Tempering of continuous and pulse current GTA welds of AISI 420 (1.4021) martensitic stainless steel

Transcription

1 Tempering of continuous and pulse current GTA welds of AISI 420 (1.4021) martensitic stainless steel T. Iamboliev Technical University Sofia, Plovdiv Branch, Department of Manufacuring Engineering, Plovdiv, Bulgaria Keywords Martensitic stainless steel, GTA welding, microstructure, tempering 1. Introduction Martensitic stainless steels /MSS/ are a desired material for structural components for several industry branches, such as the gas, petroleum and oil industries, power generation, surgery tools, etc. This steel is known for its remarkable mechanical properties when used at elevated temperatures up to 600 C, as well as for its moderate corrosion resistance at ambient temperature [1], [2]. Unlike other stainless steels the MSS can be subjected to a heat treatment in order to obtain component properties relevant to an exploitation environment. Moreover, since the MSSs are Ni-free, they are relatively low cost materials compared to the conventional Cr-Ni-stainless steels. Chromium and carbon provide both a solid-solutionstrengthening and carbide-precipitation-strengthening. A MSS containing 13 % Cr and 0.15 % to 0.40 % C exhibits an extended austenite field in the Fe-Cr diagram. Because of this most of the hot-working processes provoke a phase transformation of austenite to martensite even by air cooling [3], [4]. This has been recognized when welding the MSS. The weld metal /WM/ as well as the heat affected zone /HAZ/ of the weld become hard and brittle due to a fully or partially martensitic structure, which is prone to cold cracking known also as a hydrogen induced cracking [5]-[9]. This is the main reason for generally considering the MSS as a non-weldable structural material. Nevertheless there are some attempts to obtain quality weld joints of MSS using electric arc and laser beam as a heat source [7], [10]-[13]. In order to prevent cold cracking, preheating and/or post-weld heat treatment /PWHT/ are frequently used [5]-[7]. It is recommended that the PWHT is carried out within the range 480 to 750 C for 30 to 120 min depending on section thickness. However narrower data to be used in the welding manufacturing are still missing. Recently, pulsed laser welding of AISI 420 steel has been applied, investigating the effect of the welding parameters on the weld size and microstructure. It is reported that both retaineddelta-ferrite and austenite as well as coarse carbides M 23 C 6 have been found within the martensitic WM matrix and the HAZ. A peak hardness value of 760 HV has been measured in the HAZ causing poor toughness and ductility of the weld. After 2 types of PWHT including austenizing at 1010 C for 0.5 hours followed by tempering within C for 2 hours or tempering only, it is concluded that optimum temper temperatures are 537 C and 595 C, respectively. It is found that the PWHT resulted in a reduction of strength and hardness as well as increase of toughness and ductility [10]. However, the indicated optimum values of the temper temperature have been specified by interpolation within a wide temperature range, assuming linear temperature-hardness relationship which has not been proven before. Therefore, the temperature accuracy is not beyond doubt. Moreover, there is a lack of evidence in this investigation for the stated change of the weld toughness and strength. In a study on the effect of preheating and PWHT on the laser weldability of AISI 420 martensitic stainless steel, pulse laser welds preheated for 60 min at 300 C and PWH-treated for 75 min at 660 C have been furnace cooled down to room temperature. An EDS analysis has demonstrated a reduction of the Cr-amount in the HAZ and the WM. This finding is related to a carbide (M 23 C 6, M 7 C 3 ) precipitation due to the PWHT [11]. In another work a pulsed Nd:YAG laser welding of AISI 304 to AISI 420 stainless steel has been investigated in order to determine the appropriate beam position with respect to the joint characteristics. It has been observed that shifting the beam in the direction of AISI 420 steel has resulted in an increasing amount of the WM martensite. In addition, the AISI 420 HAZ has shown the highest microhardness value for any laser beam position, and in the tensile test fracture has Ccurred outside the weld region [12]. An investigation on a pulsed Nd:YAG laser welding of AISI 420 stainless steel to kovar alloy has been carried out. At the AISI 420 side the WM has exhibited bcc and