The study of CFRP variable feed drilling method based on sinusoidal curve

In order to reduce the delamination damage of carbon fiber–reinforced polymer (CFRP) drilling and improve the drilling quality, a variable feed drilling method based on sinusoidal curve is proposed. That means, when drilling at a distance of 1 mm from the hole exit, the feed rate of the drill would be reduced according to the rule of sinusoidal curve. The method is compared with the traditional feed drilling by experiment, and the influence of the variable feed drilling on thrust force, delamination factor, surface quality, and surface roughness of hole wall and exit hole wall morphology are analyzed. The results show that compared with constant-feed drilling, the variable-feed drilling based on sinusoidal curve reduces the thrust force near the hole exit and greatly improves the delamination factor, surface quality, and hole wall morphology at the exit. However, the influence on the hole wall roughness is not particularly obvious.


Introduction
Due to its light weight, high strength, and strong impact resistance, CFRP (carbon fiber-reinforced plastics) has been widely used in aircraft wings and fuselage and other main load-bearing structures in recent years. With the number of CFRP holes increasing rapidly during aircraft assembly, the quality and efficiency of holes are of great significance to aircraft assembly [1].
The quality problems of drilling CFRP mainly include delamination and peel-up at the exit and entrance, fiber pulling out, oversized diameter, burr on hole edge, etc., among which delamination at the exit has the greatest influence on the quality of CFRP hole [2,3]. Jia et al. [4] showed through simulation and experiment that thrust force was the main reason for delamination of CFRP exit, and the critical thrust force of CFRP drilling without delamination was given. Wu et al. [5] experimentally studied the causes of delamination in the inner wall of CFRP and believed that the inner delamination was more sensitive to the change of feed rate. The ultrasonic vibration-assisted drilling and the placement of lining plates under CFRP played a positive role in reducing delamination damage. Geier and Szalay [6] drilled unidirectional CFRP using diamond-coated tools, and the results showed that the feed rate had the greatest influence on thrust force and exit delamination, while the spindle speed had a good effect on reducing uncut fiber burr. Jia et al. [7] proposed a multi-edged tool structure to drill Ti/ CFRP stacks and showed that the multi-edged tool structure could effectively reduce CFRP exit delamination and the aperture error of Ti/CFRP stacks by experimental comparison. Some other scholars [8][9][10] used some artificial intelligence algorithms to optimize the CFRP drilling process parameters, obtained the cutting parameters corresponding to the minimum delamination damage, and found that the feed rate was positively correlated with the delamination damage.
In addition to changing cutting parameters and tool geometry, some special machining methods were also applied to the CFRP drilling. Chen et al. [11] combined ultrasonic vibration with spiral milling process to reduce thrust force and cutting heat in machining and significantly reduce the phenomenon of matrix adhesion to hole wall. Islam et al. [12] introduced EDM (electrical discharge machining) process to remove the burr of CFRP exit and improved the exit surface quality of CFRP. Yaşar and Günay [13] reduced the delamination damage at the CFRP exit through a step-change feed drilling method. Although the special machining method achieves better machining quality, its corresponding supporting equipment is not mature yet.
In brief, existing researches showed that reducing the thrust force of exit and increasing the spindle speed could suppress surface delamination damage. However, when the spindle speed remains unchanged, reducing the feed speed is bound to reduce the machining efficiency. In the paper, a variable feed drilling method based on sinusoidal curve was proposed, which has a positive effect on reducing the thrust force of exit, surface delamination damage, and the roughness of the hole wall and improving the surface morphology of hole wall by experiment comparison with traditional drilling.

