The effect of shielding gases in the Ferrite Number of austenitic stainless steels joints through GMAW

In order to better understanding the effect of shielding gases in the volume fraction of δ ferrite in welded deposits through GMAW, the microstructures of four welded joints of austenitic stainless steel produced by the MIG/MAG process with different shielding gases were studied. The deposits were produced using the same welding electrode ER309Land welding parameters, but different shielding gases from pure argon to mixtures with increasing contents of CO2. Each of the welding deposits were produced with 100% Ar, Ar+2% CO2, Ar+4% CO2 and Ar+20% CO2. The chemical compositions and the variation of the volume fractions of δ ferrite in the deposits was measured. There was an increasing pickup of carbon and decreasing volume fraction of δ ferrite in the all weld metals produced using shielding gases with increasing concentrations of CO2.The results confirm that carbon is a strong austenite stabilizer in austenitic stainless steels. Complementary techniques of microstructural analysis were used, such as optical emission spectrometry, optical microscopy and quantitative image analysis. Keywords— Austenitic Stainless Steels; Solidification Mode; Ferrite Number; GMAW.


INTRODUCTION
In addition to iron, chromium and nickel, stainless steels have other chemical elements in their composition that can stabilize ferrite and austenite. Schaeffler [1] grouped these elements into two expressions called chromium equivalent and nickel equivalent, respectively, and proposed a diagram that is shown in Figure 1, considering the ferritizing and austenitizing effects of different alloying elements.
In these diagrams, the ferrite contents of various welds had been measured experimentally by either metallography (Schaeffler) or magnetic methods (DeLongand WRC-92). [12] From the Schaeffler diagram, the first striking change was made by Delong [6], which includes the austenitizing effect of nitrogen in the nickel equivalent formula and proposed a diagram that is shown in Figure 2. O. Hammar and U. Svensson [4] showed that the addition of carbon and nitrogen decreases the volumetric fraction of ferrite in austenitic stainless steels.Taking as an example the austenitic stainless steel of type AISI 316, which usually solidifies through a ferritic-austenitic solidification mode. With the increasing of carbon and nitrogen contents as alloying elements, the solidification mode changes to austenitic-ferritic. There is, therefore, acarbon equivalent value that can change how this steel solidifies.
Ceq =% C + 0.65% N (Equation 3) Figure 3 shows the variation of the volume fraction of primary δ ferrite as a function of carbon equivalent. Kotechi [11] has pointed out that there are number of alloying elements that have not been considered in the most accurate diagram to date, the WRC -92 diagram. Elements like silicon, titanium, tungsten are not given due considerations though they are known to influence the ferrite content. He also stressed the point that cooling rate effects need to be considered more thoroughly in these constitution diagrams. [12] Table1 shows the expressions of chromium and nickel equivalents proposed by Schaeffler [1], DeLong [6] and Kotechi [11]. When the Creq/Nieq ratio <1.5, the solidification may be austenitic (mode I) or austenitic-ferritic (mode II). When the ratio 1.5 <Creq/Nieq<2.0 the solidification will be ferritic-austenitic (mode III). And finally, when Creq/Nieqratio > 2.0 the solidification will be ferritic (mode IV). [2,3,[13][14][15] The possible solidification modes in the Fe-Cr-Ni system are:

I) Austenitic solidification (LL+):
The only solid phase to form is austenite. In austenitic solidification, called solidification mode I, there is no other phase transformation at high temperature. Austenite solidifies as a primary phase in a dendritic or cellular way. As the temperature decreases, ferrite  is formed from the remaining liquid. Solidification occurs through a peritectic reaction (L+). This is called solidification mode II. [13-15] III)

Ferritic-austenitic solidification (LL+L+++):
The duplex stainless steels solidify according to ferritic-austenitic solidification (LL+L+++).  ferrite solidifies as the primary phase in dendritic or cellular fashion. As temperature decreases, austenite is formed by a peritectic (L+) or eutectic (L+) reaction. In the case of a peritectic reaction, the initially formed austenite completely surrounds the ferrite and subsequently grows into ferrite and liquid. Depending on the rate of diffusion through the austenite, the reaction may or may not be complete, and at the end of the solidification ferrite may be involved in austenite. Between the two reactions -peritectic and eutectic -the transition takes place where, during the initial formation of austenite by peritectic reaction, ferritizing elements secrete to the liquid, provoking their enrichment in these elements and consequently the simultaneous formation of ferrite and austenite by means of a eutectic reaction. This is called solidification mode III. [13][14][15][16][17][18][19][20][21]

IV) Ferritic solidification (LL+):
The only solid phase to form is ferrite. In ferritic solidification, called solidification mode IV, ferrite is the only phase to form during solidification and, depending on the chemical composition, austenite can precipitate only in the solid state in the ferritic grain boundaries. [2,3] The solidifications of austenitic stainless steels can occur according to the first three solidification modes. Depending on the conditions of solidification, the factors for the elements in the expressions of chromium and nickel can vary widely and some elements that do not influence the expressions, depending on the process, can be important when dealing with different solidification modes. Table 2 shows the expressions of chromium and nickel equivalents suggested by different researchers, considering different production process of stainless steels.

