Laser Welding of High-Strength Galvanized Steels in a Gap-Free Lap Joint Configuration Under Different Shielding Conditions

Yang Supplement December 2010:Layout 1 12/13/10 4:32 PM Page 8

Laser Welding of High-Strength Galvanized Steels in a Gap-Free Lap Joint Configuration under Different Shielding Conditions

By designing the specific shielding conditions, completely defect-free lap joints of the galvanized steels in a lap joint configuration are achieved by a single laser beam without pre- and postweld processing

BY S. YANG, B. CARLSON, AND R. KOVACEVIC

veloped at the interface of the two metal ABSTRACT sheets. The highly pressurized zinc vapor expels the liquid metal out of the molten It is a great challenge to laser weld zinc-coated steels in a gap-free lap joint con- pool and produces weld defects such as figuration due to the formation of highly pressurized zinc vapor. In this study, dif- spatter and porosity during the laser weld-

WELDING RESEARCH ferent shielding conditions were designed to mitigate the highly pressurized zinc ing process. These defects significantly de- vapor. Argon, helium, the mixture of argon and carbon dioxide, and the mixture of grade the mechanical properties of the argon and oxygen were selected as the shielding gases to study the effects of shield- weld joints. ing conditions on weld quality. The introduction of a side shielding gas not only blew In the past several decades, many ef- away the laser-induced plasma but also suppressed the instability of the molten pool forts have been made to suppress the ef- caused by the highly pressurized zinc vapor. Under the optimal setting of shielding fect of the highly pressurized zinc vapor on gas conditions, a stable keyhole was consistently formed that provided a channel to the weld quality. The American Welding vent out the zinc vapor. Under this welding condition, the laser welding process was Society requires complete removal of the very stable. Consequently, a completely defect-free lap joint was achieved in a gap- zinc coating layer at the interface of two free lap joint configuration. Experimental results demonstrated that this newly de- metal sheets along the weld interface prior veloped laser welding procedure was robust and cost effective, does not require pre- to laser welding (Refs. 1, 2). Currently, set- or postweld processing and can be directly applied in the industrial conditions. A ting a small gap between the two metal high-speed CCD camera, assisted with a green laser as the illumination source, was sheets is a common way for industries to used to monitor the behavior of the molten pool and the keyhole dynamics in real join the galvanized steels in a lap joint con- time. Energy-dispersive X-ray spectroscopy (EDS) experiments were carried out to figuration (Ref. 3). Mazumder et al. (Refs. analyze the chemical compositions in the welds. Furthermore, tensile shear tests and 4–6) developed a technique of alloying the microhardness measurements were conducted to evaluate the mechanical properties zinc with the copper before the steel is of the welds. melted. The melting point of the copper- zinc compound is 1083°C (between the melting temperature of steel and the boil- Introduction high-powered lasers to joining galvanized ing temperature of zinc). However, the steels. However, it is difficult to achieve a solubility of copper into the steel could In order to reduce fuel consumption, high-quality weld of galvanized steels in a lead to additional problems such as hot enhance passenger safety, and improve lap joint configuration with using a single cracking and corrosion (Ref. 7). Re- corrosion resistance, different grades of laser beam because of the presence of designing the lap joint to allow the zinc high-strength galvanized steels are in- highly pressurized zinc vapor. The boiling vapor to be evacuated, prior to the molten creasingly used in the automotive indus- point of zinc is 906°C, which is lower than pool reaching the interface of the two try. In the past, the high-strength galva- the melting point of steels (over 1500°C). metal sheets, has been explored in order nized steels used in the automotive During laser welding of galvanized steels to mitigate the effect of the zinc vapor industry were commonly joined with re- in a gap-free lap joint configuration, the (Refs. 8–11). In addition, Pennington et al. sistance spot welding. Considering the highly pressurized zinc vapor is easily de- (Refs. 12, 13) proposed to deposit a nickel high speed, low heat input, deep penetra- coating along the weld interface after tion, and high flexibility of laser beam stripping off the zinc coating at the inter- welding, the automotive industry has KEYWORDS face of two metal sheets. The nickel has a shown significant interest in applying melting point of 1453°C, which is higher Galvanized Steels than the boiling point of zinc. By replacing Gap-Free Lap Joint the zinc coating with a nickel coating in the S. YANG ([email protected]) is senior re- weld area, the laser welding process be- searcher, GM China Science Lab, Shanghai, P.R. Configuration Laser Beam Welding comes stable and accompanied by an as- China. B. CARLSON is lap group manager, Gen- sociated corrosion protection. Unfortu- eral Motors R&D, Warren, Mich., and R. KO- Keyhole VACEVIC is professor, Mechanical Engineering, Shielding Conditions nately, this method will impose additional Southern Methodist University, Dallas, Tex. cost and reduce productivity. Pulsed laser (Ref. 14), dual laser beam or two lasers (Refs. 15–20), and hybrid laser welding

8-s JANUARY 2011, VOL. 90 Yang Supplement December 2010:Layout 1 12/13/10 4:32 PM Page 9

A B

Fig. 1 — A — Experimental setup; B — schematic representation of lap joint configuration.

(Refs. 21–24) were also used to weld gal- A vanized steels in a gap-free lap joint configuration. Gualini et al. (Ref. 18) modified the dual beam to join the galvanized steel sheets in a gap-free lap joint configuration where the first beam cut a slot to provide an exit path for the zinc vapor and B the second beam was applied to join the metal sheets. However, experimental results demonstrated that spatter and porosity were still present in the lap joints. The laser-arc hybrid welding technique was C also used by Kim et al. (Ref. 24) to join SGCD1 galvanized low-strength steel with a 1.0-mm thickness in a gap-free lap joint configuration. It was revealed that the formation of porosity was the main concern D when using hybrid laser-arc welding of galvanized steels. In addition, they showed that process instability was the main cause of the generation of spatter and porosity in the welds. Spatter significantly dam- Fig. 2 — Lap joints obtained in Experiments 1 and 2: A — Top view of the lap joint obtained in Experi- aged the torch electrode and the porosity ment 1; B — bottom view of the lap joint obtained in Experiment 1; C — top view of the lap joint obtained in Experiment 2; D — bottom view of the lap joint obtained in Experiment 2. lowers the mechanical properties of the

