Delamination and hole wall roughness evaluation in air-cooled drilling of carbon fiber-reinforced polymer

Abstract

Drilling of carbon fiber-reinforced polymer (CFRP) is widely employed in manufacturing processes in the aeronautical, automobile, and energy industries. The evaluation of the hole region focusing on wall roughness and delamination phenomena is extremely important to predict bolted joints' performance, where at least one of the adherent is a composite material. Thus, this work performed a statistical analysis on the delamination (entrance and exit of the plate) and wall roughness of drilling holes in an 8.6 mm thick CFRP plate carried out by an uncoated carbide drill under compressed air-cooling, varying the cutting speeds and feed rates. Since air-cooling usage combines positive aspects, such as low-cost implementation and shorter process time compared to ultrasonic-assisted drilling, it turns out to be an excellent alternative for aeronautical industry. Thus, the main contribution of the present work consists on analyzing the variation of the delamination in the entrance and exit of the first and tenth holes of a CFRP plate after dry and air-cooled drilling. This variation of the delamination between the cold drill (first hole) and the heated drill due to the drilling holes' sequence (after ten holes) is investigated for 18 different combinations of parameters (runs). For instance, it is shown that if the combination of parameters values is suitable, then it is possible to reduce the mean value of adjusted delamination factor for the entrance of the tenth hole around 11% when comparing cooled-air with dry cutting.

Keywords

CFRP drilling compressed air-cooling delamination hole wall roughness

Introduction

Aeronautical, automobile and energy industries are the major boosters of composite materials application. They are focused on developing these materials to achieve better mechanical properties combining self-weight reduction, which decreases energetic costs and, consequently, reduction of pollutant gas emissions into the atmosphere. For instance, the Boeing 787-Dreamliner and the Airbus A350 possess around 50% of their weight composed of composite materials, mainly CFRP (carbon fiber reinforced polymer) parts.¹ Fasteners can join those composite parts, which require holes to assembly composite and metallic structures by using screws or rivets. However, high precision and low tolerances are critical issues during the drilling process, especially in the aeronautical industry.² In Airbus A350, more than 55 thousand holes are necessary,³ producing demountable unions.⁴ These holes are commonly made by drills⁵ or milling cutters,⁶ usually with a carbide tool.^7,8 In the particular case of drilling CFRP, the tools can fastly wear out due to the high toughness and stiffness of composite materials.⁹ Besides, delamination and cracks can occur in the composite parts during the process, mostly with the increase of the tool wear.^3,10

Considering the drilling of composite materials, there are four main types of defects, which occur during the process using helical drills in laminated composite materials, which can be identified as follows: (1) damage at the entrance of the hole (tool entrance); (2) damage at the exit of the hole (tool exit); (3) dimensional defects and (4) thermal defects.¹¹ The present work focuses on the evaluation of the first two types of defects. The damage at the hole entrance is related to the fiber debonding of the first layer due to the drill cutting edges, and this can induce delamination. In this case, the material tends to be pulled by the drill's helix (peel-up). The damage at the hole exit is due to the compression stress components caused by the thrust force at the bottom layers. When the stress state exceeds the fiber-matrix adhesion strength, initiation and propagation of cracks occur, resulting in further delamination. Hence, the material tends to be pushed (push-down), and this type of damage is harder to be avoided.¹² Furthermore, the CFRP's flexure strength decreases with the increase of the delamination area after drilling.³ In addition, some fibers can be uncut, requiring further machining even if it does not affect the mechanical properties of the final part.¹³

Regarding the scenario pointed above, different techniques have been used to detect the delamination degree, both in the hole entrance and exit. Davim and Reis¹⁴ used microscopy to evaluate the delamination promoted by drilling of CFRPs. Chen¹⁵ proposed the delamination factor concept ( $F_{d}$ ) to obtain a quantification and a delamination parameter for comparison (i.e., the delamination degree) in the drilling process of laminated composite materials. Equation (1) provides the ratio between the maximum damaged zone diameter ( $D_{\max}$ ) and the hole diameter ( $D_{O}$ ). Davim et al.¹⁶ proposed an adjusted delamination factor ( $F_{d a}$ ). Equation (2) considers $F_{d}$ and a portion that compares the damaged area ( $A_{d}$ ) with $A_{\max}$ (area related to $D_{\max}$ ) and $A_{o}$ (nominal hole area).

