Spatter reduction in deep penetration welding of pure copper using blue-IR hybrid laser

Abstract

Bead-on-plate welding of pure copper with a blue-IR hybrid laser was conducted to achieve a spatter suppression in deep penetration welding of pure copper for the performance improvement of battery and power device of e-mobility. This hybrid laser combines an infrared (IR) laser and a blue diode laser as the welding source and preheating source, respectively. Preheating increases the light absorptivity of pure copper in the IR region. A 1.5-kW blue diode laser was employed to increase the light absorptivity by changing the phase of the preheated area from solid phase to liquid and gas phase. By experimentally investigating the influence of the blue diode laser intensity, the phase effects of the preheated area on pure copper welding with a blue-IR hybrid laser are elucidated. As a result, it was found that a liquid preheated area minimizes spatter during deep penetration welding.

1 Introduction

Pure copper exhibits properties necessary to realize effective energy utilization and carbon neutrality such as high thermal and electric conductivities. Consequently, the global demand for pure copper and the need for high-quality welding of pure copper have increased [1].

Various methods are used to industrially weld pure copper. These include tungsten inert gas (TIG) arc welding [2], resistance spot welding [3], and laser welding [4]. Lasers are attracting attention as a heat source to achieve high-quality welding of pure copper as they can be focused onto a tiny spot and can realize a high-power density. Moreover, they have a high controllability. Due to these advantages, lasers can achieve deep penetration, low heat input, and high-speed welding.

Laser welding can be divided into two modes according to the laser intensity: the heat conduction mode and the keyhole mode. With the low laser intensity, heat conduction mode appears. In the heat conduction mode, laser is absorbed on the top surface of a material and converted into heat. Then, the heat is transferred by heat conduction, melting the material (i.e., forming a molten pool). The heat conduction mode has shallow penetration with a low aspect ratio [5]. In contrast, as the laser intensity increases, vaporization of the material occurs, resulting in a keyhole mode. In the keyhole mode, part of the molten pool is forced down because of the recoil pressure of vaporization and an indentation, called a keyhole, is formed. The incident laser is absorbed through multiple reflections inside the keyhole. The keyhole mode can achieve a deep penetration with a high aspect ratio [5].

Conventionally, an infrared (IR) laser with a wavelength around 1000 nm is used to weld steel, titanium, and nickel due to its high power and high beam quality [6,7,8]. However, welding pure copper with an IR laser is challenging. Because the light absorptivity of pure copper is low in the IR region [9], it rarely melts when using a low-intensity IR laser. A high-intensity IR laser is also problematic for deep penetration due to the temperature dependence of light absorptivity of pure copper in the IR region. The absorptivity gradually increases with temperature, but a jump occurs near the phase transition from a solid to a liquid [10]. This effect induces a sudden increase in the heat input, which leads to sudden evaporation of the molten pool while welding with a high-intensity IR laser. Consequently, the molten pool and keyhole become unstable in keyhole mode welding with a high-intensity IR laser. This generates spatters, which reduces the strength of the welding part and damages areas surrounding the welding part.

Blue diode lasers are attracting attention as a solution to these issues. Blue diode lasers have a wavelength of 450 nm, and pure copper has a high light absorptivity for this wavelength [9]. Morimoto et al. reported that the light absorptivity of pure copper shows a small temperature dependence for blue lasers compared to that for the IR region [11]. Due to these advantages, some researchers have reported the effectiveness of blue diode lasers for copper processing and clarified that a highly efficient heat input can be conducted for pure copper [12,13,14]. However, the output power and beam quality of current blue diode lasers are inferior to those of IR lasers such as a fiber or disk laser.

A blue-IR hybrid laser was developed to realize high-quality laser welding of pure copper. This hybrid laser combines a single-mode fiber laser with a blue diode laser. The blue diode laser serves as a preheating source to increase the light absorptivity of pure copper for the IR laser. Previously, this blue-IR system with a 200-W class blue diode laser preheating was used to weld pure copper. Although blue diode laser preheating increased the penetration depth of pure copper, spatters remained in deep penetration welding [15]. The temperature of the preheating area was around 1000 K, which is below the melting point of copper [16].

To suppress spatters in deep penetration welding, we considered that the absorptivity of pure copper must be further increased. Therefore, we used a 1.5-kW class blue diode laser as a preheating source. This allowed melting or vaporization of preheated copper as the preheated area undergoes a phase transition. If the phase of the preheated area changes to a liquid or gas phase, the light absorptivity in the IR region should be higher than that in the solid phase [10]. This should inhibit the jump in the light absorptivity that accompanies the phase change from solid to liquid when welding with an IR laser, which is one source of spatter generation. Therefore, using a 1.5-kW class blue diode laser for preheating should suppress spatters.

