Abstract

1 Introduction

The shipbuilding industry requires technological advances, such as research in welding techniques and appropriate materials to reduce processing time and costs without impairing the desired mechanical properties. Recently, researchers around the world have focused on research and development of high strength low alloy steels (HSLA), as well as the investigation of various factors influencing their weldability.

Steel plates produced by thermo-mechanical control processing (TMCP) combined with various microalloyed elements have provided an effective solution to those requirements, offering better combinations of strength and toughness without impairing the weldability ¹1 Tsukada K, Watanabe, I. The progress of thermo-mechanical control process for HSLA plate in Japan. Key Engineering Materials. 1993;84-85:22–26.^,22 Shiga C, Amano K, Enami T. Application of Multipurpose accelerated cooling systems (MACS) to the production of HSLA steel plate. In: International Conference on Technology and application of High Strength Low Alloy (HSLA) Steels . Proceedings. Philadelphia: American Society for Metals;1984.643 – 649.. Welding accounts for a one third of all services performed in the shipbuilding industry, with submerged arc welding (SAW) being one of the most widely used processes ³3 Pickering FB. Physical metallurgy and the design of steels . London: Applied Science; 1983.. Tandem submerged arc welding has been studied to increase the welding productivity.

The tandem technique process uses two or more wires aligned in the welding direction to increase the deposition rate and welding speed ⁴4 Bohme D, Nentwig AW, Knoch R. A high efficiency welding process in double wire welding. In: II Welding International Congress. Proceedings. Auckland; 1996. p 1393 – 1407..The tandem SAW has been used as an alternative to the conventional single wire SAW process, mainly in plates of medium and large thickness (greater or equal to 20 mm) ⁵5 Almqvist I, Gronbeck SB. Submerged arc welding. 2nd ed. Cambridge: The Welding Institute;1978._. Tandem SAW creates stronger and tougher weld joints due to a suitable thermal cycle in the heat affected zone, when compared with thermal cycles for a one wire welding process. The purpose of this work is to evaluate the welding of microalloyed shipbuilding steel, AH36, produced via TMCP, using flux cored arc welding (FCAW) at the root and filling with SAW with one wire and with two wires.

2 Experimental Procedure

The base metal used in this work was microalloyed shipbuilding steel, AH36, produced via TMCP, with a 20 mm thickness. Table 1shows the chemical composition and the mechanical properties of the base metal based on the ASTM A131 standard ⁶6 American Society for Testing and Materials. ASTM A131/A131M-08: Standard Specification for Structural Steel for Ships. ASTM International; 2001. and the equivalent carbon in accordance with IIW. The geometry of the joints is shown inFigure 1.

Figure 1
(a) Dimensions of plates for welding and (b) detail of the bevel adopted for the joint.

Both single wire (S-SAW) and two wires tandem SAW (T-SAW) were used in this work. The root pass, for both welds used the flux-cored arc welding E71T-1 filler metal with a diameter of 1.20 mm ⁷7 AWS CODE A5. 20/A5.20M. Carbon steel electrodes for flux cored arc welding [Internet]. American National Standard; 2005. [cited 2015 Jul 21]. Available from: https://pubs.aws.org/Download_PDFS/A5.20-A5.20M-2005PV.pdf
https://pubs.aws.org/Download_PDFS/A5.20... and 100% CO₂ as the shielding gas. The chemical compositions and mechanical properties of the FCAW root pass consumable and the SAW electrode are presented inTable 2 and Table 3. The welding parameters are typical of industrial practice and are shown in Table 4 and Table 5. The inter-pass temperature was controlled to be less than 250°C. The heat input was calculated with the conventional heat input equation using a coefficient of heat transfer into the plate of 0.95 ⁸8 Grong Ø. Metallurgical modelling of welding. 2nd ed. London: Institute of Materials; 1997..

Optical microscopy characterization was done by polishing and etching the samples using 2% Nital and Klemm 1. This reagent colors the microstructure and reveals martensite and austenite microconstituents (MA) as white and ferrite as blue. An image analysis program was used to measure the MA volume fraction. A scanning electron microscope (SEM) was used to examine the fracture surface of various specimens.

