Abstract
A comprehensive investigation was conducted into the effect of oxides on penetrations, microstructures and mechanical properties of BS700MC super steel weld bead. Boron oxide changed the penetration of weld bead by changing the Marangoni convection in the weld pool and contracting the welding arc. Chromium oxide only changed the Marangoni convection in the weld pool to increase the penetration of super steel. Thus, the super steel weld bead has higher penetration coated with flux boron oxide than that coated with chromium oxide. In other words, the activating flux TIG (A-TIG) welding with flux boron oxide has less welding heat input than the A-TIG welding with flux chromium oxide. As a result, on the one hand, there existed more fine and homogeneous acicular ferrites in the microstructure of welding heat-affected zone when the super steel was welded by A-TIG with flux boron oxide. Thus, the weld beads have higher value of low-temperature impact toughness. On the other hand, the softening degree of welding heat-affected zone, welded by A-TIG with flux boron oxide, will be decreased for the minimum value of welding heat input.
Introduction
Gas tungsten arc welding produces high-quality welds, so it is applied in almost all kinds of metal construction, but the shallow penetration restricts its ability to weld thicker structures in a single pass; this causes the productivity to be relatively low [1, 2].
Activating flux TIG was first invented in the 1960s by researchers at the Paton Electric Welding Institute in Ukraine [3]. It is a method that can improve weld penetration by brushing a thin layer of activating flux onto the surface of the weld bead before welding with conventional TIG. A-TIG welding, compared with conventional TIG welding, provides an advantage, which is that the penetration and productivity can be increased one to three times. This means that, for plates that are 3–10 mm thick, groove preparation is not needed [4, 5].
Metal oxides and nonmetal oxides are the two main kinds of activators used in A-TIG welding. The substrates could get less weld heat input coated with metal oxides or nonmetal oxides compared to conventional TIG by increasing the penetration of the weld. Thus, the microstructures and mechanical properties of the weld may be improved. Even many investigations of the mechanism and the application technology of the A-TIG process have been made and the two representative theories are arc contraction and reversal of the Marangoni convection in the weld pool [6, 7]. However, there is still no common agreement about the A-TIG mechanism.
Consequently, boron oxide and chromium oxide are chosen as the research object in this paper in addition to a systematic study into the effect of activators on penetration, microstructures and mechanical properties of weld joints.
Experimental procedures
The 800 MPa BS700MC super steel scaled 300 mm×150 mm×5 mm and 80 mm×50 mm×5 mm in bulk was used as a base plate in the study. The chemical composition of BS700MC super steel is shown in Table 1.
Chemical composition of BS700MC super steel (wt %).
| C | Si | Mn | Al | Ti | Nb | V. | Mo | P | S |
|---|---|---|---|---|---|---|---|---|---|
| 0.054 | 0.22 | 1.84 | 0.031 | 0.06 | 0.054 | 0.003 | 0.27 | 0.013 | ≤0.003 |
In order to study the effect of boron oxide and chromium oxide on molten pool convection direction, the workpieces (80 mm×50 mm×5 mm) were milled to create two 1-mm-width grooves, each of which was 1.5 mm away from the center of the plate. Because the melting point of tungsten is higher than that of high-strength steel, two 1-mm-thick tungsten plates were inserted into the grooves to block the flow of weld pool in the upper section. Three groups of activating fluxes were made, one consisting of the trace element tungsten and flux boron oxide (weight ratio 1:1), the other containing the trace element tungsten and flux chromium oxide (weight ratio 1:1), and the last one only containing the trace element tungsten. The element tungsten was selected due to its high melting point, which allowed its distribution in the weld to be easily measured.
Schematic diagram of activating flux TIG.
The substrates were grounded with sand paper and then cleaned with acetone to remove any organic elements, such as oily soil. Then, the substrates were dried to prevent the negative effect of water on the results. The fluxes were mixed with acetone and applied manually with a brush in a layer thick enough to prevent the visual observation of the base metal, as shown in Figure 1. The specimens were welded along the center line of the plate. Table 2 lists the welding parameters used in the experiments.
