[Skip to content]

TWI
Search our Site
.

Factors influencing the toughness of submerged-arc welded line pipe welds - Part 1: Oxygen and deoxidising elements

TWI Bulletin, January/February 1989

by Peter Harrison

Peter Harrison, BSc(Eng), ARSM, PhD, MIM is a Principal Research Metallurgist in the Materials Department.

The toughness of longitudinal submerged-arc seam welds in line pipe can be sensitive to small variations in weld metal chemistry brought about by dilution effects. Part 1 of this article reviews the influences of oxygen and deoxidising elements on seam weld toughness and Part 2 will discuss the importance of weld metal hardenability, weld cooling rate and weld metal A1:0 ratio.


During the last few years, consumable manufacturers have developed consumables to a point where current weld metal specification requirements can usually be met in all weld metal samples. It is now recognised, however, that small variations in the levels of deoxidising elements, introduced through dilution from the parent material, or from the welding wire, can lead to dramatic changes in weld metal toughness. This is particularly true where high heat input submerged-arc welding (SAW) is concerned, as in line pipe seam welds.

In basic flux weld metals, the best low temperature toughness ( i.e. resistance to cleavage fracture) is achieved when microstructures consisting of a high proportion of acicular ferrite, which is a fine-grained intragranularly nucleated phase, are formed ( Fig.1). [1] Studies of the development of acicular ferrite in weld metals have shown that inclusion content is of considerable importance since, under favourable conditions, acicular ferrite grains may be nucleated at or around the weld metal inclusions. The following review draws together published information on the effects of oxygen and deoxidising elements, particularly Al, Ti, and Si, on the microstructure and toughness of high dilution submerged-arc welds, typical of those found in line pipe seam welds.

Fig.1. Dependence of 40J transition temperature on weld metal acicular ferrite content for bead-in-groove welds produced with 4.9 kJ/mm arc energy.
Fig.1. Dependence of 40J transition temperature on weld metal acicular ferrite content for bead-in-groove welds produced with 4.9 kJ/mm arc energy.

Oxygen

Many studies have shown that weld metal microstructural development is strongly influenced by either the nature or number of weld metal deoxidation products ( i.e. by the weld metal oxygen content). [2-6] In submerged-arc weld metals, it has been noted that both high and low oxygen contents may lead to poor weld toughness, whereas intermediate oxygen levels often lead to high levels of toughness because of the development of tough acicular ferrite dominated microstructures. Ito and Nakanishi [2] observed optimum toughness in submerged-arc weld metals with oxygen contents of about 200-250ppm, as shown in Fig.2. A similar finding was made by Terashima and Tsuboi. [7] At the lower oxygen levels, acicular ferrite nucleation becomes difficult and less tough 'bainitic type' microstructures are formed, with a resulting loss of cleavage resistance. At the higher oxygen levels, there is a tendency to form high temperature γ to α transformation products such as pro-eutectoid ferrite and Widmanstatten sideplates, which have a large grain or colony size and therefore have poor cleavage resistance. Dilatometric measurements made by Cochrane and Kirkwood [4] showed that the γ to α transformation temperature of a high oxygen content weld metal (~ 700ppm) was some 50°C higher than that of a medium oxygen content weld metal (~300ppm) of otherwise similar composition. Harrison and Farrar [5] noted a similar trend when they degassed a 300ppm oxygen weld metal to 60ppm and found a 30°C drop in γ to α transformation temperature. These dilatometric observations indicate that increasing weld metal oxygen content has a powerful negative effect on hardenability and thus the influence of oxygen content is opposite to that of the major alloying elements.



Fig.2. Relationship between Charpy impact properties and oxygen content of weld metals (5 kJ/mm). [2]
Fig.2. Relationship between Charpy impact properties and oxygen content of weld metals (5 kJ/mm). [2]

Aluminium

Several investigations have been conducted to study the influence of aluminium on submerged-arc weld metal toughness. [8-17] Investigations by Tenkula and Heikkinen [9] and Hannerz and Werlefors [10] noted that weld metal toughness was lowered by the presence of aluminium. Yoshino and Stout [11] and Terashima and Hart [12,13] noted that the influence of aluminium varied with flux type. For basic fluxes, aluminium was detrimental, but for fluxes with lower basicity or for acid fluxes, aluminium had a beneficial influence on toughness.

Most of the above investigators noted that the detrimental influence of aluminium in basic flux welds was associated with a coarsening of the microstructure. This effect was partially quantified by Terashima and Hart, who observed a decrease in the proportion of acicular ferrite in their weld metals, accompanied by an increase in the proportion of ferrite with aligned second phase, as weld metal aluminium content increased. With acid fluxes, the effect of aluminium on microstructure is less clear, and it has been suggested by Hannerz [18] that the beneficial influence of aluminium may be because of improved deoxidation.