Damage mechanism of CFRP drilling
At present, the machining damage of CFRP cannot be repaired, and the large degree of damage will directly lead to the scrapping of parts [1,2]. Therefore, the quality of assembly hole will directly affect the assembly performance and service life of structural parts. The quality evaluation of assembly holes of CFRP structural parts mainly includes the following three aspects [14]: (1) entrance fiber peel-up and delamination; (2) hole wall surface groove; (3) exit push-up delamination. The analysis of the three damage formation mechanisms is as follows: 1. Entrance fiber peel-up and delamination: Under the action of high-speed rotation and cutting force of the tool, the main cutting edge of the tool cuts off the fibers on the surface of CFRP. When the tool is about to fully enter the workpiece (as shown in Fig. 1a), the tool acts on the fiber at the hole edge, causing fiber twist separation (Mode III) and upward peel-up (Mode I) [5]. 2. Hole wall surface groove: With the continuous drilling of the tool, the cutting heat gradually accumulates and softens the matrix epoxy resin, which will weaken the wrapping ability of the matrix to the reinforcement fiber [3]. At the same time, the cutting force causes the carbon fiber to break and peel off from the matrix, which result in the formation of groove on the surface of the hole wall (Mode II). 3. Exit push-up delamination: When the tool is about to drill out (as shown in Fig. 1b), the uncut material bends and deforms under the action of thrust force and causes interface delamination and fiber breakage (mode I and mode II).
In summary, the push-up delamination of CFRP exit is easy to occur, and it is also the main reason for the high scrap rate of structural parts [1][2][3][4][5]. Therefore, the reduction of thrust force will be an important measure to improve exit delamination damage and hole making quality.

Variable feed drilling method
Franck et al. [15] studied the thrust force causing fiber delamination under different uncut material thickness. It was considered that the formation of delamination damage could be suppressed if the thrust force was less than the critical force of delamination. The thrust force can be reduced by reducing the feed rate, but the sudden change of the feed rate is bound to cause the vibration of the cutting system. In addition, the step change of feed rate will also greatly reduce the machining efficiency. Therefore, the flexible and smooth line of sinusoidal curve provides the possibility for the achievement of high-efficiency variable feed without vibration. The specific process of sinusoidal variable feed drilling is shown in Fig. 2, and details are as follows: (III) Since the optimal feed rate of CFRP drilling is unknown, the way that feed speed decreases by half has been applied in this paper. The discrete sample points obtained by II are correlated with the feed rate to obtain each discrete feed rate. This correlation formula can be expressed as where f is the initial feed rate (mm/r), fi is the discrete feed rate of each section (i = 1,2…9).
(VI) The whole drilling process was divided into two sections; one section has low damage with traditional drilling (t 0 − t 1 ), another section has serious (1) The feed rates distributed by VI were correlated with the corresponding spindle speed. Here, the cutting time of each discrete feed is the same by default. Therefore, the cutting time of each discrete feed can be expressed as where n is the spindle speed (rpm), T is the cutting time(s) of each discrete feed, L is variable feed displacement (L = 2 mm, the distance is from the t 1 time to t 2 time). All in, all data are stored in Mitsubishi PLC controller for calling during the drilling process.

