WELDING RESEARCH cooling curves at the top surface of the fusion zone for welding velocities of 20 and 30 mm/s and a 1-mm separation distance are shown in Fig. 6. The calculated cooling rate between 1073 and 773 K increases from 21.3 to 48.4 K/s with the increasing weld speed. The corresponding cooling times are 14.1 and 6.2 s, respectively. As expected, the cooling rate increases with increasing welding speed and resulting decrease in heat input. The calculated cooling curves between 1073 and 773 K along the centerline 1.0 mm above the bottom surface of the weld for laser arc separation distances of 1 and 5 mm at a welding speed of 40 mm/s are shown in Fig. 7. This position is chosen because the width of the weld pool is the smallest at this depth, which indicates that the weld metal at this region is more prone to contain martensite. The calculated cooling rates from 1073 to 773 K are 83.3 and 66.7 K/s when the laser arc separation distance increases from 1 to 5 mm. The corresponding cooling times are 3.6 and 4.5 s, respectively. Ribic et al. (Ref. 27) reported that for hybrid laser-GTA welding with heat input of 0.10 kJ/mm, the cooling time from 1073 to 773 K increased from 0.45 to 0.75 s with the increase of laser arc separation distance from 3.5 to 9.2 mm. As can be seen from Fig. 3A, B, the weld pool length increases with the increasing laser arc separation distance. The isothermals also expand for regions beyond the liquid weld pool so that the spatial gradient of temperature decreases for the temperature range from 1073 to 773 K. Therefore, the cooling rate decreases with increasing laser arc separation distance, which is consistent with the result reported by Ribic et al. (Ref. 27). In order to validate the calculation of the cooling rates between 1073 and 773 K, the calculated cooling rates for different welding conditions are compared with the results available in the literature that examine the effects of changes in heat input and plate thickness on the cooling rate during arc welding (Ref. 42). The literature results show the cooling time from 1073 to 773 K for the welding speeds of 20, 30, and 40 mm/s are approximately 13.1, 6.0, and 3.9 s, compared with the values of 14.1, 6.2, and 3.6 s calculated by the heat transfer and fluid flow model. Therefore, the cooling rates between 1073 and 773 K obtained from the literature and the 3D heat transfer and fluid flow model show good agreement. Microstructures The comparison of the fusion zone microstructures located at the top surface 142-s WELDING JOURNAL / APRIL 2015, VOL. 94 of the weld for welding speeds of 20 and 30 mm/s at laser arc separation distance of 1 mm is shown in Fig. 8. For a welding speed of 20 mm/s, the microstructure contains 35% allotriomorphic ferrite, 11% Widmanstätten ferrite, and 54% acicular ferrite. When the welding speed increases to 30 mm/s, the amount of allotriomorphic ferrite decreases to 29%, Widmanstätten ferrite increases to 12%, and acicular ferrite increases to 59%. The average length of the acicular ferrite decreases from 13.1 to 7.1 m and the width decreases from 3.1 to 1.2 m when the welding speed increases from 20 to 30 mm/s. The hardness testing results show the microhardness value increases from 223 to 248 HV. The average microhardness of the base metal is 174 ± 11 HV. The differences in microstructure contribute to the higher microhardness of the weld metal with a higher welding speed. The comparison of the fusion zone microstructures located 1.0 mm above the bottom surface of the weld for a laser arc separation distance of 1 and 5 mm at a welding speed of 40 mm/s is shown in Fig. 9. The volume fraction of martensite is about 64% for the laser arc separation distance of 1 mm, compared with a separation distance of 5 mm when no martensite is present. The amounts of allotriomorphic ferrite and Widmanstätten ferrite are about 23 and 13%, respectively, for 1-mm separation distance. The volume fraction of acicular ferrite is about 63%, and the amounts of allotriomorphic ferrite and Widmanstätten ferrite are about 24 and 13%, respectively, for a laser arc separation distance of 5 mm. The microhardness decreases from 283 to 238 HV when the laser arc separation distance increases from 1 to 5 mm. The higher microhardness indicates martensite is present and signifies a lower toughness of the weld (Refs. 21, 43, 44), which is detrimental to the mechanical properties of the joint. Figure 10 shows the CCT diagrams computed from the TTT diagram based on the Scheil additive rule, superimposed with the cooling curves at selected locations in the four welds. The cooling curves of the welds with welding speeds of 20 and 30 mm/s at a laser arc separation distance of 1 mm both intercept with the diffusive and displacive transformation curves, so allotriomorphic ferrite, Widmanstätten ferrite, bainite, and acicular ferrite are expected. The cooling curve of the weld with a laser arc separation distance of 1 mm at the welding speed of 40 mm/s intercepts with the upper C curve and the martensite transformation line while the cooling curve for the weld with a laser arc separation distance of 5 mm at the welding speed of 40 mm/s intercepts with both the upper and lower C curves but not the martensite transformation line. As a result, martensite is expected with the laser arc separation distance of 1 mm but not for the laser arc separation distance of 5 mm, although the net heat input of the welding process is identical. The variation of the calculated volume fractions of allotriomorphic and Widmanstätten ferrite, acicular ferrite, and martensite with cooling rates is shown in Fig. 11. For the composition of DH 36 steel and austenite grain sizes observed in the experimental welds, acicular ferrite is the predominant phase, comprising nearly 60% of the microstructure up to the critical cooling rate for martensite. Widmanstätten ferrite increases with cooling rate at the expense of allotriomorphic ferrite. These two phases combine to make up 40% of the microstructure. The calculated results are consistent with the data reported in previous work (Ref. 24). Both the experimentally measured and calculated quantitative volume fractions of different phases of the four welds are shown in Fig. 12. Good agreement between the two sets of data is observed. Figure 12A shows the volume fractions of allotriomorphic ferrite and acicular ferrite slightly decrease while Widmanstätten ferrite increases with the increasing welding speed. Figure 12B shows the martensite volume fraction is about 52% when the laser arc separation distance is 1 mm, but no martensite is observed when the laser arc separation distance increases to 5 mm. A very small amount of acicular ferrite is observed experimentally for a laser arc
Welding Journal | April 2015
To see the actual publication please follow the link above