fcc crystal structures. High welding speed and low degree of pulse overlapping has increased the temperature gradient, developing coarse columnar grains at the AISI 420 fusion boundary. At lower welding speed fine equiaxed grains have been produced promoting a resistance to solidification cracking of the WM. The high microhardness in the AISI 420 HAZ has been attributed to over precipitation and coarsening of M 23 C 6 carbides in the ferrite grain boundary [13]. However, one could argue against this interpretation since carbide coarsening is known to facilitate the dislocation movement resulting in a hardness decrease. There is a growing interest in pulse welding. It is related to the significant advantages, including the possibility of precise heat input control, reduction of the residual stress, improvement of the weld microstructure and properties at lower energy consumption [14], [15]. Although the pulsed GTAW is much more accessible than the laser welding, there are few studies concerning MSS welding applying the GTAW low cost process. The effect of the constant- and pulsed-current 3

2 GTAW parameters on the microstructure and properties of the AISI 420 welds has been investigated using an experimental design approach [16]-[18]. A perfect weld appearance according to ISO 5817 and WM microstructure providing martensite, ferrite and carbide phases have been observed. The hardness of the WM and HAZ is reported to be as high as 780 to 790 HV10 providing brittleness and susceptibility to cold cracking [18]. Keeping in mind the above considerations, the goal of this work is to elaborate appropriate PWHT conditions for constant- and pulsed-current GTA welds of the AISI 420 MSS and to study the weld properties. 2. Experimental procedure Plates sized 200x70x1.6 mm of AISI 420 martensitic stainless steel with chemical composition as per Table 1 were used. Prior to welding the plates were chemically cleaned and secured in position using tack welds. In order to avoid joint distortion, the plates were securely clamped in a suitable device which was designed to ensure gas shielding of the weld bead root as well. Square butt joints of full penetration were autogenously welded applying both the GTAW and pulsed P-GTAW processes. A thoriated tungsten electrode of 2.4 mm diameter shaped at point angle 50 was used. The polarity was DCEN and the electrode-to-work distance was kept constant at 2 mm. Ar shielding gas 10 l/min was employed with a gas nozzle of diameter 12 mm. Square-wave pulse shape was used [17]. The values of the welding variables are displayed in Table 2. A heat efficiency η=0.55 was used for the heat input calculations. A power source for welding was Kempi Mastertig 2300 MLS TM ACDC. Table1. Chemical composition of AISI 420 stainless steel,. С Si Mn S P Cr Table 2. Welding variables [17]. Process I p I b t p /T f [Hz] I ave U [V] v [mm/s] q [J/mm] GTAW P-GTAW The welds were sectioned and samples sized 40x40x1,6 mm were subjected to a PWHT within the temperature range C for 15 to 150 min followed by air cooling as shown in Figure 1. The low temperature limit was chosen to be higher than the temperature M s = 300 C [7]. The temperature was measured by means of a thermocouple and recorded using a data acquisition system. After the PWHT a Vickers hardness profile was taken across the WM, HAZ and the base metal. Tensile test using 4 samples and Charpy impact test at room temperature using 3 samples of the welds and the base metal were carried out. The notch was placed in the WM. Then the samples were sectioned perpendicular to the welding direction in order to prepare metallographic specimens. The latter were ground by a disc polishing machine and subsequently polished using diamond paste with 1 µm sized particles. Etching of the specimens was done applying a HCl:HNO 3 =1:3 solution. Figure 1. Weld thermal cycle and temperature profiles of the post weld heat treatment. The microstructure of the WM, HAZ and base metal was evaluated using optical microscopy, SEM and EDX. 3. Results and discussion 3.1. Mechanical properties Hardness profile of the weld In the as welded condition, peak hardness values of 780 HV10 were established in the WM and the HAZ, whereas the hardness of the base metal was measured to be HV10 [18]. Vickers hardness profiles across the WM, HAZ and the base metal after PWHT are demonstrated in Figure 2. The highest hardness values up to 600 HV10 are measured in the WM and the HAZ, whereas a hardness decrease is established away from the HAZ falling to about 150 