welds. Additionally, Gu et al. (Ref. 25) WELDING RESEARCH also introduced the arc into the laser welding process where two heat sources share vacevic (Ref. 28) proposed a new welding this welding procedure in a highly auto- the common molten pool. They claimed procedure, which is a combination of a mated welding application. The automo- that the arc enlarges the molten pool pro- fiber laser with a gas tungsten arc welding tive industry continues to search for a new viding more space for the zinc vapor to es- (GTAW) torch used to preheat the top laser welding procedure to weld high- cape. However, spatter and porosity were surface of the galvanized metal sheets. strength galvanized steels in a gap-free lap still observed in the welds. Recently, a The GTAW preheating process burns the joint configuration with a single laser method was proposed and patented by Li zinc coating at the top surface of the metal beam without the pre- and/or postweld et al. (Refs. 26, 27) in which a thin alu- sheet, which helps in generating a thin film processing requirements. Until now, there minum foil layer was placed along the of the metal oxides (Ref. 28). The heated is no reference in the open literature on weld interface at the interface of two gal- surface with the thin film of metal oxides using a single laser beam to successfully vanized steel sheets to form an Al-Zn alloy will drastically improve the absorption of join galvanized steels in a gap-free lap con- during the laser welding process. They laser beam energy into the welded mate- figuration without the pre- and/or post- claimed that the level of the zinc vapor rial. Furthermore, the zinc coating at the processing requirements. Therefore, it is pressure was decreased through the for- interface of the two metal sheets is trans- important to develop an efficient and ro- mation of the Al-Zn alloy. In order to formed into zinc oxides, which has a bust laser welding technique to satisfy the achieve high-quality lap joints, the two higher melting point (above 1900°C) than demand from the automotive industry. metal sheets should be tightly clamped. Li that of steel (over 1500°C) resulting in less The main objective of this work was to re- et al. (Ref. 27) claimed that if a gap existed vapor generation and a more stable weld- spond to this demand and develop a cost- at the interface of two metal sheets, weld ing process. A completely defect-free, effective and easy-to-automate laser weld- defects would be produced in the welds. high-strength lap joint was achieved. ing technique. Furthermore, the weld became brittle due However, this process requires a specific In this study, the laser welding process to the dissolution of aluminum-steel alloy offset between the laser beam and GTAW was conducted to join galvanized DP980 into the weld. Recently, Yang and Ko- torch that could hinder the application of steel sheets in a gap-free lap joint config-

WELDING JOURNAL 9-s Yang Supplement December 2010:Layout 1 12/13/10 4:33 PM Page 10

Table 1 — Laser Welding Parameters

No. Coaxial Shielding Gas Side Shielding Gas Back Shielding Gas Laser Power Welding Speed (ft3/h) (W) (mm/s) Type Flow rate Type Flow rate Type Flow rate (ft3/h) (ft3/h) (ft3/h) 1 Ar 30 Ar 30 3600 40 2 He 30 Ar 30 3600 40 3 Ar 30 Ar 30 Ar 30 3600 40 4 He 30 Ar 30 Ar 30 3600 40 5 75% 30 Ar 30 Ar 30 3600 40 Ar+25% CO2 6 98% 30 Ar 30 Ar 30 3600 40 Ar+2% O2 7 90% 30 Ar 30 Ar 30 3600 40 Ar+10% CO2 8 Ar 20 Ar 30 3600 40 9 Ar 30 Ar 30 3600 40 10 Ar 40 Ar 30 3600 40 11 Ar 30 Ar 30 Ar 30 3600 30 12 Ar 30 Ar 30 Ar 30 3600 50 13 Ar 30 Ar 30 Ar 30 3600 60 WELDING RESEARCH

A A

B D

Fig. 4 — Effect of side shielding gas on the stability of the welding process and keyhole: A — Top view of the lap joint obtained in Experiment 3; B — bottom view of the lap joint obtained in Experiment 3; C — top view of the lap joint obtained in Experiment 4; D — bottom view of the lap joint obtained in Experi- ment 4.

uration. As shielding gas has a significant determine the chemical composition at effect on the stability of the welding the top surface as well as along the fusion process and the weld quality (Refs. zone of the welds. Microhardness and ten- 29–31), the gases, including pure argon, sile shear tests were carried out to evalu- C helium, and carbon dioxide as well as oxy- ate mechanical properties of the welds. gen, were combined in different ways to study the influences of the shielding con- Experimental Setup ditions on the weld quality and the keyhole stability. An optimal shielding condition The material used in this study was gal- was proposed for the welding of galva- vanized DP 980 steel sheet. The zinc coat- nized steels in a gap-free lap joint config- ing was hot dipped at the level of 60 gm/m2 uration. The mechanism of stabilizing the per side. Specimens with the dimensions of laser welding process was studied. In ad- 200 × 85 × 1.2 mm and 200 × 85 × 1.5 mm dition, a high-speed CCD camera with the were cut using an abrasive water jet. The frame rate of 4000 f/s was used for on-line 1.2-mm-thick metal sheet was selected as monitoring of the dynamic behavior of the the top sheet and the 1.5-mm-thick metal Fig. 3 — Unstable zinc vapor and laser-induced molten pool and the laser-induced plume: A — Taken with color CCD camera at sheet as the bottom sheet. The two metal 30f/s; B and C taken with high-speed camera at plasma. Energy-dispersive X-ray spec- sheets were then tightly clamped together 4000f/s assisted with a green laser. troscopy (EDS) tests were carried out to during the laser welding process and a zero

10-s JANUARY 2011, VOL. 90 Yang Supplement December 2010:Layout 1 12/13/10 4:33 PM Page 11

A A

Fig. 5 — Effect of active gases on the Marangoni D convection pattern by the surface tension gradient in the molten pool: A — The outward pattern; B— the inward pattern (Ref. 36).

