6.

Narrow Gap Welding, Electrogas - and Electroslag Welding

6. Narrow Gap Welding, Electrogas- and Electroslag Welding

77

Up to this day, there is no universal agreement about the definition of the term Narrow Gap Welding although the term is actually self-explanatory. In the international technical literature,
Process characteristics: - narrow, almost parallel weld edges. The small preparation angle has the function to compensate the distortion of the joining members - multipass technique where the weld build-up is a constant 1 or 2 beads per pass - usually very small heat affected zone (HAZ) caused by low energy input

the

process mentioned con-

characteristics 6.1

in the upper part of Figure are frequently nected with the definition

Advantages: - profitable through low consumption quantities of filler material, gas and/ or powder due to the narrow gaps - excellent quality values of the weld metal and the HAZ due to low heat input - decreased tendency to shrink

Disadvantages - higher apparatus expenditure, espacially for the control of the weld head and the wire feed device - increased risk of imperfections at large wall thicknesses due to more difficult accessibility during process control - repair possibilities more difficult

for narrow gap welding. In spite of these definition difficulties all questions about the universally valid advantages and disadvantages of the narrow gap welding method can be clearly answered.

br-er6-01e.cdr

ISF 2002

Narrow Gap Welding

Figure 6.1

The numerous variations of narrow gap welding are, in general, a further development of the conventional welding technologies. Figure 6.2 shows a classification with emphasis on several important process characteristics. Narrow gap TIG welding with cold or hot wire addition is mainly applied as an orbital process method or for the joining of highalloy as well as nonThis ferrous metals.
submerged arc electroslag narrow narrow gap welding gap welding process with straightened wire electrode (1P/L, 2P/L, 3P/L) process with oscillating wire electrode (1P/L) process with twin electrode (1P/L, 2P/L) process with lengthwise positioned strip electrode (2P/L) flat position
br-er6-02e.cdr

gas-shielded metal arc narrow gap welding

tungsten innert gas-shielded narrow gap welding

process with linearly oscillating filler wire

electrogas process with linearly oscillating wire electrode electrogas process with bent, longitudinally positioned strip electrode

process with hot wire addition (1P/L, 2P/L) MIG/MAGprocesses (1P/L,2P/L,3P/L) process with cold wire addition (1P/L, 2P/L)

method is, however, hardly applied in the practice. The other processes are more widely spread and are explained in detail in the following.

process with stripshaped filler and fusing feed

vertical up position

all welding positions

ISF 2006

Survey of Narrow Gap Welding Techniques Based on Conventional Technologies

Figure 6.2

2005

6. Narrow Gap Welding, Electrogas- and Electroslag Welding In Figure 6.3, a systematic subdivision of the various GMA narrow gap technologies is shown. In accordance with this, the fundamental distinguishing feature of the methods is whether the process is carried out with or without wire deformation. Overlaps in the structure result from the application of methods where a single or several additional wires are used. While most methods are suitable for single pass per layer welding, other methods require a weld build-up with at least two passes per layer. A further subdivision is made in accordance with the different types of arc movement.
br-er6-03e.cdr

78

GMA narrow gap welding no wire-deformation long-wire method (1 P/L, 2 P/L) thick-wire method (1 P/L, 2 P/L) twin-wire method (1 P/L)

GMA narrow gap welding wire-deformation

A
tandem-wire method (1 P/L, 2 P/L, 3 P/L) twisted wire method (1 P/L)

rotation method (1 P/L)

coiled-wire method (1 P/L) corrugated wire method with mechanical oscillator (1 P/L) corrugated wire method with oscillating rollers (1 P/L) corrugated wire method with contour roll (1 P/L) zigzag wire method (1 P/L) wire loop method (1 P/L)

explanation: P/L: Pass/Layer

A: method without forced arc movement B: method with rotating arc movement C: method with oscillating arc movement D: method with two or more filler wires
ISF 2006