Microhardness testing was performed in accordance with ASTM E384 ⁹9 ASTM E384. Standard Test Method for Knoop and Vickers Hardness of Materials [Internet]. ASTM International;2005. [cited 2015 Jul 21]. Available fron: http://www.astm.org/FULL_TEXT/E384/HTML/E384.htm.
http://www.astm.org/FULL_TEXT/E384/HTML/... , and conducted using a LECO's LM Series microhardness testing machine with a 300 gf load and a hold time of 10 s. A 0.2 mm grid spacing was used, revealing the distribution and individual hardness values in selected regions of the weld joints, as depicted in Figure 2. Charpy V impact fracture toughness tests were performed according to ASTM E23 standard ¹⁰10 ASTM International. ASTM E23; 2006. for the base metal and the weld metal. Tests were done for three samples each at -60, -40, -20, 0 and 25°C. Samples from the base metal were taken from the longitudinal rolling direction (LRD) and transverse rolling direction (TRD).

Figure 2
Weld joints showing the regions of microhardness indentations for (a) EXP1 (single wire S-SAW) and (b) EXP 2 (two wires T-SAW).

3 Results and Discussion

3.1 Microstructural characterization

Microstructural characterization of the joints is shown in Figure 3 to 8. Figure 3.a presents the microstructure of the base metal. It can be seen to be ferrite and (F) and pearlite (P). Table 6 shows a comparison between volumetric fractions of the F and P in the HAZ and BM.

Figure 3
Optical microstructure characterization of base metal. (a) 2% Nital etching and (b) Klemm I etching.

Figure 8
Optical microstructure characterization of weld metal (WM) for T-SAW. (a) 2% Nital etching and (b) Klemm I etching.

Figure 3.b shows the microstructure of the base metal etched with Klemm I reagent. This reagent is able to show ferrite as blue or brown, depending upon the crystallographic orientation. In addition, it identifies the presence of MA microconstituents as white regions. In this image, there is only ferrite in blue and pearlite as dark gray. The grain size in the base metal was measured. The results show a range from 8 to 17 µm, which agrees with past work ¹¹11 Rui F, Zhang R, Sheng LI, Kong G, Zhu X. Study on forming mechanism of lamination defect of AH36 shipbuilding plate steel. Advanced Materials Research. 2012;562-564:106-110.. Specifically, the ferrite grain size and the banded structure are due to microstructural changes that occur during TMCP processing.

Figure 4a presents the microstructure in the FGHAZ region (fine grain heat affected zone) in the welded joint for S-SAW. The observed constituents are polygonal ferrite (PF), ferrite with non-lamellar carbides (FC) and perlite (P). The mean grain size for this region was between 10 and 12 µm, which is slightly lower than that for the base metal. The temperature during the welding process is high enough to allow partial austenite transformation in the intercritical region. However, the region from the FGHAZ to the fusion line does not undergo significant grain growth of austenite, since the cooling rate drives the austenitic transformation in polygonal ferrite grains ¹²12 Poorhaydari K, Ivey DG. Estimation of cooling in the welding plates with thickness Intermediate. Welding Journal. 2005;84:149-155.. Likewise, the FGHAZ was etched with Klemm I reagent (Figure 4.b) and no MA microconstituents are present. The large amount of PF in the FGHAZ occurs because the cooling rate in this region is lower than the cooling rate for the CGHAZ region (coarse grain heat affected zone). A cooling rate between 1 and 5°C/s in the AH36 steel promotes the formation of polygonal ferrite and perlite¹³13 Ghosh SD, Chatterjee S, Ramachandrarao P. Microstructural characterization of controlled forged HSLA – 80 steel by transmission electron microscopy. Materials Characterization. 2003;50:305-315..

Figure 4
Optical microstructure characterization of FGHAZ for S-SAW. (a) 2% Nital etching and (b) Klemm I etching.

The microstructure in the CGHAZ for S-SAW (Figure 5.a) was composed of grain boundary ferrite (GF), ferrite with a second phase FS (aligned FS(A) and nonaligned FS(NA)), polygonal ferrite (PF) and ferrite with non-lamellar carbides (FC). These microstructural changes occur when the temperature in the CGHAZ region is above the A_c3 temperature for AH36 steel, thus promoting the growth of austenite grains. If the peak temperature during the last pass exceeds the A_c3 temperature, the resulting austenite after cooling may decompose in Widmanstätten ferrite, bainite and MA microconstituents, which can be seen in Figure 5.b. as white dots.

Figure 6 depicts the weld metal microstructure for S-SAW. Figure 6.a shows acicular ferrite (AF), ferrite with second phase FS (aligned FS(A) and nonaligned FS(NA)) and grain boundary ferrite (GF). In Figure 6.b one can see the MA microconstituents with a volumetric fraction comparable to the CGHAZ of S-SAW.

