Fabrication and tribological performance of a laser-textured hardmetal guiding stone for honing processes

Accepted Manuscript Fabrication and tribological performance of a laser-textured hardmetal guiding stone for honing processes Shiqi Fang, Sven Klein, Chia-Jui Hsu, Luis Llanes, Carsten Gachot, Dirk Bähre PII: S0263-4368(19)30412-3 DOI: https://doi.org/10.1016/j.ijrmhm.2019.105034 Article Number: 105034 Reference: RMHM 105034 To appear in: International Journal of Refractory Metals and Hard Materials Received date: 29 May 2019 Revised date: 5 July 2019 Accepted date: 24 July 2019 Please cite this article as: S. Fang, S. Klein, C.-J. Hsu, et al., Fabrication and tribological performance of a laser-textured hardmetal guiding stone for honing processes, International Journal of Refractory Metals and Hard Materials, https://doi.org/10.1016/ j.ijrmhm.2019.105034 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. ACCEPTED MANUSCRIPT Fabrication and tribological performance of a laser-textured hardmetal guiding stone for honing processes Shiqi Fanga,b,* [email protected], Sven Kleina, Chia-Jui Hsuc, Luis Llanesb, Carsten Gachotc, and Dirk Bährea aInstitute of Production Engineering, Saarland University, Saarbrücken, Germany bCIEFMA – Dept. Materials Science and Metallurgical Engineering, EEBE - Campus Diagonal Besòs, Universitat Politècnica de Catalunya, Barcelona, Spain cInstitute for Engineering Design and Logistics Engineering, Vienna University of Technology, Vienna, Austria *Corresponding author. Abstract. Honing is often implemented as surface finish process by improving geometrical and form precision of workpieces. Guiding stones, usually made of hard materials with requirement of excellent tribological properties, play a key role as a supporting part in the honing tool. In this paper, cemented carbides are intended for guiding application as substitute of diamond or CBN composites. Innovative surface textures, namely crossing-line patterns and dimple arrays, are fabricated by means of short pulse laser on a cemented tungsten carbide (hardmetal) grade. A laser-ablation model for cemented carbides is used for estimating optimized laser machining parameters. Implementation of these processing conditions yields satisfactory geometrical accuracy. crossing-line patterns with similar dimensional and topographic features to those exhibited by honed workpieces. The cemented carbide stones patterned by laser together with a traditional diamond guiding stone were tested on a lubricated pin-on-disc tribometer. Both crossline patterns or dimple arrays are found to improve frictional conditions compared to the ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT traditional guiding stone, but during lubricated pin-on-disc tribotests. In this regard, the former are more favorable to build up a stable sliding scenario with minimized friction than the latter. Pattern density has a more significant impact on friction, diminishing in the case of dimple arrays. Keywords: cemented carbide; honing; laser surface texturing; tribology. 1 Introduction Surface modification at the level of microns or nanometers has been proven to be an effective approach to improve tribological performance of tools and components [1-2]. Enhancement of tribological functionality, namely decreasing friction, removing heat or reducing wear, usually results from the amelioration of contact conditions and lubrication regimes [3-4]. In this regard, several technologies (i.e. coating deposition [5], electrochemical machining [6], lithography, laser and 3D printing, among others) have been successfully implemented in different materials and for a wide range of applications, e.g. in the manufacturing of bearing and tools as refs in [7-8]. Among them, pulsed laser is an emerging technology attracting large industrial interest for texturing functional surfaces. Main reasons behind it are that pulsed laser is able to produce precise (small length scale) textures with defined geometry, and at the same time to minimize side effects on the surface integrity [9,10]. Compared to other surface modification technologies, laser has the advantage to realize tiny ablation, and then to achieve high precision [11-13]. Among the abrasive machining processes, honing is often implemented as surface finish process by improving geometrical and form precision. Guiding stones play an important role as a supporting part in the honing tool [14,15]. Due to the relative movement and contact between guiding stone and workpiece surface, mandatory ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT requirements for the former are excellent tribological properties, i.e. low friction