In the following, some of the GMA narrow gap technologies are explained: Using the turning tube method, Figure 6.4, side

Survey and Structure of the Variations of Gas-Shielded Metal Arc Narrow Gap Welding

Figure 6.3

wall fusion is achieved by the turning of the contact tube; the contact tip angles are set in degrees of between 3 and 15 towards the torch axis. With an electronic stepper motor control, arbitrary transverse-arc oscillating motions with defined dwell periods of oscillation and oscillation frequencies can be realised - independent of the filler wire properties. In contrast, when the
corrugated wire method with mech. oscillator 1 1

corrugated with

wire

method

mechanical

2 3 4 5 6

2 3 4 5 6

oscillator is applied, arc oscillation is produced by the plastic, wavy deformation of the wire electrode. The deformation is ob8 - 10

1 - wire reel 2 - drive rollers 3 - wire mechanism for wire guidance 4 - inert gas shroud 5 - wire guide tube and shielding gas tube 6 - contact tip
br-er 6-04e.cdr

12 - 14

1 - wire reel 2 - mechanical oscillator for wire deformation 3 - drive rollers 4 - inert gas shroud 5 - wire feed nozzle and shielding gas tube 6 - contact tip

tained by a continuously swinging oscillator which is fixed above the wire feed rollers. Amplitude and frequency of the wave motion

Principle of GMA Narrow Gap Welding

Figure 6.4
2005

6. Narrow Gap Welding, Electrogas- and Electroslag Welding

79

can be varied over the total amplitude of oscillation and the speed of the mechanical oscilplate thickness: gap preparation: 300 mm square-butt joint, 9 mm flame cut elctrode diameter: 1.2 mm amperage: 260 A pulse frequency: 120 HZ arc voltage: 30 V welding speed: 22 cm/min -1 wire oscillation: 80 min oscillation width: 4 mm shielding gas: 80% Ar/ 20% Co2 primery gas flow: 25 l/min secondary gas flow: 50 l/min number of passes: approx. 70

lator or, also, over the wire feed speed. As the contact tube remains stationary, very narrow gaps with widths from 9 to 12 mm with plate thicknesses of up to 300 mm can be welded. Figure 6.5 shows the macro section of a GMA narrow gap welded joint with plates (thickness: 300 mm) which has been produced by the mechanical oscillator method in approx. 70 passes. A highly regular weld build-up and an almost straight fusion line with an extremely narrow heat affected zone

br-er6-05e.cdr

ISF 2002

can be noticed. Thanks to the correct setting of the oscillation parameters and the precise, centred torch manipulation no sidewall fusion

Figure 6.5

defects occurred, in spite of the low sidewall

penetration depth. A further advantage of the weave-bead technique is the high crystal restructuring rate in the weld metal and in the basemetal adjacent to the fusion line an advantage that gains good toughness properties. Two narrow-gap welding
rotation method 1 spiral wire method 1

variations with a rotating arc movement are shown in Figure 6.6. When the rotation method is applied, the arc movement is produced by an eccentrically protruding wire electrode (1.2 mm) from a contact tube nozzle which is rotating with frequencies between 100 and 150 Hz. When the wave

2 3 4 5 6

2 3 4

13 - 14

1 - wire reel 2 - drive rollers 3 - mechanism for nozzle rotation 4 - inert gas shroud 5 - shielding gas nozzle 6 - wire guiding tube
br-er 6-06e.cdr

1 - wire reel 2 - wire mechanism for wire deformation 3 - drive rollers 4 - wire feed nozzle and shielding gas supply 5 - contact piece