Welding parameters.
| Welding current | 180 A |
| Welding speed | 250 mm/min |
| Diameter of electrode | 4.0 mm |
| Vertex angle of electrode | 60° |
| Arc length | 3 mm |
| Shield gas and flow rate | 150 mL/min |
At the same welding parameters (as shown in Table 2), the images of the electric arc for TIG welding, both with and without activating, were obtained using a high-speed camera (Mega Speed CPL-MMS25K) and stored in a computer with a frame grabber. The acquisition speed was 125 f/s, and the time interval between each two pictures was 8 ms. During welding, the arc voltage was continuously monitored using an Analysator Hannover. After welding, the penetration of three specimens’ weld beams was observed using an Olympus GX51 optical microscope (OM).
The welding parameters were changed to receive the same penetration with different activators. The specimens were cut for impact tests, tensile tests and metallographic observations using corresponding standards [8, 9]. Samples were mounted, polished and etched using a 4 % nital in line with standard procedures and then subjected to further investigation using the Olympus GX51 OM and S4800 scanning electron microscope (SEM). Microstructures of sample fractures were characterized by SEM after tensile test and impact test. And the distribution of the element tungsten at both sides of the tungsten plate was measured by an electronic probe. The as-prepared standard tensile samples were measured on a WE-1000B tensile machine with 1 mm/min crossbeam velocity at room temperature. Conducted at –20 °C on a drop-weight-impact tester instrument from the XJL 300B series, the microhardness of the welding heat-affected zone was measured by an MHV2000-type digital microhardness tester with a 100 g load and 10 s dwell time; the distance between the two points was 0.2 mm.
Results and discussion
Effect of activators on penetration of weld bead
Macroscopic morphology of the weld beads: (a) A-TIG with B2O3; (b) A-TIG with Cr2O3; and (c) Conventional TIG.
The macroscopic morphology of the weld beads of the welded joints is illustrated in Figure 2. As shown in Figure 2, the values of the penetration of A-TIG welding were apparently higher than those of conventional TIG welding. The weld penetration reached the maximum under the A-TIG welding with boron oxide; thus, the 5-mm-thick steel plate could be penetrated at 180 A welding current.
Many investigations of the mechanism and application technology of the A-TIG process have been conducted, and two representative theories remained: the arc contraction theory and the reversal of the Marangoni convection in the weld pool [6, 7]. However, there is still no common agreement about the A-TIG mechanism.
Measured result of arc voltage: (a) A-TIG with B2O3; and (b) A-TIG with Cr2O3.
In this study, the arc voltage of A-TIG welding with flux boron oxide, compared with that of conventional TIG welding, increases obviously; these results are shown in Figure 3(a). However, when using chromium oxide, the arc voltage shows no notable changes during the welding process, as shown in Figure 3(b).
Arc images and schematic diagrams in welding with different flux: (a) A-TIG with B2O3; and (b) A-TIG with Cr2O3.
Many researchers have concluded that the arc voltage is higher with nonmetal oxides than with metal oxides [10, 11]. An increase in voltage first suggests that weld pool formation occurred only after the flux layer had been broken up and possibly removed from the arc column. Note that this process of displacing the flux, either by vaporization or fluid flow, is energy consuming. The electric current density will increase to achieve greater energy if the other welding parameters do not change; thus, the arc pressure is higher with nonmetal oxides than with metal oxides. Hence, the morphology of arc will migrate from the side with nonmetal oxides to the other side that does not contain activators, as shown in Figure 4(a). However, the morphology of the arc with metal oxides will not change during the welding process, as shown in Figure 4(b).
Marangoni convection mode in the weld pool: (a) ∂σ/∂T < 0 and (b) ∂σ/∂T > 0.