A further observation on the influence of aluminium was that of Devillers et al, [8] who demonstrated that the optimum or 'critical' oxygen content varied with aluminium level, as illustrated in Fig.3. This was an important observation, because it demonstrated that the main influence of aluminium on toughness was to move the optimum oxygen content for acicular ferrite development. Thus, aluminium per se is not particularly detrimental to weld metal toughness, unless perhaps excessive amounts remain in solid solution.

Fig.3. Weld metal critical oxygen content v. base metal Al content and line showing stoichiometric ratio for alumina formation (Al 2 O 3 ).
Fig.3. Weld metal critical oxygen content v. base metal Al content and line showing stoichiometric ratio for alumina formation (Al 2 O 3 ).

Devillers et al recommended that flux SiO 2 and MnO contents should be increased in order to increase weld metal oxygen contents and obtain good toughness when welding materials of high aluminium content. Saggese et al [19] observed a linear relationship between weld metal oxygen content and SiO 2 + MnO +TiO 2 oxide contents of fluxes and suggested that optimum weld metal toughness would be found when the aluminium and oxygen contents were approximately at the stoichiometric ratio for alumina formation, i.e. an aluminium/oxygen ratio of 1:12. This ratio is plotted as a dashed line in Fig.3 and is fairly close to the empirical results of Devillers. However, recent work by Thewlis [17] has suggested that acicular ferrite nucleation will be accelerated in the presence of fcc galaxite (MnO/Al 2 O 3 ) in the surface layers of inclusions. The formation of this compound was seen to be critically dependent on weld metal aluminium content, but presumably this will also depend on weld metal oxygen content during deoxidation, i.e. the Al/O ratio.

Terashima and Hart also noted that weld metal tolerance to aluminium was influenced by oxygen content, and could be increased by increasing the TiO 2 content of the flux. More recently, Harrison [20] and Suzuki et al [15] have demonstrated that tolerance to aluminium can be increased by using semi-basic or neutral fluxes with higher oxygen potentials. Terashima and Hart also noted that weld metal tolerance to aluminium was reduced as weld metal silicon content increased ( Fig.4)

Fig.4. Influence of Al and Si on submerged-arc weld metal toughness [13]
Fig.4. Influence of Al and Si on submerged-arc weld metal toughness [13]

Titanium

Many investigators have studied the influence of titanium on weld metal toughness. A review of this subject has also been undertaken by Tsuboi and Terashima. [21] In most studies, a beneficial effect has been observed for small additions of titanium, followed by a detrimental influence at higher titanium levels. Hirabayashi et al [22] observed an optimum weld metal titanium content of around 0.02-0.03wt% in submerged-arc welds made in X65 linepipe steel ( Fig.5). Nakano et al, [23] however, noted that the optimum weld metal titanium content was highly dependent on other alloying elements such as Ni, Cr and Mo and that the optimum addition decreased as the alloy level increased.Terashima et al [24] observed an increase in the optimum weld metal titanium content with increasing nitrogen levels. Nakanishi and Komizo [25] noted an increase in the optimum titanium content with increasing oxygen content ( Fig.6).

Fig.5. Relationship between Ti content in weld metal and energy at -80°C. [22]
Fig.5. Relationship between Ti content in weld metal and energy at -80°C. [22]
Fig.6. Effect of oxygen levels on the toughness of Ti-B weld metals [25]
Fig.6. Effect of oxygen levels on the toughness of Ti-B weld metals [25]

There are thought to be two main mechanisms of toughness improvement by small additions of titanium, one involving promotion of acicular ferrite development and the other, removal of 'free' nitrogen from solution in the matrix. In the former case, it was originally proposed that TiN [2,26] was responsible for enhanced acicular ferrite nucleation. However, a number of more recent investigations have suggested that TiO [27,28] promotes acicular ferrite nucleation. An investigation by Bailey [29] indicated that only approximately 0.004 wt%Ti was required to change weld metal microstructure from ferrite with aligned MAC to acicular ferrite, when welding with manganese silicate flux. A similar observation was made by Bonnet and Charpentier [30] when welding with calcium silicate fluxes.

At high levels of titanium, weld metal toughness is degraded because of matrix hardening, and it has been suggested by Thaulow [31] that either the formation of microphase regions consisting of twinned martensite, or the occurrence of fine scale precipitation of TiC in ferrite may be responsible for embrittlement.

Silicon

In recent years, the influence of silicon on submerged-arc weld metal toughness has received relatively little attention compared to titanium and aluminium. However, the investigations of Ito and Nakanishi, [2] and Hannerz [32] both indicated optimum weld metal silicon contents. The results of Ito and Nakanishi indicated that the optimum silicon level lies at about 0.15-0.25 wt%Si at 1 wt%Mn. The results of Hannerz ( Fig.7) show an optimum silicon content at about 0.25-0.35 wt%Si.