F_{d} = \frac{D_{\max}}{D_{o}}

(1)

F_{d a} = F_{d} + \frac{A_{d}}{A_{\max} - A_{o}} \cdot ({F_{d}}^{2} - F_{d})

(2)

Another important issue consists on thermal defects, which can be created due to the friction between the drill and the composite material. This affects the polymer matrix integrity and being deleterious to the finishing of the hole wall surface. In fact, the quality of the hole wall surface can be also affected by the formation of cutting chips, which consist of a group of fragile fractures that are strongly dependent on the fiber's orientation.¹⁷ Since the fracture of composites does not require large amounts of energy than shear failure in metals, the cutting force and temperature required in CFRP machining are much lower than their metal counterparts.¹⁸ Thus, for the CFRP machining process, it is possible to eliminate the possibility of occurring diffusion and/or dissolution wear mechanisms.¹⁹ However, as commented earlier, the temperatures can still affect the material properties mostly of the polymer matrix,²⁰ increasing local damages, such as debonding and pull-out of fibers.⁴ Besides, the temperatures can reach values beyond the glass transition temperature of the polymer matrix at high cutting speeds. In these cases, the softened polymer and the carbon fibers can create irregular surfaces.²¹ In general, the machined hole quality reduces with temperature increases, and therefore, it is predicted that the quality shall increase with temperature reduction. The increase of the temperature during the drilling process causes matrix stiffness reduction, resulting also in global damages, such as delamination, transverse cracking, and others.²² By one hand, one way to overcome this issue is to reduce the temperature at the hole's region using compressed air-cooling during the drilling operation aided by a vortex tube.²³ However, on the other hand, it is worth noting that the chip generated when machining CFRP can be in the form of fine particles,²⁴ and these particles can be suspended, causing danger to the operator,²⁵ mostly using compressed air-based cooling systems.

In spite of the problem described above, air-cooling usage combines positive aspects, such as low-cost implementation and shorter process time compared to ultrasonic-assisted drilling. Thus, it turns out to be an excellent alternative for aeronautical industry, where compressed air-cooling drilling of CFRP parts should be carried out in machine tools' area enclosed, mainly with an exhaustion system. Therefore, by using a machining center with an enclosed machining area, different values of combination of process parameters are investigated to evaluate delamination degree via adjusted delamination factor ( $F_{d a}$ ), and the hole wall quality by using roughness parameters R_a (average roughness), R_q (root mean squared roughness), R_z (average maximum peak-to-valley height), and R_t (total roughness). Thus, the main contribution of the present work consists on analyzing the variation of the delamination in the entrance and exit of the first and tenth holes of a CFRP plate after dry and air-cooled drilling. This variation of the delamination between the cold drill (first hole) and the heated drill due to the drilling holes' sequence (after ten holes) is investigated for 18 different combinations of parameters (runs). For instance, it is shown that if the combination of parameters values is suitable, then it is possible to reduce the mean value of adjusted delamination factor ( $F_{d a}$ ) for the entrance of the tenth hole around 11% when comparing cooled-air with dry cutting.

Materials and methods

Experiments were performed at a Romi Discovery 308 machining center with an enclosed machining area, 5.5 kW maximum power, and 4000 rpm maximum rotation. A CFRP plate with an 8.6 mm thickness manufactured by hand layup technique was used during the tests. The reinforcement was a bidirectional (BD) carbon fabric with pre-impregnated epoxy resin stacked as a ${[0]}_{38}$ laminate with around 64.1% of fiber volume fraction, $V_{f}$ . Additionally, the laminate is cured using vacuum bagging in an oven under temperature control. Additional information, prepreg material and geometry data, about the CFRP plate is shown in Table 1. For the holes' drilling, the CFRP plate was fixed at the machining center table using a polymer support plate pre-pierced with 8 mm holes (Figure 1(a)), which avoids vibrations and possible other undesired effects, such as buckling.

Table 1.

CFRP’s plate: Prepreg material and geometry data.

Prepreg material	Warp	Carbon fiber roving
Prepreg material	Weft	Carbon fiber roving
Fabric composition		Carbon fiber + Epoxy resin
Glass transition temperature		130 degrees Celsius
		Min.	Nominal	Max.
Number of threads (thread/cm)	Warp	4.75	5.00	5.25
Number of threads (thread/cm)	Weft	4.50	5.00	5.50
Weight (g/m²)		180	200	220
		Thickness	Length	Height
Geometry of the plate (mm)		8.6	130	130
Mean fiber volume fraction (%)^a		64.1
Young’s modulus in the normal direction to the plate (GPa)^a		6.78
Strength value in the normal direction to the reinforcements (MPa)^a		21.55

^aSimilar material.

The drill holes were machined by a Sandvik CoroDrill 860 H10F NM, an uncoated twist solid carbide drill with a 6.0 mm diameter, cross edging, 130° point angle, and 16.6° helix angle. The compressed air-cooling was produced by a Turbo Air Eurotools FTA-12-MC vortex tube connected to a 3 bar (300 kPa) air-line, resulting in a 6 degrees Celsius output temperature. The axial run-out error of the drill (δ) was assessed using a Starrett dial gauge with a resolution of 1.0 μm. The setup of the experimental test is shown in Figure 1(b).