and high wear resistance. Taking into account that guiding stones are usually made from hard materials, laser surface texturing becomes a potential technology to produce specific patterns on their surfaces, in order to reduce friction by changing the tribological conditions in the honing processes. Within this context, the first goal is to find out suitable surface patterns on the potential cemented carbide guiding stone, i.e. aiming to improve the tribological performances of guiding stones. In practice, honed surfaces usually exhibit crossing-line patterns, as they are supposed to improve the supply and distribution of the lubricant during the reciprocal movement, for example the honed cylinder inner wall [16, 17]. Therefore, crossing-line pattern is selected as the first target pattern to be produced by laser in this study. Among those classic tribological patterns, dimple array is one of the most studied patterns [4], since such artificially-induced cavities are easy to produce and their shape can serve as tiny reservoirs of lubricants [12,18]. Therefore, besides crossing-line patterns, dimple arrays with different densities of cavities will also be produced by laser. Once laser-textured patterns are made, a second goal of the investigation is to evaluate the tribological performance of the surface-modified hardmetal. This is done by determining frictional response as a function of sliding length, and it is conducted by using a precision tribometer under lubricating conditions. Finally, geometrical precision and surface conditions of the tested samples are assessed and compared. ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 2 Materials, experimental aspects and theoretical considerations 2.1 Materials studied In practice, the applied guiding stones are usually made of composites consisting of middle or small grain size diamonds or cubic boron nitride particles and a metallic matrix [19]. For comparison purposes, the surface topographic features of a commercial guiding stone are characterized as reference conditions. In this regard, the guiding stone D76/710/6/100 (Kadia Produktion GmbH + Co) was chosen. It consisted of 25%vol concentration of diamond grains with an average size of 76 µm embedded in a metallic binder. Figure 1(a) shows the surface topography of the inspected guiding stone. After dressing, Ra and Rz surface roughness parameters are about 3 µm and 35 µm respectively. The required use of diamond or CBN composites as guiding stones not only leads to high costs but also might induce production quality issues, as extruding grains are sometimes either dislodged or broken during the honing process. As a consequence, fragmented or erupted hard grains or debris can scratch and damage the workpiece surfaces, as shown in Figure 1(b). Within this context, it is then proposed to use more conventional hard materials as substitute of diamond or CBN composites for guiding application in the honing processes, as long as their tribological and mechanical properties meet the required service demands. ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT (a) (b) (c) Figure 1. (a) Surface topography of the guiding stone D76 with a fractured grain; (b) damage on the workpiece surface, eventually induced by broken or erupted diamond grains; and (c) microstructural aspect of hardmetal grade 15CoC. A plain WC-Co cemented carbide (also named hardmetal) grade was selected in this study as substitute material for the traditional diamond composite commonly used for guiding application. It is referred to as 15CoC, and its basic microstructural and mechanical characteristics are listed in Table 1. Microstructural aspect evidencing the composite nature of the hardmetal grade is shown in Figure 1(c). The combination of medium binder content and a relatively coarse carbide size results yields an optimal hardness-toughness relationship, suitable for its intended application as supporting part. Moreover, aiming to ameliorate its tribological performance, surface laser-texturing was implemented, as it will be detailed below (section 2.3). ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT Table 1. Microstructural and mechanical parameters of 15CoC hardmetal grade studied [20]. Binder wbinder (%wt) dWC (μm) E (GPa) Poisson’s ratio υ HV30 (GPa) Fracture toughness KIc (MPa√m) Co 15 1.7 ± 1.1 550 0.23 10.2 ± 0.1 17.0 ± 0.2 2.2 Preliminary study of the honed surface structures Honing is a technology widely used for bore finishing process. The surface of a honed part describes a functional surface with form (geometry) and dimension accuracy as well as a very high surface quality (e.g. cylindricity less than 1 µm and roughness Ra less than 1 µm). Due to kinematics implicit to honing, this material removal process produces specific crossing-line micro-patterns on the finished surface [15,19]. The honing angle, i.e. cross-angle, can be adjusted by changing the machining parameters (Figure 2(a)). It is reported that these specific patterns are beneficial to achieve effective and even distribution of lubricant as well as to reduce the friction between the contact surfaces [21]. In this regard, an additional test was conducted attempting to obtain the topographic features of the honed surfaces. It involved a steel workpiece 16MnCr5 (1.7131) with elevated hardness (HRC60) and a diamond honing tool together with the guiding stone D76 (middle grain size), under full lubrication testing conditions. The characteristic honing angle 𝛼 can be adjusted by changing oscillation (  a) and rotation (  t) speeds according to [15]: 𝛼 =2∗arctan⁡(𝑣𝑎 𝑣𝑡) (1) ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT Hence, setting na and nt values as 172 mm/s and 1000 rpm (419 mm/s) respectively, the process results in a honing angle of about 45° in case of honed bores with a diameter of 8 mm. Figure 2(b) shows the topography of the honed surface, exhibiting a crossingline pattern with the cross-angle of 45°. The lines have an interval of about 50 µm between each other and average depth of about 1 µm, a dimensional arrangement amenable to be replicated by laser surface texturing. Accordingly, this was attempted on the contact surfaces of the selected hardmetal guiding stones. (a) (b) Figure. 2. (a) Kinematics of honing processes and (b) 2D surface topography of honed steel workpiece where crossing-line patterns are evidenced. 2.3 Fabrication and characterization of textured surfaces 2.3.1 Determination of surface patterns on cemented carbide guiding stones Two different laser-induced surface patterns are investigated in this study, namely crossing-line (aiming to replicate aforementioned honed surface patterns) and dimple arrays, the latter as supplementary case study. Two array densities are chosen for ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT each pattern, i.e. crossing-line patterns with intervals of 100 µm and 50 µm [L(100) and L(50)], and dimple arrays with intervals of 100 µm and 50 µm [D(100) and D(50)]. Meanwhile, the dressed guiding stone D76 and one surface finish corresponding to polishing condition are used as references, for comparison purposes. Table 2 lists all the surface patterns studied. Table 2. Surfaces patterns (induced by either laser-texturing, polishing or dressing) considered in this investigation. Designation Material Pattern Characteristics L(100) 15CoC Crossing-lines Ra < 1 µm, interval of 100 µm L(50) 15CoC Crossing-lines Ra < 1 µm, interval of 50 µm D(100) 15CoC Dimple arrays Ra < 1 µm, interval of 100 µm D(50) 15CoC Dimple arrays Ra < 1 µm, interval of 50 µm Polished 15CoC Polishing lines Ra < 1 µm D76 Diamond Random grain exposure Ra ~ 3-5 µm 2.3.2 Laser configuration and machining parameters Crossing-line patterns and dimple arrays were fabricated using a laser machining platform (Figure 3(a)), including a solid-state Nd:YLF, Q-switched laser set-up (Explorer® One™ Spectra Physics) and a 2-axis laser beam deflection unit (Raylase GmbH). The emitted laser beams have a Gaussian profile (Figure 3(b)) and possess a wavelength of 349 nm, a pulse duration of 5 nanoseconds (FWHM), and a nominal pulse repetition frequency of 1000 Hz ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 2.4 Frictional tests on a precision tribometer In the tribological tests, a reciprocating tribometer (Basalt®-N2, Tetra) was implemented to evaluate the frictional performance of the surface topography variants. Table 3 shows the experimental configuration of the tribotest. Lubricant Kadio50 with kinematic viscosity 5 mm²/s was fed in the experiments. Spheres with diameter of 3 mm of an AISI 52100 (100Cr6) steel (material commonly used for manufacturing bearings) were applied as the counterpart, and testing conditions were room temperature and environment. Contact pressure may be predicted on the basis of Hertzian theory [27,28], according to: 𝑝0=1 𝜋(6𝐹𝐸∗2 𝑅2)1 3 where 1 𝐸∗=1−𝑣12 𝐸1+1−𝑣22 𝐸2 (4) po: maximum contact pressure, F: normal load E1, E2: modulus of elasticity of the two contacting bodies ν1, ν2: Poisson’s ratio of the two contacting bodies R: radius of sphere The contact stress was then chosen as 1.33 GPa. Considering that grains or debris usually have effective sections about several ten microns, the applied normal pressure on them may instantly reach very high levels during the cutting processes. In some few cases, the cutting process confronts a very high load due to the discrimination of microstructure, which could be up to ten times higher. However, the probability of finding such an extreme scenario is rather low. Instead, much lower normal forces, in the range of few Newtons or less, could be expected, although they still would yield relatively high Hertzian contact stresses values (at the level of GPa). Hence, a contact stress value of 1.33 GPa was used in this study. ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT Table 3. Experimental parameters of tribotest conducted. Normal Load FN (N) Trajectory Frequency (Hz) Amplitude (mm) Cycle 1 Sinus 1 1 1000 3 Results and discussion 3.1 Inspection of the laser-produced surface patterns 3.1.1 Characterization of geometrical features Figure 4(a,b,c) shows array and individual features of the laser-produced dimples on the surface of 15CoC material. Dimples exhibit diameter and depth of 20 µm and 0.7 µm approximately, and are evenly distributed along xand yaxis with a distance of 100 µm and 50 µm. They correspond to dimple density values of 100 mm-2 and 400 mm-2 respectively. Dimple depth experimentally measured using LSM are within the range between 0.6 and 0.8 m. is slightly larger than the theoretical estimation of 0.5 µm. It seems that laser ablation is sensitive to the microstructural properties of the cemented carbides, i.e. the cobalt content and the geometrical features of WC grains, including size and distribution. becomes more pronounced due to the relatively high content of the cobalt in the composite. The geometrical characteristics of the WC grains, including size and distribution, might also be important factors, as supported by the finding of deeper ablation as coarser microstructures are laser treated [25]. Figure 4(e,d,f) shows dimensional and geometrical characteristics of the crossing-lines produced on the hardmetal surface, aimed to replicate the surface topographic features of the honed workpiece. These lines cross each other with an angle ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT of 45°, exhibiting a width of 20 µm and a depth ranging from 1.0 to 1.2 µm. Compared to the dimple depth produced by single laser pulses, the groove depth is increased by the moving laser beams, i.e. the overlap of laser pulses. However, it is still difficult to predict the exact increase of the ablation depth or the ablation volume by the overlap percentage, as it can be influence by many factors, including the accumulative effect of thermal reactions resulting from continuously moving laser pulses in short time. (a) (b) (c) d) (e) (f) Figure 4. LSM images of laser-induced surface topography: dimples with 100 µm (a) and 50 µm (b) intervals, and 3D image with cross-sectional analysis (c); and crossing lines with 100 µm (d) and 50 µm (e) intervals, and 3D image with cross-sectional analysis (f). 3.1.2 Surface integrity assessment of the laser produced patterns Surface integrity of the produced dimples has been assessed by combined scanning electron microscopy (SEM) and focused ion beam (FIB) for cross-sectional analysis. 100µm ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT The shaped dimples are arranged regularly along xand y-axis and the grooves cross each other with an angle of 45°, forming the crossing-line patterns. Figure 5 shows the dimple arrays and crossing-line pattern with intervals of 100 µm and 50 µm, i.e. D(100) and L(50) as examples. (a) (b) Figure 5. SEM images of laser-induced surface topography for (a) D(100) and (b) L(50) arrays. One specific dimple on the surface of the D(100) sample is illustrated in Figure 6(a). A processed zone and a heat affected one (HAZ) may be discerned. The processed zone is located within a circle of diameter about 20 µm, i.e. close to the spot size of 19.6 µm at the focus point. It should be noticed that binder is completely removed by the evaporation, while left WC grains form a skeleton structure. Some pores can be found between adjacent WC grains. The remaining molten material, together with some micro-cracks, expands from the center to the surroundings. The expansion complies with the Gaussian profile of the laser energy, from maximum to minimum levels. The ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT HAZ situated adjacent to the processed one just outside the referred circle, is about 2 µm in width. The binder close to the processed zone is found to be slightly molten. The material in this zone is not directly ablated by the laser beams, but rather due to thermal effects such as heat diffusion from the processed zone, as often observed in laser welding processes. Hence, HAZ may be described as a