Principle of GMA Narrow Gap Welding

Figure 6.6
2005

9 - 12

6. Narrow Gap Welding, Electrogas- and Electroslag Welding

80

wire method is used, the 1.2 mm solid wire is being spiralwise deformed. This happens before it enters the rotating 3 roll wire feed device. With a turning speed of 120 to 150 revs per minute the welding wire is deformed. The effect of this is such that after leaving the contact piece the deformed wire creates a spiral diameter of 2.5 to 3.0 mm in the gap adequate enough to weld plates with thicknesses of up to 200 mm at gap widths between 9 and 12 mm with a good sidewall fusion. Figure 6.7 explains two
tandem method 1
350

twin-wire method

GMA narrow gap welding

1

2 3 4 5 6 2 3 4 5

methods which are operated without forced arc movement, where a reliable sidewall fusion is obtained either by the wire
15 - 18

1 - wire reel 2 - deflection rollers 3 - drive rollers 4 - inert gas shroud 5 - shielding gas nozzle 6 - wire feed nozzle and contact tip
br-er 6-07e.cdr

9 - 12

1 - wire reel 2 - drive rollers 3 - inert gas shroud 4 - wire feed nozzle and shielding gas supply 5 - contact tips

deflection through constant deformation (tandem wire method) or by forced wire deflection with the contact tip (twin-wire method). In both cases, two wire electrodes with thicknesses

Principle of GMA Narrow Gap Welding

Figure 6.7

between 0.8 and 1.2 mm are used. When the tandem technique is applied, these electrodes are transported to the two weld heads which are arranged inside the gap in tandem and which are indeFigure pendently selectable. When the twin-wire method is applied, two parallel switched electrodes are transported by a common wire feed unit, and, subsequently, adjusted in a common sword-type torch at an incline towards the weld edges at small spaces behind each other (approx. 8 mm) and molten. In place of the SA narrow gap welding methods, mentioned in Figure 6.2, the method with a lengthwise positioned strip electrode as well as the twin-wire method are explained in more detail, Figure 6.8. SA narrow gap welding with strip electrode is carried out in the multipass layer technique, where the strip electrode is deflected at an angle of approx. 5 towards the edge in order to avoid collisions. After completing the first fillet weld and slag removal the opposite fillet is made. Solid wire as well as cored-strip electrodes with widths between 10

2005

6. Narrow Gap Welding, Electrogas- and Electroslag Welding

81

strip electrode
SO stick out
s

and 25 mm are used. The gap width is, depending on the number of passes per layer, between 20 and 25 mm. SA twin-wire welding is, in general, carried out using two electrodes (1.2 to 1.6 mm) where one electrode is deflected towards one weld edge and the other towards the bottom of the groove or towards the opposite weld edge. Either a single pass layer or a two pass layer technique are applied. Dependent on the electrode diameter and also on the type of welding powder, is the gap width between 12 and 13 mm. Figure 6.9 shows a comparison of groove

s a x h w

gap width electrode deflection distance of strip tip to flank twisting angle bead hight bead width

so
x

twin-wire electrode

vw

vw a H z s h w p

weld speed electrode deflection stick out distance torch - flank gap width bead height bead width penetration depth

H
z

a h

p
br-er6-08e.cdr

shapes in relation to plate thickness for SA

Submerged Arc Narrow Gap Welding

welding (DIN 8551 part 4) with those for GMA welding (EN 29692) and the unstandardised,

Figure 6.8
10 7

mainly used, narrow gap welding. Depending on the plate thickness, significant differences in the weld cross-sectional dimensions occur which may lead to substantial saving of material and energy during welding. For example, when welding thicknesses of 120 mm to 200 mm with the narrow gap welding technique, 66% up to 75% of the weld metal weight are saved, compared to the SA square edge weld.
3 6 s s 3 narrow gap weld GMA-NG weld preparation (not standardised) double-U butt weld SA-DU weld preparation (8UP DIN 8551) 8 square-edge butt weld SA-SE weld preparation (3UP DIN 8551) 10
8 8 s s 16