A full-factorial design of experiments (DOE) was carried out considering the variation of cutting speed (v_c = 25, 50, 75 m/min) and feed rate (f = 25, 50, 75 µm/rev.) to compare the dry condition at room temperature (25 °C) with the air-cooled condition in CFRP drilling. Therefore, this procedure results in 3²⋅2¹ = 18 combinations of parameters (runs). Moreover, ten holes were drilled for each condition, resulting in 180 holes (Table 2).

Table 2.

Experimental parameters.

Dry drilling			Air-cooled drilling
Run	v_c (m/min)	f (µm/rev.)	Run	v_c (m/min)	f (µm/rev.)
1	25	25	10	25	25
2	25	50	11	25	50
3	25	75	12	25	75
4	50	25	13	50	25
5	50	50	14	50	50
6	50	75	15	50	75
7	75	25	16	75	25
8	75	50	17	75	50
9	75	75	18	75	75

The adjusted delamination factor was analyzed in the entry and exit of the first and tenth holes after dry and air-cooled drilling for each run (Figure 2) to verify the variation of the delamination between the cold drill (first hole) and the heated drill due to the drilling holes' sequence (after ten holes).

After drilling, images were obtained from the first and the tenth holes made in each run using a portable digital microscope Dino-Lite Pro model AM-413ZT with 50x amplification at the entrance and the exit of the drill in the CFRP plate. Essentially, those images were used for analyzing the presence or not of delamination. Based on the recorded images, the damaged area ( $A_{d}$ ) and the maximum diameter of the damaged zone ( $D_{\max}$ ) were measured using the software ImageJ, which calculates the delamination factors through pixel counting in the analyzed area. Then, the adjusted delamination factors were obtained using equation (2). The delamination degree is investigated, comparing the failure generated in the initial with the end of the process. It is important to highlight that, first delamination mechanisms take place due to the peel-up of plies at the hole entrance. After that, at the hole exit, the delamination mechanisms are generated by push-down effect.

The roughness values of the first hole drilled (cold drill) at each condition were determined through the average of three measurements performed by a Mitutoyo Surftest SJ-201P portable stylus equipment with 0.01 µm resolution, using a sampling length of l_e = 0.8 mm and an evaluating length of l_m = 4 mm. Moreover, the 2CR75 filter was used to generate the roughness parameters R_a (average roughness), R_q (root mean squared roughness), R_z (average maximum peak-to-valley height), and R_t (total roughness). In fact, based on the literature, it is expected that the parameter R_t presents the same tendency as R_a, being the most adequate for the CFRP investigations, because it can describe an average roughness and is still sensitive to fiber debonding.²⁶

Besides, additional images were obtained using a scanning electron microscope (SEM) Zeiss Evo Ma10 for improving the analysis of the hole's wall surface. It was necessary to use a Bal-tec SCD 050 thermal spray coating machine to create a thin gold skin over the surfaces.

Lastly, the variance analysis (ANOVA) of the obtained data was performed using the Minitab 18 software. It is a method to test the equality between three or more populational averages based on the sample variance analysis. According to Montgomery,²⁷ to perform the analysis mentioned above is necessary that the quality of model fit would be sufficient; this is given by the coefficient of determination (R²) above 70%. Therefore, as closer to 100% R² is, as greater the regression reliability, ensuring the statistic model's suitability and indicating low variability in the proper response results. Moreover, a confidence interval of 95% (p-value ≤ 0.05) was considered. The ANOVA of the results was performed to verify the influence of the controllable factors (cutting speed (v_c), feed rate (f), and cooling condition (cc)) on the adjusted delamination factor in the entry ( $F_{daEnt}$ ) and exit ( $F_{daEx}$ ) of the holes, and in its respective the wall roughness. The coefficient of determination (R²) was below 70% for the $F_{daEx}$ of the tenth hole for each run, as well as for R_a and R_q results.

Results and discussions

The results present the calculated delamination factor, the measured roughness values, and the performed variance analysis (ANOVA). It is worth mentioning that the tool flank wear (VB) was verified throughout the study. However, the measured wear was negligible even after the execution of all 180 holes.

Figure 3 presents the adjusted delamination factor values generated at the holes entrances ( $F_{daEnt}$ ), and Figure 4 at the holes exits ( $F_{daEx}$ ). More details about the delamination analysis are presented based on the ANOVA.

The hole wall roughness was investigated by evaluating four parameters: R_a, R_q, R_z, and R_t. R_a and R_q values are depicted in Figure 5, while R_z and R_t are shown in Figure 6. More details about the hole wall roughness analysis are presented based on the ANOVA.

