Machinability of a silicon carbide particle-reinforced metal matrix composite

Abstract

This paper presents a numerical test program to model the microstructural change associated with the machining of a particle-reinforced metal matrix composite (PRMMC). Composite materials comprising 30% and 40% volume fraction of SiC particle-reinforced 2024 aluminum alloy were modeled separately. The random particles model was adopted to analyze the micro-mechanical problems in PRMMCs. Simulation of the metal cutting process was performed using ABAQUS, an elasto-viscoplastic finite element analysis (FEA) code. Two-dimensional finite element (FE) simulations of the deformation and damage evolution in a SiC particle-reinforced aluminum alloy composite, including interphase, were performed for different microstructures and particle volume fractions of the composites. The FEA results were found to be similar to those experimentally obtained from the machined test pieces using a scanning electron microscope (SEM). On the basis of the simulation and experimental results, we obtained the effect of machined surface quality and cutting parameters on the residual stress distribution in the composite.

---

1. Introduction

Metal matrix composites (MMCs) are a relatively new class of materials characterized by lower density, greater strength, and superior wear resistance when compared to conventional materials. In particular, particulate-reinforced metal matrix composites (PRMMCs) with exceptional strength and stiffness are being widely used in diverse fields, including the aeronautic and aerospace industries.^1,2 However, the machining of MMCs is rather difficult because of the highly abrasive and intermittent nature of the reinforcements. Conventional tool materials such as high-speed steel cannot be used for machining MMCs as the cutting tool undergoes very rapid wear. On the other hand, carbides, both plain and coated, sustain significant levels of tool wear after a very short period of machining.³ To investigate the machinability of MMCs, some studies have focused on the effects of machining parameters and the associated properties of MMCs on the tool wear and the mechanism underlying this phenomenon. For instance, Karkas et al. investigated the effect of cutting speed on tool performance during the milling of B4C_p-reinforced aluminum MMCs.⁴ Zhou et al. have conducted (2D) orthogonal cutting experiments.⁵ Both of the results indicated that the cutting speed and depth had significant effects on the cutting force for MMCs. Similarly, Ding et al. compared and evaluated the MMC machining performances of different tools.⁶ According to the study reported by Chennakesavarao et al., crater wear was not appreciable in K10 tools. Besides, the tools had superior wear resistance and produced continuous chips.⁷ Qu et al. investigated the bonding mechanism, melting and expansion behavior of the SiC_p/Al composite-bismuthate.⁸ Lasri et al. focused on the effect of cutting parameters on surface damage of the machined component and wear resistance.⁹ Palanikumar analyzed the factors influencing tool wear on the machining of glass-fiber-reinforced plastic composites using coated cement carbide tools.¹⁰ Furthermore, some studies can describe many aspects of the machining process. For example, Hocheng et al. studied the effects of speed, feed, depth of cut, rake angle, and cutting fluid on the chip form, forces, wear, and surface roughness of the resulting composite.¹¹ Similarly, Chambers analyzed the tool life, surface quality, and cutting forces for MMCs.¹² Yuan and Dong investigated the effect of reinforcement volume percentage, cutting angle, feed rate, and speed on the surface integrity in ultra-precision diamond turning of MMCs.¹³ El-Gallab and Sklad compared the effectiveness of several tool materials.¹⁴ On the basis of the techniques proposed by Taguchi, Davim studied the influence of cutting speed, feed rate, and cutting time on turning MMCs (A356/20SiC_p-T6) using polycrystalline diamond (PCD) cutting tools.¹⁵ Some other parameters have also been investigated. Muthukrishnan et al. attempted to study the machinability issues of Al/SiC MMCs in turning using different grades of poly crystalline diamond (PCD) inserts.¹⁶ Palanikumar and Karthikeyan investigated the factors influencing the surface roughness during the machining of Al/SiC particulate composites using tungsten carbide tool inserts (K10).¹⁷ Soldani et al. analyzed the influence of mesh size and shape in finite element modeling of composite cutting.¹⁸ They also studied the tool geometry and its effect on composite cutting. Dabade et al. studied surface integrity as a function of process parameters and tool geometry by analyzing the cutting forces, surface finish, and microstructures of the machined surfaces of SiC/Al/10p and SiC/Al/30p composites machined using cubic boron nitride (CBN) inserts.^19,20 Most of the abovementioned studies investigated the wear characteristics of various tool materials as functions of the cutting parameters and surface finish during the machining of aluminum-based composites reinforced with SiC particles. These researches only focus on the effects of cutting parameters such as cutting speed, feed rate, and the depth of cut. However, the effects of the mass fraction of SiC particles and the particle shape on the material state and surface stress have not yet been studied.