gap was assumed. The laser welding process was performed using a 4000-W Fig. 6 — Effect of the active gas on the stability of the welding process and keyhole. A — Top view of the lap fiber laser. The multimode laser beam was joint obtained in Experiment 5; B — bottom view of the lap joint obtained in Experiment 5; C — top view brought into the laser welding head by an of the lap joint obtained in Experiment 6; D — bottom view of the lap joint obtained in Experiment 6. optical fiber with a core diameter of 0.4 mm. The laser welding head provided a examination under the optical microscope. laser beam focused to a 0.6-mm spot with a focal length of 250 mm. During the laser Results and Discussion welding process, the laser beam was focused on the top surface of the two-sheet Investigation on the Effect of Different stack-up. A high-speed CCD camera with Shielding Conditions on Weld Quality 4000 f/s and a color CCD camera with 30 f/s was applied to monitor the laser welding To study the effect of different coaxial process. In addition, a CCD color video shielding gases on the welding quality, camera was used to monitor the laser-in- pure argon and pure helium were used as duced plasma and plume. The chemical the coaxial shielding gas in Experiments 1 compositions of base metal and weld zone and 2, respectively. In addition, the pure Fig. 7 — Cross-sectional view of the sound lap joints obtained in Experiment 4 (laser power: 3600 were analyzed by EDS. A green laser with argon gas was used as the back shielding W; welding speed: 40 mm/s). the center wavelength of 532 nm and a max- gas in Experiments 1 and 2. No side shield- imum output power of 6 W was selected as ing gas was provided in these two experi- the illumination source to suppress the ments. Figure 2 shows the top and bottom laser-induced plume in order to obtain views of the laser-welded lap joints ob- the formation of a stable keyhole. The clear images of the molten pool. Further- tained in Experiments 1 and 2. As shown presence of a stable keyhole will provide more, the influence of the shielding condi- in Fig. 2A, a large amount of spatter and the channel to consistently vent out the tions on the weld quality was evaluated porosity were produced in the laser- highly pressurized zinc vapor. However, WELDING RESEARCH using different combinations of the coaxial, welded lap joints in Experiment 1. The when the coaxially delivered shielding gas side, and back shielding gases. The distance spatter scattered along the laser-beam- was argon and no side shielding gas was between the side shielding gas outlet (Fig. delivered path will absorb and block a por- utilized, a large volume of the laser- 1) and the laser spot was about 10~20 mm. tion of the laser beam energy. However, induced plasma was directly formed on In order to investigate the effect of the when the coaxially delivered shielding gas top of the molten pool, as shown in Fig. 3. coaxial shielding gas on the welding quality, was switched from pure argon to pure he- As shown in Fig. 3, the laser-induced pure argon and helium, and the mixtures of lium and the back shielding gas was main- plasma is very unstable and dynamically argon and 25% and 10% CO2, as well as the tained as pure argon (in Experiment 2), fluctuated over time when argon was used mixture of argon and 2% O2 were used as the laser welding process became very sta- as the shielding gas (Ref. 28). The unsta- the coaxial shielding gas in different weld- ble and no liquid metal was ejected from ble laser-induced plasma not only signifi- ing experiments while maintaining all other the molten pool. A sound weld with com- cantly influences the coupling efficiency of welding parameters constant. Similarly, plete penetration was achieved with he- laser beam energy into the welded mate- two kinds of gases, pure argon and pure helium, as shown in Fig. 2C and D. The large rial but also changes the direction and size lium, were selected as the side shielding gas difference in the weld quality between the of the keyhole (Ref. 34). Under this weld- to study the effect of the side shielding gas shielding conditions specified by Experi- ing condition, the keyhole tends to col- on weld quality in a sequence of experiments 1 and 2 was a result of the different lapse and the highly pressurized zinc ments on the condition that all other weld- ionization potentials of argon and helium. vapor will be trapped into the molten ma- ing parameters were kept constant. Table 1 Helium has higher ionization potential terial. As shown in Fig. 2B, only partial presents the combinations of different and better thermal conductivity than that weld penetration was achieved in the lap gases and the laser welding parameters of argon (Ref. 32). When helium was used joints obtained when argon was used as used in this study. The experimental setup as the coaxial shielding gas, the size of the shielding gas because the laser- is shown in Fig. 1. In addition, the lap joint plasma was small and stable (Ref. 32), and induced plasma and the produced spatter coupons were sectioned, ground, polished, the laser beam energy could be better cou- absorbed, scattered, and blocked the laser and etched for hardness measurement and pled into the welded material resulting in beam energy. This fact suggests that it was

WELDING JOURNAL 11-s Yang Supplement December 2010:Layout 1 12/13/10 4:34 PM Page 12

A critical to control the formation and stability of the laser-induced plasma in order to produce the stable keyhole for the laser welding process of galvanized steels in a gap-free lap joint configuration, thus achieving sound lap joints. The back side B shielding gas can cool the weld to some ex- tent to decrease the pressure level of zinc vapor because the zinc vapor pressure level is directly related to the temperature (Ref. Fig. 8 — Lap joint obtained in Experiment 7: A — Top view; B — bottom view. 32). In general, when large amounts of spatter are produced, only partial penetration can be achieved. A B As mentioned previously, the unstable plasma is one of the reasons for the collapse of the keyhole (Refs. 28, 34). Furthermore, the negative effects related to the unstable laser-induced plasma can be eliminated by applying a shielding gas with an approxi- mate flow rate (Ref. 32, 34). In order to suppress the laser-induced plasma and achieve the stable keyhole, pure argon and pure helium were selected as the side shielding gases in Experiments 3 and 4, respectively. Figure 4 shows the experimental

