Characterization of a line-shaped molten iron pool produced by a line-focused single-mode CW laser with a prism technique Open Access

An investigation of the laser-induced melting technique not only contributes to industrial applications such as welding, cutting, and surface removable techniques¹ but also extends laser matter interaction studies over a wide laser intensity range.^1–3 This technology comprises two types of melting phenomena: Heat-conduction-type melting produced by a relatively low-laser-intensity regime of ∼10⁴ W/cm² and keyhole-type melting produced by a high-laser-intensity regime of ∼10⁶ W/cm², wherein vaporization is predominant.³ The conduction type provides a shallow molten pool, and the keyhole type efficiently provides a deep molten pool during high-speed scanning irradiation of a metal target. The advantages of heat-conduction-type melting include the gentle hydrodynamics of the molten pool. In the studies on this type of melting, the molten pool behaviors have been characterized intensively in terms of laser beam shaping,^4,5 numerical simulations associated with experiments.^5–11 In keyhole-type melting, active and hazardous hydrodynamics are driven by the recoil pressure due to laser-induced vaporization,^12,13 resulting in the emission of copious sputter particles,¹³ which have been characterized for laser processing in a variety of industrial applications.^13–19 The laser irradiation control for the realization of the gentle behavior of the molten pool in scanning irradiation technique is important for the suppression of sputtering. Two candidate solutions have been demonstrated, including ring-shaped profile control²⁰ for the creation of gentle hydrodynamics around the keyhole and blue laser irradiation for increasing absorption, especially at the beginning of laser irradiation coupled with high-power near-infrared laser irradiation for successive heating.^21,22

In addition to these, a variety of laser beam shaping technologies for welding have been developed.²³ For example, rather expensive techniques such as diffractive-optical-elements (DOEs) have been demonstrated to produce a variety of irradiation profiles for welding and additive manufacturing.²⁴ In any laser shaping technique, a desired laser beam shaping technique for creating desired molten pool shaping should take into account the thermal energy transport via thermal conduction and convection especially for continuous-wave (CW) laser induced melting.

Based on the above discussion, we expect heat-conduction-type melting without scanning irradiation to be another candidate for sputterless melting. Herein, we propose a refractive-optics-based line-focusing system that includes a prism for producing a desired intensity profile along the line-focus direction as an alternative irradiation-profile-control technique without scanning irradiation. The advantage of the modified line intensity profile was proposed independently as a patent proposal.²⁵ We experimentally investigated the dynamic processes of a line-shaped molten pool produced by line-focused single-mode CW fiber laser irradiation with intensity modification using a prism. The intensity profile modification technique comprising the use of a prism was proposed for uniform-thin-film production as a patent proposal.²⁶ We irradiated a piece of a cold-rolled steel sheet as the target. High-speed cameras were used for recording the laser-intensity-dependent molten pool dynamics on the laser-irradiated and rear sides of the molten pools simultaneously synchronized with the onset of laser irradiation. With this technique, we could obtain stereographical information on the evolution of a molten pool. The experimental observations, thus, obtained indicate that if we take into consideration the heating rate induced by laser absorption and that of the extension of the heated area induced by thermal conduction in conjunction with hydrodynamic motion, we can produce the desired line-shaped molten pools on the laser-irradiated side as well as the rear side of the target. We have demonstrated a simple laser beam shaping technique for production of a desired line-shaped molten pool combined with unique high-speed camera recording technique.

Figure 1 presents the experimental setup for observing the laser-induced melting of targets using high-speed cameras. A laser beam with a power of up to 700 W was delivered using a single-mode CW fiber laser at a wavelength of 1070 nm. Throughout the experiments, the temporal shapes of laser irradiation were flat-top with rise and fall times of less than 0.1 ms. As shown in Fig. 2, the line-focused laser beam with a prism provides a spatially inverted intensity profile along the line-focus direction. It produces an elongated line-shaped molten pool, wherein the angle α in the prism is 2°, as shown in Fig. 2, and the focusing distance of the cylindrical lens is 300 mm. It should be noted that simple line-focusing optics without a prism produces a Gaussian-shaped intensity profile along the line-focus direction. When attempting to focus an intensity that is three times higher at the targets, a spherical lens with a focusing distance of 150 mm is used in addition to the cylindrical lens and prism.