Distribution of tungsten element in the weld bead (wt %).
| Group | Conventional TIG | Coated with B2O3 | Coated with Cr2O3 |
|---|---|---|---|
| [A] | 3.88 | 17.26 | 15.11 |
| [B] | 9.05 | 2.59 | 3.02 |
Generally, surface tension decreases with increasing temperature; in other words, ∂σ/∂T < 0 for a pure metal and many alloys, as shown in Figure 5(a). In the weld pool for such materials, surface tension is higher in the relatively cooler parts of the pool edge than it is in the pool center under the arc. Hence, the fluid flows from the pool center to the edge, and the heat flux easily transfers to the edge and forms a wide and shallow weld shape as shown in Figure 2(c). Negative Marangoni will result in more element tungsten distributed near the edges of the weld fusion zone than in the center, as shown in Table 3.
Heiple and Ropper [12, 13] proposed that some active elements, such as O, S and Se, can change the temperature coefficient of surface tension for iron alloys from negative to positive (∂σ/∂T > 0) when their quantity surpasses a critical value, as shown in Figure 5(b). In this case, the Marangoni convection on the pool surface is changed from an outward to an inward direction, and a relatively deep and narrow weld shape is obtained as shown in Figure 2(a) and (b). Positive Marangoni will result in the more element tungsten being distributed in the center of the weld fusion zone than that at the edges, as shown in Table 3.
Effect of activators on mechanical properties
Mechanical properties of BS700MC super steel.
| Group | Coated with B2O3 | Coated with Cr2O3 | Conventional TIG |
|---|---|---|---|
| σs/MPa | 844 | 814 | 713 |
| σb/MPa | 724 | 713 | 621 |
| AKV(–20 ℃)/J | 112 | 98 | 54 |
Table 4 illustrates the values of yield strength, tensile strength and low-temperature impact energy of welded parts with different activators. As can be seen in Table 4, the values of tensile strength, yield strength and low-temperature impact energy of conventional TIG weld beam are less than that of A-TIG with boron oxide or chromium oxide. And the values of the three mechanical properties of super steel coated with boron oxide are better than others.
Microstructure and fracture of welding heat-affected zone: (a, d) coated with boron oxide; (b, e) coated with chromium oxide; and (c) conventional TIG.
As the super steel is coated with boron oxide, the microstructures within the weld metal contains large amount of uniform and fine acicular ferrites, as shown in Figure 6(a). Acicular ferrites have large amounts of grain boundaries, free orientations, high-angle grain boundaries and dislocations, and the micro-cracks will demand more energy to cross the acicular ferrites [14, 15]. Therefore, not only are the acicular ferrites able to enhance strength, but the low-temperature impact toughness can be significantly improved. As the super steel coated with chromium oxide, the content of acicular ferrites is less than that coated with boron oxide, and some massive proeutectoid ferrites appeared in the weld metal, as shown in Figure 6(b). And the proeutectoid ferrites will distribute as thin strips around the austenite grain boundaries when the super steel was welded by conventional TIG, as shown in Figure 6(c). Thus, the microstructure of weld metal will be divided when the shape of the proeutectoid ferrites is bulk or thin strip. As a result, the value of strength and low-temperature impact energy will be reduced.
Figure 6(d), (e) and (f) shows the SEM micrographs of fractured morphologies after the impact test. Figure 6(d) and (e) presents typical ductile fracture characterized by large amounts of dimples. However, the size and distribution of dimples vary; further investigations find that many dimples in large size are observed to be present in Figure 6(d). Due to the ability of these dimples to improve the low-temperature toughness of weld metal deposits [16], it can be concluded that the sample coated with boron oxide has the highest low-temperature toughness. Meanwhile, the fractured morphology in Figure 6(f) has some cleavage fracture morphologies. So, the low-temperature impact energy of conventional TIG is less than the above two weld methods.
Microhardness distribution of the heat-affected zone: (a) overheated zone; (b) incomplete recrystallization zone; and (c) substrate.
