[Skip to content]

TWI
Search our Site
.

An initial feasibility study of moving contact arc welding

TWI Bulletin, November - December 1997

Wayne Thomas
Wayne Thomas

Wayne is a Principal Research Engineer. He joined TWI in 1983 from the steel industry and has been responsible for conception and/or development of a number of emergent technologies.





Richard Jones
Richard Jones

Richard is Manager of the Arc and Surfacing Section of TWI where he has been involved in development and application of arc welding process technologies.





Geert Verhaeghe
Geert Verhaeghe

Geert is a project leader in TWI's Arc Welding section, where he has been involved in development and application of arc welding process technologies.






Paul Woollin
Paul Woollin

Paul joined TWI in 1992 from post-doctoral research at Cambridge University. Currently he is ferrous and non-ferrous Manager in the Materials Department.

The moving contact arc welding (MCAW) technique, under development at TWI, offers an easy-to-use, alternative cladding, welding and repair method. It can be applied with restricted access and is amenable to remote operation. It is suitable for either manual or mechanised welding, as Wayne Thomas, Richard Jones, Geert Verhaeghe and Paul Woollin explain.

Fig. 1. Basic principle of moving contact arc welding with Ridgeback™ consumable electrode
Fig. 1. Basic principle of moving contact arc welding with Ridgeback™ consumable electrode

Moving Contact Arc Welding (MCAW) involves a sliding or rolling electrical contact supplying current to a shaped metal core along an almost limitless length of flux-covered lying consumable. The MCAW technology enables the distance between the electrical contact and the burning arc to be kept as short as practical. The MCAW process uses low cost portable equipment, and a Ridgeback TM consumable. This process is readily automated, simple to operate and shows great potential for application in general fabrication and shipbuilding [1] . More specifically, by using shaped electrodes, it is ideal for hardfacing and creates opportunities for pipe joining. The salient features of the process are illustrated in Fig. 1.

The technique differs from the traditional Firecracker (Elin-Hafergut) and lying consumable welding processes, because the Ridgeback TM consumable can carry greater currents and because of minimal resistive heating of the consumable, it enables almost infinite lengths of consumable to be used in one operation. Conventional MMA consumables are limited in length, because of resistive heating of the metal core during welding. This heating is inversely proportional to the area of cross-section of the metal core and directly proportional to consumable length. It is the ability to supply welding current close to the arc, which minimises the resistive heating and enables use of very long Ridgeback TM consumables, effectively reducing the number of start/stops.

The technique works as follows: current supply is made with a sliding or rolling tool to the consumable via a narrow protruding ridge in the consumable core (Fig. 2). Subsequently an arc is struck, using a fuse or fine wire wool at the arc end to ionise consumable/substrate arc gap. The thickness of the flux covering, between consumable core and substrate, ensures that a controlled arc length is maintained throughout the welding operation. The arc burns along the consumable electrode leaving a weld deposited onto the workpiece (Fig. 3).

Fig. 2. Nominal profile of Ridgeback™ consumable electrode
Fig. 2. Nominal profile of Ridgeback™ consumable electrode
Fig. 3. Welding trials with MCAW skid tool
Fig. 3. Welding trials with MCAW skid tool

Compared with conventional MMA welding the MCAW technique offers greater productivity, has reduced discontinuities and enables longer and larger consumables to be used. The MCAW technique offers an easy-to-use, alternative cladding, welding and repair method. It can be applied with restricted access and is suited to remote operation. It is suitable for either manual or mechanised welding.

Development work is currently underway at TWI to improve process control and welding productivity and to tailor the technique to meet the requirements of specific applications.

Background

Existing literature describes a number of process techniques which make use of lying electrodes. Since development of the Elin-Hafergut process, commonly known as the firecracker process, a number of attempts to improve the technique have been made [2-7] .