Lately, with the development of computer technology, several studies are focusing on the simulation and calculation of SiC PRMMC. Given that PRMMCs have reinforced particles, the simulation of the cutting of PRMMC is considered to be a simultaneous occurrence of microscopic and macroscopic processes. Therefore, conventional analytical methods and models are not considered feasible for simulating the processes. Therefore, some studies have adopted finite element method (FEM) to simulate the processes. Yuan et al. numerically analyzed the stress–strain distribution in the PRMMC SiC_p/6064AL and established a micromodel on the basis of the cohesive zone model.²¹ Aghdam developed a three-dimensional (3D) finite element micromechanical model for the failure analysis of SiC_p/Ti–6Al–4V PRMMCs.²² Using the commercial FE finite packages Third Wave Systems AdvantEdge and ABAQUS/Explicit, Dandekar et al. developed a multistep 3-D finite element model for predicting the subsurface damage after machining of PRMMCs.²³ Sun et al. presented a bridging and cohesive zone model for analyzing the composite fracture.²⁴ Based on particle distribution, Ayyar et al. established uniform distributed, stochastic distribution, and reunion distribution models to study the mechanical properties of composite materials.²⁵ Pramanik et al. performed simulations to study the effects of composite cutting tools and their enhanced particles on the formation mechanism of cutting and also studied the non-traditional machining of PRMMCs.^26,27 Ramesh et al. used a dynamic finite element method for analyzing the mechanism underlying the diamond cutting of aluminum-based SiC composites.²⁸ On the basis of these results, they derived the distribution range of cutting forces and stresses during the cutting process. Dandekar et al. developed a multiscale finite element model to predict the postmachined subsurface damage in PRMMCs subjected to laser-assisted machining.²⁹ Using the finite element method and axisymmetric model of a unit cell, Sun Chao et al. simulated the effects of a strengthened body shape (SiC_p), volume fraction, and different matrix types on the mechanical behavior of aluminum matrix composites.³⁰ Song Wei Dong et al. proposed a new FEM for simulating the real structure of particle-reinforced composites; however, it is an ideal geometric model.³¹ Therefore, the simulated model cannot be used in engineering practice.

In this study, we fully consider comparison of simulation and engineering practice for the machinability. We establish a model of SiC_p/Al composite from the mesoscale perspective. Accordingly, we propose a material model of an algorithm for volume and particle shape parameterization and make a program to obtain two-dimensional (2D) model of random distribution of particles are generated with different volume fraction. Based on the model, the cutting process is simulated. The results show the influence of particle shape, size and volume fraction on mesoscopic cutting forces, cutting stresses, and machined quality of MMCs. For example, these particles with sharp corners shape are not exfoliated easily but are only pressed into the machined surface. It will enlarge tool wear and cutting stress, but be benefits for strength material. The study explores some typical problems encountered during the cutting process with regard to factors such as the machining surface quality, cutting forces and surface cutting stresses. The outcome of this study is expected to optimize and improve machinability of SiC particle-reinforced aluminum alloy composites (SiC_p/Al composites) and lay the foundation for future studies on the performance and mechanism underlying the machining of particle-reinforced composite materials using finite element methods.

2. Mesoscopic finite element modeling of SiC_p/Al composites

The mesoscopic FEA of particle-reinforced composites usually requires a representative unit that is sufficiently large to include the typical mesostructure of composite materials. The representative unit illustrates a material point in the composite structure and hence should be sufficiently small so that the entire composite structure can be used as a continuum. Therefore, the rectangular material is 500 μm long and 300 μm wide, with an average particle size of 10 μm and a volume fraction of 30%.

To simulate the structure relation between the SiC particles and aluminum matrix more realistically and to achieve parameterized shapes and required volume fractions of the particles, we used the random sequential adsorption (RSA) method to generate particles and hybrid fiber-reinforced composite materials and to build geometric models, where the shape and volume could be determined. The modeling process is shown in Fig. 1. Three geometric models that deal with the following functions can be obtained by the abovementioned process, as shown in Fig. 2: the calculation of the particle area in an arbitrary polygon particles model; calculation of the area criterion content using the algorithm of particle interference according to the literature;³² and aggregation of the concrete content using random models. The three finite element models algorithms are based on Python, the ABAQUS scripting language.

Fig. 1 Flowchart illustrating the steps in the generate particles geometric model.

Fig. 2 SiC_p/Al composite random distribution micro-model: (a) circular particle model; (b) quadrated particle model; (c) triangle particle model; (d) arbitrary polygon particles model.

According to practical experience and existing research, the junction between the particles and matrix, which is often considered the source of cracks, is not only the weakest part of PRMMCs, but also is the part with the largest stress concentration during the cutting process. Hence, this region needs considerable attention. To guarantee precise calculations and to avoid failures in calculations due to grid distortion, while considering cost reduction, it is necessary to subdivide the substrate area above the cut line grid, especially at the particle edges. In general, SiC particles have a higher rigidity and smaller deformation considering the elastic deformation in the cutting process. Therefore, the particle mesh should be coarse, and particle mesh above the cut line grid should be more densely distributed than that at other regions. Fig. 3 shows the arbitrary polygonal particles mesh model of the composite material.

Fig. 3 Grid division of the arbitrary polygon particles model.

