Experimental evaluation on texture of flank face on tool wear in chamfer milling of stainless steel

This paper presents a novel grinding enabled texture, ‘V’ shaped texture (VST), on flank face. To implement the texture on a cutting performance, a set of chamfering experiments of stainless steel materials used as 3C product shell usually are presented. Cutting forces, surface qualities, and tool wear are measured and compared, of which results show that both smallest surface roughness and longest tool life are achieved by using a 30° VST chamfer tool. By comparing the results, a clear conclusion can be drawn that texture types and angles are not independent factors to cutting performance; therefore, a suitable combination of texture types and texture angles can provide a significant improvement of tool life and surface quality.


Introduction
Metal shell of electronics devises provides a better experience than plastic one in touch and visual perspectives, so that the mass metal shells have been used in modern 3C industry (computer, communication, and consumer electronics). Many companies, e.g. Apple, Samsung, Lenovo, and Dell, have already used metal shell in their products [1]. However, comparing with the plastic shell, the manufacturing process is more difficult, and also, the cost of metal one is higher, of which the major reason is that the machining processes of two shell are different. Here, the plastic one is produced by casting which can be performed quickly, and its surface can be improved by grinding easily. However, the metal one is produced by milling which is a time-consuming process, and its surface cannot be guaranteed as easily as the plastic one.
Metal cutting is a complex process with cutting force, vibration, and heat [2]; therefore, the surface quality is associated with many factors. (1) Machining dynamics is an essential one which is closed to the elements in process, e.g. machine tool, machining types, and tool wears [3]. (2) Chatter influencing the surface quality should be avoided [4]. (3) Machining parameters affecting cutting force [5], and (4) cutting tool: tool hangover was optimised by considering minimising cutting force and vibration [6]. Cutting tool, usually considered as a standard resource, plays a key role in machining since it contacts with machined surface directly. Therefore, the grinding texture on tool flank face, which directly contacts with machined surface, is related to the surface quality closely.
A flank face with a slant shape textures (SSTs) in chamfering was tested in previous work [7]. On top of that, this paper proposed a novel texture pattern, 'V' shaped texture (VST), which was adapted for the grinding machinability of the texture based on the surface of shark and applied to many areas successfully. In order to test the efforts, a set of experiments were carried out. The remainder of this paper is organised as follows. Literature review is described in Section 2. Experiment setups are presented in Section 3, together with the manufacture of the chamfer cutting tool with VST flank face. The experimental results of cutting force, tool wears, and surface quality are illustrated and analysed in Section 4. Finally, Section 5 concludes this paper.

Literature review
Textured surfaces can reduce friction between two contact surfaces. Shark skin pattern was applied to drag reduction, which was proved by an experiment; minimum drag reduction is roughly 7% [8,9]. A V-shaped texture was proposed and proved that the maximum drag reduction of this texture is 8% [10], and it was applied to aircraft, by which the drag reduction was 5-8% on 2D airfoils [11]. In the recent decade, a micro-or nano-fabricated ridgeline was proposed, and it was applied to enhancing surface functions in many areas [12]. Surface textures are machined in many ways, e.g. laser machining and grinding. Many researchers focused on grinding that is a straightforward method without additional processes, and they used customised grinding wheels to obtain the patterns of textured surface. The costs of the grinding wheel are much higher than those of the standard wheels. Stepién introduced texture in machining and focused on generation of surface texture by grinding [13]. Denkena et al. [14] machined riblet structures on large-scale surface by profile grinding process with multiple V-shaped profiled CBN wheel of which the height of V-shaped profile was 500 μm and the distance between peaks was 600 μm. In their works, the riblets between 20 and 120 μm were obtained by a two-step strategy involving grinding on workpiece by single profile wheel and performing the profile grinding via an axial profile offset. Shichao et al. [15] presented a new approach to obtain a surface anticipant texture by point grinding process, where the texture direction was controlled by changing the point grinding angle (β) including 0°, 5°, 10°, 15°, 20°, and 25°.
Recently, micro-textured surfaces were applied on cutting tool to reduce cutting forces and frictions in machining. Lei et al. [16] proposed that micro-pools on tool rake face could enhance lubricant performance. In their experiments, the mean of cutting forces, F f , F t , and F c , generated by micro-pool tools were reduced by 10-30% comparing with traditional tools. Enomoto and Sugihara [17] reported a set of milling experiments of A5052 with micro-textured and conventional tools. Their results showed that friction coefficient of micro-textured surface between tool and workpiece was lower than that of non-textured surface, and generated a higher anti-adhesive properties. Wu et al. [18] reported a set of experiments of machining of Ti-6Al-4V by using self-lubricating texture tools, of which results showed cutting force and cutting temperature were lower than the traditional tool. In their works, the textures were machined by laser and included circular and linear textures on both rake face and flank face. Kim et al. [19] introduced a turning test of AlSl52100 steel by four types of tools in terms of texture parameters including shape, edge distance, pitch size, and height. Their results demonstrated the cutting force generated by perpendicular texture was reduced by 6%, where the effective coefficient of friction was the lowest. Fatima and Mativenga [20] reported experimental research on influence of textured rake face on cutting tool performance, cutting force, and cutting temperature which were reduced significantly. Sun et al. [21] proposed a type of hybrid textures on rake face of cutting tool. Their results showed the cutting force was reduced by 7.1-33.3%, and surface roughness R a was reduced by 42.9-69.1%. Liu et al. [22] compared a set of textures (slot patterns) on flank face of turning tool in dry machining. The best parameters of the slot width and length was obtained.
From literature survey, compared with common cutting tool, the surface textured one delivers a better performance in terms of reducing friction, cutting force, and temperature. The reported textures on cutting tools are machined by laser  machining which is a low-efficient, high-cost process. Or the textures were ground by using the high-cost non-standard grinding wheel, and they were not used on cutting tools. Rare research is focused on texture grinding with a standard grinding wheel, especially on a flank face of chamfering tool.
In previous work, chamfering tools with and without the texture have been tested, which shows the potential function of texture on surface quality [7]. The texture patterns which are able to be ground by a standard grinding wheel are quite limited, a slant texture is the most common one. Targeting an enriched ground texture, a "V" texture, on flank face of cutting tool, this paper is therefore focused on two points: (1) the grinding strange of flank face texture on a chamfer tool and (2) the influences of texture shapes and angels on cutting force, tool wear and surface quality.