The analysis of measured $F_{d a}$ , R_z and R_t values achieved with Minitab 18 is presented in Table 3, which shows the percentage contribution (PC) of each parameter or combination of parameters. Bold values indicate that the influence of controllable input factors on the output response variables is significant for 95% confidence interval (p-value equal and lower than 0.05). Figures 7 and 8 show the curves for the most relevant parameters. For instance, in Figure 7, the mean value of adjusted delamination factor $F_{d a}$ for the entrance of the tenth hole is around 11% lower when comparing cooled-air with dry cutting.

Table 3.

ANOVA of the results.

Input variables	F_daEnt01		F_daEnt10		F_daEx01		F_daEx10		R_z		R_t
Input variables	p-value	PC(%)	p-value	PC(%)	p-value	PC(%)	p-value	PC(%)	p-value	PC(%)	p-value	PC(%)
v_c	0.026	25.0	0.003	16.3	0.637	0.6	0.881	0.1	0.309	5.3	0.113	10.8
f	0.031	23.0	0.167	2.3	0.692	0.4	0.087	18.6	0.539	1.9	0.151	8.6
cc	0.052	17.5	<0.001	43.1	0.067	11.4	0.040	29.0	0.005	64.8	0.006	46.0
v_c x v_c	0.645	0.8	0.167	2.3	0.409	2.0	0.816	0.3	0.985	0.1	0.766	0.3
f x f	0.537	1.4	0.249	1.5	0.832	0.1	0.770	0.5	0.249	7.0	0.154	8.5
v_c x f	0.126	9.9	0.025	7.3	0.575	0.9	0.919	0.1	0.762	0.4	0.698	0.6
v_c x cc	0.495	1.8	0.001	23.1	0.001	59.2	0.576	1.7	0.077	18.4	0.033	22.1
f x cc	0.038	20.6	0.075	4.1	0.448	1.6	0.369	4.5	0.518	2.1	0.368	3.1
R ²	76.0%		91.7%		76.3%		54.7%		70.7%		76.0%

The analyses of delamination showed that the level of damage at the hole entrance ( $F_{daEnt}$ ), caused by the fiber peel-up effect were always lower than the delamination observed at the exit ( $F_{daEx}$ ), in which the push-down effect rises. In some cases, the delamination at the hole entrance is almost a half at the hole exit, which is in agreement with Geng et al.²⁸

The compressed air-cooling usage significantly reduced the $F_{daEnt}$ values compared with the dry cutting ones, mainly at the tenth hole of each run, even for high cutting speeds and feed rates, demonstrating the temperature influence during the process. Ha et al.²¹ and Khairusshima et al.⁷ also observed surface quality improvement when the drilling process is performed with compressed air-cooling.

For $F_{daEnt}$ of the first hole of each run ( $F_{daEnt 01}$ ), the most influential parameter was v_c (25%), followed by f (23%) and cc (17.5%), whereas the combination f x cc (20.6%) also presented a strong influence. Krishnaraj et al.²⁹ and Khashaba et al.³⁰ showed that high cutting speeds increase the material's temperature, affecting the delamination factor. Velayudham e Krishnamurthy³¹ verified that the increase in feed rate leads to delamination growth. Thus, when using cooled-air (which reduces the delamination), combined with a low feed rate, the delamination is even more reduced. However, all values were close to each other, with the lowest ones being obtained while using cooled-air. Its usage allowed a maximum reduction on $F_{daEnt 01}$ of 4% compared to dry cutting (runs 9 and 18), since the drill does not have sufficient time to raise the temperature at the first hole entrance.

For $F_{daEnt}$ of the tenth hole of each experimental test ( $F_{daEnt 10}$ ), the most influent parameter was cc (43.1%), followed by v_c (16.3%), whereas the combinations v_c x cc (23.1%) and v_c x f (7.3%) also presented relevant influence. The combination f x cc was partially significant because it had a p-value equal to 0.075, and its contribution was 4.1%. Ha et al.²¹ observed that the temperature raises too much during machining, having peak value at the final machined portion due to the cumulated heat that is reduced when applying cooled-air.

Still for $F_{daEnt 10}$ , all values were higher than $F_{daEnt 01}$ when machining with dry cutting and similar or higher when using cooled-air, which is due to the tool's cumulated heating after uninterrupted machining several holes, raising the occurrence of damage as published by Fu et al.⁴ There is a more decisive influence of the cooled-air for $F_{daEnt 10}$ in comparison with $F_{daEnt 01}$ due to higher temperatures of the drill and not of the material. Yashiro et al.³² verified that the tool temperature is much higher than the CFRP temperature during machining.