Fig.7. Influence of Si on submerged-arc weld metal toughness [32]
Fig.7. Influence of Si on submerged-arc weld metal toughness [32]

Hill and Levine [33] observed an improvement in Charpy toughness of submerged-arc welds when the silicon content increased from 0.24-0.51 wt%. Similarly, Cochrane et al [34] reported improvement with increasing silicon from 0.19-0.28 wt%Si, and Pargeter [35] observed improvement in the range 0.35-0.42%. Abson, [36] however, observed a small decrease in toughness with increasing silicon from 0.23 to 0.57%. In all these investigations, an increase in the proportion of acicular ferrite was observed with increasingsilicon content.

Terashima and Hart, and Cochrane et al noticed that the effects of silicon and aluminium are additive with respect to their influence on weld metal microstructure. In Cochrane's work, because of the aluminium levels and flux systems studied, both elementswere seen to promote acicular ferrite development, and thus improve toughness. In the work of Terashima and Hart, it was noted that for low aluminium levels, silicon promoted acicular ferrite development and thus improved toughness,but, at high levels of aluminium, silicon reduced toughness.

Summary

Examination of the literature has shown that O, Al, Ti and Si all influence weld metal toughness, and that optimum levels of these elements exist for any given base composition and welding condition. This is thought to be closelylinked with the formation of an optimum weld metal microstructure consisting of a high proportion of acicular ferrite.

Of course acicular ferrite development is influenced by other factors, the most important of which are weld metal hardenability, weld cooling rate, and the presence or absence of suitable nucleation sites for intragranularnucleation of acicular ferrite. Part 2 will discuss these factors and present data illustrating the importance of weld metal Al:O ratio.


