Influence of MC Carbides on Pitting Corrosion Resistance of Weld Metal in Austenitic Stainless Steels

Abstract

The influence of MC-type carbide formation on pitting corrosion resistance in weld metal of austenitic stainless steel was investigated. The relationship between the microstructure such as carbides and element distribution, and pitting corrosion resistance of the simulated weld metal of austenitic stainless steel was studied. The addition of molybdenum improved the pitting corrosion resistance. The effect of niobium addition on the pitting resistance was negligible. However, the addition of titanium significantly reduced the pitting corrosion resistance. The addition of niobium or titanium induced the formation of MC-type carbides, such as TiC and NbC, at the cellular boundaries. The pits were mainly initiated near or at carbides. The chromium depletion zone was formed near M₂₃C₆ coexisting with TiC only in the specimen added titanium. Thus, TiC formed during solidification accelerated the chromium diffusion-associated M₂₃C₆ precipitation on TiC. The depletion zone deteriorated the pitting corrosion resistance of the titanium-containing specimen.

1. Introduction

Deterioration of the corrosion resistance in weld metals remains a significant challenge in austenitic stainless steels. Many studies have been conducted on the corrosion phenomena in weld heat-affected zones (HAZ), such as weld decay.¹⁾ Thus, several methods including reducing carbon content (<0.03%), incorporating niobium or titanium in base metal, using low heat input (high cooling rate), and employing solution treatment at 1050–1100°C for the decomposition of chromium carbide (such as, M₂₃C₆) have been proposed to improve the corrosion resistance of the HAZ and prevent the formation of chromium carbides. However, there are a limited number of studies on pitting corrosion resistance in the weld metal of austenitic stainless steels, even though the corrosion resistance of the weld metal tends to be worse than that of the base metal.^2,3,4)

Pitting mainly occurs at the centre of dendrites or at the interface between delta ferrite and austenite phases in weld metals. The former is caused by the depletion of chromium and molybdenum owing to solidification segregation, and the latter is caused by the formation of a chromium-depleted region around chromium carbide and the segregation of impurity elements at the interface. Thus, microstructure morphology strongly affects the pitting corrosion resistance of the weld metal in austenitic stainless steel.^5,6,7)

Alloying elements have often been added to improve the pitting corrosion resistance. Chromium and molybdenum are well-known elements used for improving corrosion resistance. However, these elements tend to segregate during solidification owing to their low partition coefficients in austenite phase.^8,9) The solidification segregation in weld metals induces pitting corrosion in fully austenitic stainless steel.^10,11) Niobium and titanium are often added to avoid the formation of chromium carbide (M₂₃C₆) because these elements preferentially form MC-type carbides (TiC and NbC) during solidification owing to their higher affinities with carbon than chromium.^12,13,14,15) In addition, Laves phases form in addition to MC-type carbides when the carbon content is low.^9,14) These secondary phases may induce the formation of chromium-depleted regions and microscopic gaps, which may influence on pitting corrosion resistance of stainless steels. The effects on NbC formation on pitting corrosion resistance have been well reported.^16,17,18) However, the influence of TiC on pitting corrosion resistance has been still under the discussion even though titanium is often added for stabilization austenitic stainless steel as same as niobium. Besides, the formation behaviour of various type of carbide such as MC and M₂₃C₆ by the addition of niobium or titanium in weld metal during welding has not been investigated in detail. The influence of type of MC-carbide on pitting corrosion resistance of weld metals with changing the chemical compositions of austenitic stainless steel is also not clear. Besides, dissimilar welding using austenitic stainless steel and nickel-based alloy will increase for high corrosion resistance and high-performance products. Thus, niobium and titanium, which are the alloying elements of these materials, are mixed into the dissimilar weld metal. It is necessary to understand the formation of carbides associated with the elements and their effects on the pitting corrosion resistance of the weld metal.

In this study, the influence of MC-type carbide formation on pitting corrosion resistance of a fully austenitic stainless steel weld metal was investigated. The Fe-18mass%Cr-22mass%Ni (hereafter mass% describes as%) alloy was employed to eliminate the influence of ferrite formation. Niobium, titanium and molybdenum were added as alloying elements. The relationship between the microstructure morphology such as carbides and element distribution, and pitting corrosion resistance of the weld metal of austenitic stainless steel was studied.