Figure 6
Optical microstructure characterization of weld metal (WM) for S-SAW. (a) 2% Nital etching and (b) Klemm I etching.

The microstructure presented in Figure 7 is similar to Figure 5. The difference is a coarser microstructure and the amount of MA microconstituent, which is larger for T-SAW. These results are likely due to the higher heat input used for T-SAW, which produced higher peak temperatures for comparable thermal cycles. These characteristics in the thermal cycles produced the changes in the microstructure observed.

Figure 7
Optical microstructure characterization of CGHAZ for T-SAW. (a) 2% Nital etching and (b) Klemm I etching.

Figure 5
Optical microstructure characterization of CGHAZ for S-SAW. (a) 2% Nital etching and (b) Klemm I etching.

Figure 8 shows the weld metal microstructure for T-SAW. The phases are identical to those in Figure 6. In Figure 8.b one can see the MA microconstituents with a volumetric fraction higher than that for the WM (weld metal) of S-SAW.

The MA microconstituent is associated with the dissolution of carbides and a decrease in the M_s and M_f temperatures. The amount of retained austenite depends on the cooling rate of the joint; the faster the cooling rate, the larger the volumetric fraction of austenite. This was observed for the case of T-SAW.

The volumetric fraction of MA microconstituent was measured by the colored etching technique. The results are presented in Figure 9. However, this metallographic technique is not sufficient to prove, in a precise way, the presence of austenite. Thus, X-ray diffraction was used to identify austenite, martensite and ferrite phases in the HAZ of the welded joint made by S-SAW and T-SAW. This technique identifies phases with at least 5% in volume fraction. X-ray diffraction was carried out in the CGHAZ regions, which according to image analysis had higher volume fractions of MA. The regions showed the presence of ferrite, austenite and martensite. Table 7 and 8 present the identification of the planes of each phase with higher intensity. The results of the diffraction patterns are depicted inFigure 10 and Figure 11.

Figure 9
Volume fraction of MA microconstituent present in the WM and regions of the HAZ for the two experiments, S-SAW and T-SAW.

Figure 10
Diffraction pattern for sample welded by S-SAW.

Figure 11
Diffraction pattern for sample welded by T-SAW.

The calculations of the crystal parameters of the phases found in the diffraction patterns are presented in Table 9, to verify the planes and phases identified in Table 7 and Table 8. The analysis by X-ray diffraction confirmed the presence of retained austenite and martensite phases in the experiments under study; in other words, the presence of MA microconstituent.

The volume fractions of the MA microconstituents for S-SAW and T-SAW, for the separate weld and HAZ regions, are shown in Figure 9. There is a larger quantity of MA in the CGHAZ for the two welding conditions, with the T-SAW sample having a fraction about 60% higher than the S-SAW sample. It is possible that the higher heat input in T-SAW is responsible for this result. Figure 12shows the heat input for each pass. The higher the heat input, the higher the peak temperature at a location in the HAZ. In addition, the cooling rate is slower, allowing more time at a given temperature in the HAZ. This increases the grain size of the regions which, together with carbide dissolution, increases the hardenability of the region in the HAZ. As a consequence, the M_sand M_f temperatures decrease, and more retained austenite is formed together with untempered martensite.

Figure 12
Heat input for each pass in S-SAW and T-SAW experiments.

3.2 Microhardness test

Figures 13 to 15 present a microhardness profiles for lines 1, 2 and 3, respectively, in Figure 2. Line 1, inFigure 13, corresponds to the filling pass and line 3, in Figure 15, is the finishing pass region. Figure 13 shows that the values of microhardness along line 1 in S-SAW are similar to line 1 in T-SAW. In both cases, reheating due to subsequent passes can produce a kind of a "localized heat treatment", keeping both hardness profiles almost the same.

Figure 13
Results of microhardness for line 1, as presented in Figure 2.

Figure 15
Results of microhardness for line 3, as presented in Figure 2.

By comparing S-SAW and T-SAW, it is possible to observe significant differences in microhardness values, in particular in the CGHAZ. Besides the heat input, the values of microhardness vary due to the different microstructures present in the welded joint and due to reheating thermal cycles during welding, which act as a kind of non-isothermal heat treatment.

Figure 14 shows the results of the Vickers microhardness measurements for line 2, the root pass for S-SAW and T-SAW. The difference in microhardness values are due to the peak temperature, which is a function of the heat input, (Figure 12).

Figure 14
Results of microhardness for line 2, as presented in Figure 2.