When comparing $F_{daEnt 10}$ for dry cutting experiments versus cooled-air ones, all results showed lower values with cooled-air, reducing up to 25% of values for dry cutting (runs 7 and 16). The lower values in dry drilling were observed for low cutting speed, v_c = 25 m/min (runs 1, 2, and 3), which did not significantly raise the tool's temperature and showed similar values when using air-cooled drilling. As Ben et al.²² and Krishnaraj et al.²⁹ presented, an increment of the speeds raises the delamination in dry cutting due to the temperature's growth. The $F_{daE 10}$ differences between dry and air-cooled drilling also increased for v_c = 75 m/min, and improvements of 25%, 13%, and 15% were registered for f = 25, 50, and 75 µm/rev., respectively.

Examples of differences of induced delamination between the first and tenth holes and between dry and cooled-air conditions are observed in Figure 9, which shows the entrance of the holes for runs 4 (dry cutting with v_c = 50 m/min and f = 25 µm/rev.) and 13 (cooled-air cutting with v_c = 50 m/min and f = 25 µm/rev.). It is verified that the reduction of the plate stiffness (after nine holes) increases the peel-up occurrence, increasing the $F_{daEnt 10}$ value.

The $F_{daEx}$ of the first hole of each run ( $F_{daEx 01}$ ) was influenced by the combination of v_c x cc (59.2%). The cutting condition was partially significant (p-value = 0.067 and PC of 11.4%). The ANOVA showed that the $F_{daEx 01}$ values presented contrary trends in terms of v_c and cc. At lower v_c value, the usage of cooled-air generates lower $F_{daS 01}$ results, but at a higher v_c value, the dry condition provides lower $F_{daEx 01}$ results.

Four runs under air-cooled condition had the lower $F_{daS 01}$ values than those achieved by dry cutting. However, for runs 8, 9 (dry) and 17, 18 (air-cooled) (v_c = 75 m/min and f = 50, 75 µm/rev.), $F_{daS 01}$ values had increased up to 22% under cooled-air. It is concluded that $F_{daS}$ is mainly influenced by the thrust force.³³ It is important to highlight that the thrust force strongly depends on drilling parameters and it is indirectly controlled. Since the feed rate is the most significant parameter that affects thrust force, then thrust force can be indirectly controlled through the feed rate.^33,34 In dry cutting, the increase of cutting speed causes the growth of the temperature in the cutting zone, reducing the polymer matrix stiffness for high thrust force values, and when this force reduces, the delamination decreases.^35,36 However, the cooled-air usage reduces the temperature and its effects, including those related to the polymer matrix's stiffness. Thus, the increment of the thrust force under compressed air-cooling is higher than the observed for dry cutting.^35,37 This explains why at high cutting speeds and feed rates tested in runs 16, 17, and 18 (v_c = 75 m/min and f = 25, 50, and 75 µm/rev.), the thrust force grew even more,²⁶ increasing the delamination for runs 7, 8, and 9 using dry cutting.

Regarding the $F_{daEx}$ of the tenth hole of each run ( $F_{daEx 10}$ ), the ANOVA had R² = 54.7% (below the recommended value of 70%), showing that some $F_{daEx 10}$ results may not be conclusive. The only relevant factor was cc with p-value = 0.040 and 29% of percentage contribution. The factor f was almost significant since it had a p-value = 0.087 and 18.6% of PC. The $F_{daEx 10}$ results presented higher values compared to $F_{daEx 01}$ that shows a similar trend to the hole entrance results, even if the delamination phenomenon being different (peel-up at entrance and push-down at the exit). The air-cooled condition did not improve $F_{daEx 10}$ compared to the dry condition, increasing the delamination until 9% at runs 8 (dry) and 17 (air-cooled) (v_c = 75 m/min and f = 50 µm/rev.). The delamination is mainly caused by the thrust forces, which grow under the compressed air-cooling.^22,37 Regarding the feed rate and the cutting speed, there is no clear trend for $F_{daEx 10}$ .

The air-cooled condition increases the maximum delamination diameter at the hole exit. The fibers remained stiff and were warped in a region far from the hole edge, increasing the push-down occurrence. Nevertheless, the cooled-air reduces the occurrences of uncut fibers (Figure 10), as demonstrated by Koklu et al.³⁸

The ANOVA of the roughness parameters indicated that, for a 95% confidence interval, the only significant influence over R_z is cc (64.8%). Besides, the combination of v_c x cc was partially significant, with 18.4% of contribution. Considering total roughness (R_t), cc (46%), and the combination of v_c x cc (22.1%) presented significant influences on R_t values. As explained earlier, the R_a and R_q parameters that did not reach R² > 70% were unconsidered. However, the parameter R_t can be useful to evaluate the surface quality.²⁶

The increase of cutting speed (v_c) harmed the surface quality because the CFRP plate used is thick, contributing to the temperature growth in the cutting zone. Thus, it increased the surface quality's adverse effects, justifying the compressed air-cooling as the most influential factor for hole wall roughness. These results agree with those observed by Abish et al.³⁷

During the dry drilling, the lower cutting speed (v_c = 25 m/min) produced better results compared to the others v_c tested due to the lower temperatures,^22,39 and thrust forces developed,³² which reduce the damage. Consequently, the increase of v_c makes the temperature rise, which harms the surface. It is noted that run 9 (v_c = 75 m/min) generated better results than the other dry runs and similar results using cooled-air. El-Hofy et al.⁴⁰ argued that this phenomenon is due to the higher feed rate (f = 75 µm/rev. in run 9) that reduces the temperature generated by high cutting speed combined with plate thickness.