References

Author Title  
1 Duncan A: 'Toughness of submerged-arc weldments in low carbon equivalent C:Mn structural steels'. Welding Institute Members Report 314/1986, September 1986. Return to text
2 Ito Y and Nakanishi M: 'Study on Charpy impact properties of weld metal with submerged-arc welding'. Sumitomo Search 1976 (15) 42-62.  
3 Abson D J, Dolby R E and Hart P H M: 'The role of non-metallic inclusions in ferrite nucleation in carbon steel weld metals'. Proc Conf 'Trends in steels and consumables for welding'. The Welding Institute, London, 1978, 75-101.  
4 Cochrane R C and Kirkwood P R: ibid, 'The effect of oxygen on weld metal microstructure'. 103-121. Return to text
5 Harrison P L and Farrar R A: 'Influence of oxygen rich inclusions on the γ to α phase transformation in high strength low alloy (HSLA) steel weld metals'. J Mat Sci 1981 16 (8) 2218. Return to text
6 Ferrante M and Farrar R A: 'The role of oxygen rich inclusions in determining the microstructure of weld metal deposits'. J Mat Sci 1982 17 3293.  
7 Terashima H and Tsuboi J: 'Submerged-arc flux for low oxygen and hydrogen weld metal'. Metal Construction 1982 14 (12) 648-654. Return to text
8 Devillers L, Keplan D, Marandet B, Ribes A and Riboud P V: 'The effect of low level concentrations of some elements on the toughness of submerged-arc welded C:Mn steel welds.' Proc Conf 'The effects of residual, impurity and microalloying elements on weldability and weldproperties' The Welding Institute, London, 1983.  
9 Tenkula J and Heikkinen V K: 'The effect of parent plate aluminium on the properties of two wire submerged-arc weld metal in Nb microalloyed steel' Welding Research International 1978 8 (5) 369-376. Return to text
10 Hannerz N E and Werlefors T: 'The influence of parent material aluminium content on microstructure inclusion distribution and mechanical properties of submerged-arc weld metal'. Proc Conf 'Weld pool chemistry and metallurgy'. The WeldingInstitute, London, 1980, 335-349. Return to text
11 Yoshino Y and Stout R D: 'Effect of microalloys on the notch toughness of line pipe seam welds'. Weld J 1979 58 59s-68s. Return to text
12 Terashima H and Hart P H M: 'Effect of aluminium on C-Mn-Nb steel submerged-arc weld metal properties'. Weld J 1984 63 (6) 173s-183s. Return to text
13 Terashima H and Hart P H M: 'Effect of flux TiO 2 and wire Ti content on tolerance to high Al content of submerged-arc welds made with basic fluxes'. Proc Conf 'The effects of residual, impurity and microalloying elements on weldability and weldproperties'. The Welding Institute, London, 1983. Return to text
14 Savchenko V A and Tertyshnaya N K: 'The influence of aluminium in low carbon steel on its weldability' Welding Production 1976 23 (1) 39-42.  
15 Suzuki M, Kunisada Y and Watanabe I: 'Effect of aluminium on high heat input submerged-arc weld metal toughness' September 1985. IIW Doc 11-1043-85, IX-1354-85. Return to text
16 Oldland R B, Mcpherson R and Brownrigg A: 'The influence of aluminium on acicular ferrite formation in welds' Proc Conf 'Metals, Materials, Manufacturing'. Australasian Institute of Metals, Ballarat C A E, Victoria, 1985.  
17 Thewlis G: 'The influence of pipe plate and consumable chemistry on the composition, microstructure and toughness of weld metal'. Proc Conf 'Welding and performance of pipelines'. The Welding Institute, London, 1986,205-227. Return to text
18 Hannerz N E: 'Weld metal properties as influenced by the deoxidation practice of the structural steel base material'. IIW Doc IX-1298-83. Return to text
19 Saggese M E, Bhatti A R, Hawkins D N and Whiteman J A: 'Factors influencing inclusion chemistry and microstructure in submerged-arc welds'. Proc Conf 'The effect of residual, impurity and micro-alloying elements on weldability and weld properties'. The WeldingInstitute, London, 1983, P15-1-P15-11. Return to text
20 Harrison P L: 'The influence of calcium treatment on submerged-arc weld metal and HAZ toughness of microalloyed C-Mn steel'. Metal Construction February 1987 19 (2) 84R-91 R. Return to text
21 Tsuboi J and Terashima H: 'Review of strength and toughness of Ti and Ti-B microalloyed deposits' Welding in the World 1983 21 (11/12) 304-317. Return to text
22 Hirabayashi K, Taira T, Yamaguchi T and Takashige K: 'Improvement of toughness of submerged-arc weld metal of low temperature service line pipe'. IIW Doc XII-A-135-77 September 1986.  
23 Nakano S, Shiga A and Tsuboi J: 'Optimising the titanium effect of weld metal toughness' IIW Doc XII-B-182-75 1975. Return to text
24 Terashima H, Nishiayama N and Tsuboi J: 'Effects of Ti, Al and N on toughness in 60 kg/mm 2 class submerged-arc weld metal'. Communication to annual assembly of JWS 1981 80. Return to text
25 Nakanishi M and Komizo Y: 'Transformation behaviour and toughness of low carbon low alloy steel weld metals'. Sumitomo Search 1983 28 53-71.  
26 North T H, Bell H B, Koukabi A and Craig J: 'Notch toughness of low oxygen content submerged-arc deposits'. Weld J 1979 58 (12) 343.  
27 Mori N; Homma H, Okita S and Wakabayashi M: 'Mechanism of notch toughness improvement in Ti-B bearing weld metals' IIW Doc IX-1196-81 May 1981. Return to text
28 Ohkita S, Homma H, Tsushima S and Mori N: 'The effect of oxide inclusion on micro-structure of Ti-B containing weld metal'. Australian Welding Journal 1984 29 (3) 29-36.  
29 Bailey N: 'Titanium flux additions during submerged-arc welding of ferritic steels'. Welding Institute Members Report 221/1983, July 1983. Return to text
30 Bonnet C and Charpentier J-P: 'Effect of deoxidation residues in wire and of some particular oxides, in CS fused fluxes on the microstructure of submerged-arc weld metals'. Proc Conf 'The effects of residual, impurity and micro-alloyingelements on weldability and weld properties'. The Welding Institute, London, 1983. Return to text
31 Thaulow C: 'Structural changes in submerged-arc weld metals of C:Mn steels containing small amounts of titanium'. Proc Conf 'Weld pool chemistry and metallurgy'. The Welding Institute, London, 1980, 17-29. Return to text
32 Hannerz N E: 'The influence of silicon on the mechanical properties and on the weldability of mild and high tensile structural steels'. IIW Doc IX-1169-80.  
33 Hill D C and Levine E: 'Effect of microstructure and inclusions on toughness of HSLA steel weld metal' in 'Physical Metallurgy of Metal Joining'. Editors R Kossowsky and M E Glicksman, AIME 1980. Return to text
34 Cochrane R C, Ward J L and Keville B R: 'The influence of deoxidation and/or desulphurisation practice on the weld metal toughness of high dilution welds'. Proc Conf 'The effects of residual, impurity and microalloying elements on weldability and weldproperties'. The Welding Institute, London, 1983, P16-1-P16-16. Return to text
35 Pargeter R J: 'Deoxidant additions to submerged-arc fluxes: Part 2 - Effects of calcium disilicide on microstructure and toughness'. Welding Institute Members Report 189/1982, August 1982. Return to text
36 Abson D J: 'A study of the influence of C and Si on the microstructure and toughness of submerged-arc welds in 25mm C-Mn steelplate'. Welding Research International 1979 9 (5) 1-23. Return to text