transition region between processed and non-affected zones. In general, application of an ultra-short pulse can effectively diminish or avoid the heat affected zone, as the thermal reaction is no longer dominant during laser ablation [29,30]. (a) (b) Figure 6. SEM image of surface topography for: (a) one dimple on D(100), and (b) cross-line intersection on L(50). Figure 6(b) shows the intersection position of the cross-line structures on the L(50) pattern. In the processed zone, molten material was oriented along the laser scan direction and formed a layer by layer stack of ‘scales’. Meanwhile, molten material spreads towards both sides and covers ablation borderlines, as marked by red lines in such figure. The measured distance between two adjacent scales is about 2.3 µm ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT approximately, close to the theoretical overlap gap of 2.4 µm, i.e. that corresponding to the overlap value of 87.8%. (a) (b) Figure 7. SEM image of (a) scale structure and (b) laser processed zone (FIB-milled cross-section) on crossing-line pattern. Figure 7(a) shows details of the ‘scale’ structures covering the processed zone, exhibiting irregular morphological features. Tiny cracks and pores are dotted among those randomly distributed ‘droplets’, resulting from re-solidification of the molten material. Cross-sectional analysis (figure 7(b)) proves that thickness of the recast layer is different depending on specific location regarding microstructure, i.e. thinner as it is linked to WC grains instead of Co binder, being about 500 nm for the latter. The laser pulse removed both Co and WC constitutive phases; however, as laser pulse penetrated deeper into the binder, this experienced more ablation than the ceramic phase. These findings agree with theoretical estimation that binder Co is more sensible to the laser beams than WC grains, as the former has lower melting and evaporation temperatures. Side effects induced by the thermal reaction, such as pores and cracks, are only evidenced at the recast layer. The applied fluence 6.7 J/cm² is able to carry out a moderate ablation of the 15CoC material by melting and evaporating both components. ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT However, the ablation mechanism of both components, especially the evaporation of WC at the corresponding fluence, should be ratified by future experiments. 3.2 Frictional performance of different surface patterns Figure 8 shows the evolution of the coefficient of friction (COF) as a function of the sliding length for the studied surface patterns during the tribotests. In general, all the surface patterns investigated, including the guiding stone, exhibit similar trends. At the beginning of the test (sliding distance lower than 250 mm), due to unstable contact conditions, a relative abrupt drop from relative high values was observed for COFs. For sliding distances longer than 250 mm, the evolution of COFs gradually slows down until reaching steady state conditions. This is not the case for the polished surface variant, an exception to the described response. Based on these observations, the frictional evolution of the surface patterns can be divided in two main stages: a runningin state, when the sliding distance goes from 0 to 250 mm, and a steady one, when the sliding distance ranges from 250 to 2000 mm. Accordingly, experimental findings are now discussed in terms of this classification. ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT Figure 8. COF as a function of the sliding length for all the surface patterns investigated.  Frictional response during the running-in state (sliding from 0 to 250 mm) It is noted that all the samples had a similar running-in state: COF curves first drop abruptly and then keep decreasing smoothly. This is not the case of the polished one, which instead increases gradually after the drop (yellow line in Figure 8). A hypothesis is here proposed to depict the phenomenon observed. During the running-in state, the applied oil was not evenly distributed on the contact surface yet. A higher amount of oil than the one predicted could be accumulated in the contact area. The temporary thicker oil film then leads to a decrease of COF in the running-in state. The film thickness at the contact area will finally fit the estimated value, only once the oil had been perfectly spread.  