**1. Introduction**

Carbon fibre reinforced polymer (CFRP) composites are widely used in the aerospace industry. In this sector, the prevalent machining process performed on CFRP components is drilling, as mechanical joining by means of rivets or bolts represents the most common joining technique for aircraft components. However, the drilling of CFRP is very challenging in comparison with the drilling of metals, as the phenomena underlying the material removal for composite materials are substantially different from those characteristic of metal machining [1]. The different properties of the reinforcement and matrix phases make the material removal mechanism highly complex, because of the heterogeneity and anisotropic behaviour of composites [1–3]. König [4] highlighted how the machining of composite materials depends on the specific properties and relative content of the reinforcement and matrix. Moreover, when machining fibre reinforced composite materials, fibre orientation plays a fundamental role, affecting the mechanism of chip formation and the cut surface quality. During the drilling process, because of the heterogeneous and anisotropic behaviour of the fibre reinforced composites, different kinds of damage can be generated in the workpiece, such as delamination, fibre pull-out, fibre breakage, matrix cracking, and thermal damage [5,6], which can affect the drilled component service life. The most relevant damage induced during CFRP drilling is delamination [7,8], which has been considered as the principal cause of notable reduction in the fatigue strength of

composite components, cutting down the long-term performance of the CFRP parts. In particular, delamination onset at the hole exit, also known as push-out delamination, is generated if the axial load exerted on the workpiece during drilling exceeds a threshold value [9]. As demonstrated by Davim and Reis [10], the delamination damage grows with increasing cutting speed and feed values. As a matter of fact, a feed rate increase raises the thrust force, with consequent delamination damage enlargement at the hole exit [11].

Moreover, tool geometry significantly affects the drilling induced damage [10,12]. Designing new tools with a better performance in terms of cost, damage reduction, and hole quality is a key factor for drilling process optimisation [12]. Piquet et al. [13] analysed the effects of drilling tool geometry, comparing the results obtained with a traditional tool and a specially designed cutting tool. Feito et al. [14] and Gaitonde et al. [15] demonstrated how the delamination increases with increasing the point angle of the twist drill bit for high speed drilling of CFRP laminates. Saoudi et al. [16] developed a model to predict the critical thrust force responsible for exit-ply delamination during the drilling of multi-directional carbon fibre-reinforced plastic laminates with core drills made of diamond grits. Cadorin et al. [17] studied the mechanisms of damage and tool wear in the drilling of three-dimensional (3D) woven composite materials using diamond-coated carbide three lips twist drills. Zitoune et al. [18] investigated the influence of the tool coating on temperature and tool wear when drilling 3D woven composite materials using three types of cutting tools, one coated with a diamond layer and the other two coated with different nano-composite multi-layers, showing that the nature of the coating is a key factor affecting the temperature of machining.

To reduce the delamination and thermal damage defects generated when drilling composites, a new process called orbital drilling (OD) has been proposed in the literature. This process is more versatile than traditional drilling, as it allows for obtaining holes with different diameters without replacing the tool. OD is a circular ramp machining process, consisting of milling with a discontinuous peripheral cut and drilling with a continuous cut along the cutting-edge axis at the same time. Hence, this technology involves the simultaneous movement along a circular path (*X*, *Y*) and an axial advancement (*Z*) with a pre-determined step [19].

In OD, the orthogonal thrust exerted on the surface is very low, resulting in a superior hole quality, especially in the case of the composite materials subject to delamination damage. This type of machining process is particularly suitable for applications in the aeronautical field, as it allows for a decrease of cutting forces and temperatures, which results in a reduction of polymer matrix damage [20]. On the other hand, the main drawbacks related to OD concern the difficulties in selecting the proper process parameters and the need to employ a three-axis machining centre to avoid vibration problems. OD operations are carried out by portable and highly flexible machines, which tend to exhibit severe chatter and forced vibrations that lead to a poor hole quality. Moreover, the thrust force responsible for delamination is reduced but still not eliminated with OD [21]. As a matter of fact, while in traditional drilling, the axial loading consists of a concentrated load at the centre of the last ply of the laminate, in orbital drilling, the axial loading is determined by an eccentric distributed load, which acts along the cutting-edge of the mill [20].

As an alternative to both traditional drilling and orbital drilling processes, this research work proposes an innovative hole making process for CFRP components, where the hole is realised by a combination of drilling and peripheral milling, carried out using the same cutting tool. The objective is to further reduce the process-induced delamination damage in comparison with traditional and orbital drilling. To perform the new combined drill-milling process, an innovative drill-milling tool was developed. The main results of the experimental drill-milling tests on the CFRP composite laminates are reported in this paper, showing encouraging results in terms of the hole surface finish. The new drill-milling process is performed under dry conditions in a green technology perspective, offering advantages because of the absence of cutting fluids and a lower environmental impact.

#### **2. Proposed Drill-Milling Process**

In conventional drilling of fibre reinforced composite materials, in accordance with the model of linear elastic fracture mechanics, the delamination onset at the hole exit, also known as push-out, occurs if the axial load acting on the workpiece exceeds a threshold value [9,22]. The axial load performs a work equal to the sum of the energy necessary to deform the last ply of the laminate and the energy required to generate a new fracture surface [23].