The interface of a composite material plays a key role in determining its mechanical properties. According to previous studies, PRMMCs prepared by liquid penetration method have higher interfacial strength, whereas the strength of general particles prepared by powder metallurgy and extrusion processes is lesser at the interface.³³ To simulate the bond relation between the particles and matrix and the instantaneous separation of particles from the matrix in the cutting process, the interface is considered to be welded up between the particles and substrate and to have a high strength equivalent to a part of the particle. When the particles around the base unit do not work, the particles tend to separate from the matrix, which defines the contact friction relation between the particles and matrix and also that between the particles. Therefore, the forcing passes could be achieved during the cutting process.

The matrix material considered in this study is 2024Al. We selected the Johnson–Cook (J–C) materials constitutive equations of thermoviscoplastic deformation to describe the material under large deformation, high temperature, and high strain rate.³⁴ The materials are cut along the shear surface under continuous shear slippage. Hence, the definition of damage in the simulation is shear damage. The hardness and stiffness of the SiC particles are much higher than that of the matrix metal, and which has a smaller size in the model. Therefore, we can consider that the constitutive relation of the SiC particles is isotropic linear elastic. The mechanical and physical parameters of 2024Al and SiC_p and the parameters of the J–C model are shown in the Tables 1–3.

Table 1 Physical parameters of 2024Al

Density (g mm^−3)	Elastic modulus (MPa)	Hardness (HB)	Young's modulus (MPa)	Plastic hardening modulus (MPa)	Elongation rate	Poisson's ratio	Yield stress (MPa)	Heat capacity (J kg °C^−1)
2.82 × 10^−3	70 000	120	68 000	1480	18	0.3	355	660

---

Table 2 2024Al Johnson–Cook material parameters

A (MP)	B (MP)	m	N	C	ε0	Tmelt (K)	Ttransition (K)
218	546	0.38	3.73	0.14	1	775	298

---

Table 3 Physical parameters of SiC_p

Density (g mm^−3)	Expansion coefficient (μm m^−1 k^−1)	Young's modulus (MPa)	Hardness (HB)	Plastic hardening modulus (MPa)	Poisson's ratio	Yield stress (MPa)	Heat capacity (J kg^−1 °C^−1)
3.22 × 10^−3	4.4	410 000	580	0.14	0.35	670	1600

---

The actual microstructure of SiC_p/Al (Fig. 4(a)), observed using SEM, is similar to that of the FE arbitrary polygon particles model. Therefore, we applied the arbitrary polygon particles model to analyze the cutting process. To simplify the simulation, the cutting tool of PRMMC is defined as a rigid body. During the tool loading process, the node of the materials is constrained on the left and at the bottom to simulate the chucking of the workpiece such as that in real cutting. The cutting speed is set as 300 m s⁻¹, the cutting depth a_p is 0.1 mm, and the orthogonal rake angle of tool r is 20°. The orthogonal clearance of the tool is 10°, the radius of the edge is 5 μm, and the volume fraction is 40%. The average dimension of the particles is 20 μm.

Fig. 4 Comparison of FE model and the representative SEM image: (a) SEM photograph of SiC_p/Al composite; (b) FE model of the cutting process (volume fraction is 40%).

3. Simulation of the cutting of SiC_p/Al composites

Fig. 5 shows the stress distribution during the cutting process. As shown, the overall stress distribution is similar to that of homogeneous materials. The stress is mainly located in three major deformation zones, with the SiC particles being the primary cause of stress. Under the delivery of the matrix, the stress in the first deformation zone (shear front) extends to the machined surface bottom, with the first deformation zone featuring higher stress on the left.

Fig. 5 Distribution of stress during the cutting process.

The maximum values of the PRMMCs stress occurred at the interface between the particles and matrix in the shear plane during the cutting process usually, so the cutting workpiece through shear sliding would be difficult to complete and continuous. The results of the simulation showed that the cutting chips are blocky or powder granular. The results of comparison between homogeneous materials and^35,36 the PRMMCs chips are shown in Fig. 5(b).

Compared with the first and second deformation zones, the stress in the third deformation zone is relatively smaller. The stress field is complex during the cutting process, wherein the particle shape and distribution influence the delivery direction and force size. Given the SiC particle exhibit higher hardness, the particles exist around the blade equivalent to that of tool extension during the cutting process and tend to change the instant cutting parameters in a disguised form. Therefore, the stress field of the cutting material will change with variations in the shape and distribution of the particles and with changes in the relative positions between the particles and cutting tool.

During the cutting process, the particles located above the cutting path, which are under the extrusion friction effect of rake face, tend to roll and peel off from the matrix. Subsequently, the particles move to the second deformation zone with the cutting chips and generate great compressive stress under the squeezing action of the cutting tool and other particles and under matrix compression. This will extend the shear crack but will lead to the deformation of particles or crushing and abrasive wear. The other parts of the particles are on the cutting path, which flip and slide under the action of the blade. As a result, the particles are stripped off the matrix, leading to the formation of machined surface pits, as shown in Fig. 6(a).

Fig. 6 Simulation of the cutting transient state: (a) machined surface pits; (b) interaction between particles and cutter.

During the cutting process, the particles perform the role of a cutting tip to produce more uneven surfaces in addition to changing the stress distribution in the workpiece material.