Essentially, the advantages of the lying electrode technique can be seen in those application areas with limited access. In addition, its predictable arc gap and its inherent low skill requirement make the process user-friendly and readily mechanised. The use of firecracker welding for butt, fillet, groove and multipass welds has been reported to have achieved acceptable welds [2] .

It has also been established that the arc transfer characteristics and the formation of the weld bead are influenced by the point of connection of the welding current return supply. Arc control is aided by the work return supply being positioned up-stream, that is, welding has to be carried out in the direction away from the current return supply to reduce adverse electromagnetic forces that can develop and cause excessive spatter loss [3] .

However, the firecracker type of lying electrode has limitations, one of which is resistive heating. Because of this resistive heating, the length of the consumable is restricted. A number of attempts to overcome this length limitation problem have been made with limited practical success [2,5,6] .

The basic concept of the moving contact approach potentially provides a more elegant solution to welding comparatively long, almost limitless, lengths of lying consumable. However, it was not until the consumable profile was significantly changed to the RidgebackTM type geometry that acceptable welds could be achieved without use of additional granulated flux or gas shielding [1] .

The Ridgeback TM type electrode significantly reduces the amount of metallic core exposed. In addition, re-entrant features either side of the comparatively thin metal core contact ridge provide locations for additional amounts of flux on the upper surface of the consumable. These additional regions of flux subsequently provide the necessary slag protection to prevent atmospheric contamination and to facilitate metal transfer to the underlying weld pool [1] .

Experimental Procedure

Consumable electrodes used in this work programme were produced from 6mm diameter BS 970 grade 316L stainless steel for convenience. To suit the Ridgeback TM geometry these bars were machined nominally to the profile shown in Fig. 2. The substrate material comprised 6mm thick by 30mm wide by 500mm long BS 970 grade 304L stainless steel.

Fig. 4. MCAW wheel tool
Fig. 4. MCAW wheel tool

Flux composition of the Ridgeback TM consumable was chosen to suit the required weld metal chemistry. Both an AC transformer and a DC power source were used for welding. The welding tool which supplied welding current to the consumable ridge was positioned about 50mm from the arc end, and consisted of a rectangular copper block mounted upon an insulated steel bar. The welding tool resembles a bicycle handle bar and is of simple and robust construction. To minimise friction between the welding tool and electrode, a copper wheel can also be chosen instead of the skid (Fig. 4).

Welding was carried out manually, mainly in the flat position. However, welding in position was also possible providing the right choice of process conditions, flux composition, core material and consumable geometry was selected. For the purpose of this paper, bead-on plate welds only have been reported.

Welding performance was assessed in the form of arc stability, level of spatter, ease of slag-removal, weld bead profile, etc. The completed MCAW welds were subject to both a visual inspection and a metallurgical examination. Furthermore a Vickers hardness test was carried out to EN 288 Pt3 specification whereby weld metal, heat-affected zone (HAZ) and fusion line were examined closely.

Metallographic sections were taken transverse to the weld and prepared to a 1µm finish, followed by electrolytic etching in an aqueous 20% H 2SO 4 solution with 0.1g/1 NH 4CNS. The sections were examined under an optical microscope and selected photomicrographs taken to show typical microstructural features. A Vickers hardness survey was performed using a 5kg indenting weight.

Additionally, the welds were subject to simple bend testing, to check the weld soundness.

Results and Discussion

Fig. 5. Example of surface appearance of MCAW with Ridgeback™ consumable electrode, BS970 grade 316L deposit onto BS970 grade 304L substrate
Fig. 5. Example of surface appearance of MCAW with Ridgeback™ consumable electrode, BS970 grade 316L deposit onto BS970 grade 304L substrate

Trials with the Ridgeback TM type consumable were carried out and acceptable weld deposits that satisfied mechanical and metallurgical evaluation were achieved. The typical visual appearance of a deposit produced with a Ridgeback TM consumable is shown in Fig. 5, which displays a well-defined weld surface ripple and is generally similar in appearance to those weld deposits produced by conventional MMA electrodes.

