A Comparison Of Gas Metal Arc Welding With Flux-cored Wires And Solid Wires Using Shielding Gas

Int J Adv Manuf Technol (1999) 15:49–53  1999 Springer-Verlag London Limited

A Comparison of Gas Metal Arc Welding with Flux-Cored Wires and Solid Wires Using Shielding Gas M. T. Liao1 and W. J. Chen2 1

Department of Mechanical Manufacturing Engineering, and 2Department of Mechanical Materials Engineering, National Huwei Institute of Technology, Yunlin, Taiwan

In the present work, gas metal arc welding (GMAW) with fluxcored wires and solid wires using shielding gas has been adopted for welding stainless steel. Five different compositions of shielding gas are used with flux-cored wire and three with solid wire. Spatter rates, chemical compositions, tensile strength and elongation tests have been performed and are reported. The spatter rates of the sample made using fluxcored wires are less than that for the sample made using solid wire. The ultimate tensile strength and elongation are not influenced by the composition of the shielding gas. Keywords: Austenitic stainless steels; Flux-cored wires; Gas metal arc welding (GMAW); Property; Shielding gas; Solid wires

1.

Because of the wide applications of the austenitic stainless steels under both mild and severe corrosive conditions at cryogenic or elevated temperatures, the properties and microstructures of welded joints in these steels have received attention from researchers [3–5]. Nowadays, rationalisation of the welding processes demands more and more gas metal arc welding of stainless steel. Ar-He-CO2, Ar-He, Ar-N2, Ar-CO2 and Ar-CO2-N2 mixtures are mainly used as the shielding gas at present [6–8]. The selection of the shielding gas for stainless steel welds has been reported [2,6,7]. Although the use of solid and flux-cored welding wires is common, there seem to be only a few published investigations of the influence of shielding gas on the metal properties in stainless steel welding [8]. Hence, this study aims to compare the use of solid wire (AWS ER308L) and flux-cored wire (AWS E308LT-1) when using different shielding gases.

Introduction

Recently, the gas metal arc welding (GMAW) process, with either solid or flux-cored welding wires, has become popular, because high-quality and economical welds can be obtained [1]. In the processes, a shielding gas must be used. The primary function of the shielding gas is to protect the molten metal from atmospheric nitrogen and oxygen as the weld pool is being formed. The gas also promotes a stable arc and uniform metal transfer. The quality, the efficiency, and the overall operating acceptance of the welding operation are strongly dependent on the shielding gas, since it dominates the mode of the metal transfer. The shielding gas not only affects the properties of the weld but also determines the shape and penetration pattern. During welding, the shielding gas interacts with the welding wire to form a strong and tough corrosion-resistant weld. The shielding gas also affects the residual content of hydrogen, nitrogen and oxygen dissolved in the weld metal [2]. Correspondence and offprint requests to: Dr W.-J. Chen, National Huwei Institute of Technology, Department of Mechanical Materials Engineering, Huwei, Yunlin, 632 Taiwan. E-mail: wjchen얀sparc.nhit.edu.tw

2. Experimental Procedure Premixed gases are used as a shielding gas in this study. For solid wire, the composition of the shielding gas is 98%Ar+2%CO2(M1), 90%Ar+10%CO2 (M2), and 80%Ar+20%CO2 (M3) (These group samples are designed as MAG). For flux-cored wire, the composition of the shielding 60%Ar+40%CO2 (F2), gas is 80%Ar+20%CO2(F1), 40%Ar+60%CO2 (F3), 20%Ar+80%CO2 (F4), and 100%CO2 (F5) (These group samples are designed as FCAW). Wires used in the study are of 1.2 mm diameter and conform to the AWS specifications. A 12 mm(t) × 50 mm(w) × 200 mm(l) base metal plate of AISI type 304 stainless steel was used for the down welding. In order to evaluate the properties of the deposited metal, mild steel plates (19 mm(t) × 125 mm(w) × 300 mm(l)) with a V-shaped groove were used and a multipass procedure was carried out. In accordance with the AWS A5.22–80 specification, the V-groove surface was coated with a JIS 304L stainless steel before the multipass procedure was performed, as shown in Fig. 1(a). A Panasonic YD-356KRI welding machine and a cantilever beam supporting frame comprise the automatic welding system. A constant-potential power supply was used and the welding

50

M. T. Liao and W. J. Chen

Fig. 1. Details of test assembly for all–weld-metal tension test.

