between key parameters (tool angle, cutting edge, texture width, texture depth) and the response variable (subplane damage). This model enables efficient prediction of subplane damage depth and reduce the cost.

(3) We introduced the beluga genetic optimization algorithm by combining genetic and Beluga algorithms. This algorithm was employed to minimize subplane damage, resulting in optimal parameter combination: a front knife angle of 25°, blade radius of 0.005 μm, texture width of 50 μm, and texture depth of 15 μm. This combination significantly reduces underplane damage by 60.35% compared to ordinary textured tools and 58.14% compared to micro-texture tools.

(4) Although the developed model demonstrated high accuracy, its applicability remains somewhat limited. Future research should aim to further refine the model to improve its generalizability. Additionally, researchers could explore a wider range of evaluation metrics for optimizing micro-groove machining processes.

Overall, this study provides practical guidelines for utilizing micro-texturing tools in CFRP cutting applications. The established prediction model and optimized tool parameters improve the machining quality and efficiency.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This project is supported by National Natural Science Foundation of China.

ORCID iDs

Peiwei Yao

Liangfei Hu

Data Availability Statement

Data sharing not applicable to this article as no datasets were generated or analyzed during the current study.

References

Liu

Chen

. An approach to predict the mechanical properties of CFRP based on cross-scale simulation. Compos Struct 2019; 210: 339–347.

Karpat

Bahtiyar

Değer

. Mechanistic force modeling for milling of unidirectional carbon fiber reinforced polymer laminates. Int J Mach Tool Manufact 2012; 56: 79–93.

Hou

Yao

Zhang

, et al. Chatter stability and surface quality in milling of unidirectional carbon fiber reinforced polymer. Compos Struct 2021; 271(12): 114131.

Ouyang

Jiao

, et al. Experimental study on CFRP drilling with the picosecond laser “double rotation” cutting technique. Opt Laser Technol 2021; 142(4): 107238.

Bogenfeld

Gorsky

Wille

. An experimental damage tolerance investigation of CFRP composites on a substructural level. Comp Part C: Open Access 2022; 8: 100267.

Jin

Nam

Jeon

, et al. A study on machining performances of micro-drilling of multi-directional carbon fiber reinforced plastic (MD-CFRP) based on nano-solid dry lubrication using graphene NanoPlatelets. Materials 2021; 14(3): 685.

Alessandro

Khamis

. 3D micro-mechanical modelling of orthogonal cutting of UD-CFRP using smoothed particle hydrodynamics and finite element methods. Compos Struct 2019; 218: 174.

Qin

Wang

, et al. Simulation of effect of fiber direction on cutting process of CFRP. Mechanical Science and Technology for Aerospace Engineering 2016; 35(03): 472–476, (in Chinese).

Linde

Boer

. Modelling of inter-rivet buckling of hybrid composites. Compos Struct 2006; 73(2): 221–228. DOI: 10.1016/j.compstruct.2005.11.062.

10.

Geng

Ming

, et al. Simulation machining of fiber-reinforced composites: a review. Int J Adv Des Manuf Technol 2021; 117: 1.

11.

Meng

Cai

Cheng

, et al. Investigation of CFRP cutting mechanism variation and the induced effects on cutting response and damage distribution. Int J Adv Manuf Technol 2020; 106(7): 2893–2907.

12.

Rong

, et al. Cutting simulation of carbon fiber reinforced resin matrix composite material based on macroscopic anisotropy. Acta Mater Compos Sin 2014; 31(3): 584–590, (in Chinese).

13.

Zitoune

Collombet

Lopez

. Experimental and analytical study of the influence of HexFit? glass fibre composite manufacturing process on delamination during drilling. Int J Mach Mach Mater 2008; 3(3/4): 326. DOI: 10.1504/ijmmm.2008.020967.

14.

Fang

Obikawa

. Cooling performance of micro-texture at the tool flank face under high pressure jet coolant assistance. Precis Eng 2017; 49: 41–51.

15.