Herein, bead-on-plate welding was conducted with a blue-IR hybrid laser using a 1.5-kW class blue diode laser preheating to realize spatter suppression in deep penetration welding of pure copper. The effect of the phase of the preheated area while welding pure copper with a blue-IR hybrid laser was experimentally investigated. Specifically, the phase of the preheated area was changed by varying the intensity of the blue diode laser. Then, the molten pool behavior and spatters were observed during welding. Additionally, a cross-sectional image of the bead was analyzed after welding.

2 Experimental setup

Figure 1 schematically depicts the experimental setup. A 1.5-kW single-mode fiber laser (Furukawa Electric) was used as a welding source, and a 1.5-kW blue diode laser (Shimadzu) was used as a preheating source. The specifications of this blue diode laser are reported in the literature [17]. The single-mode fiber laser was collimated by an f50-mm plano-convex lens, focused by an f150-mm plano-convex lens, and irradiated vertically onto a sample. Additionally, the blue diode laser was collimated by an f100-mm achromatic lens and focused by an f75-mm achromatic lens. The blue diode laser was irradiated onto the sample at an incident angle of 45°.

Fig. 1

Schematic of the experimental setup

Table 1 shows the experimental conditions. In the experiments, each laser was focused on the top surface of the sample at the focal point. The spot size was set at x, 55 µm and y, 54 µm, for the single-mode fiber laser and x, 424 µm and y, 300 µm, for the blue diode laser. Each laser spot was centered and combined at the processing point. The intensity of the single-mode fiber laser was set at 0 or 42.1 MW/cm², while that of the blue diode laser was varied from 0 to 1.14 MW/cm². Both lasers were irradiated simultaneously by a pulse generator (DG535, Stanford Research Systems).

Table 1 Experimental conditions

Figure 2 schematically shows the bead-on-plate welding test of a pure copper plate. While irradiating, the laser spots were scanned for 25 mm on a 2-mm-thick pure copper sample by a linear stage at a scanning speed of 100 mm/s. High-speed video camera 1 (Q1v, NAC) at an angle of 45° from the horizontal observed a molten pool during laser irradiation. To observe the molten pool, laser light entering high-speed video camera 1 was cut by equipping it with a long-pass filter at a 560-nm wavelength and a short-pass filter at a 900-nm wavelength. High-speed video camera 2 (HX-3, NAC) from a direction perpendicular to the laser scanning direction observed the spatters. Spatters generated during welding were counted one by one from the captured image. The spatter rate was determined by dividing the total number of spatters generated while welding by the laser scanning length. Table 2 lists the observation conditions of each camera. To analyze the spatter generation behavior in more detail, the mass of the pure copper sample was measured before and after laser welding to evaluate the mass loss.

Fig. 2

Schematic of the bead-on-plate welding test

Table 2 Observation conditions of the high-speed video cameras

The intensity of the blue diode laser was varied during the welding test. In the preliminary experiment, pure copper was irradiated with only a blue diode laser to identify the intensity of the blue diode laser that induces a phase transition in the preheated area. In the subsequent experiment, a single-mode fiber laser was combined with a blue diode laser to conduct blue-IR hybrid laser welding. Afterwards, the sample was cut, polished, and etched with 10% hydrochloric acid and iron(III) chloride. Then, the cross-section of the bead was observed with an optical microscope (VHX-5000, KEYENCE).

After the bead-on-plate welding test, lap welding of two 1-mm-thick pure copper plates was demonstrated with the blue-IR hybrid laser. Figure 3 depicts the lap welded sample. The pure copper plate measured 30 mm^w × 100 mm^l × 1 mm^t. The laser was scanned on the center of lapped part for 25 mm at 100 mm/s. The intensity of the single-mode fiber laser was set at 50.5 MW/cm², while the intensity of the blue diode laser was varied. Three samples were prepared for each condition.

Fig. 3

Schematic of the lap joined pure copper sample

Next, the ultimate shear strength (USS) was determined. First, the joined samples were cut into 15 mm^w × 60 mm^l × 1 mm^t pieces. Second, USS was measured using a universal tensile testing machine (Model 1185, Instron) at a speed of 0.083 mm/s. Third, an optical microscope evaluated the fracture surface. Finally, the measured shear strength in N was divided by the fracture surface area to give the USS in MPa.