2. Materials and Methods

2.1. Specimen Preparation and Microstructural Observation

Fe-18%Cr-22%Ni (18Cr-22Ni) stainless steel samples with varying amounts of niobium, titanium and molybdenum were employed to obtain fully austenitic solidification (to avoid ferrite formation during solidification). To modify the secondary phase formation, 0.5% niobium or 0.3% titanium were added as alloying elements. Additionally, 3% molybdenum was added in addition to niobium and titanium to modify the pitting resistance equivalent (PRE =%Cr + 3.3×%Mo+ 16×%N) value. The specimens were fabricated by gas tungsten arc (GTA) button melting in order to simulate the welding thermal cycle of weld metal as shown in Fig. 1. 30 g of the base metal (Type 304L) and raw materials (chromium, nickel, niobium, titanium and carbon) with different weights were melted in the copper mould (φ32 × 20 mm) by GTA arc with a current of 250 A and argon gas shielding. The cooling rate during GTA button melting under the conditions were nearly the same as the rate in weld metal during GTA welding. The chemical compositions of the specimens are shown in Table 1.

Fig. 1. Schematic illustration of GTA button melting for specimen fabrication. (Online version in color.)

Table 1. Chemical compositions of specimens (mass%).

	C	Si	Mn	P	Cr	Ni	Nb	Ti	Mo	PRE
18Cr-22Ni	0.05	0.390	1.31	0.018	18.7	22.8	―	―	0.177	19.3
3Mo	0.05	0.364	1.19	0.019	18.2	23.8	―	―	3.32	29.2
0.5Nb	0.05	0.359	1.26	0.031	18.1	22.1	0.553	―	0.156	18.6
0.5Nb-3Mo	0.05	0.342	1.10	0.021	17.3	22.8	0.559	―	3.36	28.4
0.3Ti	0.05	0.398	1.20	0.025	17.6	22.1	―	0.314	0.161	18.1
0.3Ti-3Mo	0.05	0.386	1.13	0.023	18.3	22.6	―	0.253	3.15	28.7

The cross-sectional microstructures of the specimens were observed by scanning electron microscopy (SEM; Hitachi SU-1510). The specimen surface was treated for the observation by electrochemical etching using a 10% oxalic acid solution (50°C, 3 V, 10 s) after final polishing with colloidal silica. Electron probe microanalysis (EPMA; JEOL JXA-8530F) was performed to investigate the elemental distribution. Additionally, transmission electron microscopy (TEM; JEOL JEM-2100F) and energy dispersive X-ray spectroscopy (EDS) were performed to identify the detailed distribution of the secondary phases and elements. The specimens for TEM observation were prepared via an extraction replica method using an evaporated carbon film and a focused ion beam (FIB).¹⁹⁾

2.2. Evaluation of Pitting Corrosion Resistance

Pitting potential measurements and ferric chloride tests were conducted to evaluate the pitting corrosion resistance.^20,21) For the pitting potential measurement (according to JIS G 0577), 3.5% NaCl was used as the test solution at a controlled temperature of 30 ± 1°C. The solution was degassed by purging with N₂ gas for an hour prior to testing. A platinum electrode was used as the counter electrode, and an Ag/AgCl electrode (KCl sat.) was used as the reference electrode. The specimen was cut to dimensions of 12 mm × 12 mm × 5 mm, and the surface was ground using a #600 abrasive paper before the test. The specimen was immersed in the solution at open circuit potential for 10 min. Anodic polarisation was performed at a scan rate of 20 mV/min until the current density reached 500 μA/cm². The potential at 100 μA/cm² in the anodic polarisation curve was defined as the pitting corrosion potential based on JIS G 0577. The potential was measured four times for each specimen. For the ferric chloride test (according to ASTM G48), 6 mass% FeCl₃·6H₂O + N/20 HCl was used as the test solution at a controlled temperature of 50 ± 1°C. The specimen geometry and pre-treatment were the same as those for the pitting potential measurement. The weight loss after immersion in the solution for 24 h was measured as the evaluation index of the pitting corrosion rate.

3. Results and Discussion

3.1. Microstructural Observation

Figure 2 shows the SEM images of the cross-sectional microstructure of the specimens before the corrosion test. The cell structure of the austenite phase is mainly observed in 18Cr-22Ni. Microconstituents such as secondary phase are observed at the cellular boundaries in the specimens, including molybdenum, niobium, and titanium. Lamellar-type microconstituents are observed at the cellular boundaries of 3Mo. In the specimen containing 0.5% Nb, microconstituents in the form of film-type phases are observed whereas in the 0.5Nb-3Mo specimen, both film- and lamellar-type phases are observed at the cellular boundaries. However, in the specimen containing 0.3% Ti, film-type microconstituents at the cellular boundary and blocky-type microconstituents within the primary austenite phase are observed. The geometries of the constituents depend on the type and content of alloy elements. MnS and oxides were not observed. Molybdenum induces the formation of lamellar-type constituents, as shown in the cross sectional SEM images of 0.5Nb-3Mo and 0.3Ti-3Mo.