HV10 in the base metal. At PWHT temperature up to 550 C the WM and HAZ are neither ductile nor machineable irrespective of the holding time as obvious from Figure 2 a, b, but hard and brittle and susceptible to cold cracking. The upper hardness limit of 350 HV10 [19], as well as lower values of the WM and HAZ is achieved at temperatures of 650 C for 30 min and 700 C for any holding time, as illustrated in Figure 2, c, d. Within the hardness range HV10 there is a ductility and resistivity guarantee against cold cracking as well as machineability available as stated in [7], [8]. These results show that the maximum hardness of the WM and HAZ is first of all temperature-dependent and the effect of the holding time within the range 30 min to 150 min is weaker. Similar results are reported in previous works. Peak microhardness values of 620 HV0.1 in the WM and 780 HV0.1 in the HAZ in the as welded condition have been established. After PWHT at 600 C for 2 hours hardness of 355 HV10 in the WM and 350 to 250 HV10 in the HAZ have been measured, whereas PWHT at 650 C has been found to result in 320 HV10 WM hardness [10]. Preheating at 300 C for 60 min and PWHT at 660 C for 75 min has been shown to decrease the hardness of the WM and HAZ to 280 and 290 HV0.1, respectively [11] Tensile test The tensile test results of the welds after PWHT at different conditions are compared with those of the base metal in Table 3. Except for the GMAW sample treated at 600 C for 150 min the tensile strength of all the other welds is up to 106 % superior to that of the base metal. The tensile strength of the pulsed current 4

3 15 min. All the weld samples fractured in the base metal as shown in Figure 3, hence, covering the requirements of [19]. Table 3. Tensile test results after PWHT. Sample PWHT [ºC/min] R m [МPа] A Ψ Fracture location BM GTAW P-GTAW 600/ / / / / / / / Base metal Base metal Figure 3. Tensile test samples fractured in the base metal: a) GTAW, 600 С/150 min; b) P-GTAW 700 С/15 min Charpy impact test The Charpy impact strength value of the base metal samples was found to be 48 J. Table 4 displays the test results of the weld samples in the as-welded condition and after PWHT. The impact strength increases with the temperature, achieving 55 J at 700 C for 15min of both the GTAW and P-GTAW weld samples, confirming weld ductility. Table 4. Charpy test results, J. Figure 2. Hardness profile of welds after PWHT at different temperature: a) 380 ºC; b) 550 ºC; c) 650 ºC; d) 700 ºC; FL fusion line welds was about 5 % higher than the strength of the constant current welds. The elongation of the weld samples was 153 % to 169 % compared to the one of the base metal. An exception to this is provided by a GTAW sample treated at 700 C for 15 min and a P-GTAW sample treated at 600 C for 150 min with an elongation of the base metal of 85 %. A transverse contraction of the welds was measured, which was % higher than that of the base metal. In conclusion, the best strength and ductility were performed by the P-GTAW sample treated at 700 C for Sample Aswelded After PWHT at 450 ºС/60 min 600 ºС/60 min 700 ºС/15 min GTAW P-GTAW Microstructure A common view of the P-GTAW WM is shown in Figure 4. A full penetration weld with a smooth appearance and a lack of weld imperfections is obvious. The WM microstructure in the as-welded condition is built of martensite, both vermicular and interdendritic types of residual ferrite, a small amount of retained austenite and (Cr, Fe) 23 C 6 carbides as shown in Figure 5, a [18]. This phase morphology of the WM is a result of primary δ-ferrite solidification, followed 5

4 by eutectic deposition of δ + γ crystals from the melt in the last stage of solidification [7]. The eutectic reaction is assisted by Figure 4. Macrostructure of a P-GTA weld metal obtained under 76 J/mm heat input. the significant carbon content of 0.26 % C, which might give rise to a C enrichment in the space between the primary δ-ferrite dendrites. An epitaxial growth of cellular dendrites at the fusion line and columnar dendrite growth at the weld axis were identified as a response to the decreasing temperature gradient during solidification in the WM [8]. On cooling, the austenite transforms into martensite due to the high rate of cooling in the WM and the HAZ. respectively. The primary outlines of the dendrites are partially lost and the amount of carbides has grown, as seen in Figure 5, b. A further increase of the tempering temperature to 700 C provides a less distinctive morphology as exhibited in Figure 