Eventually, the particles fly along the rake face, as shown in Fig. 6(b). Some particles that exist below or among the cutting path are pressed into the matrix, producing stress under the effect of the cutting tool. These particles, in turn, contribute to abrasive action on the flank of the cutting tool. The particles that are pressed into the matrix cause the failure of the matrix unit. Moreover, the irreversible plastic deformation generates a processing residual strain, while the elastic deformation of the particles reduces after the removal of the cutting tool.

To analyze the influence of the SiC particles on the parameters of the SiC_p/Al composite during the cutting process, such as cutting stress and cutting force, it is necessary to understand the parameters of the material. Fig. 7 presents the simulation results for the particles with aver. Comparison of Fig. 7(a), (c) and (e) clearly indicates that the defects become evident when particle size increases; however, these simulated surfaces undergo exfoliation causing them to form concave pit defects. Besides, the surface also deteriorates, with the chip's form changing from whole mixed powder to massive chips. Fig. 7(b), (d) and (f) represent the corresponding stress distributions in the cutting area. In all the three models, the particles, especially on the shearing surface, bear the major stress. The largest stress is experienced at the knife point of the first deformation area and the rake surfaces near the particles of the second deformation area. However, these areas have some differences such as the residual stress distribution on the lower area of the machined surface. For the same base and a similar volume fraction, a reduction in particle size increases the amount of particles. Therefore, in the case of random distribution, the particles' distribution inside the material will be uniform.

Fig. 7 Effect of particle size on the stress of PRMMC during cutting simulation for average particle size of (a and b) 10 μm, (c and d) 20 μm, (e and f) 30 μm, the shape of the particles being arbitrary polygon with random distribution and volume fraction of 30%.

Thus, a reduction in the heterogeneity defects of multiphase materials tends to increase the intensity of the material. In other words, during the PRMMC cutting process, exfoliation produced by the particles and base interface is one of the important mechanisms of chip information. When the particles refine and increase in amount, the plastic flow of the base material gains a larger limit, thereby enhancing the strength of the material. According to the simulation result, the smaller the particle size, the larger is the cutting force.

When the particle amount decreases, the residual stress increases after processing, thereby improving the surface quality. On the other hand, when the particles expand, the residual stress decreases after processing, leading to the deterioration of surface quality. There is a large amount of small expanding particles despite the presence of problems in the cutting process, including cut off, detachment, and exfoliation. This is because the size of the particles is small and there are no pits on the surface. Furthermore, the strength of the material becomes higher under this situation, making it difficult for the cracks evoked by the particles and the defects around the interface to expand. Overall, the refinement of the particles is beneficial for improving the quality of the PRMMC machined surface. During the refinement of mixed material particles, many particles in the path of the cutting edge will be stripped and will roll along the machined surface on the flank, leading to stability and continuity of stress in the machined surface. Thus, when the material moves on the cutting tool, it will still produce a larger stress distribution area. However, when the size of the particles is big, the frequency of stripping of the particles decreases. The shedding particles are either pressed into the base or fly with the chips. Hence, stable and continuous rolling will not be formed. Moreover, the stripping of the particles will cause the material to release some compressive stress. Therefore, when the particles expand in size, only some parts of the materials will produce residual stress, the distribution of which can be correlated with the position of the particles.

According to previous studies, the mechanical properties of mixed particles tend to significantly affect the shape of the particles.^8,18–20 Reinforcements with different shapes will cause different resistance of base deformation and load bearing capacity. Consequently, particles with different shape will cause different strengthen effects in mixed particles, which in turn will affect the cutting properties of the material. According to this study, the optimum average particle size is 20 μm, as determined by simulating the PRMMC cutting process using random distribution circular, quadrilateral, and triangular particle models with a volume fraction of 30%.

As shown in Fig. 8(a), (c) and (e), the shapes of the particles in the model do not affect the cutting stress distribution. The greatest stress distribution is located inside the shearing area and the most stress is endured by the particles. The greatest stress mainly occurs either on a side of the cutting tool or at the interface between the particles and base located on the cutting area analyzing the circled area shown in Fig. 8, it can be observed that the largest stress generated during the complete cutting process is at a certain corner of contact of the cutting tool and particles. Comparing the simulation results shown in Fig. 8(b), (d) and (f), it can be realized that there is a sequential decrease in the stress distribution area of the three machining models. The cutting surface of the triangular particles model was the smoothest with least number of pits.

Fig. 8 Influence of particles size towards the PRMMC cutting simulation stress for (a and b) circular particles model, (c and d) quadrilateral particles model, (e and f) triangular particles model.

The triangular particles model, compared with the other two particles models, has triangular particles with certain sharp corners. When the particles exfoliated by the cutting tool during the cutting process are rapidly inserted into the surface of the melting machining base, the sharp corners of the particles appear on the cutting path. These particles with sharp corners are not exfoliated easily but are only pressed into the machined surface. As for the stress distribution of the machining area, it was seen that the triangular particles are more beneficial in terms of transmitting the load. Thus, they reduce stress concentration and machined defects, thereby improving the machine properties compared with the other two models.

