Welding stainless in the nineties - an update

TWI Bulletin, November/December 1993

Brian Ginn joined TWI in 1960 following a 5 year student apprenticeship as a foundry engineer, training in the production of castings in iron, ferritic steel and stainless steel and has a Higher National Certificate in mechanical engineering.

Currently he is a project leader in the Stainless Steels and Corrosion Section of the Materials Department where his work has included studies of plasma cutting, high alloy stainless steels, hard zones in line pipes, hyperbaric welding of duplex stainless steel and more recently low cost stainless steels.

Trevor Gooch joined TWI in 1965, starting work in the stainless steel section after graduating in industrial metallurgy at the University of Birmingham. He subsequently obtained the degrees of MSc(Eng) and PhD from the University of London. He was appointed group leader in 1967 and became Head of the Materials Department in 1980.

Initially, he worked on the properties and cracking propensity of welded joints in all grades of stainless steel and later on cryogenic materials and both nickel and copper alloys. As Head of the Materials Department his work now embraces the welding characteristics of almost all metallic materials.

New grades of stainless steels have presented new manufacturing and joining problems in recent years. Brian Ginn and Trevor Gooch consider these increasingly popular materials, and outline the welding processes best suited to fulfil their service requirements.

To the layman, stainless steels may well be considered to be a modern product. This is understandable, given that a prime outlet for the material is in the domestic consumer market. However, stainless steels, defined as iron alloys having a minimum chromium content of 11-12%, have been in use for many decades following independent developments, especially in the UK and Germany, before the first World War. Indeed, the first application of stainless steel in a passenger motor vehicle was recorded in 1928, in the Ford Model A. ^[1] Since then world production of stainless steel has increased rapidly, and in 1991 was estimated to be 10.6 million metric tonnes with, at that time, an annual growth rate of approximately 5%. Corresponding figures for carbon steel production were 700 million metric tonnes worldwide, but with a falling demand.

Development of new types and grades of stainless steel inevitably brings new problems in manufacture and joining. This is particularly true for welding where the desired material properties carefully produced by the steelmaker can be compromised by a process which locally melts and recasts part of the workpiece. There is a continuing demand for increased productivity, while maintaining the parent material properties.

Material changes

Increases in the number of applications for stainless steels would not occur without appropriate development programmes to obtain alloys meeting specific requirements, especially for corrosion resistance. This development continues with production in the last decade of the 22%Cr ferritic/austenitic duplex stainless steels from which the 25% 'superduplex' grades with higher pitting resistance have evolved. During this time, high molybdenum 'superaustenitic' and other complex alloy stainless steels have also emerged and are frequently specified for service in aggressive media, see the Table.

Typical composition of conventional, duplex and high alloy austenitic stainless steels

When welding duplex steels, the weld area becomes ferritic at high temperature, with reformation of austenite on cooling ( Fig.1). These changes are dependent on the material composition and total weld thermal cycle, and dominate welding behaviour. Adequate austenite reformation is essential to obtain good mechanical and corrosion properties, and this need has significantly affected alloy design. Modern materials are remarkably tolerant to changes in arc welding procedure, and, although welding must be carefully controlled, as with any other alloy system, welded joints have so far performed well in a range of plant. ^[2]

As a casting, weld metals inevitably display segregation of alloying elements. The effect is significant in highly alloyed superaustenitic grades; the dendrite cores are depleted in molybdenum in particular, and this reduces the local resistance to pitting attack to below the level of the parent steel ( Fig.2). In consequence, special welding procedures are necessary, with use of an 'overalloyed' nickel-base filler, but, provided this is recognised, again the materials have given excellent service in welded fabrications. ^[3]

Welding processes

Electron beam welding

Using a focused, high energy electron beam (EB), materials can be welded in a single pass typically from <lmm to >200mm thickness. Normally carried out in a vacuum, the process offers precision and virtual freedom from atmospheric contamination. ^[4]

In EB welding under vacuum, there is some loss of volatile elements, particularly dissolved gases such as oxygen and nitrogen. Evolution of gas from the weld pool itself can, in extreme cases, disrupt the stability of the capillary and cause cavity defects to form. Furthermore, nitrogen is increasingly being added to austenitic and duplex stainless steels to give higher strength, greater austenite stability at cryogenic temperatures and enhanced corrosion resistance. Any loss of this element during welding can therefore impair properties in the weld region.