Durairaj

Guo

Aramcharoen

, et al. An experimental study into the effect of micro-textures on the performance of cutting tool. Int J Adv Manuf Technol 2018; 98(1-4): 1011–1030.

16.

Zhang

Yang

Chen

, et al. Study on the influence of micro-textures on wear mechanism of cemented carbide tools. Int J Adv Manuf Technol 2020; 108(6): 1701–1712.

17.

Anis

Paul

. A comparative study on cutting performance of rake-flank face structured cutting tool in orthogonal cutting of aisi/sae 4140. Int J Adv Des Manuf Technol 2015; 78(9-12): 2097–2106.

18.

You

Zhao

Chu

, et al. Design, preparation and cutting performance of bionic cutting tools based on head microstructures of dung beetle. J Manuf Process 2020; 58: 129–135.

19.

Duan

Deng

Lei

, et al. Effect of derivative cutting on machining performance of micro textured tools. J Manuf Process 2019; 45: 544–556.

20.

jin

XIE

Luo

, et al. Micro-grinding of micro-groove array on tool rake surface for dry cutting of titanium alloy. Int J Precis Eng Manuf 2012; 13(10): 1845–1852.

21.

Sun

Zhou

Deng

, et al. Effect of hybrid texture combining micro-pits and micro-grooves on cutting performance of wc/co-based tools. Int J Adv Manuf Technol 2016; 86(9-12): 3383–3394.

22.

Xie

Luo

, et al. Experimental study on cutting temperature and cutting force in dry turning of titanium alloy using a non-coated micro-grooved tool. Int J Mach Tool Manufact 2013; 73(1): 25–36.

23.

Liu

, et al. Performances of micro-textured wc-10ni 3 al cemented carbides cutting tool in turning of ti6al4v. Int J Refract Metals Hard Mater 2019; 84: 104987.

24.

Sugihara

Enomoto

. Performance of cutting tools with dimple textured surfaces: a comparative study of different texture patterns. Precis Eng 2017; 49: 52.

25.

You

Xie

Chu

, et al. Cutting Performance of Bionic Cutting tools Based on Surface Microstructures of Blood Clams in Dry Cutting of CFRP. The International Journal of Advanced Manufacturing Technology. 2020; 25: 129–135.

26.

Chen

Guo

Zhang

, et al. Study on the surface quality of cfrp machined by micro-textured milling tools. J Manuf Process 2019; 37: 114–123.

27.

Cheng

Zhang

, et al. Research on CFRP cutting mechanism by the micro-textured tool using macroscopic and microscopic numerical simulations. J Reinforc Plast Compos 2023; 42(11-12): 577–588.

28.

Yan

Reiner

Bacca

, et al. A study of energy dissipating mechanisms in orthogonal cutting of UD-CFRP composites. Compos Struct 2019; 220: 460–472.

29.

Hejjaji

Rédouane

Bouvet

, et al. New tool for reduction of harmful particulate dispersion and to improve machining quality when trimming carbon/epoxy composites. Composites Part A Applied Science and Manufacturing 2020; 131: 105806. DOI: 10.1016/j.compositesa.2020.105806.

30.

Samsudeensadham

Vijayan

Zitoune

. Investigations on drilling of Al/CFRP/Ti stacks using micro grooved drills. In: International conference on engineering research and applications, Tajoura, Libya, 15 to 17 October 2024, pp. 102–113.

31.

Alessandro

Sein

Khamis

. Modelling the orthogonal cutting of UD-CFRP composites: development of a novel cohesive zone model. Compos Struct 2017; 168: 65.

32.

Calzada

Kapoor

DeVor

, et al. Modeling and interpretation of fiber orientation-based failure mechanisms in machining of carbon fiber-reinforced polymer composites. J Manuf Process 2012; 14: 141–149.

33.

Xianren

Xin

Lingxin

, et al. Teaching reform of experimental design and statistical analysis for aricultural machinery specialty. Agricultural Engineering. 2017; 7(3): 142–145.