WELDING RESEARCH results. Neither spatter nor porosity were present in the laser welded lap joints and Fig. 9 — X-ray transmission images of keyhole and transverse sections for increased weld penetration by completely penetrated lap joints were adding oxygen into the shielding gas (fiber laser power: 7 kW; welding speed: 1 m/min) (Ref. 43). achieved with both of these laser welding processes. This fact indicates that the laser welding process of galvanized steels was A stabilized by the introduction of the side shielding gas. During the laser welding process, the side shielding gas will blow away the laser-induced plasma and plume. Furthermore, the use of side shielding gas stabilized the turbulent molten pool caused B by the highly pressurized zinc vapor resulting in the flat surface of the welds (Refs. 30, 34), which will be shown in the following section by the high-speed camera. Com- pared with Experiment 1, the absorption C efficiency of the laser beam energy was increased and the laser beam was relatively uniform enabling it to be coupled into the welded material and generate a stable keyhole. Similar to Experiment 2, the stable keyhole mitigates the highly pressurized D zinc vapor. Marangoni convection driven by the surface tension gradients is one of the main factors to influence the keyhole shape and its dynamics (Ref. 35). It has been revealed E that the addition of active gases such as O2 and CO2 into the argon shielding gas can change the Marangoni convection from outward to inward, as shown in Fig. 5 (Ref. 36). In addition, the surface tension of a molten pool can be lowered (Ref. 37). The F outward and inward Marangoni convection are shown in Fig. 5. When active gases such as O2 or CO2 are introduced into the shielding gas, it is possible to deepen and enlarge the keyhole, compared with the Fig. 10 — Lap joints obtained with different flow rates in Experiments 8–10: A — Top view of lap joints with a side flow rate of 20 ft3/h; B — bottom view of lap joints with a side flow rate of 20 ft3/h; C — top case using pure inert gas as the shielding view of lap joints with a side flow rate of 30 ft3/h; D — bottom view of lap joints with a side flow rate of gas (Ref. 37). In order to investigate 30 ft3/h; E — top view of lap joints with a side flow rate of 40 ft3/h; F — bottom view of lap joints with a whether adding the active gases O2 or CO2 side flow rate of 40 ft3/h (laser power: 3600 W; welding speed: 40 mm/s). into the inert gas could facilitate formation

12-s JANUARY 2011, VOL. 90 Yang Supplement December 2010:Layout 1 12/13/10 4:34 PM Page 13

Fig. 12 — Penetration depth vs. intensity for laser C beam welding process (Ref. 47).

ure 7 shows the cross-sectional view of the lap joints obtained in Experiment 5. As D shown in Fig. 7, no porosity is present in the lap joints. By direct observation of the laser welding process, it was found that the welding process in Experiment 5 with the mixture Fig. 11 — The lap joints obtained at different welding speeds in Experiments 12 and 13: A — Top view of 75% Ar + 25% CO was the most sta- of lap joint obtained at 60 mm/s welding speed; B — bottom view of lap joint at 60 mm/s welding speed; 2 C — top view of lap joint obtained at 50 mm/s welding speed; D — bottom view of lap joint at 50 mm/s ble among all the gas mixtures tested. In welding speed (laser power: 3600 W). addition, the laser welding processes in Experiments 4 and 6 exhibited a more stable process than that of Experiment 3. When using pure argon as the coaxial shielding gas in Experiment 3, the welding process was slightly unstable and a small amount of spatter was observed during the laser welding process. More severely, some porosity was produced in the welds, suggesting that when the coaxially deliv- A BC ered shielding gas contains either CO2 or O2 or is pure helium, the laser welding C process obtains the greatest degree of stability. Therefore, it is recommended to in- troduce CO2 or O2 gas into the shielding gases or use pure helium or the mixture of He and Ar instead of pure argon to stabilize the laser welding process for galvanized steels. Trials were also carried out to D EF join the galvanized steels in a gap-free lap joint configuration with 90% Ar + 10% WELDING RESEARCH CO2. Figure 8 shows the experiment results. As shown in Fig. 8, complete penetration was achieved and the weld bead was continuous without the presence of spatter or porosity. Further studies will be performed to explore the process window for different gas ratios in different shield- G H ing conditions.

Fig. 13 — Images of the molten pool successively obtained with the unstable laser welding process (laser Mechanism for Enhanced Stability of the power: 3600 W; welding speed: 40 mm/s; the pure argon shielding gas used in the coaxial and back sides). Laser Welding Process by the Introduction of Active Gases

As discussed in the previous section, the of the keyhole and enhance its stability for than 25% in order to maintain the same introduction of an active gas (O2 or CO2) laser welding of galvanized steels in a gap- properties in the weld as the base material into the shielding gas can suppress the free lap joint configuration, the mixture of (Ref. 38). In addition, the mixture of 98% highly pressurized zinc vapor and stabilize argon + CO2 and argon + O2 was further argon and 2% O2 is commonly used in in- the laser welding process. During laser tested in Experiments 5–7. The high CO2 dustry for welding steel. Therefore, the welding of galvanized steels, the active or O2 content in the shielding gas may de- mixtures of 75% Ar + 25% CO2 and 98% gases play these possible roles as follows: crease the mechanical properties of welds Ar + 2% O2 were used in the experiments. 1. During the laser welding process, (Ref. 30). When welding steel, the per- Figure 6 shows the experimental results. dissociation and ionization processes take centage of O2 and CO2 in the inert shield- As shown in Fig. 7, high-quality welds with place in the gases (Ref. 39). The chemical ing gas is recommended to be no more complete penetration were achieved. Fig- reaction between Zn vapor and the active

WELDING JOURNAL 13-s Yang Supplement December 2010:Layout 1 12/13/10 4:35 PM Page 14

gases O2 and CO2 will form ZnO according ¡ ¡ ¡ to 2Zn + O2 ZnO and CO2 C + O2 Zn+ O ¡ ZnO. This chemical reaction could help stabilize the welding process. 2. A thin film of metal oxides (mainly iron oxides) could form in the front of the molten pool. The formation of metal oxides as well as heated surface material increased the AB C coupling efficiency of laser beam energy into the welded material (Refs. 40, 41). Under this condition, the keyhole was readily formed, which provided the channel for the highly pressurized zinc vapor to be vented out. When the surface tension had an outward pattern, weld penetration was shallow (Ref. 42). Inversely, deep weld penetration E F D can be obtained when the surface tension has an inward pattern. By introduction of active shielding gases such as O2 and CO2, the surface tension changed from an outward pattern to an inward pattern (Ref. 43). Con- sequently, the keyhole was enlarged and deepened (Ref. 43), as shown in Fig. 9. 3. The viscosity of molten metal was de- G H creased through introduction of active

