2020 年 61 巻 2 号 p. 244-250
Femtosecond laser machining was employed to trim the surface roughness of diamond coating and to sharpen the edge corner of diamond coated punch toward fine piercing and embossing processes in metal forming. CVD-diamond coated WC (Co) punches with the head diameter of 2 mm were prepared for this laser-trimming process. The initial coated diamond film was 9 µm thick; the curvature of coated punch shoulder was 12 µm. The rotating jig was utilized to attain the homogeneous trimming of punch side surface as well as punch head. SEM (Scanning Electron Microscopy) and Raman spectroscopy were employed to describe the microstructure evolution with trimming as well as the effect of trimming on the nanostructure of diamonds. Fine embossing process into 0.1 mm thick AISI304 sheets with narrow clearance, was utilized to describe their plastic flow around the punch shoulder as well as the geometric distortion by elastic recovery. In addition, EBSD (Electron Back-Scattering Diffraction) was also employed to analyze the difference in plastic flow of AISI304 sheets during micro-embossing between the untrimmed and trimmed punches. The applied load to stroke relation, the plastic strain distribution and the phase mapping as well as the shear droop were compared between two diamond-coated punches to investigate the effect of laser-trimming on the quality of embossed product.
Fig. 11 Comparison of the embossing load and stroke relations between the untrimmed and trimmed diamond-coated pinches.
A diamond coating has grown up as an essential means to protect the WC (Co) tools and silicon parts from wear and friction and to work as a semi-conductor at elevated temperature. In particular, the diamond CVD (Chemical Vapor Deposition)-coated WC (Co) punches and dies are developed to put the dry technology of plasticity into practice without wet lubrication.1) Due to the three dimensional growth of diamond crystals on WC (Co) tool surface, the normal diamond film has high surface roughness in its maximum up to 1 to 3 µm.2) This high maximum surface roughness is not allowed in the precise stamping and forging; a post-treatment of as-coated diamond films is needed to adjust the surface roughness as well as the dimensional inaccuracy in geometry into the tailored surfaces of tools.3) Plasma oxidation processing has been utilized not only to ash the used DLC (Diamond Like Carbon) films coated onto the tools and dies4–6) but also to etch and polish the rough diamond films.7–9) Although these processes are effective to remove the used DLC and diamond coatings and to reduce their surface roughness, it is difficult to adjust the geometric inaccuracies of diamond films on the tool substrate within the tolerance of dimensions required for fine piercing punch in precise stamping of lead frames and fuel-cell separators.10–12)
The diamond film CVD-coated onto the tools, still has rough surface roughness out of the tolerance for fine piercing and embossing.13) As illustrated in Fig. 1(a), the original as-deposited diamond film with rough surface condition, must be trimmed to adjust the coated-punch diameter and height within the tolerance of defined dimensions. Next, the trimmed surface profile is adjusted and polished by reducing the maximum roughness as depicted in Fig. 1(b). Authors14) proposed the excimer-laser trimming process to adjust the geometry and dimension of diamond coated WC (Co) punch within the demanded tolerance for precise piercing of Cu–Be alloy sheets. This pulse-laser irradiation with the short wave length was effective to trim the unnecessary diamond coated films and to successfully adjust the side surface roughness of rectangular punch within 1 µm. More flexible laser processing method is necessary to trim the cylindrical punches as well as the complex-shaped punch array.15)
Illustration of CVD diamond-coated tool before and after the laser trimming. (a) Rough as-deposited diamond coating, and, (b) geometrically trimmed and adjusted diamond coating.
In the present paper, a femtosecond laser machining is employed to trim the geometric inaccuracies and to reduce the punch shoulder curvature as well as the surface roughness on the CVD-diamond films. CVD-diamond coated WC (Co) punches with the head diameter of 2 mm are prepared for this laser-trimming process. The initial coated diamond film is 9 µm thick; the curvature of coated punch shoulder is 12 µm. The rotating jig is also utilized to attain the homogeneous trimming of punch side surface as well as the punch head. SEM (Scanning Electron Microscopy) and Raman spectroscopy are employed to describe the microstructure evolution with trimming as well as the effect of trimming on the nanostructure of diamonds. Fine embossing process into 0.1 mm thick AISI304 sheets with narrow clearance is utilized to describe their plastic flow around the punch shoulder as well as the geometric distortion by elastic recovery. In addition, EBSD (Electron Back-Scattering Diffraction) is also employed to analyze the difference in plastic flow of AISI304 sheets during micro-embossing between the untrimmed and trimmed punches. The shear droop, the phase mapping and the plastic strain distribution around the punch edge are compared between the pierced AISI304 sheets by the un-trimmed and trimmed punches to investigate the effect of laser-trimming on the quality of pierced product.