Hardness, one of the most important indices of mechanical properties, is closely related to its corresponding microstructures [17]. The heat-affected zone directly influences the quality of weld joints, given that it is the most sensitive part of the weld joints. Figure 7 shows the microhardness of the heat-affected zone. As can be seen, the microhardness of each of the relevant three samples obeys the same law. (1) The value of the microhardness of the heat-affected zone is less than the substrate. (2) The heat-affected zone has two softened zones, the first one is the overheated zone near the fusion line and the second one is the incomplete recrystallization zone near the substrate. In addition, Figure 7 shows that microhardness of the heat-affected zone reaches the maximum when the super steel was welded by A-TIG welding with flux boron oxide. The microhardness of the heat-affected zone reaches the minimum when the super steel was welded by conventional TIG welding.
Super steel has excellent mechanical properties for grain refinement and quenching and tempering treatment. However, during the welding process, the grain of the welding heat-affected zone will grow during the welding thermal cycle, and the effect of quenching and tempering treatment will also be decreased [18]. Hence, the microhardness and other properties of the welding heat-affected zone will decrease accordingly.
The heat input (KJ/mm) was evaluated using eq. (1) [19]:
In expression (1), V is arc voltage, I is arc current, S is welding speed and 0.9 is the arc efficiency.
Minimum current for fusion penetration and grain size of the welding heat-affected zone.
V and S are taken as the constant values in this experiment, and the value of welding heat input depends on the welding current. Figure 8 shows the minimum current required for fusion penetration of three welding methods. The welding heat input reaches the minimum when the super steel is welded by A-TIG welding with flux boron oxide. And the conventional TIG has the maximum welding heat input. Thus, the grain size of welding heat-affected zone after conventional TIG has the maximum among the three welding methods.
The softening of the overheated zone, as shown in Figure 7, occurs mainly because the peak temperature of the overheated zone is too high, and the residence time is so long that the amount of melted alloy elements and their compounds becomes higher. As a result, the degree of segregation of the alloy elements and their compounds becomes higher after refrigeration. The segregation of alloy elements, their carbides and a large amount of dislocations will weaken resistance to deformations and cracks, so the microhardness of the overheated zone is lower than the substrate [20]. The softening of the incomplete recrystallization zone is caused by the following two reasons [21]: (1) The dislocation density that is used to control rolling and low-temperature phase transition declines and (2) a large amount of deformed grains is transformed into equiaxed grains by recrystallization under the welding heat, although phase transition of the microstructure in the grains does not occur, and the size of the grains changes little. Therefore, the effect of the work expended for the hardening and deformation–strengthening of parts of the grains by thermodynamics rolling technology is weakened to a great extent. So, the softening area coverage of the welding heat-affected zone is increased.
Conclusions
Under investigation are the penetration, microstructural evolution, strength, low-temperature toughness and microhardness of BS700MC super steel weld beads when subjected to different activators. The major conclusions are summarized as follows:
Boron oxide can increase the weld penetration by contracting welding arc and reversing the Marangoni convection in the weld pool. Chromium oxide only changes the Marangoni convection in weld pool to increase the penetration of super steel.
The microstructure of welding heat-affected zone generates a large amount of acicular ferrites when the super steel is coated with boron oxide, causing the weld joints to have the maximum value of low-temperature impact toughness.
Adding boron oxide and chromium oxide could significantly reduce the minimum current required for fusion penetration. As a result, the softening degree of the welding heat-affected zone will be reduced for the minimum value of welding heat input.
Acknowledgments
The authors wish to acknowledge Mr Wang Fei and Mr Feng Yue Qiao for their useful technical discussions and support during the process of this project.