The force generated by the triangular particles is the largest, followed by the quadrilateral particles. Among the three models, the force generated by circular particles is the smallest. The shape of the particles determines the direction of plastic flow when subjected to external force and stress by the particles. During the cutting process, the function of the cutting tool with quadrilateral and triangular particles, apart from aforementioned particles' position, had uncertainty in shape correlation. According to the cutting results, when the vertex of the polygon is in contact either with the cutting tool or the other particles, the stress concentration will be higher. Conversely, when there is contact between the particles or surface contact of the particles with the cutting tool, the stress concentration will be smaller. In case of circular particles with symmetrical structures, when the particles are in contact with the cutting tool or other particles, there exists a tangential relation. This will produce more normal stress, which is smaller than the stress concentration in the particles with sharp corners. Overall, the cutting force generated by circular particles tends to be smaller than the other two models.

The volume fraction of the reinforced particles is one of the most important factors determining the mechanical properties of the SiC_p/Al composite. The difference in volume fraction of the SiC_p/Al composite is expected to bring different specialties during the cutting process. Fig. 9 shows the simulation results of the cutting process for the arbitrary polygon particles of radius 20 μm, with volume fractions of 30% and 40%.

Fig. 9 Simulation showing the influence of the volume fraction of particles on the PRMMC cutting stress for the volume fraction of (a and b) 20%, (c and d) 40%.

Comparing Fig. 9(a) and (c), it can be realized that the volume fraction expands with reduction in the chip deformation. Meanwhile, comparing the simulations shown in Fig. 9(b) and (d), it can be realized that the amount of deformed concave pit is greater in surfaces with higher volume fractions than in surfaces with smaller volume fractions because of separation of particles. For particles of similar size, when the volume fraction is higher, the amount of particles is also greater. Consequently, the amount of particles on the cutting path will also be greater, which naturally generates more shedding, crushing, pushing, even plowing phenomena during actual processing. Following the increase in volume fraction, there is an increase in the powdered degree of the chips. This could be attributed to the fact that the materials in the cutting area change into chips as a result of shear sliding. At the same time, with an increase in the amount of particles, the distance between the particles tends to decrease. Therefore, because of the effect of shear stress, the particles, stripping boundaries, and associated expansion of cracks easily appear in the surface.

4. Experimental study on the machinability of the SiC_p/Al matrix

In the present study, we performed a series of tests, including cutting force test, surface microtopography test, surface stress test, and surface roughness test to obtain the relation between the different machined parameters and physical properties of the SiC_p/Al composite.

The cutting force test was performed on a machine tool with milling force measurement equipment. In the typical process, the SiC_p/Al specimens were mounted on the cutting force test platform, which was connected with the CNC machined platform. Each workpiece was milled under different cutting parameters, and the corresponding average cutting force was obtained as shown in Fig. S1.†

As shown in Fig. 10(a), the cutting speed improves 10 folds from 50 to 550 m min⁻¹ with a 30% decrease in average cutting force from 155 to 110 N along the X- and Y-directions.

Fig. 10 Average cutting force along the X- and Y-directions: (a) cutting speed; (b) cutting depth.

When the SiC_p/Al composite is cut at a low speed, the major source of chip formation is the isolation of crushing particles and interfacial bonding. On the other hand, plastic deformation becomes the main characteristic for the chipping of the composite material at a higher cutting force. With an increase in the cutting speed, there is an increase in the cutting temperature, resulting in the softening of the matrix material. Consequently, there is a decrease in the plastic strain, cutting energy consumption of the composite material, and the cutting force. Under the same cutting conditions, when the volume fraction is 30%, with a cutting speed of 100 m min⁻¹, the average cutting force is 150 N. Similarly, when the volume fraction is 40%, the average cutting force is 125 N. With an increase in the volume fraction, the amount of SiC particles expands. Consequently, the intensified effect among the particles, cutting tool, and matrix will lead to an upward trend in cutting. As shown in Fig. 10(b), with an increase in the cutting depth from 0.1 mm to 0.6 mm, the cutting force also rises from 120 to 420 N. This signifies an almost four fold increase in the average cutting force. Therefore, it can be concluded that a higher cutting force results in an enhancement of the cutting depth.

When milled under different conditions of cutting parameters, the surface of the SiC_p/Al composite undergoes various phenomena, including shedding, peeling, crushing, pressing, and plowing. These phenomena often lead to defects on the surface at the locality. Therefore, the surface morphology and surface roughness are determined via scanning electron microscopy, as shown in Fig. S2.†

Fig. 11(a) shows the overall morphology of the milled surface, which is observed on the right hand side of the white bar. Fig. 11(b), (c) and (d) show the high-magnification images of the regions marked in Fig. 11(a).

Fig. 11 Surface morphology of the machined SiC_p/Al composite (v = 100 m min⁻¹, a_p = 0.2 mm, f_z = 0.1 mm). (a) Overall morphology; (b) milling texture and crushed particles; (c) ironing; (d) groove.