WELDING RESEARCH shielding gases such as O and CO in com- Fig. 14 — The images of the molten pool successively obtained in the stable laser welding process (laser 2 2 3 parison with the laser welding experiments power: 3600 W; welding speed: 40 mm/s; the coaxial 90% Ar +10% CO2 shielding gas: 30 ft /h; the back and side pure argon shielding gas: 30 ft3/h). where pure argon was used as the shielding gas (Ref. 44). Therefore, the shear force ex- erted on the keyhole wall by the highly pressurized zinc vapor could be reduced. Thus, A B the potential for formation of spatter and porosity was subsequently reduced. During the laser welding process, carbon dioxide can be dissociated into CO+ 1 ⁄2O2 or C+O2, depending upon the dissociation potential. If the dissociation potential is high enough, the CO is further trans- 1 formed into C + ⁄2O2. The large amount of carbon dissolving into the weld metal re- duces the corrosion resistance of the low- carbon grades of steels (Ref. 45). Further- more, the strength of the welds may be decreased by reducing the amount of deox- idating alloying elements such as manganese and silicon, which react with oxygen Fig. 15 — EDS analysis of the top surface of weld; A — EDS result in the measured spot; B — SEM image in the shielding gas (Ref. 45). In like man- of the top surface of weld. Welds obtained by the following welding parameters: Laser power, 3600 W; weld- 3 ner, the addition of oxygen into the shielding speed, 40 mm/s; the coaxial shielding gas, 30 ft /h of the mixture of 75% Ar + 25% CO2; back and side shielding gas, 30 ft3/h of pure argon. ing gas may pose a risk for the reduction of the weld strength. Therefore, the issue on control of the percentage of CO2 or O2 in the shielding gas may be of concern when A B using the mixtures of Ar + CO2 or Ar + O2 in the shielding gas for laser welding of galvanized steels. On the other hand, since the laser welding process was carried out at the high welding speed, the cooling rate was fast. As a result, the interaction time between the carbon or oxygen and the deoxi- dizable alloys was extremely short such that the issues mentioned previously may not be of concern. Further studies are planned to explore these phenomena.

Influence of the Side Shielding Gas Flow Rate on Weld Quality Fig. 16 —EDS analysis of the cross section of the weld: A — EDS result in the measured spot; B — SEM image of the cross section of the weld. In order to study the influence of the

14-s JANUARY 2011, VOL. 90 Yang Supplement December 2010:Layout 1 12/13/10 4:36 PM Page 15

side shielding gas flow rate on the weld A B quality, laser welding experiments were carried out with shielding conditions as follows: no coaxial shielding gas, a constant 30 ft3/h pure argon back shielding gas flow rate, pure argon side shielding gas flow rate varying from 20 to 40 ft3/h. Figure 8 shows the experimental results on the weld quality. As shown in Fig. 8, no spatters or porosity were generated in the welds and complete penetration was achieved in the welds when the flow rate of the side shielding gas was 20 and 30 ft3/h. When the flow rate of the side shielding gas was increased to 40 ft3/h, the welding process became dramatically unstable and some spatter and poros- Fig. 17 — EDS analysis of the cross section of base material: A — EDS result in the measured spot; B — ity were produced in the welds, as shown in SEM image of the cross section of the base material. Fig. 10E. In addition, only partial penetration was achieved in the welds. This phe- nomenon can be explained by the fact that when the flow rate of the side shielding gas was increased to a sufficiently high level, the argon gas enabled the formation of more plasma when the laser beam penetrated through the argon gas than what laser-induced plasma was blown off for the given laser power and welding speed. It is worth mentioning that the argon side shielding gas can be replaced by another shielding gas such as helium to suppress the laser-induced plasma with the optimal flow rate to achieve a sound lap joint.

Influence of Welding Speed on Weld Quality

In order to study the influence of the welding speed on weld quality, three trial Fig. 18 — Microhardness profile of a weld. (Welds obtained in Experiment 6.) tests were conducted where the welding speed was varied from 30 to 60 mm/s in in- crements of 10 mm/s and all other welding parameters were identical to those in Ex- periment 4. Experimental results demonstrated that for a given laser power of 3600 W, a sound weld could be achieved at welding speeds of 30, 40, and 50 mm/s. However, WELDING RESEARCH the laser welding process tended to become unstable, producing spatter and porosity in the welds at the welding speed of 60 mm/s, as shown in Fig. 11. The instability of the laser welding process at 60 mm/s welding Fig. 19 — Schematic diagram of the geometry of the tensile shear test sample. speed can be explained by the fact that when the welding speed was increased, the keyhole became unstable and tended to collapse (Ref. 46). Keyhole formation re- AB quired that the laser beam energy density was beyond some threshold value (Ref. 47), as shown in Fig. 12. The energy density can be described by the following equation:

Ed = d * (Pd/V) (1)

where d is the focused spot size, Pd is the laser power at the focus point, and V is the welding speed. When the speed was increased, the energy density was decreased and the keyhole tended to collapse. The collapsed keyhole failed to vent out the Fig. 20 — Failure location of the tensile shear test samples with different welding conditions: A — Ten- highly pressurized zinc vapor, which led to sile shear test fracture location of lap joints obtained in Experiment 5; B — tensile shear test fracture location of lap joints obtained in Experiment 7.

WELDING JOURNAL 15-s Yang Supplement December 2010:Layout 1 12/13/10 4:36 PM Page 16

Furthermore, for all of the measured points in the cross section of the weld, the presence of Fe and Mn elements without the O element was identified from the peaks present in the EDS analysis results. Based on the absence of the oxygen element in the weld, the conclusion may be made that no oxygen contamination has occurred in the weld during the laser welding process with the introduction of oxygen and carbon dioxide. Based on the weight percentages of various elements in Figs. 16 and 17, the amount of the essential alloying element of Mn in the weld is not de- graded during the laser welding process even if a high percentage of CO2 is used in the coaxial shielding gas. This fact indi- Fig. 21 — Tensile test results. cates that the issue of the loss of oxidizing alloy elements such as manganese may not a large amount of spatter and porosity in spatter and porosity in the welds dramati- be of concern for the high-speed laser the welds. The spatter flying in the laser cally reduced the strength of the lap joints. welding process with the addition of car- beam path absorbed and reflected the In contrast, a stable keyhole was consis- bon dioxide or oxygen into the coaxial laser beam energy. In addition, the multi- tently formed and kept open during the shielding gas. However, further study is reflection will not occur in the collapsed entire laser welding process, which miti- still required for the resistance test. Since keyhole. So, when the keyhole is col- gates the highly pressurized zinc vapor, the iron oxides can influence the metallur-