Fig. 2. SEM images of cross sectional microstructure of specimens fabricated by GTA button melting. (Online version in color.)

The elemental distributions of the specimens were analysed by the EPMA as shown in Fig. 3. Iron, chromium, and nickel segregated within the cells or cellular boundaries in 18Cr-22Ni. In the case of 3Mo, molybdenum, carbon, and chromium are strongly distributed in identical areas at the cell boundary. Thus, these phases are identified as carbides. Niobium and carbon are concentrated in identical areas of the film-type secondary phase in 0.5Nb. Thus, the phase is identified as NbC. However, the film-like constituents are characterized as carbides containing chromium and titanium because of the detection of these elements in 0.3Ti. Titanium and nitrogen are also detected in the blocky phase.

Fig. 3. EPMA results of specimens. (Online version in color.)

TEM-EDS observations and analysis of the diffraction pattern for each specimen based on the extracted replica method were performed to identify the secondary phases. The TEM-EDS results and diffraction patterns of 0.3Ti are shown in Fig. 4. Titanium and nitrogen are detected in the area indicated by Arrow 1. In addition, the diffraction pattern of the points reveals that the phase to be TiN. Chromium and iron are detected in the area indicated by Arrow 2, and the pattern reveals an M₂₃C₆ phase. In addition, Ti is distributed near the M₂₃C₆ phase. The phases and the position in each specimen which were identified by the same analysis are listed in Table 2. The Cr–Mo carbide formed at the cell boundary in 3Mo was identified as the secondary phase. In 0.5Nb, the film-like secondary phase formed at the cell boundary was identified as NbC. An additional molybdenum-containing phase was induced to form Cr–Mo carbide in addition to NbC in 0.5Nb-3Mo. However, in 0.3Ti, M₂₃C₆ and TiC formed at the cell boundary and the blocky TiN formed within the primary austenite phase. In 0.3Ti-3Mo, Cr–Mo carbide was identified in addition to M₂₃C₆ and TiC and TiN.

Fig. 4. TEM-EDS results and diffraction patterns for 0.3Ti. (Online version in color.)

Table 2. Type of microconstituent and the position identified by TEM.

	Phase	Position
18Cr-22Ni	＼	＼
3Mo	Cr–Mo carbide	Cell boundary
0.5Nb	NbC	Cell boundary
0.5Nb-3Mo	NbC, Cr–Mo carbide	Cell boundary
0.3Ti	M23C6, TiC	Cell boundary
	TiN	Intercell
0.3Ti-3Mo	M23C6, TiC, Cr–Mo carbide	Cell boundary
	TiN	Intercell

3.2. Pitting Corrosion Resistance

The pitting potential of each specimen was measured. Figure 5 shows the anodic polarisation curves during the pitting potential measurement. The current density of each specimen increases significantly after remaining below the passive current density with the increasing potential. The pitting potentials at a current density of 100 μA/cm² are 186 mV for 18Cr-22Ni, 465 mV for 3Mo, 202 mV for 0.5Nb, 603 mV for 0.5Nb-3Mo, 89 mV for 0.3Ti and 161 mV for 0.3Ti-3Mo.

Fig. 5. Anodic polarisation curve of specimens. (Online version in color.)

SEM images of pitting after pitting potential measurement is shown in Fig. 6. Pitting corrosions are mainly found on the specimen surface in each specimen. The size of the pit in 0.3Ti-3Mo is relatively large and deep compared to 3Mo and 0.5Nb-3Mo. Besides, crevice corrosions were not found in all specimens. Thus, the increase of the current density must be caused by the pitting initiation.

Fig. 6. SEM images of pitting after pitting potential measurement.

The pitting potential of each specimen, which indicates the index for pitting initiation, was measured and investigated using the PRE values, as shown in Fig. 7. The increase in the PRE value due to the addition of molybdenum increases the potential of the specimens containing each alloying element. This tendency is similar to the previous results.²²⁾ The potentials of the specimens containing niobium, 0.5Nb and 0.5Nb-3Mo, are higher than those of 18Cr-22Ni. In contrast, the potentials of the specimens containing titanium, 0.3Ti and 0.3Ti-3Mo, shown in blue plots are considerably smaller than those of the other specimens. In addition, the increase in potential by molybdenum addition is also small compared with those of the other specimens.

