Table 5 — Composition of Different Weld Metals (wt%) Derived from Dilution and Creq, Nieq, Creq / Nieq Ratio and Martensite Transformation Temperatures (Ms and Mεs) Sample Specification J1 J2 J3 J4 C 0.0219 ±0.0004 0.022 ±0.0007 0.022 ±0.0009 0.022 ±0.001 Si 0.499 ±0.002 0.499 ±0.003 0.498 ±0.004 0.498 ±0.004 Mn 1.268 ±0.02 1.262 ±0.04 1.259 ±0.05 1.257 ±0.06 P 0.020 ±0.0005 0.020 ±0.0008 0.020 ±0.0009 0.020 ±0.001 S 0.023 ±0.0005 0.022 ±0.0007 0.022 ±0.0009 0.022 ±0.001 Cr 15.63 ±0.25 15.57 ±0.4 15.540 ±0.48 15.530 ±0.55 Ni 5.122 ±0.26 5.058 ±0.43 5.025 ±0.52 5.012 ±0.6 Mo 0.078 ±0.003 0.078 ±0.004 0.077 ±0.005 0.077 ±0.005 Cu 0.056 ±0.002 0.056 ±0.002 0.056 ±0.003 0.0554 ±0.004 Nb 0.022 ±0.0003 0.022 ±0.0004 0.0215 ±0.0005 0.0217 ±0.0005 N 0.0331 ±0.0013 0.0328 ±0.002 0.0326 ±0.002 0.0325 ±0.003 Creq 15.72 ±0.25 15.69 ±0.23 15.63 ±0.32 15.60 ±0.45 Nieq 6.655 ±0.36 6.52 ±0.1 6.47 ±0.27 6.58 ±0.62 Creq/ Nieq 2.36 ±0.1 2.40 ±0.15 2.415 ±0.17 2.37 ±0.12 MS K(˚C) 368.3 (95.3) ±19.1 377.04 (104.04) ±20.7 379.5 (106.5) ±23.9 370.2 (97.2) ±17.4 MεS K(˚C) 358.8 (85.8) ±8.91 360.5 (87.5) ±15.1 361.4 (88.4) ±18.23 360.3 (87.3) ±11.8 found out from Equations 1, 2, and 3, respectively. The effect of shielding gas compositions on the geometrical characteristics of the weld with respect to AWD, ABF,, and DL is given in Table 4. As expected, weld metal prepared with pure Ar shielding (J1) leads to lower dilution. Other weld metals have comparatively higher DL% and it increases with the increase in CO2 content. Weld metal compositions obtained from dilution calculation are shown in Table 5. Furthermore, optical emission spectroscopy (as per ASTM 1086-94) of the weld metal was carried out to ascertain the compositions obtained from dilution. Only the spectroscopic result of carbon content, as shown in Fig. 5, differs from the dilution calculation, due to the fact that in dilution calculation the absorption of carbon from dissociated shielding gas mixtures (during welding) was not considered. The final weld metal compositions were then used to calculate the chromium equivalent (Creq) and nickel equivalent (Nieq) values using Equations 4 and 5 along with Creq/Nieq ratios, and the values are also given in Table 5. The variation in Creq and Nieq among four different welds (J1, J2, J3, and J4) is due to significant variation in wt-% of Cr and Ni (Table 5). The Creq and Nieq values were then incorporated in the modified WRC-1992 diagram (Ref. 18) to predict the ferrite number of the welds. The predicted ferrite (d) content of the entire weld varied with shielding gas composition used. It is interesting to note that the predicted ferrite (d) content and the measured ferrite (d) content for different weld metals are very close to each other as shown in Fig. 6A. Present investigation reveals that welds J2 and J3 contain comparatively a higher amount of ferrite (d) than welds J1 and J4. This may be due to the higher Creq/Nieq ratio of J2 and J3 welds compared to J1 and J4 as shown in Table 5. Higher Creq/Nieq ratio increases the stability of the d-ferrite by shifting the solidification line away from the triple point (i.e., L + g + d zone) into the d-ferrite region. In order to understand the phase transformation in different welds, the presence of - and -martensite, - martensite start temperature (Ms), and -martensite start temperature (Ms), i.e., martensite transformation temperatures, were calculated using Equations 6 and 7, respectively, and the values are given in Table 5. The welds in general have lower Ms temperature values compared to Ms temperature (Table 5), and therefore, g transformation will take a predominant role over g transformation (Ref. 12). However, Ms and Ms values for J1 and J4 being very close (less than 10°C) to each other compared to welds J2 and J3, the possibility of the -martensite present in JI and J4 welds will be greater. Nevertheless, it is well known that g transformation has more thermodynamical stability over g transformation (Ref. 19) and thus under any stress such as residual stress during welding, -martensite will transform into -martensite (Ref. 20). Therefore, it can be assumed that the weld metals should contain -martensite along with some -martensite in their final microstructure. Also, martensite transformation temperatures have been correlated with Creq/Nieq ratios of different welds as shown in Fig. 6B. The weld metal compositions in terms of Creq/Nieq ratio have created a variation in the martensite transformation temperatures among the welds, and hence, the amount of martensite laths. Accordingly, J2 and J3 having higher Creq/Nieq ratio and martensite transformation temperatures than welds J1 and J4 (Table 5) should provide higher amount of martensite laths. To validate the possibility of a higher amount of martensite formation in welds J2 and J3, -martensite content in mass % has been determined and the values are presented in Fig. 6C. It indicates that J2 and J3 have higher amounts of -martensite than welds J1 and J4. This observation can also be supported by the X-ray diffraction patterns of welds as shown in Fig. 7, which depicts major peak intensities of bcc phase (), whereas austenite (fcc) phase does not differ significantly, indicating primary ferrite solidification. Therefore, it is exciting to note that, under the same heat input or cooling rate (Table 2), the amount of solid-state phase transformation is solely dependant upon the chemical composition of the weld metal, which is the result of variation in shielding gas compositions. WELDING RESEARCH 106-s WELDING JOURNAL / APRIL 2015, VOL. 94
Welding Journal | April 2015
To see the actual publication please follow the link above