For the runs with higher v_c and f, the compressed-air cooling (runs 13 to 17) resulted in lower roughness values compared to dry cutting (runs 4 to 8), showing up to 58% of decrement for R_a, 52% for R_z, 57% for R_q and 57% for R_t. The air-cooled condition reduced the damage related to the temperature for high speeds, which generated better results since the high speed is beneficial to the hole wall roughness.^26,41 Joshi et al.³⁵ and Abish et al.³⁷ also obtained lower roughness values using air-cooled than dry drilling. Run 18 had similar results to run 9, with a slight reduction in the values of roughness parameters in comparison with dry cutting (5% for R_a, 11% for R_z, 7% for R_q and R_t) due to high cutting speed and feed rate (v_c = 75 m/min and f = 75 µm/rev.). Joshi et al.³⁵ observed that the cutting zone's temperature negatively affects the surface finish for high cutting speed because the cooled-air could not reduce the temperature, showing similar results when in dry drilling.

Figure 11 shows SEM images of the holes' walls for runs 4 (dry) and 13 (air-cooled). It is noticed a high variation between dry and air-cooled drilling. Various cavities were identified in dry drilling, as well as fiber pull-out from the matrix, with high variation between the laminate plies. A high degradation of the polymer matrix in dry drilling is noticed, presenting the fibers exposed without polymer matrix. This difference between the machined surfaces using dry and air-cooled drilling was observed, not only for runs 4 and 13, but also in all runs for v_c = 50 and 75 m/min.

Low temperatures facilitate the occurrence of fibers' breakage since they reduce their resilience under air-cooled. These ruptures avoid the pull-out phenomenon, generating fewer rough surfaces after drilling. In this case, the occurrence of fiber pull-out generates high cavities on the surface. In dry drilling, the fiber pull-out and the fiber-matrix separation increase with the temperature, deteriorating the machined surface.^{4,35,37,42,43}

Another observation about the hole wall roughness was the high variability in the same hole for three measurements with different profiles found. Eneyew and Ramulu²⁶ explained that these variations could occur in locations where the carbon fibers pulled out, creating valleys, splinters, and cracks.

Figure 12 shows the lowest and the highest roughness values for the three measured profiles considering runs 4 (dry) and 13 (air-cooled) in the drilling with v_c = 50 m/min and f = 25 µm/rev. Figure 12(a) presents the high variability observed for dry drilling. Also, the air-cooled condition results showed similar roughness profiles throughout the hole wall surface (Figure 12(b)).

Figure 13 presents the runs 6 (dry) and 15 (air-cooled) with v_c = 50 m/min and f = 75 µm/rev., showing the differences between the first and tenth holes. It is observed that for both runs (dry and air-cooled), the surface damage increases with the number of uninterrupted holes, probably due to the cumulated heat. This phenomenon is analogous to delamination occurrences.

Conclusions

The present work investigated delamination and hole wall quality of CFRP plate drilling under different cooling conditions (cc), cutting speeds (v_c), and feed rates (f). After measuring data to calculate the adjusted delamination factor ( $F_{d a}$ ) as well as roughness parameters (R_a, R_q, R_z, R_t), the ANOVA was performed, and relevant aspects were observed:

Regarding the $F_{d a}$ in the hole entrance, it was concluded that v_c, f, cc, and the combination f x cc were influent factors in the first hole for a 95% confidence interval. The cooled-air application helped decrease the tenth hole's delamination, but the increase in v_c showed a negative influence. Therefore, higher values of v_c can produce more delamination.

Concerning $F_{d a}$ in the hole exit, the combination v_c x cc was significantly influent in the first hole, while the tenth hole was influenced significantly by cc, but they showed R² below 60%. Besides, in this case, the delamination is highly dependent on the temperature. In fact, it is recognized that the temperature can promote changes in terms of the mechanical properties at the polymer matrix of the composite plate. For the cooled drilling, the authors understand that the temperature is not necessary to be measured. However, for the dry drilling, future investigations are required.