3 Results and discussion

3.1 Pure copper welding with a blue diode laser

To assess the impact of the blue diode laser intensity on the phase transition in the preheated area, a bead-on-plate welding test was conducted using pure copper and a blue diode laser. Figure 4 shows the results for different laser intensities and times. Pure copper did not melt when irradiating with a blue diode laser at (a) 0.21 MW/cm² or (b) 0.42 MW/cm². Under these conditions, preheating was within the solid temperature range of copper. In contrast, a blue diode laser at (c) 0.64 MW/cm² or (d) 0.85 MW/cm² melted pure copper, forming a molten pool in the heat conduction mode.

Fig. 4

Time elapses of the blue diode laser irradiation spot with the intensity of a 0.21 MW/cm², b 0.42 MW/cm², c 0.64 MW/cm², d 0.85 MW/cm², e 1.06 MW/cm², and f 1.14 × 10⁶ W/cm.²

Irradiating with a blue diode laser at (e) 1.06 MW/cm² or (f) 1.14 MW/cm², a keyhole was formed, and intermittent plume ejections were observed at 0.04 s and 0.07 s with the 1.14-MW/cm² blue diode laser. Due to keyhole formation, the preheated area was in the gas phase. It should be noted that spatter was not observed while welding with only the blue diode laser within the range of the present experimental conditions. Hence, varying the blue diode laser intensity can change the phase of blue diode laser irradiation spot to solid, liquid, or gas without spatter.

3.2 Pure copper welding with a blue-IR hybrid laser

3.2.1 Cross-section evaluation

A single-mode fiber laser was combined with each intensity of the blue diode laser that realized a phase transition in the preheated area. Then, bead-on-plate welding of pure copper was conducted with the blue-IR hybrid laser. The penetration depth and the molten area were measured after laser irradiation. Figure 5 shows the cross-sectional image of the bead after irradiation of (a) 42.1-MW/cm² single-mode fiber laser only and (b) 42.1-MW/cm² single-mode fiber laser with 1.14-MW/cm² blue diode laser. All conditions yielded narrow and sharp wine-cup shape beads, which are characteristic of deep penetration welding.

Fig. 5

Cross-sectional image of the bead after irradiation with a 42.1-MW/cm² single-mode fiber laser only and (b) 42.1-MW/cm² single-mode fiber laser and 1.14-MW/cm² blue diode laser

Figure 6 shows the penetration depth of pure copper as a function of the blue diode laser intensity. The penetration depth increased as the blue diode laser intensity increased. Figure 7 shows the correlation between the molten area of pure copper and the blue diode laser intensity. The circles and triangles represent the results using the blue-IR hybrid laser and blue diode laser only, respectively.

Fig. 6

Correlation between the penetration depth of pure copper and blue diode laser intensity

Fig. 7

Correlation between the molten area of the cross-section and blue diode laser intensity

Both the hybrid and blue diode lasers produced a larger molten area as the blue diode laser intensity increased. For a given intensity, the hybrid laser formed a larger molten area than the blue diode laser alone. Therefore, more than 95% of the molten area generated by the hybrid laser is attributed to the heat input by the single-mode fiber laser.

The melting efficiency $\eta$ was obtained from the acquired molten area. $\eta$ is defined as

$$\eta =\frac{v{S}_{blue+IR}}{{P}_{IR}+{P}_{blue}}$$

(1)

where $v$, ${P}_{IR},$ and ${P}_{blue}$ are the laser scanning speed, output power of the single-mode fiber laser, and output power of the blue diode laser, respectively. ${S}_{IR+blue}$ is the molten area of the hybrid laser.

Figure 8 shows the results of $\eta$ for each blue diode laser intensity. $\eta$ increased with the blue diode laser intensity. Each $\eta$ was classified according to the phase of the preheated area based on initial bead-on-plate welding test with a blue diode laser. When the preheated area was in the solid phase (blue diode laser intensity, 0 ~ 0.43 MW/cm²), the average $\eta$ was 0.015 mm³/J, but when in the liquid phase (blue diode laser intensity, 0.64 ~ 0.85 MW/cm²) and gas phase (1.06 ~ 1.14 MW/cm²), the average value of $\eta$ increased to about 0.021 mm³/J and 0.027 mm³/J, respectively. This effect is attributed to the increased absorptivity of pure copper for the IR laser due to the phase transition of the preheated area.

