In�uence of the �bre laser parameters on the surface texturing of 420 stainless steel

The surface of an annealed 420 stainless steel plate was textured using a �bre laser in this work. Four hundred textures were produced from eighty different combinations of processing parameters (laser power, scan speed, line spacing and number of passes). The most promising textures (fourteen combinations of the processing parameters) were obtained with laser power values of 16 and 64 %, scanning speeds up to 2,000 mm/s, number of passes of 8 or higher and line spacings of 40 and 50 µm. The lower the laser power, the lower the scanning speed and the high number of passes and line spacings were required for a suitable ablation process. High laser powers were responsible for the dissolution of the chromium carbides in the laser tracks (top and edges of the unmachined areas). The hardness of these regions was 320 ± 11 HV against 255 ± 5 HV for the unmachined zones.


Introduction
AISI 420 martensitic stainless steel is a material widely used in the mould industry as well as having other tooling applications such as surgical tools [1][2][3] due to its high tensile strength, hardness, and corrosion properties [1][2][3][4][5]. In the annealing state, its structure consists of a ferritic matrix with a homogeneous dispersion of M 23 C 6 spheroidised carbides with a high ductility [3]. The tensile strength and hardness properties are improved by a quenching heat treatment in which a martensite structure is formed [3]. The presence of alloyed chromium increases the corrosion resistance [1,6]. However, this material has some limitations where good tribological properties are required, i.e. where controlled friction, and increased wear resistance and lubrication are needed.
Different approaches have been used to modify the surface properties of steel components such as sandblasting, chemical etching, deposition of coatings, electrode discharge machining (EDM), electrochemical machining (ECM), and texturing by electron beam, electric arc, or laser ablation [7][8][9][10]. Laser surface texturing, in particular, has been an e cient surface modi cation technique to improve the thermal e ciency of components, as well as their tribological behaviour [11,12]. However, just one study on the laser surface texturing of 420 stainless steel was found in the literature. In that study, Pan et al. [13] used a picosecond laser for the fabrication of a superhydrophobic antibacterial textured surface on 420 stainless steel plates.
Several types of lasers such as CO 2 (10.64 µm), YAG (1064 nm) and diode (848 nm) can be used for machining or to texture surfaces. The type of laser and wavelength to be used for a speci c application must be selected according to the material to be textured since it in uences the energy absorbed [14,15].
This work concerns the texturing of a 420 stainless steel by an Nd: YVO 4 bre laser. A detailed study of the in uence of the laser parameters on the quality of the textures produced is presented and discussed. Four hundred textures were produced from eighty different combinations of processing parameters (power, scan speed, and number of passes).

Experimental Procedure
Square 420 stainless steel (35 mm ⋅ 3 mm) samples were textured using an Nd: YVO 4 bre laser (Model XM-30D Fibre Laser Marking Machine) with a wavelength of 1,064 nm, maximum power of 30 W, spot size of 10 µm, and pulse width of 10 µs. The processing parameters (laser and drawing parameters) used in this work are shown in Table 1. The texturisation process is presented in Figure 1. The number of lines varied from 10 to 50 depending on the drawing parameters. The greater the line spacing the lower the number of lines. Table 1 Summary of the processing parameters used for the laser machining of the 420 stainless steel surface.

Laser parameters Drawing parameters
Laser power -P (%) Scan speed -s (mm/s) Number of passes -n Line spacing -l (µm) The set of parameters presented in Table 1 correspond to different laser energy uences (F), de ned as the laser energy divided by the area (J/mm 2 ). This parameter is calculated considering the laser power (P ) in W, the scanning speed (s) in mm/s, the number of passes/loops (n), and the line spacing/interspacing between adjacent lines (l) in mm, according to the equation: In this work, the laser energy uence value varied from ~0 to 7,680 J/mm 2 .
The dimensional/topological analysis of the textures (surface width and depth of the grooves) was done using the scanning electron microscopy (SEM) images (NanoSEM-FEI Nova 200 (FEG/SEM)), by means of an image analysis software (Image J), and by 3D pro lometry (In niteFocus from Bruker alicona).
Prior to the observations, the samples were polished with silicon carbide abrasive papers with successive grades till 4,000 mesh. The in uence of the laser parameters on the microstructure and grain size of the 420 stainless steel was assessed by optical metallography after etching the samples with Vilella's reagent. X-ray diffraction (XRD) (PANalytical X'Pert PRO MPD with Co Kα radiation) was used for the structural analysis. Hardness of the samples was evaluated by means of Shimadzu HMV-2 equipment with a load of 10 gf and a dwell time of 15 s.