This is particularly the case for duplex steels in which the austenite/ferrite phase balance is critical in ensuring that optimum strength, toughness and corrosion properties are achieved. During autogenous EB welding, the rapid weld thermal cycle dictates that cooling occurs too quickly for austenite to develop, and this is exacerbated by loss of nitrogen. In consequence EB welds in duplex grades ( Fig.3) tend to develop structures lean in austenite, and can suffer from gross porosity because of nitrogen outgassing. Recent work at TWI has demonstrated that by filler additions to the weld pool, via either wires or interlayers, satisfactory phase balance can be achieved and gross defects avoided, even in steel containing fairly high levels of nitrogen ( i.e. 0.2 wt%). Figure 4 shows a weld made with addition of a small quantity of pure nickel to the weld fusion zone, in this case a 0.15mm thick interlayer.

The potential for using EB welding in stainless steel fabrication has long been recognised and demonstrated, particularly in the aerospace (martensitic and high strength precipitation hardened stainless steels) and nuclear power generation industries (austenitic stainless steels). Examples of applications of the EB process which exploit one or more of its characteristics are cited by Wyatt. ^[5] These range from production welding of gas turbine components in which a relatively slow traverse speed ensures a crack free deposit, to manufacture of small bellows assemblies which necessitates a low heat input and negligible distortion. The EB process has been applied extensively to the common grades of austenitic stainless steel such as Types 304 and 316, exploiting the advantages of minimal distortion and high joining rate. Travel speeds can be as high as 10m/min for material below 1mm thickness, and 100mm/min for 200mm thickness steel.

Laser processing

The industrial laser has become a popular alternative to traditional materials processing techniques for welding, cutting and surface modification. ^[6] The laser beam, which is a high-intensity, monochromatic light source, can be focused to achieve very high power densities of the order of 10 ⁶ W/cm ². This compares with the TIG process which has a power density of approximately 10 ² W/cm ² and plasma welding and cutting with a power density of 10 ³ W/cm ².

This high power density results in rapid heating, melting and vaporisation of the material to form a vapour cavity or 'keyhole'. The molten material flows around this keyhole during welding, solidifying at the rear. The result is a weld that can be made at very high travel speeds and has a narrow, deep penetration fusion zone with near parallel sides as in electron beam welding. The extremely high power density also provides very high cooling rates, resulting in a very narrow heat affected zone and minimal distortion. In laser cutting, which uses a gas jet to remove the molten material, the high power density produces very narrow kerfs, recast layers and heat-affected zones.

Nearly all types of stainless steel are easily processed using an industrial laser ( Fig.5). Because of the low thermal conductivity of common austenitic steels as compared with carbon steels, even higher welding speeds and/or deeper welds are possible. For example, a 2.5kW CO ₂ laser can weld 3mm thickness type 304 stainless steel at greater than 4m/min. In microelectronic applications, laser-beam pulsing allows placement of the weld close to sensitive components without danger of thermal damage.

Laser cutting of stainless steels can also offer advantages over traditional thermal and mechanical cutting processes. Compared with mechanical cutting techniques, the processing rates can be significantly greater, cutting speeds up to 2m/min for 4mm stainless steel with a 1kW CO ₂ laser and oxygen assist. A relatively new cutting technique, termed 'clean cut', uses a non-oxidising gas such as nitrogen at high pressure to remove the molten material from the kerf. ^[7] The resulting cut edge is oxide-free with very high cut quality and the component can be welded or used as a 'show' surface with little or no post-cut processing. The process can be CN controlled, so that the part can be cut to final dimensions, in intricate shapes, to very close tolerances.

Friction welding

Rotational friction welding is a one-shot joining process involving localised heating by friction and pressure between two components, followed by a forging action. One or both components may be rotated. Joining time is short, only a few seconds, and the method can be applied to virtually all grades of stainless steel, including ferritic alloys which have poor arc weldability. Radial friction welding is a modified version of this process, developed at TWI for welding pipelines. It operates on the principle of rotating and compressing a ring about two stationary pipe ends, producing a solid state bond ( Fig.6 and 7). It has been demonstrated that radial friction welding is capable of producing high integrity bonds in duplex stainless steel with balanced ferrite/austenite microstructure in the as-welded condition. ^[8] In addition, friction welding can be used for superaustenitic alloys, and, since joining is carried out in the solid state, loss of corrosion resistance from segregation is avoided. ^[9]

Conventional friction welding methods, including the radial technique, are restricted to components having circular symmetry. ^[10] Studies with orbital motion have led to development of linear friction welding. ^[11] Relative movement between the parts is provided by moving one across the other, through a small amplitude of a suitable frequency, while under axial load. The possibility now exists of joining complex assemblies rapidly, especially where a number of welds are required.