The laser-trimming and polishing process is instrumented into an experimental setup with notes on the uniform machining and laser beam pass control. Figure 2 illustrates a standard setup for present trimming experiments. The side surface of punch head is first trimmed by using the laser beam controlling in mode-1 where the laser is scanned along the central axis of WC (Co)-punch till the total reduction in thickness reaches to 2 µm. The top surface is secondly processed by using the mode-2. The laser beam is moved from the center to the end of punch head. In both modes of experiment, the end of punch is held in jig to rotate with the constant velocity.
Femtosecond laser trimming experiment set-up with use of mode-1 and -2 operations for trimming the side and top surfaces at the vicinity of punch edge, respectively.
Figure 3 depicts the femtosecond laser machining system with the pulse width in the order of 100 femtoseconds. Since the focused spot of work materials is subjected to ultra-high power irradiation, how to scan the beam spot becomes more important when using this laser machining system. Higher repetition frequency of laser beams as well as higher scanning speed result in fast-rate, dimensionally accurate machining. Table 1 summarizes the characteristic features of the present femtosecond laser machining system. The beam spot in practice is strongly dependent on the controllability of laser beam.
Femtosecond laser machining system.
CVD-diamond coated WC (Co) punch with the head diameter of 2 mm was prepared for laser trimming test. The diamond film thickness is 9 µm. The austenitic stainless steel type A304 sheet with the thickness of 0.1 mm was utilized as a work material. Its average grain size was 7.5 µm.
2.3 CNC stampingThe diamond coated WC (Co) punch with and without trimming was fixed into a cassette die set for embossing experiment under narrow clearance. Figure 4 depicts a typical experimental set up where the stroke is controllable by every 1 µm with in situ measurement of embossing load and punch stroke.
Precise CNC-stamper for micro-embossing of the diamond coated WC (Co) punch into AISI304.
SEM (Scanning Electron Microscopy; Ricoh, Co., Ltd.) was utilized to describe the CVD-diamond films before and after trimming process. Three dimensional profilometer (Infinite-Focus; Alicona Imajing GmbH.) was also used to measure the dimensional change by this trimming process. Raman spectroscopy (Renishaw, Co., Ltd.) was also utilized to describe the nanostructure of as-deposited and laser-trimmed diamond films. EBSD (Electron Back-Scattering Diffraction) was employed to describe the difference in the crystallographic change, the equivalent plastic strain and the induced phase transformation by micro-embossing through direct measurement of the crystallographic orientations, the KAM (Kernel Angle Misorientation) and the phase mapping, respectively.
As-deposited diamond coated WC (Co) punch is depicted in Fig. 5(a) with its SEM image around the punch edge in Fig. 5(b). The punch head and shoulder are uniformly coated by diamond film with the maximum roughness of 2 µm.
An initial diamond-coated punch. (a) Outlook of punch, and, (b) Surface profile around the punch edge.
Figure 6 shows the Raman spectrum of diamond film coated onto the WC (Co) punch substrate. In general, the CVD-coated diamond film is characterized by the sp3 peak at 1330 cm−1 as well as the sp2 peak around 1600 cm−1. The measured spectrum is deconvoluted into two broad peak pairs and one single peak. In each pair, one broad peak is detected at 1330 cm−1 and the other, around 1600 cm−1. This implies that as-deposited diamond film has mainly nanostructure of sp3–sp2 mixture with slightly different contents. In the latter, a single peak is detected at 1330 cm−1; sp3-rich sub-nanostructure is embedded into sp3-sp2 mixture. To be noted, this single peak is sometimes un-detected in the present measurements; this CVD-diamond film is mainly characterized by sp3-sp2 nano-mixture.
Characterization of the as-deposited diamond film by the Raman spectroscopy.
The diamond-coated punch is fixed to have a skew angle of 40° to the surface plate for measurement of the curvature at the punch shoulder edge. The height profile of punch head and side surfaces is continuously measured by using three dimensional surface profilometer along the line A to B in Fig. 7(a). Figure 7(b) depicts the height distribution from A to B. The circular template is incrementally fitted to the curve at the edge for measurement. The curvature (R) of untrimmed diamond film is measured to be R = 12 µm.
Measurement of the curvature at the punch shoulder edge. (a) Coordination of measurement line, and, (b) Height distribution from A to B.