WELDING RESEARCH lapsed, a relatively lower level of laser thus producing a stable laser welding gical properties of welds (Ref. 39), the per- beam energy is absorbed by the welded process. As shown in Fig. 14, the keyhole centage of oxygen or carbon dioxide in the material, resulting in partial penetration. appears as a black spot located in front of shielding gas mixture should be controlled Therefore, it is important to control weld- the molten pool. Due to the mitigation of in order to maintain the same chemical ing speed for obtaining a stable keyhole. zinc vapor through the stable open key- composition and mechanical properties of In addition, when welding speed was rela- hole, the molten pool was stable and no the welds as the base material. tively lower, for the given laser power and liquid metal was rejected from it. Conse- shielding gas flow rate, it was more possi- quently, sound lap joints were achieved in Mechanical Tests ble to achieve complete penetration. the stable laser welding process. Further- When complete penetration was more, the coupling efficiency of laser Vickers hardness measurements were achieved, a portion of the highly pressur- beam energy into the welded material was conducted on a cross section of the weld ized zinc vapor can escape from the bot- enhanced by multireflection into the key- along a line 0.25 mm under the top surface tom of specimens producing a more stable hole. Therefore, deep weld penetration is using a load of 200 g and a 10-s dwell time. laser welding process. attained. Figure 18 shows the measured line and the hardness profiles across the weld zone and Direct Observation of Dynamic Behaviors Energy-Dispersive X-Ray Spectroscopy the heat-affected zone (HAZ) as well as of the Molten Pool with the Machine (EDS) Analysis the base material. The microhardness val- Vision System ues are a function of the distance from the In order to analyze the chemical ele- weld center. In the fusion zone, the weld To study the behaviors of the molten ments of the weld and ensure that the use has a higher hardness value than that of the base material due to the fast cooling. pool and the keyhole dynamic, a high- of high levels of CO2 would not cause the speed (4000 f/s) CCD camera was applied loss of the alloying elements in the weld, The hardness then decreases toward the to monitor the laser welding process in energy-dispersive X-ray spectroscopy fusion boundary. Because of the effect of real time. Figures 13 and 14 show the (EDS) experiments were carried out at dif- softening, the HAZ is characterized by the molten pool and the keyhole obtained ferent locations on the top surface and lowest hardness value, lower than that in during the unstable and stable laser weld- along the cross-sectional zone of the welds. both the weld zone and the base material. ing processes. During the unstable laser Figures 15 and 16 present typical EDS Tensile shear tests were performed to welding process, the keyhole has difficulty analysis results at the measured points on determine the strength of the welds. The forming, as shown in Fig. 13. In addition, the top surface and along the cross section sample geometry is shown in Fig. 19. The the molten metal flow in the rear part of of the weld. Figure 17 shows the EDS tensile shear strength is determined by the the molten pool dramatically fluctuates analysis results at the cross section of the average value of two series of values taken over time. When the welding process was base material. As shown in Fig. 15, the O, on the same specimen welded under a spe- unstable, the highly pressurized zinc vapor Fe, Mn, and Zn elements are presented in cific set of welding parameters. The weld developed at the interface of two metal the EDS analysis results detected on the width at the interface measured from the sheets ejects the liquid metal out of the top surface of the weld. It is conjectured weld cross section, which was changed by rear portion of the molten pool. This ex- that the zinc oxide, iron oxides, and man- different welding conditions, was used to pulsive liquid metal condensed in the air ganese oxide are probably produced on the calculate the tensile strength. Figure 20 and became spatter that was deposited on top surface of the welds due to the dissoci- shows the weld fracture location after the tensile shear test. As shown in Fig. 20, both the weld surface or weld zone in random ation of CO2 into the formations of CO + of the test specimens are broken under the directions. A large amount of spatter not O2 or C + O2. The oxides of ZnO, MnO, only damaged the optical lens but also and FeO are produced according to the shear force at the fusion zone, for a coaxial shielding gas mixture of 90% Ar + 10% produced poor quality lap joints and poor following reactions: Zn+ ½O2 ZnO; CO . However, when the coaxial shielding surface appearance. The presence of the Fe+ ½O2 FeO; and Mn+ ½O2 MnO. 2

16-s JANUARY 2011, VOL. 90 Yang Supplement December 2010:Layout 1 12/13/10 4:37 PM Page 17

gas is switched to the mixture of 75% Ar A + 25% CO2 and the other welding parameters are kept the same as in the case of using the 90% Ar + 10% CO2 , the fracture of test samples happens in the HAZ, as illustrated in Fig. 20B. Figure 21 sum- marizes the tensile test results. The average shear strength for the welds obtained with the welding parameters in Fig. 20A is 860 MPa and the average tensile strength B for the welds obtained with the welding parameters in Fig. 20B is 800 MPa. The strength of the base material is 980 MPa. Comparison of the strength of the welds with that of the base materials reveals that the strength is slightly reduced when using Fig. 22 — Completely defect-free laser-welded lap joint with a side shielding gas only: A — Top view; B the mixtures of 75% Ar + 25% CO2 and — bottom view (laser power: 3600 W; welding speed: 30 mm/s; side shielding gas: pure argon; side shield- 90% Ar + 10% CO2 as the shielding gases. ing gas flow rate: 30 ft3/h).