Welding conditions used to produce acceptable Ridgeback TM weld deposits in the flat position ranged between the following:

  • Current 110 - 130 Amps
  • Arc voltage 23 - 24V
  • Welding traverse speed 150 - 170 mm/min
  • Consumable wire cross-sectional area 21mm 2

A number of comparative weld trials were also made using lying electrodes produced to a configuration as described in the earlier literature [6] . Standard MMA electrodes were used, with the upper layer of flux removed to provide electrical contact for the current supply. This contrasted with the very narrow ridge of core wire exposed for the Ridgeback TM type consumable. Trials with these partially exposed round core consumables proved unsatisfactory. This resulted in a very unstable arc, which was frequently broken. In addition to the high level of spatter, the final weld deposit profile was not uniform.

In contrast, the weld performance of the Ridgeback TM type consumables was much better. Flux on either side of the (core) metal ridge deflected the arc downwards and minimised the level of spatter. Flux underneath the consumable core ensured electrical insulation between base material and the Ridgeback TM consumable and maintained a controlled arc length throughout the welding operation. The arc length can be changed by altering the thickness of the flux or by changing the shape of the metal core. Several fluxes were used to determine the optimum for welding in the flat position on the stainless steel substrate.

Fig. 6. Example of part consumed electrodes Ridgeback™ profile on left-hand side round diameter exposed wire (prior art) on right-hand side
Fig. 6. Example of part consumed electrodes Ridgeback™ profile on left-hand side round diameter exposed wire (prior art) on right-hand side

It should be noted that different flux compositions of the Ridgeback TM consumable will be required for welding in position and to suit a particular welding application, much in the same way conventional electrodes are produced. The Ridgeback TM consumable provided a stable arc, smooth metal transfer with good weld metal deposition characteristics. In addition, the slag was self-removable and the overall result was a weld bead with a very smooth surface appearance. Figure 6 clearly shows that the two electrode types had been consumed in a different manner. During welding it is evident that the round section exposed core wire electrode (prior art) arc burns in a more exposed manner resulting in a very unstable arc welding technique. Moreover, the latter is more susceptible to atmospheric contamination. For the latter electrode type to work successfully, additional flux or gas shielding would be necessary.

As previous researches have indicated [3] , arc control can also be aided by positioning the work return supply up stream of the arc ie welding away from the current return connector. In this respect the inherent magnetic fields created by the welding current tend to deflect the arc directly towards the substrate, thus minimising the amount of weld metal lost through spatter. Moreover, it is sometimes convenient, although not always necessary, to add a further electrical connection from the electrode supply to the extremity of the core wire away from the arcing end. In that way, unintentional loss of contact of the welding tool with the electrode will not result in momentary arc interruptions.


Deposit evaluation

Fig. 7. Transverse section through typical MCAW
Fig. 7. Transverse section through typical MCAW

The macrophotograph in Fig. 7 illustrates modest penetration of weld metal into the base steel, which is comparable with that achievable with an MMA electrode of a similar size. Microstructurally, the weld metal is predominantly single-phase austenite, with a cored dendritic structure (Figs. 8a, 8b). Visual assessment indicates little, if any, delta ferrite in the bulk of the weld metal. Some change in structure near to the fusion line was noted, with an increased volume fraction of delta ferrite (visually assessed as 3-4%) being present within around 5Omm of the fusion boundary (Fig. 8c). The HAZ area showed no significant change in structure when compared to the base steel (Fig. 8c). For further development work with 316L stainless steel it may be necessary to design the chemical composition of the consumable to provide higher levels of weld metal delta ferrite, to offset any tendancy for solidification cracking.