Regarding the hole wall roughness, higher v_c values improved its quality but provided worse results for $F_{d a}$ in the hole entrance. Moreover, R_z values were influenced by cc, and R_t values were influenced by cc and the combination v_c x cc.

Finally, on the one hand, the compressed air-cooling application reduced the delamination at the entrance (which is confirmed by the literature⁴⁴) of the holes. Therefore, the hole wall roughness is improved since the fibers' resilience is modified due to the air-cooled, facilitating the drilling and reducing the fiber pull-out and damage due to high cutting speeds. On the other hand, the delamination at the hole exit is increased when push-down effects increase compared to dry cutting. This may occur due to the change in the material properties, mainly polymer matrix, under low temperatures. However, the variation of cutting speed (v_c) combined to the variation of feed rate (f) (at the same time) in drilling is a promising alternative to decrease the delamination at the hole exit. Thus, if small values for v_c are used during the hole opening, and, after that, values of v_c are increased and values of f are reduced when the drill point approaches to the end of the drilled hole, then better results can be obtained in terms of reduction of delamination.

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: The authors thank Sandvik Coromant Co. for the cutting tool, Green Service Co. for the vortex tube, and CAPES (grant no. 88882.346387/2019-01) and CNPq (process no. 310656/2018-4 and 134587/2018-9) for financial support.

ORCID iDs

Nícholas Hoffmann

Gabriel SC Souza

Volnei Tita

References

Beaumont

PWR

Soutis

Hodzic

(eds). The structural integrity of carbon fiber composites: fifty years of progress and achievement of the science, development, and applications. Switzerland: Springer International Publishing, 2017.

Alizadeh Ashrafi

Miller

Wandro

, et al. Characterization and effects of fiber pull-outs in hole quality of carbon fiber reinforced plastics composite. Materials 2016; 9: 828.

Haeger

Schoen

Lissek

, et al. Non-destructive detection of drilling-induced delamination in CFRP and its effect on mechanical properties. Proc Eng 2016; 149: 130–142.

Jia

Wang

, et al. Drill-exit temperature characteristics in drilling of UD and MD CFRP composites based on infrared thermography. Int J Mach Tools Manuf 2018; 135: 24–37.

Iliescu

DMFD

Gehin

Gutierrez

, et al. Modeling and tool wear in drilling of CFRP. Int J Mach Tools Manuf 2010; 50: 204–213.

Amini

Baraheni

Hakimi

Enhancing dimensional accuracy and surface integrity by helical milling of carbon fiber reinforced polymers. Int J Lightweight Mater Manuf 2019; 2: 362–372.

Khairusshima

Hassan

Jaharah

, et al. Effect of chilled air on tool wear and workpiece quality during milling of carbon fibre-reinforced plastic. Wear 2013; 302: 1113–1123.

Faraz

Biermann

Weinert

Cutting edge rounding: an innovative tool wear criterion in drilling CFRP composite laminates. Int J Mach Tools Manuf 2009; 49: 1185–1196.

Wang

Kwon

Sturtevant

Kim , et al. Comparative tool wear study based on drilling experiments on CFRP/Ti stack and its individual layers. Wear 2014; 317: 265–276.

10.

Rawat

Attia

Wear mechanisms and tool life management of WC–Co drills during dry high speed drilling of woven carbon fibre composites. Wear 2009; 267: 1022–1030.

11.

Liu

Tang

Cong

WL.

A review of mechanical drilling for composite laminates. Compos Struct 2012; 94: 1265–1279.

12.

Durão

LMP

Tavares

JMRS

de Albuquerque

VHC

, et al. Drilling damage in composite material. Materials (Basel) 2014; 7: 3802–3819.

13.

Geier

Davim

Szalay

Advanced cutting tools and technologies for drilling carbon fibre reinforced polymer (CFRP) composites: a review. Compos A: Appl Sci Manuf 2019; 125: 105552.

14.

Davim

Reis

Drilling carbon fiber reinforced plastics manufactured by autoclave – experimental and statistical study. Mater Des 2003; 24: 315–324.

15.

Chen

WC.

Some experimental investigations in the drilling of carbon fiber-reinforced plastic (CFRP) composite laminates. Int J Mach Tools Manuf 1997; 37: 1097–1108.

16.

Davim

Rubio

Abrão

AM.

A novel approach based on digital image analysis to evaluate the delamination factor after drilling composite laminates. Compos Sci Technol 2007; 67: 1939–1945.

17.

Koplev

Lystrup

Vorm

The cutting process, chips, and cutting forces in machining CFRP. Composites 1983; 14: 371–376.

18.

El-Hofy

Soo

Aspinwall

, et al. Tool temperature in slotting of CFRP composites. Proc Manuf 2017; 10: 371–381.

19.

Wang

Kwon

Sturtevant

, et al. Tool wear of coated drills in drilling CFRP. J Manuf Process 2013; 15: 127–135.