The SEM image shown in Fig. 11(b) clearly shows the cutting tool lines, pits caused by the dispersing fragmented particles, peeled particles, and plowed particles. With an increase in temperature in the three main deformation zones during the cutting process, the aluminum matrix in the cutting zone under heating condition undergoes plastic deformation. Either after the stripping of particle or after the composite effect of cutting tool and particle, parts of the aluminum matrix move separately with the chip to the nearby cutting edge in the cutting path. Furthermore, following the ironing press of the blunt cutting edge, these parts apply on the machined surface, as shown in Fig. 11(c). The cutting groove on the side of the morphology is shown in Fig. 11(d). As shown, the morphology of the aluminum matrix is homogeneous and, at the same time, the internal side is inlaid completely or comprises broken SiC particles. Due to the occurrence of particle peeling, holes and pits were formed on the section, thereby unflattening the section.

The main reasons underlying the poor surface of the machined PRMMC could be attributed to the formation of holes, pits, cracks, and scratches, as shown in Fig. 12(a).

Fig. 12 Morphology of the machined surface (v = 400 rpm, a_p = 0.2 mm, v(d) = 220 rpm, a_p(d) = 0.6 mm, f_z(a/b/c/d) = 0.1 mm): (a) surface defects; (b) surface pits; (c) surface ironing; (d) surface bonding.

The stress concentration at the SiC particles and matrix interface leads plowing and peeling of the particles to form holes or pits. On the other hand, the strong interaction among the particles, cutting tool, and matrix during the cutting process causes parts of the particles to get crushed and cut. The fragmented SiC particles formed under the effect of the tool flank in the machined surface forms scratched surfaces and pits. Finally, some parts get pressed into the machined surface. The machined surfaces of the PRMMCs have numerous pits with 60 μm diameter, as shown in Fig. 12(b). The formation of such bigger pits could be attributed to the cracks between the outcrops. Some small cracked particles could be observed on the pit, which are the leftovers of the pressed–peeled particle matrix.

The melting point of the aluminum matrix is low, and therefore, during the cutting process, the heat produced by the cutting edge and the effect of strong rapid press and friction lead to softening of the matrix, resulting in a molten condition. Furthermore, as a result of the ironing flank effect, the surface defects tend be concealed or erased, leading to a more flat surface, as shown in Fig. 12(c). Therefore, the “ironing” phenomenon, to a certain extent, improves the surface quality. Consequently, most machined surfaces have a re-melted and ironing pressure structure. An increase in the cutting speed will not affect the surface roughness. As can be seen from the large coating of the aluminum matrix alloy layer shown in Fig. 12(d), the surfaces adhere to block structures. These structures mainly wrap the molten aluminum in the SiC particles composite. This could be attributed to the following two reasons: (1) the aluminum chips formed during the milling process do not discharge along the rake face while continuing to remain at the processing surfaces at high temperature; (2) under the intense action, the particles and the cutting tool, after the plastic flow, will instantaneously cool and solidify.

Irrespective of whether common metals or composite materials are machined and whether the process is turning or milling, the existence of corner radius in the tool nose prevents complete excision of the workpiece during feeding. In fact, the cutting tool will go forward by one unit of f_z with the tool turning one circle during the feeding process. A small portion of the workpiece still remains because of the circular profile of the knife point. This explains the formation of residual height in the sample.

Table 4 shows the parameters of the cutting tool and Table 5 presents the relation between the milling speed and surface roughness during the milling of SiC_p/Al matrix composites. As shown, the surface roughness decreases with an increase in milling speed. When the cutting speed is increased respectively from 100 to 600 m min⁻¹, the surface roughness is initially higher, but gradually decreases with a further increase in milling speed. According to the analysis, this may be related to the temperature and material properties during the high-speed milling process. During this process, the SiC particles play an enhancement role and aluminum plays a connecting role by actually transmitting the force. An increase in the milling speed inevitably increases the wearing of the tool, processing, and friction, leading to an increase in the milling temperature. The higher the milling speed, the higher is the temperature. The high temperature generated as a result of the high milling speed can soften the aluminum-based material and reduce its hardness, thereby contributing to a smoother surface.

Table 4 Parameters of cutting tool

Diameter, mm	Number of teeth	Material of tool	Radius of edge, μm	Tool orthogonal rake, °	Tool clearance, °	Helix angle, °
12	4	Solid carbide	1	15	15	45

---

Table 5 Cutting parameters and results of measurement

No.	Volume fraction, %	Feed rate per tooth, mm	Cutting depth, mm	Cutting width, mm	Cutting speed, m min^−1	Main shaft speed, rpm	Maximum cutting force, N		Surface roughness, μm
							X	Y
1	30	0.1	0.2	12	100	2653.93	212.2	247.3	4.84
2	30	0.1	0.2	12	200	5307.86	207.3	239.4	5.27
3	30	0.1	0.2	12	300	7961.78	204.4	231.0	5.31
4	30	0.15	0.2	12	400	10 615.71	428.3	437.9	6.98
5	30	0.15	0.2	12	500	132.69.64	356.1	386.7	8.53
6	30	0.15	0.2	12	600	15 923.57	325.2	339.5	10.76
7	40	0.1	0.2	12	200	5307.86	252.4	250.2	5.15
8	40	0.1	0.3	12	200	5307.86	291.9	303.1	5.83
9	40	0.1	0.4	12	200	5307.86	397.8	402.9	5.25
10	40	0.15	0.5	12	200	5307.86	476.0	466.0	6.15
11	40	0.15	0.6	12	200	5307.86	511.6	492.6	6.43
12	40	0.15	0.7	12	200	5307.86	542.5	538.7	6.84