The diamond-coated punch after laser trimming in mode-1 and -2, is shown in Fig. 8(a). The side surface with the length of 150 µm from the punch head was trimmed to remove the film by 2 µm. The punch head surface roughness is also reduced by trimming in mode-2. Figure 8(b) shows the SEM image of trimmed punch around the edge. Although a thin periphery of punch head surface is left even after trimming in mode-2, the sharp edge is formed by this laser trimming.
The diamond-coated punch after laser trimming process. (a) Surface profile at the vicinity of punch edge and (b) SEM image of trimmed punch surface.
Raman spectroscopy was also utilized to describe the difference of nanostructure in the diamond film before and after laser trimming. Figure 9 depicts the measured Raman spectrum as well as its deconvoluted peak profiles. A single small peak detected at 1330 cm−1 in Fig. 6 disappeared after trimming; sp3-rich nanostructure at the surface is ablated by laser irradiation. Other two broad peak pears detected around 1330 cm−1 and 1600 cm−1 in Fig. 9, is detected to be same as those in Fig. 6. Main nanostructure of CVD-diamond film is not affected by laser ablation in this trimming process.
Characterization of the trimmed diamond film by the Raman spectroscopy.
The curvature of punch shoulder edge was measured in the same manner to Fig. 7. As depicted in Fig. 10, the measured curvature by optimum fitting is R = 2.75 µm. The curvature in the as-deposited film is significantly reduced from 12 µm down to 2.75 µm by the present laser trimming.
Measurement of the curvature at the laser-trimmed punch edge. (a) Coordination of measurement line, and (b) local height distribution.
Micro-embossing experiment was employed to describe the effect of curvature at the diamond-coated punch shoulder edge on the shearing behavior. Figure 11 compares the relationship of the applied load to stroke between the untrimmed and trimmed diamond-coated punches. The applied load increases more rapidly with the stroke and the maximum load (Pmax) is also enhanced when using the trimmed punch. In fact, the average of Pmax is 437 N with its standard deviation (σ) of 4.2 N among ten embossing experiments when using the untrimmed punch. In case of the trimmed punch, this average of Pmax reaches to 465 N with σ = 7.5 N. This difference in the load–stroke relations implies that embossing behavior is significantly affected by the edge curvature of diamond film coated on the embossing punch.
Comparison of the embossing load and stroke relations between the untrimmed and trimmed diamond-coated pinches.
Figure 12 compares the SEM image on the cross-sectional microstructure of embossed AISI304 sheets between the untrimmed and trimmed punches. Although the applied stroke is slightly different between two, the affected zone along the shearing line by the plastic flow is narrowed by using the trimmed punch. This suggests that the plastic flow in shearing is forced to concentrate on the vicinity of shearing line when embossing the sharp diamond coating edge into the work materials. As demonstrated in Ref. 16), the measured KAM distribution and phase mapping analyzed on the cross-section of pierced work materials in the EBSD analysis, provided a well-defined information to describe the actual polycrystalline plastic flow during the fine piercing. In the following experiments, EBSD is also employed to analyze the difference in the crystallographic orientation distribution, the KAM distribution and the phase mapping on the cross-sections of embossed materials in Fig. 12.
Comparison on the cross-sectional microstructure of embossed AISI304 sheets by the untrimmed and trimmed diamond-coated punches. (a) Untrimmed punch, and (b) trimmed punch.
Figure 13 depicts the EBSD results of embossed work materials in Fig. 12(a) by using the untrimmed diamond-coated punch. In correspondence to findings in Ref. 16), relatively large KAM and more volume fraction of martensitic phase are noticed along the shearing line. This implies that most of crystals in work are subjected to larger plastic flow together with phase transformation from the original austenite to martensite during the micro-embossing process. On the other hand, these KAM and volume fraction of martensite are reduced along the shearing line during the micro-embossing when using the trimmed diamond-coated punch, as shown in Fig. 14. This reduction of plastic flow, crystallographic spin-rotation and induced phase transformation in embossed work materials reveals that the plastic flow of work materials concentrates at the diamond film edge and along the shearing line by using the sharped edge of diamond coating.
EBSD analysis of the KAM distribution and phase mapping along the shearing line on the cross-section of embossed AISI304 sheets by the untrimmed diamond-coated punch. (a) Inverse pole figure in the normal direction, (b) KAM distribution and (c) phase mapping.
EBSD analysis of the KAM distribution and phase mapping along the shearing line on the cross-section of embossed AISI304 sheets by the trimmed diamond-coated punch. (a) Inverse pole figure in the normal direction, (b) KAM distribution and (c) phase mapping.