Fig. 8a Dendritic, fully austenitic structure near the weld cap
Fig. 8a Dendritic, fully austenitic structure near the weld cap
Fig. 8b Dendritic, fully austenitic weld metal near the middle of the deposit (MCAW)
Fig. 8b Dendritic, fully austenitic weld metal near the middle of the deposit (MCAW)
Fig. 8c Fusion boundary area of (MCAW) weld showing variation of etching response
Fig. 8c Fusion boundary area of (MCAW) weld showing variation of etching response

The weld microstructural features observed and hardness measured are broadly typical of an arc weld produced by any of the common processes in an austenitic stainless steel.

Results of the hardness survey are given in the Table and show that the parent steel (167 HV 5) was harder than both the weld metal (154 HV 5) and HAZ (150 HV 5), which suggests that the parent steel might have a small degree of cold work. The weld metal immediately adjacent to the fusion line, which showed the highest volume fraction of delta ferrite, gave 153 HV 5, ie very similar to the bulk weld metal. Bend tests confirmed the soundness of the weld deposit.

Table 1: Results of Vickers hardness survey

Area measured Vickers hardness, HV 5 (presented as min-max/average)
min max average
Substrate parent 304L stainless steel 160 177 167
Weld metal 316L stainless steel 153 157 154
Fusion line 147 158 153
HAZ 144 158 150

Potential applications

The Moving Contact Arc Welding technique could be carried out using hollow-core consumable electrodes, allowing additional wire consumable to be fed into the weld pool. This wire feed technique permits introduction of considerable amounts of additional metal to the weld deposit and at the same time provides a conduit for metallic powder delivery and/or active or inert shielding gases.

This technique is an alternative cladding method with little restriction in length or size of the consumable. Also, the geometry of the consumable can easily be tailored to fit the application. For instance, for cladding large surface areas, a flat strip instead of a wire can be used as the consumable core.

Fig. 9. MCAW local habitat for dry under water welding
Fig. 9. MCAW local habitat for dry under water welding

Restricted access and remote operation capability of the MCAW process are two major advantages unique to this technique. Their process features may be exploited for underwater welding, under dry conditions in a local habitat (Fig. 9). The practicability of restricted access conditions has been demonstrated using a local habitat chamber, trials are currently being carried out at TWI which will further endorse the potential advantages of this technique. Repair welding in hazardous surroundings such as the nuclear industry, and welding or repair inside tubes and pipes, are likely to provide ideal applications of the process technique.

Conclusions

Investigations at TWI have confirmed the feasibility of the Moving Contact Arc Welding process for bead-on plate welding in the flat position using Ridgeback TM electrodes. Trials have shown that stable welddeposition conditions can be achieved and sound weld deposit can be produced. Metallurgical examination of Ridgeback TM weld deposits showed a structure consistent with similar stainless steel composition depositsproduced by a conventional MMA welding.

References

No. Author Title & reference  
1 Thomas W M and Jones R L 'Low skill arc welding development shows its potential' Connect, 1992 October/November.  
2 Evans R M, Meister R P, Brayton WC 'Firecracker welding for shipyard application' Welding Journal, 1976, 55 (7) 555-565  
3 Yurikov V N 'On the use of high-quality standard electrodes for building-up in a lying position' Svar Proiz, 1972, 8 16-17  
4 Kikuta Y, Miyao N, Okuto K, Ikeda M and lgarashi S 'Application of lay down process in automatic fabrication of steel structures,' Osaka University, Yamada-Kami, Suita, Osaka 565, Japan, Paper No 2-1-(2)  
5 Sumikin Welding Electrode Co Ltd 'Method and apparatus for firecracker arc welding' British Patent 1 282 461. Filed 8 July 1969. Return to text
6 La Soudure Electrique Autogène SA (Belgium) Patent No. 920 769, Procedure for electric arc welding, patented since 4 January 1947, authorised by the Ministry of Industrial Production, France.  
7 Sidoruk V, Dudko D A, Kirichenko A V 'Submerged - Arc welding with suspended electrode', (E 0 Paton Welding Institute, Kiev) Return to text