---

With the softening of the base material used to connect and transmit the force, the hardness of the composite material reduces. As a result, the tool will have a flat ironing surface at high temperatures. In simple terms, the surface roughness is reduced overall. This seems apparent because the surface roughness and milling speed are inversely proportional during the milling of the SiC_p/Al matrix composite. In order to study the impact of milling depth on the surface roughness of the composite, we performed experiments at a constant milling speed of 200 m min⁻¹ and feed rate per tooth of 0.1 mm and 0.15 mm with varying milling depth. Results show the relation between the milling depth and surface roughness.

As can be seen from Table 5, when the milling depth increases from 0.2 mm to 0.7 mm, the surface roughness shows an upward trend, but with little changes. More quantitatively, the surface roughness of 5.15 μm at a milling depth of 0.2 mm increases to 6.84 μm at a milling depth of 0.7 mm. Thus, the milling depth has a very minor impact on the surface roughness of the SiC/Al particles composite material. However, the milling depth has a great impact on the milling force, as can be observed from the data no. 7 to no. 12 in Table 5.

The residual stress state in the machined layers of the workpiece was analyzed by X-ray diffraction using the sincp² method.³⁷ The residual stress measurement was performed on X-ray stress measurement equipment. Prior to analysis, the equipment was calibrated using a stress-free standard sample (Table 6). The collimator was modified, and the distance was determined and set between the probe and the sample. The residual stress in the cutting surface of each specimen was measured at 3 points after milling under different cutting parameters. Each workpiece was milled under 6 milling parameters. The machined surface of each workpiece was cleaned using acetone, and the two milling speed directions were marked on each specimen. Subsequently, the two specimens were placed on the test platform, and the same direction of the milling feed was maintained for both the specimens. The specific sample positions, measured point locations, and the experimental setup are shown Fig. S3.†

Table 6 Parameters used for X-ray measurement

Name	Parameters
Type of equipment	Proto-IXRD-MGR40
The collimator	Standard linear 512 channel detectors
Anode target	Cr
Measured material	SiCp/Al matrix
Error of measurement	10 MPa
Collimating pipe size	2 mm × 5 mm
Cooling system	Solid state detector Ge
Work voltage	30 kV
Current	4 mA
Measurement of Φ	0, +45°, +90°
Range of β	−45° to +45°
Time of exposure	2 s
Each of point measurement times	5 times
Distance between probe and sample	40.32 mm

---

As shown in Fig. 13(a), with increase in the cutting speed, there is an increase in the residual stress at the cutting surface from 150 to 410 MPa. A considerable amount of friction occurs between the SiC particles and tool. The volume fraction from 30% to 40% increase residual stress about 25% in average. The cutting heat generated during the SiC_p/Al machining process greatly affects the cutting residual stress. The resulting surface residual stress is totally different from unitary aluminum or steel.³⁸ The cutting heat for the surface residual stress of aluminum alloy is not high because of the transient squeezing effect. Thus, cutting speed can be visualized as an appreciable factor affecting the surface residual stress during the SiC_p/Al machining process.

Fig. 13 Surface cutting stress in the work-piece milled under different cutting parameters: (a) cutting speed; (b) cutting depth.

Additionally, the influence of the cutting depth on the cutting surface residual stress is not a marked factor, as observed in Fig. 13(b). The cutting residual stresses show little change with increase in the cutting depth from 0.1 to 0.6 mm. Thus, it appears that the cutting depth plays a variable role in affecting the cutting force and surface residual stress in SiC_p/Al composites.

5. Conclusion

This study reports extensive simulation and experimental studies on the theory underlying machinability of the representative SiC_p/Al composites. The results indicate the relation of the cutting stress distribution, surface state, and cutting force, with the machining parameters. The principle and different machinability phenomena associated with the cutting of SiC_p/Al composites have been studied. The following conclusions of machinability have been drawn:

(1) Surface machined state (roughness and cutting residual stress) is improved and optimized by the machining parameters. The phenomena of poor machinability such as ironing, groove and crushed particles and so on are explained.

(2) The results show particle is one of key factors of surface machine quality. The size, shape and volume fraction of particles directly affect machinability for SiC_p/Al composites.

(3) The cutting force of SiC_p/Al composites is greater and fluctuates irregularly compared to that of homogeneous material. To control fluctuation of cutting force will improve machined surface quality.

Acknowledgements

This study was financially supported by National Natural Science Foundation of China (Grant No. 51105025), and International Science & Technology Cooperation Program of China (Grant No. 2013DFB70110), Aeronautical Science Funds of China (Grant No. 20130451010).