Experimental method
3.1 Experimental setups Figure 1 shows the experimental setups of chamfering, where a cuboid workpiece is used to carry out the cutting force experiment ( Fig. 1(a)), and a set of pipe ones are used for tool x (x 1 ) y y 1 q 1 Fig. 2 Transformation of coordinate systems for cutting edges wear experiment ( Fig. 1(b)

Chamfering tools
Four types of textured flank faces, referring to 2 SSTs and 2 VSTs, are applied to the chamfering cutting tools of which geometrical parameters are shown in Table 1. The tools were machined in a five-axis grinding centre, SAACKE UWIF, by following a series of grinding codes that are generated based on grinding algorithm and grinding strategy.

Grinding path models
Cutting tool models, analytic geometry, and differential geometry can be established by two methods [23,24]. Analytic geometry is applied to establishing mathematical models of cutting tool. Cutting tool is modelled in coordinate system xyz where z-axis is central axis of cutting tool, and the origin of system xyz is on tool tip, as shown in Fig. 2. The coordinate system x 1 y 1 z 1 can be obtained by the translation t x of xyz along x-axis. The system x 2 y 2 z 2 can be generated by rotation of x 1 y 1 z 1 around y 1 through an angle of α. By rotating x 2 y 2 z 2 around x 2 axis through an angle of β, the system x 3 y 3 z 3 can be obtained. The system x 4 y 4 z 4 is generated after moving x 3 y 3 z 3 along y 3 -axis by t y . Finally, the system x 5 y 5 z 5 is obtained by rotating x 4 y 4 z 4 around x 4 -axis through an angle of ŋ. The where M xt is a transformation matrix from xyz to x 1 y 1 z 1 , M xŋ a transformation matrix from x 1 y 1 z 1 to x 2 y 2 z 2 , M xα a transformation matrix between x 2 y 2 z 2 and x 3 y 3 z 3 , M yt a transformation matrix between x 3 y 3 z 3 and x 4 y 4 z 4 , and M yt a transformation matrix from x 4 y 4 z 4 to x 5 y 5 z 5 . For the chamfer tool, the cutting edge is a straight line and its model is established easily in the system xyz, as shown in Eq. 2.
where t is the parameter of the model and a and b are the constraints of t.
The second flank face is a three-dimensional surface which is constituted by cutting edge and the grinding path of the second flank face. A straight line in x 5 y 5 z 5 (Eq. 3) is selected as the grinding path of the second flank face. Therefore, the path in xyz is calculated by substituting Eq. 3 into Eq. 1.
where t 1 is the parameter of the model and c and d are the constraints of t 1 .

Grinding strategies
The diameter of grinding wheel used in the machining of cutting tool is bigger than 100 mm in general, while the width of the first flank face is under 1 mm. Therefore, the texture of flank face is a cycle critically; however, a part of a more than 100-mm diameter cycle on a less than 1-mm face can be considered as a straight line which composites the textures in this paper. On top of that, the grinding strategies are designed, as shown in Fig. 3 grinded on the upper right area of grinding wheel by a single pass, and the distance between grooves is controlled by using a same feed rate. V-30 and V-45 are generated by two-pass grinding including the same parts for S-30 and S-45 and the downright parts on grinding wheel. Here, the VST is not a flat face, i.e. up and down grooves (Fig. 3) are not on the same flat face, since the latter pass removes the texture generated by the former paper pass, if they are on a flat face.