CVD-diamond film is characterized by the broad peak pair around 1330 cm−1 and 1600 cm−1 in the Raman spectroscopy. Besides for a sharp peak at 1330 cm−1 in the Raman spectrum for the as-deposited diamond film surface, both Raman spectra before and after laser trimming have nearly the same broad peak pair in Figs. 6 and 9, respectively. In fact, the D- and G-peaks of as-deposited diamond film are detected at 1345 cm−1 and 1540 cm−1 in average while these peaks are detected at 1350 cm−1 and 1550 cm−1 in the trimmed diamond film. After 17), D-peak area (rD) ratio is calculated to describe the difference of broad peak pair in both films; e.g., rD = 62% in the untrimmed diamond film while rD = 65% in the trimmed film. This less difference in Raman characterization proves that the diamond film is not essentially modified by the laser trimming process.
The difference in the micro-embossing behavior in Figs. 11 to 14 between the untrimmed and trimmed diamond films, must come from the curvature of diamond coating at the punch shoulder edge. Let us consider two extreme cases in the embossing process. When using the dull-edge diamond coating, each AISI304 crystal plastically deforms around the diamond film edge with large spin rotation. This significant plastic flow of crystals is mechanically characterized by slow increase of embossing load in Fig. 11. Most of crystals far from the shearing line are subjected to plastic flow by work hardening as seen in Figs. 12(a), 13(a) and (b). The transformed martensitic phase zone surrounds the shearing line in large just as detected in Fig. 13(c). This large crystallographic distortion accompanies with large elastic recovery in each AISI304 crystal after embossing and results in the residual strains left in the work materials as a damage by embossing.
When using the sharp-edge diamond coating with nearly zero curvature, the plastic strain concentrates in work materials at the vicinity of diamond edge corner. Since the work shears by this plastic concentration at the sharp edge corner, the shearing deformation advances into the depth of work with indentation of the punch head into the work. Assuming that the rigid punch is indented into the rigid-plastic work material, its plastic behavior is analyzed by the Prandtl slip-line theory.18) The diamond punch head starts to indent into the work materials when the applied pressure exceeds the critical limit (Pc). Due to this rigid-plastic calculation, the relationship between the applied load and stroke in this extreme case is estimated by the dotted curve in Fig. 11. After elasto-plastic loading in the initial stage, the applied load is held constant by Pmax = Pc and followed by the elastic recovery in the unloading. During the embossing process, the plastic concentration in work materials accompanies with the separation of materials under the compressive static pressure. The burnished surface area ratio increases with indentation; the shear droop is minimized by this indentation.
With reduction of the edge corner curvature in the diamond coating, the measured load–stroke curve as well as the crystallographic plastic straining, approach to the above extreme state. The sharpening of edge corner curvature in the diamond coated punch has much importance on the shearing process during embossing. The diamond coating with the sharp edge corner drives the shearing process of work materials in elasto-plasticity by the plastic flow concentration at its edge.
The femtosecond laser trimming is favored for significant reduction of the edge corner curvature in the diamond-coated WC (Co) punch. No essential change takes in the sp2-sp3 nanostructure of CVD-diamonds after this trimming process. Embossing experiment is performed to describe the shearing process with and without this trimming. Relatively large crystalline plastic flow in the embossed AISI304 sheet is characterized by the wider plastic zone and significant γ to α′ phase transformation in the EBSD analysis when using the diamond-coated punch with dull edge corner. This plastic zone as well as the volume fraction of α′-phase zones are narrowed in the embossed AISI304 by the trimmed diamond-coated punch. The measured load–stroke diagram during embossing process, approaches to the extreme case in the rigid-plastic indentation into AISI304 sheet with decreasing this punch edge curvature. This implies that the plastically strained zone as well as the transformed zone are significantly reduced in the embossed AISI304 sheet by using the trimmed diamond-coated punch.
This reduction of edge-curvature by laser trimming the diamond coating, has also much influence on the shearing process during the piercing process of metallic sheets. Since the affected zone area by piercing process is reduced in the work metals, the wrought damage in work as well as the geometric distortion by elastic recovery are expected to be reduced. This results in high qualification of the pierced products in the metal forming.
The authors would like to express their gratitude to Mr. H. Aihara (Komatsu-Seiki Kousakusho, Co., Ltd.) and Mr. T. Sanbon-Matsu (LPS, Co., Ltd.) for their help in experiments. This study was financially supported in part by the METI-Program on the Supporting Industries in 2019.