3.3 Charpy V notch impact test

The results obtained in the Charpy V notch impact test are depicted in Figure 16. An ANOVA analysis, with 95% confidence, showed no difference in weld metal absorbed energy, comparing samples from the two conditions (S-SAW and T-SAW). However, when comparing different regions of the welded joint, the weld metal had the lowest impact toughness. This deleterious behavior is due to an as-solidified weld metal with a coarse grain size and also the presence of microsegregation.¹⁴14 Svensson LE. Control of microstructures and properties in steel arc welds. Cambridge - U.K: Materials Science & Technology;1993. 233 p.

Figure 16
Results of Charpy V tests at 0°C for different regions of the welded joint; weld metal, 1 mm away from the fusion line and 5 mm from the fusion line.

According to the fracture images obtained through scanning electron microscopy, shown in Figure 17, it is possible to identify transgranular fractures and dimples. This is the typical appearance of specimens that suffered ductile fracture. The Charpy impact test, for a notch situated 5 mm away from fusion line (FL+5), offers the highest average value of absorbed energy (220 J) in the FG HAZ region for S-SAW. That is a zone of grain refinement, which underwent reheating by subsequent welding passes. The average grain size is 5.5 μm and the predominant microstructure consists of polygonal ferrite with a smaller fraction of pearlite. The same zone observed in T-SAW not only consists of polygonal ferrite and pearlite, but also a microstructure with zones of Widmanstätten ferrite (WF). The presence of those morphologies has contributed to a slight reduction in the amount of absorbed energy, which is about 10% lower than that for S-SAW.

Figure 17
(a) and (b) SEM fractographs of Charpy samples at 0°C of the weld metal for S-SAW; (c) and (d) SEM fractographs of Charpy fractographs at 0°C of the weld metal for T-SAW.

The results of the ANOVA analysis, with 95% confidence, presented no difference between the absorbed energy values in the Charpy impact test for the region situated 1 mm away from the fusion line towards the HAZ (FL+1) for both S-SAW and T-SAW. The same results were observed when comparing zones (FL+5) and (FL+ 1), i.e., both experiments presented no difference in absorbed energy values. These results can be confirmed when comparing the (FL+1) zone microstructure, which corresponds to the CGHAZ, to the (FL+5) zone microstructure, which corresponds to the FGHAZ. The microstructural changes for both experiments are virtually the same. It can be noticed by means of microstructural characterization that the (FL+1) zone for both experiments have MA constituents, which does not occur in the (LF+5) FGHAZ zone.

4 Conclusions

Based on the aforementioned experiments and their respective results, we may draw the following conclusions:

1. 1

   The microstructure of the HAZ for both experiments includes polygonal ferrite, grain boundary ferrite, ferrite with second aligned/non-aligned phase, ferrite with carbides and martensite/austenite microconstituents. In particular, the CGHAZ region in the T-SAW (tandem welding technique) has more MA greater than the same region in S-SAW (single wire condition).

2. 2

   The microhardness values for the regions analyzed fluctuate, not only because of the phases and inclusions present, but also because of the different heat input used in each pass. This is the case of the filling pass for T-SAW (line 3 in Figure 2) condition with a hardness of 240 HV. However, the heat input for the pass corresponding to line 3 in the T-SAW was 33% lower than that in single wire condition (S-SAW). The CGHAZ region in T-SAW was harder than the one in S-SAW.

3. 3

   The weld metal for both conditions reached an average absorbed energy value 63% lower then the energy of the grain refined and the base metal regions. In summary, two factors are associated with this deleterious behavior in the weld metal (WM). The first one is the microstructural characteristics of the weld metal, which consists AF, GF and FS, where the crack follows the grain boundaries of FG and spreads easily between FS plates separated by low angle. The second one is the fracture pattern. Fractography of the weld metal exhibit transgranular fractures and dimples.

4. 4

   According to the ANOVA analysis, the CGHAZ for T-SAW had the same absorbed energy as S-SAW, in the same notch position in the HAZ regions.

Acknowledgements

The authors wish to acknowledge funding provided by the Emerging Leaders in the Americas Program (ELAP) and the Natural Science and Engineering Research Council of Canada. The use of research facilities at EVRAZ Inc. NA as well as the supply of steel for this research from Usiminas Company at Cubatão - SP are also acknowledged. Finally, the assistance of EVRAZ Inc. NA staff, especially Dr. Laurie Collins, Muhammad Rashid, and Jon Jackson in carrying out some of the test work is also acknowledged.

References

Publication Dates

History