For traditional drilling, the axial load evaluation generally refers to a concentrated load at the centre of the last ply of the laminate, as studied in the delamination analysis by Hocheng et al. [9,24]. Instead, for orbital drilling, an energy criterion characterized by the application of an eccentric distributed load, which acts along the cutting-edge of the mill, is adopted. This means that the thrust force causing delamination is reduced but not eliminated, as reported in Figure 1a,b [20].

To fully eliminate the drilling-induced delamination, a different original technique for CFRP hole making is proposed in this work, where the hole is realised by means of a combination of drilling and peripheral milling processes. In Figure 1c, the two phases of the process are illustrated, namely: in the first phase traditional drilling is employed to produce a hole with a smaller diameter than the final hole, and then, in the second phase, the material around the first hole is removed by peripheral milling until the final diameter is achieved. These two phases are sequentially combined in a single process carried out using the same cutting tool.

**Figure 1.** Representation of (**a**) traditional drilling; (**b**) orbital drilling [20] and (**c**) drill-milling processes. The red arrows indicate the load.

To develop this technique, an innovative drill-milling tool was designed and realized, and experimental cutting tests on CFRP laminates were performed to characterize the process. During the machining tests, the thrust force and radial force signals were detected through a sensor system to allow for the analysis of the cutting forces occurring during the process. Moreover, the produced holes were characterised via metrological and roughness measurement procedures to evaluate the hole quality in terms of size, roundness, and internal surface finish.

#### **3. Materials and Methods**

#### *3.1. Preparation of the Specimen*

The work-material employed for the drill-milling tests consisted of CFRP laminates composed of 26 prepreg plies, made of CYCOM 977-2, Cytec Industries Inc., Woodland Park, NJ, USA epoxy matrix, and Toray T300 unidirectional carbon fibres (Toray Industries, Inc., Tokyo, Japan), with stacking sequence [±45/0/±45/90/±45/0/−45/90/45/90]s. The main mechanical properties of the single unidirectional plies are reported in Table 1.


**Table 1.** Main mechanical properties of the unidirectional plies.

The nominal thickness of each CFRP laminate was 5 mm, and a thin fiberglass/epoxy ply reinforced with 0◦/90◦ fabric was laid on the laminate front and back surfaces. In the aeronautical sector, it is common practice to apply one or more lightweight fiberglass fabric plies on the front and back surfaces of the laminate in order to limit the corrosion occurring as a result of the connection of materials of a different nature. The laminates were fabricated according to the procedure employed in the aeronautical industry, consisting of hand lay-up of prepreg plies, vacuum bag moulding, and curing cycle in autoclave. The curing cycle was controlled through the use of sensors within the autoclave, and it was performed according to the cycle suggested by the CYCOM 977–2 resin matrix producer. Heating was carried out up to 177 ◦C with speed equal to 2 ◦C/min, then, the laminate was held at 177 ◦C for 2 h, and eventually, a natural cooling in autoclave was accomplished.

Because of the fabrication process, the bag side laminate surface was irregular compared to the smooth mould side surface. To perform the drill-milling tests, specimens of 30 mm × 400 mm were cut from the original CFRP laminates.

#### *3.2. Equipment*

The drill-milling tests were performed under dry conditions on a five-axis Cosmec Conquest 3200 NC machining centre (Poggibonsi, Siena, Italy), equipped with a sensor system to detect the cutting force signals generated during the process (Figure 2). The signals of the thrust force, Fz, and radial force, Fx, were detected by a Kistler (Kistler Group, Winterthur, Switzerland) three-axis stationary dynamometer model 9257A, positioned under the workpiece fixture. The sensor system included a Kistler charge amplifier model 5007, and a National Instruments data acquisition board model 9239 (Austin, TX, USA) that digitized the Fx and Fz signals at 10 kS/s.

**Figure 2.** Equipment employed for the experimental drill-milling tests.

The radial force can be resolved into two components acting along the *X* and *Y* dynamometer directions, respectively. As the process is symmetrical in the CFRP laminate plane, the values of the radial forces were similar, so that it was decided to report only the values recorded along the *X* direction.

All of the tests were carried out using ISO CNC programming. According to the new drill-milling process, the first operation is the drilling of a pilot hole with a nominal diameter of 6 mm, followed by a peripheral milling operation. Specifically, once the pilot hole is realised, the tool follows a spiral milling path to increase the hole diameter by 0.10 mm per revolution, up to a hole diameter of 6.9 mm. Finally, three boring revolutions at constant diameter are executed. The process phases are shown in Figure 3.

In the experimental tests, the parameters selected for the drilling phase were chosen so as to limit the drilling-induced delamination damage to an extent that is lower than the radius of the material to be removed by the peripheral milling. In particular, the drilling speed was set to 10,000 rpm and the axial feed to 375 mm/min, while the milling speed was set to 14,000 rpm and the axial feed to 375 mm/min.

The drill-milling tests were conducted on 60 consecutive holes to analyse the process behaviour with a tool wear increase.

**Figure 3.** Representation of the following (**a**) tool parts and process phases; (**b**) drilling and (**c**) milling.