Fig. 8

Melting efficiency $\eta$ for different blue diode laser intensities

When the phase of preheated area changes from solid to liquid, the absorptivity of pure copper for IR laser jumps [10]. When the preheated area was in the gas phase, a keyhole formed on the molten pool. Consequently, the incident light was absorbed by multiple reflections inside the keyhole, further enhancing the absorptivity [18]. Therefore, $\eta$ increased as the phase of the preheated area changed from a solid to a liquid and then a gas.

3.2.2 Melting and solidification dynamics of pure copper

While welding with blue-IR hybrid laser, spatters were observed from the horizontal direction by the high-speed video camera. Figure 9 shows the spatter rate and the mass loss of sample for different blue diode laser intensities. The spatter rate decreased as the blue diode laser intensity increased from 0 to 0.85 MW/cm². Once the laser intensity reached 0.85 MW/cm², the phase of the preheated area is a liquid, the spatter rate decreased about 90% compared to that of 0 MW/cm². However, an increased spatter rate accompanied the phase transition of the preheated area from a liquid to a gas. The mass loss trend was consistent with that of the spatter rate.

Fig. 9

Spatter rate and mass loss for different blue diode laser intensities while blue-IR hybrid laser welding of pure copper

Figure 10 shows the melting and solidification dynamics of pure copper while welding with the blue-IR hybrid laser at a blue diode laser intensity of (a) 0 MW/cm² (i.e., the single-mode fiber laser only), (b) 0.21 MW/cm², (c) 0.85 MW/cm², and (d) 1.14 MW/cm². It should be noted that these values coincide with a phase change of the preheated area when welding with a blue diode laser only. When the single-mode fiber laser was irradiated onto the pure copper with (a) 0-MW/cm² and (b) 0.21-MW/cm² blue diode laser, the preheated area was in the solid phase and spatters along with the ejection of the molten pool were observed. Using (c) 0.85-MW/cm² blue diode laser, the preheated area was liquid. An ejection of molten pool was not occurred, and only a few spatters were observed. Using (d) 1.14-MW/cm² blue diode laser, the preheated area was in the gas phase, and the fluctuations of the molten pool surface increased. In addition, the keyhole aperture repeatedly expanded and contracted. Not only was spatter generation observed but a plume also formed intermittently while welding.

Fig. 10

Melting and solidification dynamics of pure copper welding at a scanning speed of 100 mm/s with a blue-IR hybrid laser and parameters of 42.1 MW/cm² for the single-mode fiber laser and a 0 MW/cm², b 0.21 MW/cm², c 0.85 MW/cm², and d 1.14 MW/cm² for the blue diode laser

From the captured images, the width of the keyhole aperture (${W}_{ka}$) parallel to the laser scanning direction was measured at different times. The fluctuation rate of ${W}_{ka}$ (FR) was obtained to evaluate the fluctuation of the keyhole aperture for a given blue diode laser intensity. FR is defined as

$$FR=\frac{1}{8}{\sum }_{q=1}^{8}\left|\frac{{W}_{ka, q}-{W}_{ave}}{{W}_{ave}}\right|\times 100$$

(2)

${W}_{ave}$ is the average value of ${W}_{ka}$ from 0.01 to 0.08 s. ${W}_{ka, 1}$, ${W}_{ka, 2}$, …, and ${W}_{ka, 8}$ mean ${W}_{ka}$ at 0.01 s, 0.02 s, …, and 0.08 s, respectively. Figure 11 shows FR for different blue diode laser intensities along with the phase of the preheated area. As the blue diode laser intensity increased, FR decreased in the region from 0 to 0.85 MW/cm² but increased from 0.85 to 1.14 MW/cm². The average value of FR for each phase of the preheated area was 23.2%, 9.2%, and 19.8% for the solid, liquid, and gas phase, respectively. Hence, FR decreased when the preheated area was a liquid. However, FR increased when the keyhole was formed and the preheated area was in the gas phase.

Fig. 11

Fluctuation rate of ${W}_{ka}$ within the time variation from 0.01 to 0.08 s for each blue diode laser intensity during blue-IR hybrid laser welding of pure copper

The trends of the spatter rate and FR for each phase of the preheated area were consistent. Next, we considered the variation mechanism of FR and spatter generation with respect to the phase transition of the preheated area.

Both FR and the spatter rate decreased when the preheated area transitioned from a solid to a liquid. This is due to the increased light absorptivity of pure copper for the single-mode fiber laser. Previously, bulging of keyhole was identified as the origin of the keyhole aperture fluctuation [19]. Additionally, Miyagi et al. reported that the sudden vaporization of the material inside the keyhole is responsible for bulging [20]. Here, we considered that the sudden vaporization of the material is caused by the rapid increase of the heat input into the material from the laser beam. It is caused by the jump in the light absorptivity of pure copper in the IR region accompanies the phase transition from solid to liquid while welding.