One further recent development with considerable potential is metal deposition, solid phase cladding, by friction welding. ^[12] In this process, a rotating consumable bar is applied under an axial load to a laterally moving substrate, leaving a deposit bonded to the base metal ( Fig.8). The method can be used to deposit stainless steel cladding, with no dilution from the substrate, thus avoiding problems of intermixing encountered with arc cladding, and producing a sound overlay of defined composition.

A recent study ^[8] of the relative costs of the production of a 64 kilometre long, 305mm diameter X 9.5mm wall thickness duplex stainless steel pipeline revealed that savings of approximately 10% would accrue from use of friction welding compared with conventional TIG/MIG manufacturing techniques. This study showed that savings result from increased production rates and reduced labour costs.

Arc welding

Cored wire consumables have been available for the common stainless steels for a number of years, but recently there has been significant interest and growth in practical application of small diameter flux-cored wires for arc welding in all positions ( Fig.9). Improvements in wire manufacturing technology have enabled stainless tubular wires to be drawn down to small diameters. Availability of tubular wire diameters of 1.2mm and less, together with incorporation of a rapidly freezing rutile slag system, has allowed positional welding with stainless steel flux-cored wires to become a practical reality. Such consumables combine the operating efficiency of MIG welding with the ease of use and high standard of weld finish normally associated with high quality MMA coated electrodes. In addition, deposition rates can exceed those of MIG with similar size wires, particularly at higher arc energy conditions, see Fig.10. ^[13] Flux-cored wires are finding applications ranging from fabrication of architectural components through to pressure vessels.

Small diameter flux-cored wires are available in standard grades, e.g. 308L, 316L, 309L and, more recently, in duplex alloys. These are designed to operate using argon CO ₂ shielding gas mixtures or 100%CO ₂ when weld metal carbon level is of less concern. The Ar/CO ₂ mixtures produce less spatter, whereas slight increases in weld penetration may be obtained with 100% CO ₂ gas. The welding performance of these consumables is very similar to their ferrous counterparts and they are associated with comparable potential benefits, i.e. ease of use and high productivity. Arc and process stability can be maintained over a wide range of welding parameters making them suitable for welding plate thicknesses from 3mm upwards. By virtue of the rutile slag system, spray metal transfer characteristics can be used for positional welding, considerably reducing the risk of lack of fusion defects. The rapidly freezing slag also makes weld pool control easier for out of position welding. Deposited weld beads are bright, smooth, largely free of spatter and are very easily de-slagged. In most cases, welds can be left in the as-deposited condition, eliminating the need for expensive post-weld grinding and dressing.

Small diameter flux-cored wires can be used with conventional MIG welding equipment although the wire delivery requirements to achieve consistent and reliable feeding are more stringent than with solid filler wires. Careful attention to wire feed rolls, pressure, gun liners and so on must be given with small diameter wires because of their tubular structure.

As for MMA welding, effective fume extraction and adequate ventilation are necessary to provide a safe working environment. However, it can be expected that, with increased fabricator confidence in their use, further development and continued growth in use of small diameter flux-cored wires will be sustained.

In conclusion

Even against a background over the last decade of increased stainless steel production, the growth in use of superaustenitic and, more especially, duplex grades is noteworthy. This trend is likely to continue and, while a number of problems remain, the salient welding characteristics are now well understood and the materials can be reliably joined by welding.

In large part, stainless steels will continue to be welded using established arc processes. However, from experience with ferritic steel fabrications, cored wire welding will become more common in the thrust for increased productivity. The other welding processes considered above will involve a greater departure from conventional production practice, and will necessitate careful evaluation of equipment costs. However, they have the common feature of producing joints rapidly in a single operation, and, given appropriate workpiece geometry and production runs, may well prove to be of economic advantage.

References