Simulation and manufacturing of chamfer tool
The grinding processes, as shown in Fig. 4, are simulated on NUMROTO Plus considering the safety reasons. In general, this grinding process includes five steps, rake face, edge side, second flank face, first flank face (positive), and first flank face (negative), as shown in Fig. 4a, b. Here, an SST on the first flank face is obtained by one pass (Fig. 4(c)), while a VST needs two passes (Fig. 4(c, d)). The textures on the machined chamfer tool are shown in Fig. 5, including an S-30 and an S-45( Fig. 5(a)) and a V-30 and a V-45 ( Fig. 5(b)).  Figure 7 shows the relationship between tool flank face wear V B and numbers of chamfered holes. Tool wears on the four tools are increased with chamfered hole increasing in general. The tool wears, between 0 and 120 μm, are increased with chamfering hole increasing. The incensement of V-30 tool wear is slowest, followed by V-45, S-45, and S-30. More than 900 holes are chamfered by V-30 tool, followed by 500 (S-30), 500 (V45), and 700 (S-45), respectively. Here, V-45 wear speed was much lower than that of S-30 and S-45 before 400 holes; however, cutting edge is broken suddenly at chamfering 500 holes. Similarly, the cutting edges of all tools were broken. Comparing SST with VST, the wear speed of SST is higher than that of VST, before holes 400; the wear speeds of V-30 and V-45 is slower than those of S-30 and S-45. Wear resistance of VST is higher than that of SST. Within VST, V-30 wear resistance is higher than that of V-45, while within SST, S-30 wear resistance is lower than that of S-45. A smooth surface is usually considered to cause a lower wear than a rough one. However, according to the research on liquid lubricants, the performance of a well-designed rough surface was better than that of the smooth one [25]. Similarly, the friction coefficient of a grinded textured tool flank face was smaller than that of a conventional one [26]. A little coolant liquid is stored in the small grooves, as shown in Fig. 8(b), which reduces the cutting temperature during cutting. Since the volumes of the grooves of four type textures are almost the same ideally, there are not significant observed differences between tool wears regarding texture types or texture angles. However, a good combination of type and angle of texture can provide a significant improvement of tool performance, i.e. the tool life of V-30 is 80% higher than that of V-45 and S-30. Figure 9 shows the comparison of the surface forms of the chamfered holes. In general, there are two major issues on the machined surface: (1) burn point caused by high temperature at a single point and (2) coarse texture due to unsuitable status at tool left machined surface. The burn issue has been reduced significantly compared with 0°S

Surface quality
ST, 15°SST, and 60°SST, of which result is shown in [7]. Compared with SST tools, the surfaces generated by VST tools are better in terms of the two issues. The number of burn points on the surface of SST is more than that of VST, while the generated coarse texture of SST is similar with that of VST.
Surface roughness is a quite sensitive criterion which is influenced by many uncertain factors; therefore, the standard deviation, the error bar, is calculated at each roughness R a measurement at each sampling point, 100 chamfered holes, as shown in Fig. 10, where standard deviations of measured values of surface roughness are quite high, especially 700 holes of S-45 and 900 holes of V-30. The lowest surface roughness is obtained by V-30, followed by S-30 (S-45) and V45, where the values of R a by S-45 and S-30 are similar. R a of V-30 and V-45 is increased with hole number increasing slightly between 0.25 and 0.45 μm. For S-30 and S-45, R a is reduced firstly at 200 and 300 holes, respectively, and then is increased between 0.3 and 0.5 μm. There is an interesting phenomenon that R a is unstable, the large error bars, when the tool edges of S-45 and V-30 are broken.
The texture on flank face changes the material flow status on the workpiece surface during machining, especially the direction of the flow, which causes the differences of contact stress between the selected textures, as shown in Fig. 11. Machined surface is generated by material returning against the workpiece after tool tip passing (similar phenomenon was found in machining of superalloys [27]), as shown in areas A and B of Fig. 11. The material flows on SST ( Fig. 11(a)) are changed into another direction, by which the material build up is delayed compared with a normal flank face. However, the material may be still built in area  Fig. 11, which can be improved by a VST by changing the material flows. Therefore, the surface quality can be improved.

Conclusions
This paper presents a set of experiments to evaluate the influence of textured flank face on tool wear and surface qualities in chamfering holes. Four types of textured flank face are tested, including two SSTs and two VSTs. The grinding strategies are developed for the tools, as well as grinding path models. The generated grinding programmes were used to grind the tools in a five-axis grinding machine. Then, cutting forces, tool wear, and surface quality are observed and measured based on the four machined tools. The performance of chamfering tool with textured flank face is improved. The contributions of this work are summarised as follows: (1) A suitable combination of texture types and texture angles can provide a significant improvement of tool life. The V-30 textured flank face produces a best resistance against tool wear, and the tool life is 80% longer than the other three. (2) The material flow could be changed by texture on flank face, and a VST could be changed second time potentially, which can improve the surface quality conditionally. In this case, V-30 shows the best performance in terms of surface roughness. The results of this paper show that VST is able to provide an improvement on surface quality. However, the surface quality does not only rely on VST but also rely on many other factors, e.g. material properties and cutting parameters, which is a major limitation that is not covered in this paper. Those factors may raise the rejection rate of the textured tool. To address the issues, the future work is therefore planned: (1) a comprehensive analysis on surface and subsurface quality and (2) a robust grinding algorithm to provide a stable textured chamfering tool.