Fig. 3. Oxygen and nitrogen content in the weld with the use of solid wires and flux-cored wires.

voltage was held at 30 V during welding. A constant wirefeeding rate was maintained for all the welds. However, a slight variation in current was observed owing to a small change in the cover gas supply. Consequently, for the solid wires, the welding current was in the range of 200 ⫾ 10 A. For the flux-cored wires, the welding current was in the range of 180 ⫾ 10 A. Cylindrical tensile specimens were machined from the deposited metal. These specimens and were taken parallel to the weld direction at a fixed distance from the weld centre, as shown in Fig. 1(b). Tensile tests were performed at room temperature. The fracture surface was investigated by scanning electron microscopy (SEM) to observe the fracture mode. Chemical microanalysis was carried out using an SEM equipped with an EDAX detector. Magne-gauge readings were taken to determine the ferrite content of the deposited metal. The spatter rates for each mixture, were measured by collecting spatter within a fixed time period.

3. Results and Discussion

Fig. 2. (a) The spatter rates as a function of the amounts of CO2 in the Ar + CO2 mixtures. (b) Spatter of the F1 sample. (c) Spatter of the M3 sample.

Spatter, with solid and flux-cored welding wires, was generated in the down welding position when using all the mixtures. The spatter rate depends on the amounts of CO2 in the Ar + CO2 mixture and is shown in Fig. 2(a). The spatter rates for the sample made using flux-cored wire are less than those for the sample made using the solid wires. The photographs in Fig. 2(b) and 2(c) are from the flux-cored wire samples (F1) and the solid wire samples (M3), respectively. With solid wire, the spatter creates larger particles when the CO2 content is higher. With flux-cored wires, the spatter is small for all mixtures. The difference of the spatter and spatter rates between these two group samples is due to the difference in the welding wires used in this study. The reason is that the FCAW tests have flux, whereas the MAG do not have flux. In the FCAW tests, the flux alters the metal transfer mode, and make the droplets become smaller, and causes the spatter to become extremely low [9]. Thus, the spatter rates and spatter are only influenced by the flux. They are not affected by the compositions of the shielding gas.

Gas Metal Arc Welding with Flux-Cored and Solid Wires

51

Table 1. Chemical composition and ferrite content of all weld metal. MAG Chemical composition of MAG wires Chemical composition C (%) Si (%) Mn (%) Ni (%) Cr (%) Mo (%) P (%) S (%) N (p.p.m.) O (p.p.m.) Nickel equivalent (NiEq)* Chromium equivalent (CrEq)† Ferrite content (FN)

0.045 0.45 1.92 9.15 20.37 0.17 0.029 0.009

M1

FCAW M2

M3

F1

F2

F3

F4

F5

98%Ar+ 2%CO2

90%Ar+ 10%CO2

80%Ar+ 20%CO 2

80%Ar+ 20%CO2

60%Ar+ 40%CO2

40%Ar+ 60%CO2

20%Ar+ 80%CO2

100%CO2

0.040 0.44 1.73 9.4 19.3 0.19 0.022 0.007 412 332

0.050 0.38 1.61 9.4 19.3 0.18 0.022 0.007 378 251

0.070 0.39 1.60 9.4 19.0 0.18 0.022 0.007 361 245

0.050 0.70 1.60 10.4 19.9 0.05 0.018 0.004 123 466

0.040 0.70 1.56 10.4 19.8 0.06 0.019 0.006 145 513

0.040 0.66 1.51 10.4 19.6 0.06 0.018 0.004 148 521

0.040 0.64 1.47 10.3 19.5 0.05 0.018 0.004 144 516

0.040 0.61 1.42 10.2 19.5 0.05 0.018 0.003 171 494

12.67

12.91

13.5

13.07

12.82

12.80

12.67

12.61

20.15

20.05

19.77

19.16

19.97

20.65

20.51

20.47

8.5

7.35

5.75

10.5

10.4

10.0

9.1

8.8

*Nickel equivalent (NiEq) = %Ni+30 × %C+30 × %N+0.5 × %Mn †Chromium equivalent (CrEq) = %Cr+%Mo+1.5 × %Si+0.5 × %Nb

Fig. 4. The ferrite contents of all the deposited metal as a function of the amounts of CO2 in the Ar + CO2 mixtures.

Figure 3 illustrates a comparison of oxygen and nitrogen content in the weld when using solid wires and flux-cored wire. The oxygen content of the weld when using flux-cored wire is higher than that when using solid wires. This is due to the action of acidic substances such as rutile and silica sand in the flux. The nitrogen content of the weld when using fluxcored wire is lower than that when using the solid wire. This may be due to the fact that the flux and shielding gas are more effective in protecting the molten metal from atmospheric nitrogen as the weld pool is being formed. Therefore, the nitrogen content of weld is low in the FCAW samples. The chemical compositions of all the deposited metals and their ferrite contents, using either solid or flux-cored welding wires, and measured by the Magne gauge for each shielding