Fig. 7. Relationship between pitting potential and PRE. (Online version in color.)

For further investigation, the corrosion rate, which indicates the index for the initiation and growth of pitting, was measured after the ferric chloride test, as shown in Fig. 8. Pitting corrosions were mainly found on the specimen surface in each specimen. The corrosion rates of 18Cr-22Ni, 3Mo, 0.5Nb and 0.5Nb-3Mo are nearly the same at approximately 9 g/m²·h. This indicates that the growth rates of pitting are similar in these specimens, even though there is a large difference in the pitting potentials corresponding to the PRE value. In the case of specimens containing titanium, the corrosion rates are significantly higher (approximately 3.5 times) than those of the other specimens. The results of the pitting potential and corrosion rate reveal that the addition of titanium induces the deterioration of the pitting corrosion resistance.

Fig. 8. Corrosion rate measured by ferric chloride test. (Online version in color.)

3.3. Pitting Corrosion Behaviour

The initiation point of the pit in each specimen was observed using SEM as shown in Fig. 9. The specimens except for 3Mo and 0.5Nb-3Mo were prepared such that potential scanning was stopped as soon as the current density reached 100 μA/cm² during the pitting potential measurement. The specimens 3Mo and 0.5Nb-3Mo were also prepared by immersion for 600 to 1200 s of the ferric chloride test because the occurrence of pitting by the electrochemical test was not enough to investigate the initiation. The pit initiates in the primary austenite phase of 18Cr-22Ni. However, the pits mainly initiate near the film-like secondary phase NbC in 0.5Nb and M₂₃C₆ in 0.3Ti, which exist along cellular boundaries. The same tendency is observed in the specimens containing molybdenum, such as 0.5Nb-3Mo and 0.3Ti-3Mo. As described in Figs. 7 and 8, the addition of titanium deteriorated the pitting corrosion properties, such as the pitting potential and corrosion rate, compared with niobium. Thus, the type of secondary phase corresponding to alloying elements such as niobium and titanium is attributed to the deterioration of the pitting corrosion resistance.

Fig. 9. SEM images of pitting initiation in each specimen after pitting potential measurement. Arrows indicate pitting. (Online version in color.)

TEM-EDS analysis using the FIB specimen was conducted for a more detailed investigation of the secondary phases of NbC in 0.5Nb, and TiC and M₂₃C₆ in 0.3Ti. Figure 10 shows the bright-field images and element distributions. The NbC phase exists independently in 0.5 Nb. Niobium and carbon are concentrated in the carbide. However, in 0.3Ti, both the carbides of TiC (red arrow in Fig. 10(b)) and M₂₃C₆ (blue arrow in Fig. 10(b)) are observed in the film-like microconstituents. Because titanium and chromium are distributed in each carbide, they form independently and exist consecutively. Line EDS analysis of the cross direction of each secondary phase (carbide) was performed to investigate chromium distribution near the carbides. The chromium concentration in the austenite matrix phase is 18 to 20%, and the change in the concentration is small in the vicinity of NbC in 0.5Nb. In contrast, in the vicinity of TiC and M₂₃C₆, the chromium concentration in the austenite matrix phase decreases near the carbide phases. This tendency is more significant on the side of M₂₃C₆, and the concentration drops locally by up to 12%. Therefore, the chromium depletion zone is formed by the formation of M₂₃C₆ via the diffusion of chromium. In addition, the depletion zone formed near M₂₃C₆ in 0.3Ti induces the deterioration of the pitting corrosion resistance. The chromium concentration in the TiC region indicates high as shown in line analysis results (Fig. 10(b)). It is considered that M₂₃C₆ phase existed at the backside of the TiC was detected because the TiC phase is tiny and thin.

Fig. 10. TEM bright field images and the element distributions. (Online version in color.)