We irradiate a few-micrometers-thick nickel-coated cold-rolled steel sheet target of size 25 × 20 mm and thickness of 0.3 mm. The targets contained carbon, magnesium, phosphorus, and sulfur at concentrations of less than 0.08%, 0.45%, 0.03%, and 0.03%, respectively.²⁷ The thin layer of nickel prevented the oxidation of the steel layer. The melting temperatures of nickel and iron were almost identical at ∼1500 °C, and the temperature was suitable for the present laser power for observing the melting, vaporization, and resolidification processes. The targets were fixed using a metallic support. We observed the dynamic interaction processes on laser irradiation and rear sides of the target simultaneously and characterized the evolution of the molten pools based on high-speed camera recordings. Both cameras were synchronized with laser irradiation. The high-speed camera used to observe the laser-irradiation side was operated at a frame rate of 1000 frames/s, synchronized with an illumination lamp at a wavelength of 850 nm. In addition, each image was monochromatized using a wavelength-selective filter to reduce undesired emissions and clarify the melting, vaporization, and resolidification dynamics with sputter emission. The filtered camera for observing the rear side was operated at a rate of 4000 frames/s and was synchronized with an illumination lamp.

The intensity-dependent interaction processes were observed under the cylindrical lens and prism combination coupled with and without the spherical lens with a focal distance of 150 mm, providing approximate intensities of 6 × 10⁴ and 2 × 10⁴ W/cm², respectively. We did not use nitrogen shield gas for melting experiments except for that described in Sec. III C. High-speed camera recordings obtained using the prism, cylindrical lens, and spherical lens are presented in Fig. 4 (Multimedia available online). In the early part of laser irradiation, two separate molten pools appear, as shown in Fig. 4(a). These molten pools move toward the central part of the line focus, followed by coalescence to form a single line-shaped molten pool, as shown in Figs. 4(b) and 4(c). It should be noted that, in Figs. 4 and 5 (Multimedia available online), the scales in the vertical and horizontal directions depend on the observation angles of the high-speed cameras subtended to the target. The scales were calibrated with a method to compare a snapshot of the high-speed camera recording at the resolidified stage with a corresponding photograph of an irradiation mark taken by a microscope with a scale.

The temporal evolution of the molten pool without the prism is presented in Fig. 5 (Multimedia available online). In the early period of laser irradiation on the laser-irradiated side, a single molten pool appears at the central part of the line focus, and it evolves with increasing length and width owing to the thermal conduction from the high-temperature region, wherein the intensity at the central part of the focusing pattern is at a maximum. This results in the pool appearing far from the laser irradiation profile owing to the larger laser absorption energy and stronger thermal conduction effect than those in the other part of the laser-irradiated region. Similar dynamics were observed on the rear side, with a delay caused by the penetration time of the molten pool.

With high-speed recordings, we plotted the evolution of molten pools as a function of time quantitatively. Using the optical configuration described in Sec. III A, photographs of the irradiation marks were obtained after laser irradiation with an optical microscope, as shown in Figs. 6(a)–7(d). The sizes of the resolidified molten pool in these photographs correspond to those in the final state of the resolidified molten pools in the high-speed camera recordings. We can then estimate the sizes of the molten pool at each frame in the recordings. Figures 7(a) and 7(b) present schematic views of the molten pools produced with the prism, and Fig. 7(c) presents those without the prism. The notations in Figs. 8, 9, and 11–13 correspond to those described in the caption of Fig. 7. Note that we estimated two kinds of errors including a time origin and a reading error of the molten pool sizes as shown in Figs. 8, 9, 11, 12, and 15. The jitter of the time origin was measured with an oscilloscope that displays both laser signals and triggering electric signals for the high-speed cameras to be 0.05 ms, which was shorter than the individual time step of the high-speed camera frames. The error was negligible. The reading error arose when a solid to liquid boundary on each frame of the high-speed camera records was unclear. Specifically, it happened when the molten pools grew rapidly. In this region, the error bars were roughly estimated to be <0.4 mm. While in the flat regions, the error bars were <0.2 mm. The shot-to-shot reproducibility of the times and sizes among three identical target shootings are approximately 20 and 0.2 mm, respectively.