gas are shown in Table 1. The ferrite contents of all the deposited metal as a function of the amounts of CO2 in the Ar + CO2 mixtures are shown in Fig. 4. For the solid wire samples, the ferrite content is decreased by increasing the amount of CO2 in the Ar + CO2 mixtures for CO2 from 2% to 20%. The increase of the percentage of CO2 in the mixtures will raise the carbon content in the deposited metal, therefore the ferrite content will be decreased and the Ni equivalent will become larger. Also, the increase of the percentage of CO2 will increase the consumption of Cr and Si owing to oxidation, and make the Cr equivalent smaller [8]. For the flux-cored wire samples, the ferrite content is also decreased by increasing the amount of CO2 in the Ar+CO2 mixtures for CO2 from 20% to 100%. The reason is not known. The ultimate tensile strength (UTS) and elongation of the deposited metal for each of the samples (MAG and FCAW) is given in Table 2. The comparison of the ultimate tensile strength between MAG and FCAW shows that MAG samples are of higher strength. This may be due to the fact that the inclusion content of MAG is less than that of FCAW. (This can be seen from the fracture surface.) The fracture surface of the F5 and M3 tensile specimens are shown in the SEM images in Figs 5(a) and 5(b), respectively. Similar fracture surface morphologies were also observed for FCWA from F1 to F4 and for MAG from M1 to M2. In Figs. 5(a) and 5(b), it is clear that the fracture surface morphology is dimple rupture. The dimples are associated with many impurity particles (inclusions) which are generally round and of various sizes. With flux-cored wire, the results of the EDX analyses indicated that the inclusions contained manganese, iron, and

52

M. T. Liao and W. J. Chen

Table 2. Tensile strength and elongation data for FCAW and MAG. MAG

FCAW

M1

M2

M3

F1

F2

F3

F4

F5

98%Ar+ 2%CO2

90%Ar+ 10%CO2

80%Ar+ 20%CO2

80%Ar+ 20%CO2

60%Ar+ 40%CO2

40%Ar+ 60%CO2

20%Ar+ 80%CO2

100%CO2

Tensile strength (N/mm2)

613

602

610

573

569

567

571

567

Elongation (%)

38.6

38.2

37.6

36.6

38.6

37.2

36.4

38.0

Fig. 5. Fracture morphology (SEM image) of tensile specimen fractured at room temperature, (a) F5 sample, (b) M3 sample. The inclusion is indicated by the arrow. The EDAX spectra of the inclusions is shown in (c) and (d). (c) Typical EDAX spectra of the inclusions in (a). (d) Typical EDAX spectra of the inclusions in (b).

chromium, as shown in Fig. 5(c). With solid wires, the results of the EDX analyses indicated that inclusions contained silicon, manganese, iron, and chromium, as shown in Fig. 5(d). This indicates that inclusions are silicon oxides and manganese oxides, etc. The difference in composition of the inclusions is due to the different sources of the inclusions. In MAG, the impurities mainly arose from the shielding gas. In FCAW, the impurities mainly arose from the flux and shielding gas.

4. Conclusions The compositions of the shielding gases have significant effects on the properties of the stainless steel weld metal using solid welding wire. The composition of the shielding gas has only a slight effect on the properties of the stainless steel weld metal using flux-cored wire. In FCAW samples, the spatter rates and spatter are influenced only by the flux. They are not

Gas Metal Arc Welding with Flux-Cored and Solid Wires

affected by the compositions of the shielding gas. The ferrite content is decreased by increasing the amount of CO2 in the Ar + CO2 mixtures for both group samples. Acknowledgements

The authors wish to thank Goodweld corporation for providing welding wire and Air Liquide Far Eastern corporation for providing premixed gas. References 1. N. Stenbacka and K. A. Persson, “Shielding gases for gas metal arc welding”, Welding Journal, 68, pp. 41–47, November 1989. 2. K. A. Lyttle and W. F. G. Stapon, “Select the best shielding gas blend for the application”, Welding Journal, 69, pp. 21–27, November 1990.

53

3. E. Folkhard, in Welding Metallurgy of Stainless Steel, SpringerVerlag, 1988. 4. J. A. Brooks and A. W. Thompson, “Microstructural development and solidification cracking susceptibility of austenitic stainless steel welds”, International Materials Reviews, 36 (1) pp. 16–43, 1991. 5. P. Bilme, A. Gonzalez and C. L. Lorente et al., “Effect of ferrite solidification morphology of austenitic stainless steel weld metal on properties of welded joints”, Welding International, 10 (10), pp. 797–808, October 1996. 6. R. Petersens, I. Ballingal and O. Runnerstam, “Selecting shielding gases for welding of stainless steels”, Welding Review International, pp. 152–158, August 1993. 7. K. A. Lyttle, “Shielding gases”, in Material Handbook of ASM, vol. 6, pp. 64–69, 1993. 8. T. Kobayashi and T. Sugiyama, “The effect of shielding gas composition on the characteristics of stainless steel weld metal”, IIW Doc.XIZ-E-33–82.XII-B-25–82, 1982. 9. M. Sato, K. Suda and H. Nagasaki, “How to weld using flux-cored wires”, Welding International, 11(4), pp. 264–272, April 1997.