Notes and references

1. J. E. Allision and G. S. Gole, J. Min. Met. Mater. Sci., 1993, 45, 19–24.
2. F. E. Kennedy, A. C. Balbahadur and D. S. Lashmore, Wear, 1997, 203–204, 715–721.
3. L. A. Looney, J. M. Monagham, P. O'Reilly and D. M. R. Taplin, J. Mater. Process. Technol., 1992, 33, 453–468.
4. M. S. Karkas, A. Adem, Ü. Mustafa and Ö. Bilgehan, J. Mater. Process. Technol., 2006, 178, 241–246.
5. L. Zhou, S. T. Huang and D. Wang, Int. J. Adv. Des. Manuf. Technol., 2011, 52, 619–626.
6. X. Ding, W. Y. H. Liewb and X. D. Liu, Wear, 2005, 259, 1225–1234.
7. B. Chennakesavarao, P. N. Reddy, N. Komariah and R. D. Malla, In Proceedings of the Seventeenth AIMTDR Conference REC Warangal, 2000, pp. 53–56.
8. S. Qu, H. Lou and X. Li, RSC Adv., 2015, 5(64), 52167–52173.
9. L. Lasri, M. Nouari and M. E. Mansori, Wear, 2011, 271, 2542–2548.
10. K. Palanikumar and J. P. Davim, J. Mater. Process. Technol., 2009, 209, 511–519.
11. H. Hocheng, S. B. Yen, T. Ishihara and B. K. Yen, Composites, Part A, 1997, 28, 883–890.
12. A. R. Chambers, Composites, Part A, 1996, 27, 143–147.
13. Z. J. Yuan, L. Geng and S. Dong, CIRP Annals - Manufacturing Technology, 1993, 42, 107–109.
14. M. Ei-Gallab and M. Sklad, J. Mater. Process. Technol., 1998, 83, 151–158.
15. J. P. Davim, J. Mater. Process. Technol., 2003, 132, 340–344.
16. N. Muthukrishnan, M. Murugan and K. P. Rao, Int. J. Adv. Des. Manuf. Technol., 2008, 38, 447–454.
17. K. Palanikumar and R. Karthikeyan, Mater. Des., 2007, 28, 1584–1591.
18. X. Soldani, C. Santiuste, A. Muñoz-Sánchez and M. H. Miguélez, Composites, Part A, 2011, 42, 1205–1216.
19. U. A. Dabade, S. S. Joshi, R. Balasubramaniam and V. V. Bhanuprasad, J. Mater. Process. Technol., 2007, 192–193, 166–174.
20. A. C. Basheer, U. A. Dabade, S. S. Joshi, V. V. Bhanuprasad and V. M. Gadre, J. Mater. Process. Technol., 2008, 197(1–3), 439–444.
21. M. N. Yuan, Y. Q. Yang, C. Li, P. Y. Heng and L. Z. Li, Mater. Des., 2012, 38, 1–6.
22. M. M. Aghdam and S. R. Falahatgar, Compos. Struct., 2004, 66, 415–420.
23. C. R. Dandekar and Y. C. Shin, Composites, Part A, 2009, 40, 1231–1239.
24. C. T. Sun and Z. H. Jin, Compos. Sci. Technol., 2006, 66, 1297–1302.
25. A. Ayyar, G. A. Crawford, J. J. Williams and N. Chawla, Comput. Mater. Sci., 2008, 44, 496–506.
26. A. Pramanik, L. C. Zhang and J. A. Arsecularatne, Int. J. Mach. Tool Manufact., 2007, 47, 1497–1506.
27. A. Pramanik, Int. J. Mach. Tool Manufact., 2014, 86, 44–61.
28. M. V. Ramesh, K. C. Chan, W. B. Lee and C. F. Cheung, Compos. Sci. Technol., 2001, 61, 1449–1456.
29. C. R. Dandekar and Y. C. Shin, J. Mater. Process. Technol., 2013, 213, 153–160.
30. C. Sun, R. J. Shen and M. Song, Trans. Nonferrous Met. Soc. China, 2012, 22, 476–484.
31. W. D. Song and J. Wang, Trans. Beijing Inst. Technol., 2009, 29, 501–505.
32. Z. G. Gao and G. T. Liu, J. Tsinghua Univ., Sci. Technol., 2003, 43, 710–714.
33. Q. Jin and J. J. Qin, Acta Mater. Compositae Sin., 2006, 23, 14–20.
34. J. P. Wang, P. Zeng and L. P. Lei, J. Plast. Egn., 2008, 15, 102–104.
35. M. Issa, K. Saanouni, C. Labergere and A. Rassineux, International Journal of Machining and Machinability of Materials, 2011, 9, 295–315.
36. Q. Wu and D. P. Li, Int. J. Precis. Eng. Manuf., 2014, 15, 1499–1506.
37. O. T. Lancu, D. Munz, B. Eigenmann, B. Scholtes and E. Macherauch, J. Am. Ceram. Soc., 1990, 73, 1144–1149.
38. R. M'Saoubi, J. C. Outeiro, B. Changeux, J. L. Lebrun and A. M. Dias, J. Mater. Process. Technol., 1999, 96, 225–233.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra00340k

---