Figure 8 presents the size evolution of the molten pools produced with the prism on the laser-irradiated and rear sides as a function of time from the start of laser irradiation. The melting rise time and coalescence time of the two counter-propagating molten pools were 5 and 40 ms, respectively, on the laser-irradiated side and 75 and 125 ms, respectively, on the rear side. The molten pools do not propagate isotropically; their propagation is restricted to the line-focused region.

Figure 9 presents the size evolution of the molten pools as a function of time at which the laser irradiation begins without prism for performing a comparison of the laser-irradiated and rear sides. The melting rise times on the laser-irradiated and rear sides were 5 and 40 ms, respectively. The error range was the same as that shown in Fig. 8.

We also performed an experiment on a longer line focus using a prism and a cylindrical lens without a spherical lens to characterize lower laser intensity and longer line focus condition. Photographs of the irradiation marks were captured using an optical microscope, as shown in Figs. 10(a)–10(d).

Figure 11 presents the size evolution of the molten pools produced with the prism as a function of time from the onset of laser irradiation on the laser-irradiated side and those on the rear side. The melting rise times and coalescence time of the two counterpropagating molten pools were 40 and 230 ms, respectively, on the laser-irradiated side, and those on the rear side were 230 and 550 ms, respectively. The error bars are identical to those in Fig. 8.

Figure 12 presents the size evolution of the molten pools on the laser-irradiated and rear sides as a function of time from the onset of laser irradiation without prism for comparison. The melting rise times on the laser-irradiated and rear sides were 30 and 120 ms, respectively, which are earlier than those in Fig. 11. The error bars are identical to those in Fig. 8.

We performed an experiment to determine the effect of the nitrogen shield gas. We determined a suitable gas flow configuration for the shielding effect against oxidation, wherein the shielding gas flowed through the walls placed on the left-hand side of the laser-irradiated region, and the gas flowed successively through the other wall, as shown schematically in Fig. 13. The optical configuration of laser irradiation comprised a prism, cylindrical lens, and spherical lens with a focal length of 150 mm. The gas flow rates between 1 and 20 l/s maintained almost an identical shielding effect in the present experiment. Under such gas flow conditions, we expect that nitrogen gas was occupied instead of air in the laser-irradiated region and that the gas prevented oxidation on this surface. In the experiment, the gas flow without walls did not provide a stable shielding effect. We also note that the compressed air flowed over the laser-irradiated targets with the same configuration and gas flow rates as those of the nitrogen gas flow. As a result, the dynamics of the molten pools observed with the high-speed camera were almost identical to those without the gas flow.

The irradiation marks on the laser-irradiated and rear sides are presented in Figs. 14(a) and 14(b), respectively. The shield gas flow maintains a metallic luster on the irradiation mark on the laser-irradiated side. The irradiation marks on the rear side of the target were similar to those shown in Figs. 6 and 10.

The data obtained from the high-speed camera recordings are presented in Fig. 15 and show the size evolution of the molten pools as a function of time from the onset of laser irradiation on the laser-irradiated side. The flow rate was set as 2 l/min. The rise and coalescence times were 26 and 160 ms, respectively, which are 21 and 120 ms later than those without the shield-gas flow condition, owing to the lower heating energy without oxidation. The error range was identical to that presented in Fig. 8.

The material characterization of the irradiated targets described in this section was performed using an electron microscope. The absorption at room temperature was measured using a spectrophotometer at an observation angle of 15° from the target normal. The details are described in the Discussion section.

From the high-speed camera recordings shown in Figs. 4 and 5 (Multimedia available online), the hydrodynamic motions of the molten pools were gentle, and no sputters were observed in the line-shaped molten pools under the present experimental conditions. In this study, we experimentally characterized the interaction dynamics between line-shaped laser irradiation and molten pools.

First, we considered the temperature-dependent laser absorption on the surfaces of a few-micrometers-thick nickel-coated cold-rolled steel sheet targets having a thickness of 0.3 mm. The laser absorption efficiency of a metal target consists of free electron contribution in the conduction band and interband contribution, which depends on the energy band structure of a specific material. It also depends on the surface conditions such as contamination and roughness. The interband contribution is predominant in the visible wavelength range, as described in chapter 2 of Ref. 2. The laser absorption due to free electrons was calculated from the temperature-dependent electric resistivity using the Drude model.²⁹ An expected scenario for the absorption is as follows. At room temperature, the laser absorption of nickel is sufficiently high to initiate a temperature increase on the laser-irradiated surface. Successive laser irradiation increases the temperature, which is followed by an increase in the electrical resistivity that causes an increase in the laser absorption via the free electron contribution, as described previously. Figure 16 presents a schematic of temperature-dependent absorption according to Refs. 7 and 30. The absorption of the nickel layer at room temperature is indicated by a circle that was measured as ∼0.4 in our reflectivity measurement. According to the Wiedemann–Franz law,³¹ the thermal conductivity decreases when the temperature increases, and the absorbed energy tends to be confined to the heated region. Both these processes cause an increase in absorption when the temperature increases. This is favorable for efficient heating and melting.