Experimental design and conditions
The workpiece used in the trials was provided by Weihai Guangwei Composite Material Co., Ltd. The carbon fiber workpiece (Brand: T300 / 7901) was overlaid by 24 layers of unidirectional prepreg. The detailed properties of unidirectional prepreg can be referred in Table 1. The layup sequence of the CFRP was [45/0/ − 45/90] 3S and was cured to a thickness of 3.6 mm. The total workpiece with Variable feed drilling method association process dimensions of 100 mm (length) × 40 mm (width) × 3.6 mm (thickness) was cut by water jet machining. Uncoated tungsten carbide (K20) twist drills without lubrication were used for drilling trials. All drills had the same geometry in terms of diameter (6 mm), flute length (25 mm), point angle (135°) and helix angle (30°). The general full factorial design was used to design the experiment intent. The experimental parameters that meet CFRP drilling requirements are shown in Table 2, where "0" represents traditional feed strategy, and "1" represents variable feed strategy based on sine curve. All the trials were conducted without coolant. It can be seen from Table 2 that the various levels of all experimental factors were combined to form 18 different experimental conditions. Two independent repeated experiments were carried out under each experimental condition. A total of 36 experiments need to be carried out, and these experiments were performed in random order. In order to minimize the impact of tool wear, the tool was cooled and cleaned after drilling and was replaced after 10 holes. In addition, the discrete data of variable feed rate under different feed parameters are presented in Table 3.
All experiments were carried out on the drilling experiment platform, which was developed by the authors' lab. The drilling experiment platform consists of the feed platform, the fixture, the tool, the workpiece, and the sensor, and it can be seen in Fig. 4. The feed platform was equipped with a servo motor (PKE569AC, Oriental Motor), which can set multiple feed speeds that are convenient for the control system to call and provide convenience for variable feed drilling process. All feed data were calculated in advance and stored in the servo motor and the control system. The dynamometer (DYZ-101, DAYSENSOR) was used to measure the thrust force in the drilling process. The thrust forces were transmitted from the dynamometer to the block of analog input (FX3U-4AD, Mitsubishi), then to Programmable Logic Controller (FX3U, Mitsubishi) and recorded on a computer using Lab-View data acquisition software (Fig. 3).
After the experiment, a numerical microscope (VHX-2000, KEYENCE) was used to measure the surface delamination damage and hole wall morphology at the exit. The surface delamination damage can be expressed by dimensionless one-dimensional delamination factor F d [9,13] (see Fig. 4) and its formula is as follows: where D max is the maximum diameter including all delamination areas, D nom is the nominal diameter of the hole. In order to obtain the difference in hole wall roughness of different drilling strategies, the workpiece was cut by water jet machining, and surface roughness Ra was measured from median line to exit (see Fig. 5) with portable roughness meter (TR200, TIME.) following the 0.8-mm stroke.

Comparison and analysis of thrust forces in different strategies
The thrust force is an important factor that affects the exit surface delamination, and lower thrust force helps to reduce exit delamination damage [16,17]. Figure 6 shows the relationship between thrust force and feed displacement under f = 0.04 mm/r, n = 2100 rpm cutting parameters. It can be seen that the thrust force at the exit of the variable feed was always lowered than that of the traditional feed. The change trend of the difference between the two thrust forces at the same drilling position is as follows: first, gradually increase and then slowly decrease. The main reason is the reduction of thrust force caused by the variable feed when the tool chisel is 1 mm away from the exit with a sinusoidal rule to reduce the feed rate. From the surface quality of the two different drilling strategies in Fig. 6, it can be seen that the surface quality of variable-feed drilling (lower fiber delamination and burrs) is better than traditional drilling.
In summary, the change law of the sine curve (trend: small-large-small) was consistent with the change of the thrust force difference between the two-drilling strategies, which verified the correctness of the application of the proposed method on processing equipment. In addition, the excellent surface quality of the variable feed achieved the expectation of reducing the delamination damage on the exit by reducing the thrust force. Figure 7 shows the measurement results of the delamination factor. It can be seen that the delamination factor increases with the increase in the feed rate, and an appropriate spindle speed is also beneficial to reducing the delamination factor. In addition, the relative errors of the delamination factors under the two different drilling strategies are basically positively correlated with feed rate, which indicates that the variable feed drilling method has a better performance at a higher material removal rate.

Delamination factor comparative analysis
We also noticed that under the cutting conditions: n = 2400 rpm, f = 0.08 mm/r, the delamination factor of traditional feed drilling was lower than that of variable feed. At n = 2400 rpm, the effect of feed rate on the delamination factor is not particularly significant than others. This is mainly because n = 2400 rpm is the optimal speed (see Fig. 8), and the thrust force is smaller than the critical thrust force at different feed rates, which results in no significant difference in the delamination factor between conventional drilling and variable feed drilling. Figure 8 shows the main effect plot of the delamination factor. In the figure, 0 in black represents the traditional feeding strategy, and 1 in green represents the variable feeding strategy. Using the relative error to represent the variation law of the difference between two drilling strategies, it can be seen from the figure that the parameter that has the greatest influence on the mean value of the delamination factor is the feed rate, followed by the drilling strategy. The variable feed drilling process reduces the delamination factor from 1.6 to 1.3. The influence of cutting parameters on the delamination factor is as follows. The feed rate was reduced in a sinusoidal rule at the exit, which reduced the cutting thickness of each tooth of the tool, and weakened the effect of thrust force on the uncut material. The action will reduce the delamination tear damage along the surface of the hole. In addition, the spindle speed increased the friction between the tool and the composite material and generated a large amount of cutting heat. As the tool fed, the cutting heat gradually accumulated, which caused the resin to soften and reduced the effect of the thrust force.