5, c. It is a result of the martensite decomposition leading to deposition of a mix of fine carbides and ferrite phase in the matrix. Based on this, the degree of tetragonality of the martensite crystal lattice decreases, followed by internal stress relief. The deposited carbides coagulate and become spheroidised in order to minimize their surface area. In the high alloyed steels, having a high content of carbide formers, such as Cr, the coagulation takes place at temperatures above 450 C [20]. The martensite decomposition leads to decreasing britlleness and hardness as observed in the impact test and in Figure 2, c-d, as well as to increasing toughness and ductility, hence, excluding the risk of cold cracking in the WM and HAZ. In the sample tempered at 700 C for 15 min there was a light spot phase detected along the fusion line as seen in Figure 5, c. According to the results of the X-ray analysis, the composition of the light phase is identical with that of the WM. A phase of similar morphology has been studied earlier by means of X-ray diffraction analysis. It is reported that the light phase is an untempered martensite originating from the retained austenite prior to the PWHT [7]. Based on the weld mechanical properties obtained in this investigation, it seems that the untempered martensite has no significant adverse effect on the ductility and toughness of the welds, which are superior to those of the base metal. The fracture surface of the samples after the tensile test is illustrated in Figure 6. Tested prior to welding, the base metal exhibited a ductile fracture. In the as welded condition fracture there are no dimples but smooth surface regions shown by Figure 5. Microstructure of a P-GTAW sample welded with 76 J/mm heat input: a) as-welded condition, fusion line; b) weld metal after PWHT at 550 ºС/30 min; c) fusion line after PWHT at 700 ºС/15 min; d) EDX-ray spectrum of the light spots in c). The carbides most frequently outline the grain boundary, however, a smaller amount of carbides is dispersed within the WM grains as well. The carbides precipitate following a decrease of carbon solubility in both the austenite and ferrite below 1000 C. This carbide precipitation takes place in the grain interior or at the grain boundary, leaving chromium depleted regions around. These regions seem to be responsible for the moderate corrosion resistance of the martensitic stainless steel welds. There is a coarse grain region at the fusion line corresponding to the upper temperature range of the HAZ where carbides are difficult to be seen because of their dissolution into the austenite. Adjacent to the coarse grain region there is a fine grain region situated in the lower HAZ temperature range. The WM microstructure after PWHT at 550 C for 30 min as well as at 700 C for 15 min is displayed in Figure 5, b and c, Figure 6. Fracture surface, x1600: a) base metal; b) P-GTAW weld metal in the as welded condition; c) P-GTAW weld metal after PWHT at 700 ºС/15 min; d) GTAW weld metal after PWHT at 700 ºС/15 min. the arrows in Figure 6, b, thus confirming completely brittle WM structure. After PWHT both the GTA and P-GTA WM demonstrate dimpled fracture surface as a consequence of the increased ductility following the martensite decomposition. In addition, the fracture surface of the P-GTA WM seems to be superior to that of the GTA WM since there are some smooth 6

5 surface spots visible in Figure 6, d, seeming to represent less ductile metal. Nevertheless it is obvious that their effect on the ductility values obtained in the mechanical test results is negligible Energy consumption Based on the results of this investigation the following values of the PWHT parameters can be applied for tempering of both the GTA and P-GTA welds: 600 C/150 min, 650 C/60 min, 680 C/30 min, 700 C/15 min. The energy consumption of the furnace used for each PWHT condition is displayed in Table 5. The lowest energy consumption is kwh and corresponds to tempering at 700 C/15 min. Table 5. Energy consumption for PWHT, kwh. Temperature, [ C] Time [min] Energy [kwh] Conclusions 1. The hardness limit of 350 HV10 of the weld metal and HAZ, enabling ductility as seen from fracture surface, as well as machineability, is achieved applying the following PWHT conditions: 600 C/150 min, 650 C/60 min and 680 C/30 min. 2. Hardness in the range HV 10 of the weld metal and HAZ can be obtained under the PWHT conditions: 650oC/150 min and 700 C/15 min. 3. The tensile test revealed that the welds tempered at 650 C/60 min and 700 C/15 min performed a strength, elongation and transverse contraction which are 5 %, % and up to 204 % higher than those of the base metal, respectively. Fracture occurred in the base metal. 