The energy-dispersive x-ray spectrometry analysis of the irradiated targets revealed that almost no nickel atoms were observed in the resolidified regions, while in the other regions, dense nickel atoms were observed even in the discolored area, wherein the oxygen concentration was a few times higher than that in the area far from the laser irradiation region.

When a prism was used, two molten pools appeared at the right and left edges of the inverted intensity line focus, as shown in Fig. 4 (Multimedia available online). Both counterpropagated to form a molten pool. The separation distance is determined by the angle of deviation of the prism and the focal lengths of the cylindrical and spherical lenses, as shown in Figs. 2 and 3. Without prism, however, a single molten pool arises at the central part of the line focus, and it propagates in opposite directions from the central part along the line focus, as shown in Fig. 5 (Multimedia available online). The molten pools also expand in the translational directions, especially in the central part of the line focus. The resolidified shapes of the molten pools with the prism are presented in Figs. 6(a), 6(b), 10(a), and 10(b), wherein the line-shaped profiles with circular edges can be observed. However, without the prism, as shown in (c) and (d) of Figs. 6 and 10, the central parts of the molten pools grow into an elliptical shape because of the thermal conduction and the convection in the transverse direction at the central parts of the line focus, which results in a shape far from the line shape. We also highlighted that the controllable laser parameters for shaping the molten pool on the irradiation surface are the laser irradiation period and intensity in the present experimental arrangement. The controllable parameters of the prism technique include the apical angle of the prism, beam diameter, and focal length of the spherical lens, which determine the separation distance of the two molten pools in their early stage and the laser intensity profile presented schematically in Figs. 2 and 3. The laser absorption increased with increasing temperature on the thin nickel layer toward its melting point.

Figures 8, 9, 11, and 12 present the evolution of the molten pools during the laser-irradiated and resolidification periods observed at laser intensities of 6 × 10⁴ and 2 × 10⁴ W/cm², respectively. Figures 8 and 11 show that the lines indicated as LL appear to be almost flat as a function of time, which indicates that the heat flow by thermal conduction coupled with laser absorption energy along the line focus contributes to the molten pool formation. The other parts that were affected by thermal conduction without laser irradiation did not reach the melting temperature. We identified two characteristic times: The rise time of melting and coalescence time of the two molten pools. For optimizing the laser parameters for the shaping of a molten pool, we also take into consideration the shape of a molten pool on the rear surface. The rise and coalescence times of melting on the rear surface were delayed from those on the laser-irradiated side by 70 and 85 ms at an intensity of 6 × 10⁴ W/cm² and 190 and 320 ms at an intensity of 2 × 10⁴ W/cm², as shown in Figs. 8 and 11, respectively. For comparison, without using prism, the rise times of melting, as shown in Figs. 9 and 12, are delayed by 30 and 90 ms, respectively.

The thermal time constants and measured delay time constants are listed in Table I. Although the estimated thermal time constant of the 0.3 mm-thick target was 1–2 ms, the observed delay times given by the time interval between the rise time of the molten pool on the laser-irradiated side and that on the rear side listed in Table II were an order of magnitude longer than those estimated from the penetration time based on thermal conduction. We expect that, in the thermal conduction melting regime, the convection driven by the Marangoni effect owing to the gradient of the surface tension on the molten pool dominates to bring thermal energy from a heated region to a colder region in the molten pool rather than conduction, where the expected Péclet number is much larger than 1.^6,7,10 For example, in experiment, the real time observation of convection in an aluminum molten pool was successfully demonstrated^34,35 using ultrabright quasimonochromatic x-ray source at SPring-8 facility. While in theoretical and computational analyses, the speed of convection is on the order of ∼1 m/s, as computed by Paul and DebRoy⁹ and Ebrahim et al.⁵ under similar irradiation conditions, and the absorbed laser energy is carried mainly by convection in the molten pools having a length of several millimeters, <1 mm width, and 0.3 mm depth within the laser irradiation period of 10–100 ms. Shape of a molten pool tends to be wider and shallower,^7,8,11 which causes longer penetration time from the laser-irradiated surface to the rear surface.