20.

Akay

Aspects of dynamic mechanical analysis in polymeric composites. Compos Sci Technol 1993; 47: 419–423.

21.

Kim

Yang

, et al. Influence of cutting temperature on carbon fiber-reinforced plastic composites in high-speed machining. J Mech Sci Technol 2017; 31: 1861–1867.

22.

Ben

Hang

Quan

, et al. Influence of cutting heat on quality of drilling of carbon/epoxy composites. Mater Manuf Process 2012; 27: 968–972.

23.

Dixit

Sarma

Davim

JP.

Environmentally friendly machining. New York: Springer-Verlag, 2012.

24.

Nguyen-Dinh

Hejjaji

Zitoune

, et al. New tool for reduction of harmful particulate dispersion and to improve machining quality when trimming carbon/epoxy composites. Compos A: Appl Sci Manuf 2020; 131: 105806.

25.

Ngoc

Hue

Van Hung

, et al. Dust emission during machining of CFRP composite: a calculation of the number and mass of the thoracic particles. In: International conference on engineering research and applications. Cham: Springer, 2020, pp.341–349.

26.

Eneyew

Ramulu

Experimental study of surface quality and damage when drilling unidirectional CFRP composites. J Mater Res Technol 2014; 3: 354–362.

27.

Montgomery

DC.

Design and analysis of experiments. 8th. ed. Hoboken, NJ: John Wiley & Sons, 2013.

28.

Geng

Liu

Shao

, et al. Delamination formation, evaluation and suppression during drilling of composite laminates: a review. Compos Struct 2019; 216: 168–186.

29.

Krishnaraj

Prabukarthi

Ramanathan

, et al. Optimization of machining parameters at high speed drilling of carbon fiber reinforced plastic (CFRP) laminates. Compos B: Eng 2012; 43: 1791–1799.

30.

Khashaba

El-Sonbaty

Selmy

, et al. Drilling analysis of woven glass fiber-reinforced/epoxy composites. J Compos Mater 2013; 47: 191–205.

31.

Velayudham

Krishnamurthy

Effect of point geometry and their influence on thrust and delamination in drilling of polymeric composites. J Mater Process Technol 2007; 185: 204–209.

32.

Yashiro

Ogawa

Sasahara

Temperature measurement of cutting tool and machined surface layer in milling of CFRP. Int J Mach Tools Manuf 2013; 70: 63–69.

33.

Sorrentino

Turchetta

Bellini

A new method to reduce delaminations during drilling of FRP laminates by feed rate control. Compos Struct 2018; 186: 154–164.

34.

Karimi

Heidary

Minak

Critical thrust and feed prediction models in drilling of composite laminates. Compos Struct 2016; 148: 19–26.

35.

Joshi

Rawat

Balan

ASS.

A novel approach to predict the delamination factor for dry and cryogenic drilling of CFRP. J Mater Process Technol 2018; 262: 521–531.

36.

Karnik

Gaitonde

Rubio

, et al. Delamination analysis in high speed drilling of carbon fiber reinforced plastics (CFRP) using artificial neural network model. Mater Des 2008; 29: 1768–1776.

37.

Abish

Samal

Narenther

, et al. Assessment of drilling-induced damage in CFRP under chilled air environment. Mater Manuf Process 2018; 33: 1361–1368.

38.

Koklu

Morkavuk

Cryogenic drilling of carbon fiber-reinforced composite (CFRP). Surf Rev Lett 2019; 26: 1950060.

39.

Pérez-Pacheco

Moreno-Chulim

Valadez-González

, et al. Effect of the interphase microstructure on the behavior of carbon fiber/epoxy resin model composite in a thermal environment. J Mater Sci 2011; 46: 4026–4033.

40.

El-Hofy

Soo

Aspinwall

, et al. Factors affecting workpiece surface integrity in slotting of CFRP. Proc Eng 2011; 19: 94–99.

41.

Liu

Xie

Sun

, et al. Investigations on micro-cutting mechanism and surface quality of carbon fiber-reinforced plastic composites. Int J Adv Manuf Technol 2018; 94: 3655–3664.

42.

Hou

Qiu

Zhang

, et al. Comparative tool wear and hole quality investigation in drilling of aerospace grade T800 CFRP using different external cooling lubricants. Int J Adv Manuf Technol 2020; 106: 937–951.

43.

Cai

, et al. Drilling machinability evaluation on new developed high-strength T800S/250F CFRP laminates. Int J Precis Eng Manuf 2013; 14: 1687–1696.

44.

Campos Rubio

Abrão

Eustáquio Faria

, et al. Delamination in high speed drilling of carbon fiber reinforced plastic (CFRP). J Compos Mater 2008; 42: 1523–1532.