Frictional response during the steady state (sliding from 250 to 2000 mm) It is found that COFs achieved by the non-laser textured surfaces, i.e. guiding stone (0.14) and polished one (0.12), are higher than those obtained by the laser-textured ones, 0500 1000 1500 2000 0.10 0.11 0.12 0.13 0.14 0.15 0.16 0.17 0.18 0.19 0.20 0.21 0.22 0.23 0.24 COF Sliding Length (mm) D(100) D(50) L(100) L(50) Guiding Stone Polished ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT i.e. dimple arrays (0.11) and crossing lines (0.10) for both densities. Once sliding response gets into steady state, crossing-line pattern achieved the lowest friction during the whole test, followed by the dimple arrays and then the polished surface. On the other hand, the guiding stone achieved the highest COF, as it exhibits a quite higher surface roughness values. The interval of the dimples, i.e. pattern density, has an impact on the frictional performance. For example, higher dimple density on the contact surface results in a more obvious drop of COF at the beginning of the frictional test. However, this correlation does not apply for the case of crossing-lines. In this tribological system, under non-continuous lubricant feeding, crossing line-like pattern is more beneficial to maintain a stable tribological performance, as the lubricant reservoirs are connected to each other. Thus, it becomes easier to compensate the lubricant volume as it is being consumed. On the other hand, dimples are not connected to each other. Accordingly, as certain dimples run out of lubricant, they cannot be refilled as there is not any channel-like connection among adjacent dimples; therefore, contact becomes rough and unstable. Lack of lubricant is more obvious when dimple density is higher. This viewpoint is supported by the instability occurred during the tests in the cases of the guiding stone and dimple arrays with 50 µm interval, directly related to severe changes discerned in the corresponding COF curves. They could be rationalized on the basis of the significant changes taking place in surface condition, related to either eruption of abrasive grains or dimples running out of lubricant. The curve of the polished surface exhibits a slow increase at the beginning of the steady state, yielding then a higher value of COF as compared to those determined for the laser patterned samples. According to Hamrock and Dowson [31] a lambda value, describing the ratio of minimum film thickness divided by surface roughness, can be evaluated. The calculation of minimum film thickness is given by: ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 𝐻𝑚=3.63𝐺0.49𝑈0.68𝑊−0.073(1−𝑒−0.68𝑘) (5) where G, U and W are the dimensionless parameters of material properties, speed and load, respectively. The lambda value of the test of the polished surface was evaluated as 0.002, indicating that the average oil film thickness is much less than surface roughness (previous fine polishing reaches a surface roughness about 0.1 µm). Accordingly, the lubrication condition can be determined as the boundary lubrication regime [32]. In this condition, asperities are involved in the contact mechanism that causes unavoidable wear [33]. As a consequence, the gradual increase of COF of the reference sample can be attributed to a transition of the boundary lubrication before reaching the steady state. However, the laser patterns, providing extra lubricants on the surface, changed the lubrication condition of the sliding test. The structure produced by laser, performing as lubricant reservoirs, can effectively increase the volume of oil storage in the contact area. The resulting lubrication condition was modified subsequently to the mixed lubrication regime due to the surface structuring, which improved the tribological performance. Therefore, lower values of friction coefficient were observed compared to the specimen without patterns. 3.3 Inspection of surface condition after the tribotest After rubbing for 1000 cycles, sample surfaces were examined to evaluate the condition of the structure produced by laser surface treatment. In general, no measurable wear damage was found by white light interferometry. Figure 9 shows surface aspect of rubbed area for laser-textured hardmetals used as guiding stones in tribotests conducted. Along the orientation pointed by the arrow, it is hardly to observe any wear tracks. It ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT [33] S.M. Hsu, R.S. Gates, Boundary lubricating films: formation and lubrication mechanism, Tribol. Int. 38 (2005) 305–312. doi:10.1016/j.triboint.2004.08.021. ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT  Reproduction of honed surface topography on cemented carbide grade  Satisfactory surface integrity and geometrical precision obtained by nanosecond laser  Crossing-line patterns achieved better frictional conditions than dimple arrays  Pattern density has a significant impact on frictional performance ACCEPTED MANUSCRIPT