The configuration of the nitrogen shield gas flow is presented in Fig. 13. On the laser-irradiated side, a metallic luster irradiation mark accompanying the minimum discolored area can be observed, as shown in Fig. 14. The photographs indicate that no significant oxidation occurred on the laser-irradiated surface. Using energy dispersive x-ray spectroscopy, we observed that the resolidified metallic luster regions contained an order of magnitude less oxygen concentration compared with the resolidified regions without the shield gas. These results confirm the effect of nitrogen shield gas against oxidation. Weak-metallic-luster resolidified regions with a much larger discolored area are observed on the rear side, where the nitrogen gas does not flow. The resolidified iron regions are observed to be an order of magnitude higher oxygen concentration than that in the other regions.

The molten pool evolution revealed that the rise of the molten pool (26 ms) was slower than that without the shield gas (5 ms). This indicates that the oxidation energy accelerated the molten pool evolution accompanying surface oxidation. The dynamics of the molten pools under compressed air-flow conditions did not change in those without gas flow.

We describe the experimental characterization of the dynamics of line-shaped molten pools induced in 2–3 μm-thick nickel-coated 0.3 mm-thick cold-rolled steel sheet targets using a modified line-focused laser irradiation technique consisting of a prism coupled with a cylindrical lens and spherical lens. With high-speed cameras synchronized with laser irradiation, we obtained time-resolved observations of the laser-irradiated molten pools from the laser-irradiated and rear sides simultaneously at average focusing intensities of 2 × 10⁴ and 6 × 10⁴ W/cm² for comparison.

During the experiments, we did not observe any significant sputtering emissions from the molten pools at these intensities. The absorption at room temperature was sufficiently high to initiate successive temperature increases, followed by a further increase in absorption toward melting. Two shallow molten pools appeared, which counterpropagated along the line-focused region to form a single molten pool. The rise time of melting and the coalescence time of both sides were measured at different intensities and line-focusing lengths for quantitative characterization. We believe that not only thermal conduction but also convection driven by the Marangoni effect plays an important role in this molten pool. The penetration of the molten pool appeared to be decelerated by convection. When the nitrogen shield gas flowed on the laser-irradiated surface, the surface oxidization was minimized. At the same time, the required duration of laser irradiation for melting was longer, and the speed of the molten pool evolution was slower than that without the shield gas flow.

We have demonstrated the technique for creating sputterless line-shaped molten pools without scanning irradiation. We should point out that the optimization of irradiation time under conduction mode melting is important for any applications. For future studies, the present study is expected to extend higher power laser induced longer line-shaped shallow molten pools keeping the upper limit of laser intensity and minimizing laser irradiation time, leading to the industrial applications such as the fields of laser welding,^1,2 powder bed fusion,²³ cutting,¹ and surface cleaning.^1,2

We are indebted to H. Hasegawa, M. Tsunekane, T. Ago, and S. Uegaki of Panasonic Corporation; K. Yamamoto and H. Yoshida of the University of Osaka; and Y. Hironaka of EX-Fusion Inc. for their support and fruitful discussions. We also thank K. Ebisawa, S. Matsuura, C. Nakatsuji, and A. Kameyama of the University of Osaka for their assistance. We would like to acknowledge Editage for English language editing.

The authors have no conflicts to disclose.

Hiroyuki Daido: Conceptualization (equal); Data curation (lead); Formal analysis (equal); Investigation (equal); Visualization (equal); Writing – original draft (lead); Writing – review & editing (equal). Takayuki Hirose: Conceptualization (equal); Data curation (equal); Formal analysis (equal); Investigation (equal); Project administration (equal); Writing – review & editing (equal). Daisuke Tanaka: Data curation (equal); Investigation (equal); Writing – review & editing (equal). Takumi Sato: Data curation (equal); Investigation (equal); Writing – review & editing (equal). Keisuke Shigemori: Conceptualization (equal); Funding acquisition (equal); Supervision (equal); Writing – review & editing (equal).