II. EXPERIMENTAL
Four welded joints of austenitic stainless steel produced by the MIG/MAG process with different shielding gases were studied. The deposits were produced using the same welding electrode ER309L1,2 mm according to AWS 5.9, and welding parameters, but different shielding gases from pure argon to mixtures with increasing contents of CO2. Each of the welding deposits were called sample 1, sample 2, sample 3 and sample 4 and produced, respectively,using100% Ar, Ar+2% CO2, Ar+4% CO2 and Ar+20% CO2, as the shielding gases. The GMAWwelding machine was adjusted to allow a stable welding for the four shielding gases.
After adjustment of optimum welding parameters to have arc stability with the different shielding gases, an automatic welding tractor was used to guarantee the correct travel speeds to have similar heat inputs for all the four samples.
In order to minimize the effect of base metal chemical composition, 6 layers of 5 beads each were deposited.Overlapping passes were used, depositing approximately 25 mm on the base metal that was an AISI 304L type stainless steel.The weld pads were cut in longitudinal and transversal directions. Chemical analyzes were carried out in all samples at 20 mm from the base metal, by means of an optical emission spectrometer, according to ASTM E 1086-08. [22] Afterwards, the samples transversal and longitudinal samples were embedded in hot-cure resin (bakelite). The conventional manual polishing was applied using water slicks (100, 240, 320, 400, 600 and 1000 mesh) in order to standardize the surface finish of the samples. A cloth polishing with 9, 3 and 1 μm diamond abrasive paste was carried out in this sequence.  Table 3 presents the welding parameters used to weld the samples.It is important to emphasize that the welding wire used to produce samples 1, 2, 3 and 4 wasthe ER309L according to AWS 5.9, 1.2 mm diameter.     Table 5 presents the calculated values Creq,Nieq and Creq/Nieq ratio according to the expressions of chromium and nickel equivalents proposed by Schaeffler [1], DeLong [6] and Kotechi [11]. The calculations of Creq,Nieq and Creq/Nieq ratio were done using formulas taken from Table 1. The results presented on table 4 and 5, show that increasing the concentration of CO2 in the shielding gases increases the concentration of C in the deposited metal. The results obtained suggest that with an increase in the concentration of CO2 in the shielding gases, a decrease in the Creq of the alloys occurs due to the selective oxidation of the elements Cr and Si. Figure 5 shows the contents of C, Si, Mn and Cr (% by weight) of filler metal ER 309L and welded chemical pads.  Despite the selective oxidation of Mn, the observed trend is of increasing of the Nieq. Figure 6 shows the variations of the Creq and Nieq values (% by weight) and the Creq/Nieq ratio of the filler metal ER 309L and the welded chemical pads according to the expressions of chromium and nickel equivalents proposed by Schaeffler, DeLongand Kotechi.     The volumetric fractions of δ ferrite verified in the longitudinal direction are smaller than those verified for the transversal direction in the four welded specimens. The metallographic analysis of the welded specimens revealed that the increase in the concentration of CO2 in the shielding gases decreases the volume fraction of δ ferrite in the deposited metal. As discussed earlier, these results are consistent with chemical analysis of chemical weld pads. In Figure 6, it can be seen, that with the increase of the CO2 concentration in the shielding gases, there is a decrease in Creq and an increase in Nieq, resulting in a decrease in the Creq / Nieq ratio of the alloys.

IV. CONCLUSIONS
The increase in the concentration of CO2 in shielding gases increases the Nieqof the alloys by increasing thecontent of carbon in the deposited all weld metal.
The results obtained suggest that increasing the concentration of CO2 in the shielding gases decreases in the Creq of the alloys due to the selective oxidation ofCr and Si elements.
Increasing the carbon content of the alloy decreases the volume fraction of δ ferrite in the deposited all weld metal.
The results obtained suggest that solidification of the studied alloys is ferritic-austenitic (mode III).
The increase in concentration of CO2 in shielding gases decreases Creq and increase Nieq, in the deposited all weld metal, resulting in the decrease in the Creq/Nieq ratio of the alloys.