TiC and NbC phases mainly form during solidification, and M₂₃C₆ precipitates during the solid state after solidification.^13,14,23) Besides, this tendency was confirmed by thermodynamics simulation results using Themo-Calc software in the chemical compositions of 0.3Ti and 0.5Nb. Thus, the precipitation of M₂₃C₆ during solid state (650–850°C) tends to induce the formation of a chromium depletion zone because M₂₃C₆ grows with the diffusion of chromium. TEM observations indicate that TiC coexists with M₂₃C₆ and NbC exist individually, as shown in Fig. 9. Chromium depletion occurred in the austenite phase near M₂₃C₆ even though the diffusion rate of chromium in the austenite matrix phase was small in 0.3Ti. Thus, the presence of TiC accelerates the formation of the chromium depletion zone, corresponding to the precipitation of M₂₃C₆ on TiC. In contrast, NbC exists as single phase and M₂₃C₆ and the chromium depletion zone are hardly observed in 0.5Nb. The addition of titanium deteriorates the pitting corrosion resistance compared to niobium, as shown in Figs. 7 and 8.

Kesternich et al. have reported that the coprecipitation of M₂₃C₆ on primary TiC is caused by radiation-induced carbon segregation in Alloy1.4970.²⁴⁾ However, a clear trigger for the heterogeneous nucleation was not observed. TiC is also known as a heterogeneous nucleus for M₇C₃ precipitation because of a low lattice misfit of 10% between (110)_TiC and (010)_M7C3.^25,26) Both NbC and TiC have face-centred cubic structures (NaCl model). The lattice constants of TiC and NbC are 0.432 and 0.447 nm, respectively, and the difference is small.^25,27) M₂₃C₆ has a face-centred cubic structure with lattice constants of 1.0599 to 1.06599 nm.²⁸⁾ The lattice constant of M₂₃C₆ is more than twice than those of TiC and NbC. Thus, it is difficult to reveal the difference in the advantage of acting as heterogeneous nuclei and the apparent orientation relationship between MC and M₂₃C₆.

Titanium is often applied as a typical alloying element for austenitic stainless steels to modify the sensitisation in weld heat affected zone (weld decay) owing to its higher affinity for carbon than chromium. However, to undergo the solidification during welding process in weld metal of austenitic stainless steel containing titanium, induces TiC formation during weld solidification. Besides, TiC promote the precipitation of M₂₃C₆ during solid state and attribute to the sensitization as shown in Figs. 7, 8 and 10. This tendency did not confirmed in the austenitic stainless steel containing niobium even though niobium also accelerate NbC formation during solidification in weld metal. Therefore, a detailed mechanism of the heterogeneous nucleation of M₂₃C₆ from TiC is required to improve the pitting corrosion resistance in the weld metal of austenitic stainless steel with added titanium.

4. Conclusions

The influence of MC-type carbide formation on pitting corrosion resistance of a fully austenitic stainless steel weld metal was investigated using different alloying elements as niobium and titanium. The pitting potential of 0.5Nb was equivalent to that of 18Cr-22Ni, and pitting potential of 0.5Nb-3Mo increased significantly. However, the pitting potentials of 0.3Ti and 0.3Ti-3Mo decreased. Thus, the addition of molybdenum improved the pitting corrosion resistance. The effect of niobium addition on the pitting resistance was negligible. However, the addition of titanium significantly reduced the pitting corrosion resistance. The addition of niobium, titanium and molybdenum induced the formation of MC-type carbides, such as TiC and NbC, at the cellular boundaries of the primary austenite phases. Pitting was mainly initiated near or at carbides. Comparing 0.5Nb and 0.3Ti, a chromium depletion zone was formed near M₂₃C₆ only in 0.3Ti. Thus, TiC accelerated the chromium diffusion-associated M₂₃C₆ precipitation on TiC. The depletion zone deteriorated the pitting corrosion resistance of the titanium-containing specimen.

Statement for Conflict of Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

This work was performed under the joint researcher program of the Joint Usage/Research Center on Joining and Welding, Osaka University.