Figure 12 shows a schematic diagram of the keyhole dynamics during pure copper welding with a blue-IR hybrid laser. Figure 12(a-1) to (a-4) depicts the keyhole dynamics with a low-intensity blue diode laser preheating. With a low-intensity blue diode laser of 0 or 0.21 MW/cm², the phase of the preheated area remains a solid (a-1). If the single-mode fiber laser is irradiated onto the solid preheated area, a jump of absorptivity occurs as the phase transitions from a solid to a liquid (a-2), leading to a sudden heat input to the material and keyhole bulging. Keyhole bulging pushes the molten pool upward, inducing spatter and ejection of the molten pool with the expansion of keyhole aperture (a-3). After molten pool ejection, the previous keyhole remains as a void and a new keyhole is created with a small keyhole aperture (a-4).

Fig. 12

Schematics of the keyhole dynamics while welding pure copper with a blue-IR hybrid laser when the preheated area is (a) solid, (b) liquid, and (c) gas

In contrast, the phase of the preheated area is in the liquid phase with 0.64- or 0.85-MW/cm² blue diode laser preheating (b-1). If a single-mode fiber laser is irradiated onto the preheated area, which is already a liquid, then the jump in the light absorptivity while welding is avoided (b-2). This should reduce the sudden increase in the heat input and consequently prevent keyhole bulging, resulting in stable welding (b-3).

Both FR and the spatter rate increased when accompanied by a phase transition of the preheated area from a liquid to a gas. One reason is intermittent plume generation while welding, as seen in Fig. 10d. Figure 12(c-1) to (c-4) schematically diagram the keyhole dynamics when the preheated area is in the gas phase. A high-intensity blue diode laser forms a keyhole, and the preheated area transforms into the gas phase. When a keyhole is formed, a plume is ejected intermittently from the keyhole aperture (c-1). Because this plume interacts with the laser based on absorption and scattering [21], the incident laser energy for preheating fluctuates with the generation of plume (c-2). Due to this effect, the keyhole aperture becomes unstable as it repeatedly contracts (c-3) and expands (c-4). The velocity of metal vapor from the keyhole fluctuates when the keyhole aperture fluctuates, causing an unstable molten metal flow around the keyhole [22]. Therefore, the instability of the keyhole aperture is responsible for the unstable molten pool behavior and leads to spatters.

3.3 Pure copper lap welding

To investigate the phase effect of the preheated area on the actual joining process of copper, pure copper plates were lap welded with the blue-IR hybrid laser. Then, tensile testing was performed on a lap welded sample measuring 15 mm^w × 60 mm^l × 1 mm^t. Figure 13 shows USS of the samples lap welded at different blue diode laser intensities. Laser intensities of 0, 0.75, and 1.34 MW/cm² correspond to the phase transition from solid to liquid and gas, respectively. USS of the lap welded sample increased as the blue diode laser intensity increased.

Fig. 13

Ultimate shear strength of the lap welded pure copper sample as a function of the blue diode laser intensity

4 Conclusions

A bead-on-plate welding test of pure copper was conducted with a blue-IR hybrid laser, which combined a single-mode fiber laser for welding with a blue diode laser for preheating. A high-power blue diode laser was used to increase the light absorptivity of pure copper in the IR region by modifying the phase of the preheated area. The effects of the phase of the preheated area on the penetration depth, molten area, and welding dynamics were examined. As the blue diode laser intensity increased, both the penetration depth and molten area of pure copper increased. The melting efficiency $\eta$ also increased with the phase transition of the preheated area from solid, liquid to gas, indicating that the light absorptivity of pure copper for the IR laser increased with the phase transition at the preheated area.

The spatter rate and FR decreased when the phase of the preheated area transitioned from a solid to a liquid. This is attributed to the reduced bulging inside the keyhole and is associated with increased light absorptivity of pure copper for the IR laser accompanying the phase transition from a solid to a liquid. On the contrary, the spatter rate and FR increased as the phase of the preheated area went from liquid to gas due to the fluctuations of the keyhole aperture and molten pool driven by the occurrence of plume. Hence, the phase of the preheated area is a critical factor of the welding phenomenon of pure copper with blue-IR hybrid laser. From the result, it is assumed that deep penetration welding of pure copper with low spatter can be achieved when the preheated area is in the liquid phase.