Laser Welding of Galvanized Steels in a Gap-Free Lap Joint Configuration with a sorption efficiency of laser beam is in- Coated Steels. Miami, Fla.: American Welding Single Side Shielding Gas creased, and it is better coupled to the ma- Society. terials to be welded. It is not necessary to 3. Graham, M. P., Hirak, D. M., Kerr, H. W., In order to study the influence of a sin- use the bottom shielding gas during the and Weckman, D. C. 1994. Nd:YAG laser laser welding process. welding of coated sheet steel. Journal of Laser gle side shielding gas on the weld quality, a Applications 3. The keyhole could be stabilized, en- 6(4): 212–222. further trial test was conducted where the 4. Mazumder, J., Dasgupta, A., and Bem- welding speed was 30 mm/s, laser power larged, and deepened by the addition of benek, M. 2002. Alloy based laser welding. U.S. was 3600 W, the side shielding gas flow rate active gases (O2 and CO2) during the laser Patent No. 6,479,168. was 30 ft3/h, and no back and coaxial welding process. Compared with the case 5. Dasgupta, A., Mazumder, J., and Bem- shielding gases were used. As shown in Fig. of using pure argon, deeper penetration benek, M. 2000. Alloying based laser welding of 22, a sound lap joint with complete pene- could be achieved when adding the active galvanized steel. Proceedings of International Conference on Applications of Lasers and Elec- tration was also achieved. No spatter and gas (O2 and CO2) into the shielding gas. porosity were produced on the welds. 4. Lower travel speeds that results in tro Optics, Laser Institute of America, Dear- born, Mich. Therefore, it is also possible to just use a complete penetration produced sound welds whereas the partial penetration 6. Dasgupta, A., and Mazumder, J. 2006. A single side shielding gas for laser welding novel method for lap welding of automotive of galvanized steels in a gap-free configu- weld made at the fastest travel speed was sheet steel using high power CW CO2 laser, ration to obtain sound lap joints. Further defective. Proceedings of the 4th International Congress on studies on the influences of side shielding 5. Alloying elements such as Mn are Laser Advanced Materials Processing. gas composition, size/shape of the side not reduced by the introduction of the ac- 7. Pieters, R. R. G. M., Bakels, J. G., Her- shielding gas nozzle, distance of the side tive gas (O2 and CO2) into the coaxial mans, M. J. M., and den Ouden, G. 2006. Laser shielding gas nozzle from the keyhole, and shielding gas. welding of zinc coated steels in an edge lap con- the angle of the nozzle with respect to the 6. The fracture location of the welds figuration. Journal of Laser Applications 18(3): plane of the top sheet on the weld quality varies from the fusion zone to the heat- 199–204. affected zone (HAZ) during the tensile 8. Graham, M. P., Hirak, D. M., Kerr, H. W., are needed to be completed for a given and Weckman, D. C. 1994. Nd:YAG laser weld- tests with an increase in the percentage of WELDING RESEARCH welding speed and laser power. In this ing of coated sheet steel. Journal of Laser Ap- CO in the mixture of argon + CO . In ad- paper, we will not discuss these issues. 2 2 plications 6(4): 212–222. dition, the HAZ is softened during the 9. Sacks, J. L., Davis, H., and Spilchen, W. Conclusions laser welding process. Furthermore, EDS 2007. Corrosion resistant wire products and analysis results do not detect a loss of Mn method of making same. U.S. Patent No. in the weld during the laser welding 20070119715. The following conclusions can be drawn process. 10. Graham, M. P., Hirak, D. M., Kerr, H. from this work: W., and Weckman, D. C. 1996. Nd:YAG laser 1. With the specific shielding condi- Acknowledgments beam welding of coated steels using a modified tions, a completely defect-free lap joint of lap joint geometry. Welding Journal 75(5): 162- galvanized steels in a gap-free configura- This work was funded by General Mo- s- to 170-s. tion can be achieved by a single laser beam tors and NSF grant No. EEC-0541952. 11. Graham, M. P., Hirak, D. M., Kerr, H. W., without pre- and postweld processing. The The authors acknowledge Andrew Socha, and Weckman, D. C. 1996. Laser welding of Zn- proposed laser welding procedure, which research engineer; Junjie Ma, PhD stu- coated sheet steels. Proceeding of SPIE: Interna- is robust, can be conducted at high speed dent; and Dr. Xinfeng Li, visiting scholar tional Society for Optical Engineering, V. 2703. 12. Pennington, E. J. 1987. Laser welding of to obtain completely penetrated, high- at Research Center for Advanced Manu- galvanized steel, U.S. Patent No. 4642446. strength lap joints. facturing at Southern Methodist Univer- 13. Williams, S. W., Salter, P. L., Scott, G., 2. The introduction of the side shield- sity for their assistance in performing the and Harris, S. J. 1993. New welding process for ing gas can not only suppress the laser- experiments. galvanized steel. Proc. 26th Int. Symp. Automo- induced plasma and plume but also stabi- References tive Technology and Automation, 49–56, Aachen, lize the turbulent molten pool caused by Germany. the highly pressurized zinc vapor resulting 1. Akhter, R., Steen, W. M., and Watkins, K. 14. Teng, Y. F. 1999. Pulsed Nd:YAG laser in a stable keyhole, which allows the highly G. 1991. Welding zinc-coated steel with a laser seam welding of zinc-coated steel. Welding Jour- pressurized zinc vapor to be vented out. and the properties of the weldment. Journal of nal 78(7): 238-s to 244-s. With the formation of the keyhole, the ab- Laser Applications 3(2): 9–20. 15. Forrest, M. G., and Lu, F. 2005. Devel- 2. AWS WZC/D19.0-72, Welding Zinc- opment of an advanced dual beam head for

WELDING JOURNAL 17-s Yang Supplement December 2010:Layout 1 12/13/10 4:38 PM Page 18