Results And Discussion
Raw material The raw 420 stainless steel plate was formed by ferrite and spheroidised chromium carbides (Fig. 2), characteristic of an annealed state.
Texturing Figure 3 shows the SEM images of the textured 420 stainless steel produced with different processing parameters. The quality of the textures was evaluated considering two factors: continuity of the machined grooves and dimension of cavities produced. Fourteen different combinations of processing parameters were found to give rise to continuous textures (marked in red in Fig. 3). Another fourteen conditions (marked in yellow) corresponding to low laser powers and high scanning speed did not reveal any visible cavities using SEM cross-section analysis, probably due to the accumulation of material at the site where the supposed laser ablation occurs or lack of removal capacity (low laser energy uence) [16,17].
The de nition and quality of the grooves depends on the processing parameters. For low laser powers, low scanning speeds and a high number of passes and line spacings are required for a suitable ablation process. Low laser powers (1 and 16%) combined with high scan speeds (2,000 and 5,000 mm/s), and line spacings of 40 and 50 µm, led to low-quality and non-uniform textures, characterised by discrete machined areas (Fig. 4a). This phenomenon is due to the low laser energy uence obtained for these parameter combinations. Furthermore, for low laser power (1%) and mainly for a low number of passes, it is possible to observe a continuous action of the laser, but non-machined areas are also observed (Fig.   4b). Average power laser values of 16 and 64%, scanning speeds up to 2,000 mm/s, 8 passes or more and line spacings of 40 and 50 µm gave rise to the largest number of good quality grooves (Fig. 4c). In fact, the de nition of the machined lines increased as the line spacing increased, because of the reduced interaction between successive passes of the laser.
For the maximum laser power (100%), high scanning speeds and greater line spacings are required to produce suitable textures. The number of passes plays an important role in this case: for n = 1, the tracks are shallow whilst for n > 8 an excessive remelting of the base material occurs. The high energy leads to an exaggerated increase in the surface temperature, which causes the material around the area being irradiated by the laser to melt, therefore impeding its ablation. This phenomenon was also detected for high laser powers and low scan speeds and leads to some material accumulating inside the grooves. This is particularly true for a high number of passes (Fig. 5).
In fact, the amount of laser energy absorved by the material decreases signi cantly throughout the thickness and has a great in uence on the surface outcome. Therefore, if the energy is high enough the material is ablated, and it creates a textured surface. This may occur when combining medium to high laser power with a medium scan speed. On the other hand, a remelted surface occurs when the amount of energy is insu cient for ablation or is too high and causes the material to overheat, which occurs when high laser powers and low scan speeds are used.
The morphology of the abovementioned fourteen promising textures was analysed by 3D pro lometry (Fig. 6). The corresponding width and depth values were determined in the middle of the textured grooves and are presented in Figure 7.
From Figures 6 and 7 it is possible to conclude that: (i) the groove width grows with the increase in the number of passes for medium and high laser powers and scanning speeds (16-64% and 500-2,000 mm/s, respectively). For low laser powers (1%) and scanning speeds (100 mm/s), increasing the number of passes is no longer bene cial to increase the width. Regarding the depth, as expected, it tends to increase with the number of passes; (ii) the increase in laser power has a signi cant impact on width and depth (cases analysed: P64 s2,000 n8 l50 and P100 s2,000 n8 l50). The percentage difference veri ed in the case of the width is 20% while in the case of the depth it is 16%; (iii) in the case of scanning speed, two cases could be analysed: the use of low scanning speeds (100 and 500 mm/s) and the use of high scanning speeds (2,000 and 5,000 mm/s). In the former, the increase in scanning speed led to a sharp decrease in width, while in the latter the variation in width is inverse. On the other hand, in the case of depth, a low scanning speed has a no signi cant impact while a high scanning speed increases this value. Therefore, for low laser powers (16%) and scanning speeds (100 and 500 mm/s), increasing the scanning speed has a strong negative impact on width, whereas for high laser powers (100%) and scanning speeds (2,000 and 5,000 mm/s), increasing the scanning speed has a strong positive impact on width; (iv) The line spacing proved not to be a determinant factor regarding the width and depth of the textured grooves for the parameters studied.
(P = laser power in %, s = scanning speed in mm/s, n = number of passes, l = line spacing in µm) Figure 8 shows the SEM images of the cross-section of the grooves. For low laser powers (1 and 16%), the grooves produced present a waveform, with no noticeable differences when the other parameters are changed (scan speed, number of passes, and line spacing). On the other hand, when the laser power is increased (64 and 100%) the cavity has a more de ned shape. In this case, the increase in the number of passes led to the formation of a cavity with a more funneled shape. Therefore, it is possible to conclude that the width, depth, and shape of the cavities are in uenced not only by the laser energy uence but also by each individual process parameter. For the same laser energy uence values, the change in laser power and scanning speed was shown to affect these cavities' characteristics signi cantly (width, depth, and shape). Comparing the results obtained by 3D pro lometry and SEM (Figs. 7 and 8, respectively), a discrepancy in width and, mainly, depth of the grooves is observed. This can be explained by Figure 9 which shows that the depth of the cavity along the groove is not constant, it is greater at the beginning of the ablation process. This is due to the excessive increase in the surface temperature and with the formation of a great amount of the liquid phase that impedes the penetration of the laser beam and consequent ablation of the material.
The in uence of the processing parameters on the microstructure of the different textured samples was assessed by optical metallography from their cross-section. Figure 10 shows the two typical examples of textures obtained from different processing parameters.
For low laser powers, the scanning speed did not in uence the microstructure of the base material (Fig. 10a), formed by ferrite + chromium carbides. However, for higher laser power values, and number of passes higher than 1, white regions with no carbides could be detected on the top and the edges of the unmachined areas, corresponding to remelted base material (Fig. 10b). This means that high laser powers are responsible for the dissolution of the chromium carbides. Microhardness evaluation was performed in these regions and far from the machined tracks. Values of 320 ± 11 HV and 255 ± 5 HV were obtained, respectively. The dissolution of the chromium carbides led to the incorporation of chromium in the austenite at a high temperature with the corresponding hardening of the microstructure after cooling.