Analysis of exit surface morphology
The main surface damages in CFRP processing are delamination, fiber tearing, and burrs (see Fig. 9) [18]. Delamination is caused by fiber cracking at the hole edge due to the thrust force being greater than the critical force between the CFRP layers; fiber tearing is the surface damage caused by the delamination and cracking of the fiber along the fiber direction under the action of thrust force; burrs are left because the material at the edge of the exit hole cannot be effectively cut. Therefore, the difference of surface damage represents the bearing capacity of the hole and service life [19]. Figure 10 is a comparison of the surface morphology of different drilling strategies. It can be seen that for traditional feed, as the feed rate increases, the surface tearing and delamination become more serious and the damage area gradually increases. The increase of the spindle speed can effectively remove the burrs on the hole edge, but it cannot reduce the fiber tearing and delamination of the exit. In general, a good surface quality requires a lower feed rate and an appropriate spindle speed.
For variable feed, the overall surface topography quality is better than traditional feed. However, the feed rate is still the main factor that determines the drilling damage of exit. The drilling damage at the variable feed exit is mainly uncut burrs, small-scale fiber tearing, and low delamination. The main reason is that the feed rate is reduced in a sine curve, so that the thrust force of exit is less than the critical force of delamination, which avoids serious delamination damage at the exit. The demonstrative explanation of the effect of the thrust force on the surface damage can be seen in Fig. 11. Coupled with cutting heat, the thrust force caused by the drill chisel edge separates the reinforcement fiber from the matrix, which results in delamination between uncut material and cut material. At the same cutting position near the CFRP exit, the thrust of the variable feed is lower than that of the conventional feed ( F1 < F0 ), which results in a smaller delamination damage diameter of the variable feed ( D1 < D0 ). Therefore, the surface quality of the variable feed drilling is better than that of conventional drilling.

Surface roughness of hole wall
Damage to the surface of the hole wall will lead to uneven stress distribution during the connection of the components, resulting in a decrease in the load-bearing capacity [16,17]. Due to the different cutting mechanisms of different fiber angles in CFRP (see Fig. 12), CFRP hole wall roughness is often poor and difficult to measure [8]. The roughness measurement result (see Fig. 13) shows that two drilling strategies under different feed rates have a minor difference on the roughness.
Combined with the main effect plot of roughness (Fig. 14), it can be seen that the increase of the feed rate results in an increase in the roughness of the hole wall. The feed rate increases from 0.08 to 0.12 mm/r, and the average roughness increases greatly. The main reason is that the higher cutting force of the tool causes the severe peeling of the fiber from the resin matrix, resulting in the worse hole wall roughness [1,3]. The increase of the spindle speed results in an increase in the roughness of the hole wall (see Fig. 14). The main reason is that the increase of the tool spindle speed leads to a sharp increase in cutting heat, which exceeds the resin curing temperature and forms defects such as fiber pullout and matrix cracks under the action of force [11]. In addition, under the action of cutting heat, the fiber debris adheres to the hole wall to form plow-shaped particles, resulting in poor hole wall roughness [13].
In summary, the factors that have a greater impact on roughness are feed rate > spindle speed > drilling strategy. Combining Figs. 13 and 14, it can be concluded that the variable-feed drilling process has a positive effect on the improvement of the surface roughness of the CFRP hole wall compared with the traditional feed.