4. After PWHT at 700 C/15 min the impact strength of the welds is superior to that of the base metal. 5. The as-welded weld metal microstructure is built of martensite, retained vermicular and interdendritic ferrite, retained austenite and carbides. After PWHT the martensite matrix decomposes, leaving a ferrite-carbide mix in the weld metal and HAZ. It is likely, that some untempered martensite is available along the fusion line without any adverse effect on the strength and ductility. 6. After PWHT at 650 C/60 min, 680 C/30 min and 700 C/15 min both the GTA and P-GTA weld metal displayed dimpled fracture surfaces, confirming sufficient ductility. 7. The properties of the full penetration GTA welds and the P-GTA welds seem to be equal. The energy consumption used for the P-GTA welds is 12 % less than that used for the conventional constant current GTAW weds. 8. The lowest energy consumption among all the PWHT conditions ensuring a hardness of the weld metal and HAZ equal or below 350 HV10 is kwh corresponding to tempering at 700 C/15 min. Acknowledgements The author would like to express his gratitude to Ass. Prof. T. Petrov, Mrs. I. Borisova and Mr. S. Yakimov for taking part in the experiments. The financial support of the Research Fund of the TU Sofia under contract 102ni051-24/2010 is highly appreciated. References [1]. Heuser, H., Jochum C., Bendick W., Hahn B., Welding of dissimilar joints of new power plant steels. Int. congress Safety and Reliability of Welded Components in Energy and Processing Industry, Graz, 2008, pp [2]. Vekeman, J., Huysmans S., Cold weld repair of T91. Int. congress Safety and Reliability of Welded Components in Energy and Processing Industry, Graz, 2008, pp [3]. Balevski, A, Metaloznanie /Physical metallurgy/, Sofia, Tehnika, 1972 /in Bulg/. [4]. Guliaev, A. Metallovedenie /Physical metallurgy/, Moskau, Mashinostroenie, 1977 /in Russian/. [5]. Velkov, K. Tehnologia na zavariavaneto /Welding technology/. Sofia, VMEI, 1987 /in Bulg/. [6]. Zhelev, A., Materialoznanie. Tehnika i tehnologia. T. 2: Tehnologichni procesi i obrabotvaemost /Physical metallurgy and processability of materials/. Sofia, Bulvest, 2003 /in Bulg/. [7]. Lippold, J., Kotecki, D., Welding metallurgy and weldability of stainless steels, John Willey & Sons, Inc., Hoboken, NJ, [8]. Kou S., Welding metallurgy, John Wiley & Sons, Madison, Wisconsin, 2nd ed., [9]. Hristov, S., Cold cracks in welding and precautions. Sofia, Prof. M. Drinov, 2011 /in Bulg/. [10]. Baghjari, S. H., AkbariMousavi, S.A.A. Effects of pulsed Nd:YAG laser welding parameters and subsequent post-weld heat treatment on microstructure and hardness of AISI 420 stainless steel. Materials and Design 43 (2013), pp [11]. Kose, C., Kacar R., The effect of preheat & post weld heat treatment on the laser weldability of AISI 420 martensitic stainless steel. Materials and Design 64 (2014), pp [12]. Berretta, J. R., Rossi, W., Neves, D. M., Almeida, I., Vieira, N. D. Jr., Pulsed Nd:YAG laser welding of AISI 304 to AISI 420 stainless steels. Optics and Lasers in Engineering 45 (2007), pp [13]. Baghjari, S. H., Akbari-Mousavi, S.A.A., Experimental investigation on dissimilar pulsed Nd:YAG laser welding of AISI 420 stainless steel to kovar alloy. Materials and Design 57 (2014), pp [14]. Becker, D., C. Adams, Jr., The role of pulsed GTA welding variables in solidification and grain refinement. Welding Journal, 1979, 5, pp. 143-s 152-s. [15]. Tseng, K., Chou C., The effect of pulsed GTA welding on the residual stress of a stainless steel weldment. Journal of Materials Processing Technology, 123, 2002, pp [16]. Iamboliev. T., GTA welding of martensitic stainless steel X26Cr13. Mashinostroene i mashinoznanie, 18, 2013, 1, pp /in Bulg./. [17]. Iamboliev, Т., Design experiment of pulsed GTA welding of martensitic stainless steel X26Cr13, Mashinostroene i mashinoznanie, 18, 2013, 1, pp /in Bulg./. [18]. Iamboliev, T., Microstructure analyses of weld joints of martensitic steel Х26Cr13. Proc. of the 3rd nat. conf. MHANS, IMS-BAN, Sofia, , pp /in Bulg./. [19]. ISO 15614:2011 Specification and qualification of welded procedures for metallic materials. [20]. Rashkov, N., Termichna obrabotka na stomanite /Heat treatment of steels/, Sofia, Tehnika, 1972 /in Bulg./. Copyright IIW rd IIW SEENET Int. Congress Reproduced with kind permission of the IIW. 7