Conclusions
An Nd: YVO4 bre laser was successfully used to texture an annealed 420 stainless steel plate. The following main conclusions can be drawn from this study: The processing parameters in uenced the ablation process.
The most promising textures were obtained with laser power values of 16 and 64%, scanning speeds of up to 2,000 mm/s, number passes of 8 or higher and line spacings of 40 and 50 µm.
For low values of laser power, low scanning speeds and a high number of passes and greater line spacings were required to obtain a good quality texture. The inverse situation was observed for high laser power values, which was responsible for microstructural changes on the top and edges of the regions adjacent to the machined ones (chromium carbide dissolution), with an increase in hardness.

Con icts of interests/Competing interests
The authors declare that they have no known competing nancial interests or personal relationships that could have appeared to in uence the work reported in this paper.
Availability of data and material    SEM images of the 420 stainless steel textured samples processed with different processing parameters (laser power, scan speed, number of passes and line spacing).

Figure 5
Example of the accumulation of material inside a groove when high laser energy is used (P = 100 %, s = 100 mm/s, n = 128).  Depth pro le of the 420 stainless steel textured with P = 64 %, s = 2,000 mm/s, n = 8, l = 40 µm,.

Figure 10
Typical examples of the microstructure of 420 stainless steel after texturing using (a) low laser power and scanning speed (P = 16 %, s = 500 mm/s, n = 256) and (b) high laser power and scanning speed (P = 64 %, s = 2,000 mm/s, n = 32). The line spacing in both cases was 50 µm.