laser lap joining of zinc coated steel sheet with- welding of galvanized steel sheet. 20th Interna- ern Technology (1st edition). Springer Press. out gap at the interface. 24th International Con- tional Congress on Applications of lasers & Elec- 36. Lu, S. P., Fuji, H., Nogi, K., and Sato, T. gress on Applications of Lasers and Electro-Op- tro-Optics, ICALEO: 13–19. 2004. Marangoni convection and weld shape tics, ICALEO: 1069–1074. 26. Li, X. G., Lawson, W. H. S., and Zhou, variations in Ar-O2 and Ar-CO2 shielded GTA 16. Chen, W., Ackerson, P., and Molian, P. Y. N. 2008. Lap welding of steel articles having welding. Material Science Engineering A, Vol. 2009. CO2 laser welding of galvanized steel a corrosion resisting metallic coating. U.S. 380, pp. 290–297. sheets using vent holes. Material & Design, Vol. Patent No. 2008/0035615 A1. 37. Dong, W. C., Liu, S. P., Li, D. Z., and Li, 30, pp. 245–251. 27. Li, X., Lawson, S., and Zhou, Y. 2007. Y. Y. 2009. Modeling of the weld shape devel- 17. Xie, J. 2002. Dual beam laser welding. Novel technique for laser lap welding of zinc opment during the autogenous welding process Welding Journal 81(10): 223-s to 230-s. coated sheet steels. Journal of Laser Applications by coupling welding arc with weld pool. Journal 18. Gualini, M. M. S., Iqbal, S., and Grassi, 19(4): 259–264. of Material Engineering and Performance, pp. F. 2006. Modified dual-beam method for weld- 28. Yang, S. L., and Kovacevic, R. 2009. Laser 1–9. ing galvanized steel sheets in lap configuration. welding of galvanized DP 980 steel assisted by the 38. Smith, A. A. 1970. CO2 Welding of Steel Journal of Laser Applications 18(3): 185–191. GTAW preheating in a gap-free lap joint config- (3rd Edition). The Welding Institute. 19. Forrest, M. G., and Lu, F. 2005. Devel- uration. Journal of Laser Application, Vol. 21, No. 39. Glowacki, M. H. 1995. The effects of the opment of an advanced dual beam head for 3, pp. 139–148. use of different shielding gas mixtures in laser laser lap joining of zinc coated steel sheet with- 29. Natio, Y., Mizutani, M., and Katayama, welding of metals. Journal of Physics D: Applied out gap at the interface. 24th International Con- S. 2006. Effect of oxygen in ambient atmos- Physics, Vol. 28, pp. 2051–2059. gress on Applications of Lasers and Electro-Op- phere on penetration characteristics in single 40. Grigoruants, A. G. 1997. Basics of Laser tics, pp. 1069–1074. yttrium-aluminum-garnet laser and hybrid Material Processing. CRC Press. 20. Milberg, J., and Trautmann, A. 2009. welding. Journal of Laser Applications. Vol. 18, 41. Xie, J., and Kar, A. 1999. Laser welding Defect-free joining of zinc-coated steels by bi- pp. 21–27. of thin sheet steel with surface oxidation. Weld- focal hybrid laser welding. Production Engineer- 30. Olsen, F. O. 2009. Hybrid Laser-Arc ing Journal 78(10): 343-s to 348-s. ing Research Development, Vol. 3, pp. 9–15. Welding. CRC Press. 42. Fukuyama, H., and Waseda, Y. 2009. 21. Choi, H. W., Farson, D. F., and Cho, M. 31. Zhang, L. J., Zhang, J. X., Wang, R., High-Temperature Measurements of Materials H. 2006. Using a hybrid laser plus GMAW Yao, W., and Gong, S. L. 2005. Effects of side (Advances in Materials Research), Springer process for controlling the bead humping de- assist gas on the CO2 laser welding process of Press. fect. Welding Journal 85(8): 174-s to 179-s. thin stainless plate. Applied Laser, Vol. 25, pp. 43. Zhao, L., Tsukamoto, S., Arakane, G., WELDING RESEARCH 22. Mueller, G. H. 2004. Hybrid welding of 217–221. and Sugino, T. 2009. Influence of oxygen on galvanized steel sheet. European Patent EP 32. Ready, J. F., and Farson, D. F. 2001. LIA weld geometry in fibre laser welding. ICALEO 1454701. Handbook of Laser Materials Processing, Laser 2009, pp. 759–765. 23. Kusch, M., and Thurner, S. 2008. Appli- Institute of America. 44. Welding Handbook, Vol. 2. Miami, Fla.: cation of the plasma-MIG technology for the 33. Shackelford, J. F., and Alexander, W. American Welding Society. joining of galvanized steel materials. Welding 2000. Materials Science and Engineering Hand- 45. Jeffus, L. 2002. Welding Principles and and Cutting 7(1): 54–59. book (third edition). CRC Press. Applications, fifth edition, Thomson Delmar 24. Kim, C., Choi, W., Kim, J., and Rhee, S. 34. Yang, S. L. 2009. Hybrid laser-arc weld- Learning Press. 2008. Relationship between the weldability and ing of galvanized steels in a gap-free lap joint 46. Zhang, Y. M., and Zhang, S. B. 1999. the process parameters for laser-TIG hybrid configuration. Dissertation. Southern Methodist Observation of the keyhole during plasma arc welding of galvanized steel sheets. Materials University. welding. Welding Journal 78(2): 53-s to 58-s. Transactions 49(1): 179–186. 35. Dowden, J. 2009. The Theory of Laser Ma- 47. Olsen, F. O. 2009. Hybrid Laser-Arc 25. Gu, H., and Mueller, R. 2001. Hybrid terials Processing: Heat and Mass Transfer in Mod- Welding. CRC Press.

WELDING JOURNAL Instructions and Suggestions for Preparation of Feature Articles Text labeled with captions • approximately 1500–3500 words in length • references/bibliography should be included at the end of the article • submit hard copy Editorial Deadline • submissions via disk or electronic transmission — preferred format is • January issue deadline is November 21 Mac but common PC files are also acceptable • February issue deadline is December 19 • acceptable disks include zip, CD, and DVD. • March issue deadline is January 23 Format • April issue deadline is February 20 • include a title • May issue deadline is March 23 • include a subtitle or “blurb” highlighting major point or idea • June issue deadline is April 20 • include all author names, titles, affiliations, geographic locations • July issue deadline is May 22 • separate paper into sections with headings • August issue deadline is June 22 • September issue deadline is July 24 Photos/Illustrations/Figures • October issue deadline is August 21 • glossy prints, slides, or transparencies are acceptable • November issue deadline is September 21 • black and white and color photos must be scanned at a minimum of • December issue deadline is October 22 300 dpi • line art should be scanned at 1000 dpi Suggested topics for articles • photos must include a description of action/object/person and • case studies, specific projects relevance for use as a caption • new procedures, “how to” • prints must be a minimum size of 4 in. x 6 in., making certain the photo • applied technology is sharp • do not embed the figures or photos in the text Mail to: • acceptable electronic format for photos and figures are EPS, JPEG, Mary Ruth Johnsen and TIFF. TIFF format is preferred. Editor, Welding Journal 550 NW LeJeune Road Other Miami, FL 33126 • illustrations should accompany article (305) 443-9353, x 238; FAX (305) 443-7404 • drawings, tables, and graphs should be legible for reproduction and [email protected]

18-s JANUARY 2011, VOL. 90