Comparative analysis of hole wall morphology
The composition of the composite material and the cutting mechanism determine that it is a difficult-to-machine brittle material, and the hole wall morphology also has a greater impact on the assembly performance of CFRP components [19,20]. The main defects of CFRP hole wall morphology are exit fiber delamination, hole wall surface groove, and fiber pull-out [5,14]. The fiber delamination is mainly caused by the excessive thrust force of the chisel blade. Surface cavities are caused by fiber breakage and peeling due to the angle of −45° between the tool and the fiber direction (see Fig. 15). The fiber pull-out is due to the combined action of the thrust force and the cutting heat, and the fiber is separated from the matrix, leaving a rough surface with fiber pull-out. Figure 16 shows a comparison diagram of the morphology of the hole wall exit. It can be seen that the feed rate has a greater impact on the damage of the CFRP hole wall, and the increase of the spindle speed improves the improvement of surface cavitation, fiber pull-out, and delamination damage. Observation of the morphology shows that the quality of the hole wall obtained under the condition of n = 0.04 mm/r is better than other conditions as a whole.     Therefore, choosing a lower feed rate and a higher spindle speed when drilling carbon fiber composite materials will effectively reduce hole wall defects (delamination damage, surface cavities, and fiber pull-out) and improve drilling quality.
In addition, it can be seen in Fig. 16 that the variablefeed drilling process has better performance in reducing exit delamination and fiber pull-out compared with traditional feeding. However, the surface cavities are not significantly improved, which is mainly because −45 cutting mechanism cannot be eliminated between the tool and the fiber direction. It can also be seen from Fig. 16 that variable feed drilling has tolerance behavior for drilling with larger feeds (that is, better exit quality can be maintained under larger feeds).
In brief, no matter what kind of drilling strategy, it cannot completely eliminate the drilling damage at the exit of the composite material. The thrust force is still the key factor for the damage of delamination. On the premise of sacrificing less processing efficiency, the variable feed drilling strategy of reducing the exit feed rate and thus reducing the thrust force provides a new idea for improving the drilling quality of composite materials.

Conclusion
In this paper, the damage mechanisms and factors affecting the quality of CFRP drilling were analyzed. The influence of feed rate, spindle speed, and drilling strategy on the factors of hole quality (thrust force, delamination factor, surface roughness, and hole wall surface morphology) was focused on. Conclusions are as below: 1. In order to reduce the thrust force of the exit, a variable feed drilling method based on the sinusoidal curve setting feed rate was proposed. That means the feed rate was set according to the sinusoidal law at a distance of 1 mm from the tool to the exit. The variable feed drilling strategy was carried out and compared with traditional drilling in the hole wall roughness and the surface delamination damage on the exit. 2. It can be seen from the thrust force of the tool at the same position that the thrust force of the variable feed was lower than that of the traditional feed, which verified the correctness of the application of the proposed method. Through the quantitative analysis of the delamination factor and hole wall roughness, it can be known that the variable feed drilling process has a more significant increase in the delamination factor than the traditional feed. However, the difference on the roughness is not particularly obvious. 3. By comparing the surface of the exit and the morphology of the hole wall, it can be known that the variable feed process can effectively improve the surface drilling quality (delamination, fiber tearing, burr) of the composite material to obtain a smooth hole wall. Overall, under the premise of reducing the small processing efficiency, the variable feed drilling method provides a new idea for improving the drilling quality of composite materials.
Author contribution Material preparation, data collection, and analysis were performed by Wang Shuaipu and Liang Jie. The first draft of the manuscript was written by Wang Shuaipu, and all the authors commented on previous versions of the manuscript. All the authors read and approved the final manuscript.
Availability of data and material All data generated or analyzed during this study are included in this published article.

Declarations
Ethics approval Not applicable.

Consent to participate Not applicable.
Consent for publication Not applicable.

Conflict of interest
The authors declare no competing interests.