References
[1] J.E. Rodriguez-Sanchez, F. Perez-Guerrero, S. Liu, A. Rodriguez-Castellanos and A. Albiter-Hernandez, Fatigue Fract. Eng. Mater. Struct., 37 (2014) 637–644.10.1111/ffe.12146Search in Google Scholar
[2] K.D. Ramkumar, D. Mishra, B.G. Raj, M.K. Vignesh, G. Thiruvengatam, S.P. Sudharshan, N. Arivazhagan, N. Sivashanmugam and A.M. Rabel, Mater. Des., 66 (2015) 356–365.10.1016/j.matdes.2014.10.084Search in Google Scholar
[3] C.L. Yang, S.B. Lin, F.Y. Liu, L. Wu and Q.T. Zhang, J. Mater. Sci. Technol., 19 (2003) 225–227.Search in Google Scholar
[4] A. Berthier, P. Paillard, M. Carin, F. Valensi and S. Pellerin, Sci. Technol. Weld. Joi., 17 (2012) 609–615.10.1179/1362171812Y.0000000024Search in Google Scholar
[5] P. Vasantharaja and M. Vasudevan, J. Nucl. Mater., 421 (2012) 117–123.10.1016/j.jnucmat.2011.11.062Search in Google Scholar
[6] R.H. Zhang, J.L. Pan and S. Katayama, Front. Mater. Sci., 5 (2011) 109–118.10.1007/s11706-011-0125-5Search in Google Scholar
[7] M. Vasudevan, V. Arunkumar, N. Chandrasekhar and V. Maduraimuthu, Sci. Technol. Weld. Joi., 15 (2010) 117–123.10.1179/136217109X12577814486773Search in Google Scholar
[8] ISO 6892-1: Metallic Materials – Tensile Testing – Part 1: Method of Test at Room Temperature (2009).Search in Google Scholar
[9] ISO 9016: Destructive Tests on Welds in Metallic Materials – Impact Tests: Test Specimen Location, Notch Orientation and Examination (2001).Search in Google Scholar
[10] K.A. Yushchenko, D.V. Kovalenko, I.V. Krivtsun, V.F. Demchenko, I.V. Kovalenko and A.B. Lesnoy, Weld. World., 53 (2009) 253–263.10.1007/BF03321137Search in Google Scholar
[11] Q.M. Li, X.H. Wang, Z.D. Zhou and J. Wu, Trans. Nonferrous Met. Soc., 17 (2007) 486–490.10.1016/S1003-6326(07)60120-4Search in Google Scholar
[12] C.R. Heiple, J.R. Roper and R.T. Stagner, Weld. J., 62 (1983) 72–77.Search in Google Scholar
[13] C.R. Heiple and P. Burgardt, Weld. J., 64 (1985) 159–162.Search in Google Scholar
[14] Z. Chunguo, S. Xuding, Lu. Pengmin and Hu. Xiaozhi, Mater. Des., 36 (2012) 233–242.10.1016/j.matdes.2011.11.016Search in Google Scholar
[15] R.A. Ricks, J. Mater. Sci., 17 (1982) 732–740.10.1007/BF00540369Search in Google Scholar
[16] R. Cao, S.S. Zhu, W. Feng, Y. Peng, F. Jiang, W.S. Du, Z.L. Tian, J.H. Chen, J. Mater. Process. Technol., 211 (2011) 759–772.10.1016/j.jmatprotec.2010.12.011Search in Google Scholar
[17] E. Gharibshahiyan, A. Honarbakhsh Raouf, N. Parvin and M. Rahimian, Mater. Des., 32 (2011) 2042–2048.10.1016/j.matdes.2010.11.056Search in Google Scholar
[18] I.I. Frantov, A.A. Velichko, A.N. Bortsov and I.Y. Utkin, Weld. J., 93 (2014) 23S–29S.Search in Google Scholar
[19] B. Arivazhagan and M. Vasudevan, J. Mater. Eng. Perform., 22 (2013) 3708–3716.10.1007/s11665-013-0694-9Search in Google Scholar
[20] M. Rahmani, A. Eghlimi and M. Shamanian, J. Mater. Eng. Perform., 23 (2014) 3745–3753.10.1007/s11665-014-1136-zSearch in Google Scholar
[21] J. Moon, C.H. Lee, T.H. Lee and H.C. Kim, Metall. Mater. Trans. A, 46A (2015) 156–163.10.1007/s11661-014-2623-4Search in Google Scholar
©2016 by De Gruyter
This article is distributed under the terms of the Creative Commons Attribution Non-Commercial License, which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