References

• 1)  Corrosion of Weldments, ed. by J. R. Davis, ASM International, (2006), 47. ISBN: 978-1627083591
• 2)   C.  Garcia,  F.  Martin,  P.  De Tiedra,  Y.  Blanco and  M.  Lopez: Corros. Sci., 50 (2008), 1184. https://doi.org/10.1016/j.corsci.2007.11.028
• 3)   C.  Garcia,  M. P.  De Tiedra,  Y.  Blanco,  O.  Martin and  F.  Martin: Corros. Sci., 50 (2008), 2390. https://doi.org/10.1016/j.corsci.2008.06.016
• 4)   M.  Dadfar,  M. H.  Fathi,  F.  Karimzadeh,  M. R.  Dadfar and  A.  Saatchi: Mater. Lett., 61 (2007), 2343. https://doi.org/10.1016/j.matlet.2006.09.008
• 5)   Y.  Cui and  C. D.  Lundin: Mater. Lett., 59 (2005), 1542. https://doi.org/10.1016/j.matdes.2005.05.022
• 6)   Y.  Cui and  C. D.  Lundin: Mater. Des., 28 (2007), 324. https://doi.org/10.1016/j.matdes.2005.05.022
• 7)   E.  Zumelzu,  J.  Sepulveda and  M.  Ibarra: J. Mater. Process. Technol., 94 (1999), 36. https://doi.org/10.1016/S0924-0136(98)00450-6
• 8)   S.  Kou: Welding Metallurgy, John Wiley & Sons, New Jersey, (2003), 243. ISBN: 978-1119524816
• 9)   J. C.  Lippold and  D.  Kotecki: Welding Metallurgy and Weldability of Stainless Steels, John Wiley & Sons, New Jersey, (2005), 208. ISBN: 978-0471473794
• 10)   T. G.  Gooch: Weld. Res., 75 (1996), 135s.
• 11)   P. I.  Marshall and  T. G.  Gooch: Corros., 49 (1993), 514. https://doi.org/10.5006/1.3316080
• 12)   K.  Kadoi,  S.  Ueda,  S.  Tokita and  H.  Inoue: J. Alloy. Compd., 28 (2020), 154423. https://doi.org/10.1016/j.jallcom.2020.154423
• 13)   J. N.  Dupont,  C. V.  Robino and  A. R.  Marder: Weld. J., 77 (1998), 417s.
• 14)   J. N.  DuPont,  C. V.  Robino and  A. R.  Marder: Sci. Technol. Weld. Join., 4 (1999), 1. https://doi.org/10.1179/stw.1999.4.1.1
• 15)   K.  Kadoi and  K.  Shinozaki: Metall. Mater. Trans. A, 48 (2017), 5860. https://doi.org/10.1007/s11661-017-4340-2
• 16)   A. F.  Padilha,  I. F.  Machado and  R. L.  Plaut: J. Mater. Process. Technol., 170 (2005), 89. https://doi.org/10.1016/j.jmatprotec.2005.05.002
• 17)   G.  Bai,  S.  Lu,  D.  Li and  Y.  Li: Corros. Sci., 108 (2016), 111. https://doi.org/10.1016/j.corsci.2016.03.009
• 18)   A. S.  Hamdy,  E.  El-Shenawy and  T.  El-Bitar: Mater. Lett., 61 (2001), 2827. https://doi.org/10.1016/j.matlet.2006.10.043
• 19)   T.  Akutagawa,  I.  Uchiyama and  A.  Fukami: Tetsu-to-Hagané, 43 (1957), 1222 (in Japanese). https://doi.org/10.2355/tetsutohagazne1955.43.11_1222
• 20)   N. J.  Laycock and  R. C.  Newman: Corros. Sci., 40 (1998), 887. https://doi.org/10.1016/S0010-938X(98)00020-1
• 21)   P. C.  Pistorius and  G. T.  Burstein: Corros. Sci., 36 (1994), 525. https://doi.org/10.1016/0010-938X(94)90041-8
• 22)   T.  Ogawa,  S.  Aoki,  T.  Sakamoto and  T.  Zaizen: Weld. J., 61 (1982), 139s.
• 23)   J.  Dupont,  J.  Lippold and  S.  Kiser: Welding Metallurgy and Weldability of Nickel-Base Alloys, John Wiley and Sons, New Jersey, (2009), 59. ISBN:978-0470087145
• 24)   W.  Kesternich and  R. V.  Nandedkar: J. Nucl. Mater., 179 (1991), 1015. https://doi.org/10.1016/0022-3115(91)90263-7
• 25)   X.  Zhi,  J.  Xing,  H.  Fu and  Y.  Gao: Mater. Charact., 59 (2008), 1221. https://doi.org/10.1016/j.matchar.2007.10.010
• 26)   Z.  Xu,  G.  Liang,  Q.  Guan and  Q.  Jiang: Mater. Res. Bull., 39 (2004), 457. https://doi.org/10.1016/j.materresbull.2003.10.014
• 27)   G.  Will and  R.  Platzbecker: J. Inorg. Gen. Chem., 627 (2001), 2207. https://doi.org/10.1002/1521-3749(200109)627:9<2207::AID-ZAAC2207>3.0C O;2-5
• 28)   H. L.  Yakel: Acta Crystallogr., B43 (1987), 230. https://doi.org/10.1107/S0108768187097957