The practical application of SA narrow gap welding for the production of a flanged calotte joint for a reactor pressure vessel cover is depicted in Figure 6.10. The inner diameter of the pressure vessel is more than 5,000 mm with Figure 6.9
2005
double-U butt weld GMA-DU weld preparation (Indexno. 2.7.7 DIN EN 29692)
br-er6-09e.cdr

Comparison of the Weld Groove Shape

6. Narrow Gap Welding, Electrogas- and Electroslag Welding

82

wall thicknesses of 400 mm and with a height of 40,000 mm. The total weight is 3,000 tons. The weld depth at the joint was 670 mm, so it had been necessary to select a gap width of at least 35 mm and to work in the three pass layer technique.

workpiece

wire guide

electrode shielding gas arc weld pool Cu-shoe weld advance weld metal water +

designation: gas-shielded metal arc welding (GMAW acc. DIN 1910 T.4) position: vertical (width deviations of up to 45 ) plate thickness: 6 - 30 mm square-butt joint or V weld seam 30 mm double-V weld seam materials: unalloyed, lowalloy and highalloy steels gap width: 8 - 20 mm electrodes: only 1 (flux-cored wire, for slag formation between copper shoe and weld surface) 1.6 - 3.2 mm 350 - 650 A amperage: 28 - 45 V voltage: weld speed: 2 - 12 m/h shielding gas: unalloyed and lowalloy steels CO2 or mixed gas (Ar 60% and 40% Co2) highalloy steels: argon or helium
br-er6-10e_sw.cdr ISF 2002
br-er6-11e.cdr

Electrogas Welding

Figure 6.10

Figure 6.11

Electrogas welding (EG) is characterised by a vertical groove which is bound by two watercooled copper shoes. In the groove, a filler wire electrode which is fed through a copper nozzle, is melted by a shielded arc, Figure 6.11. During this process, are groove edges fused. In relation with the ascending rate of the weld pool volume, the welding nozzle and the copper shoes are pulled upwards. In order to avoid poor fusion at the beginning of the welding, the process has to be started on a run-up plate which closes the bottom end of the groove. The shrinkholes forming at the weld end are transferred onto the run-off plate. If possible, any interruptions of the welding process should be avoided. Suitable power sources are rectifiers with a slightly dropping static characteristic. The electrode has a positive polarity.

2005

6. Narrow Gap Welding, Electrogas- and Electroslag Welding The application of electrogas welding for lowalloyed steels is more often than not - limited, as the toughness of the heat affected zone with the complex coarse grain formation does not meet sophisticated demands. Long-time exposure to temperatures of more than 1500 and C low crystallisation rates are responsible for this. The same applies to the weld metal. For a more wide-spread application of electrogas welding, the High-Speed Electrogas Welding Method has been developed in the ISF. In this process, the gap cross-section is reduced and additional metal powder is added to increase the deposition rate. By the increase of the welding speed, the dwell times of weldadjacent regions above critical temperatures and thus the brittleness effects are significantly reduced. Figure 6.12
designation: position: plate thickness: gap width: materials: electrodes:
6. copper shoe 7. water cooling 8. weld seam 9. Run-up plate 1 2 3 4 5 6 7 8 9

83

1. base metal 2. welding boom 3. filler metal 4. slag pool 5. metal pool

resistance fusion welding vertical (and deviation of up to 45 ) 30 mm (up to 2,000 mm) 24 - 28 mm unalloyed, lowalloy and highalloy steels 1 or more solid or cored wires 2.0 - 3.2 mm plate thickness range per electrode: fixed 30 - 50 mm oscillated: up to 150 mm amperage: 550 - 800 A per electrode voltage: 35 - 52 V welding speed: 0.5 - 2 m/h slag hight: 30 - 50 mm
br-er6-12e.cdr

Electroslag Welding

Figure 6.12 shows the process principle of Electroslag Welding. Heating and melting of the groove faces occurs in a slag bath. The temperature of the slag bath must always exceed the melting temperature of the metal. The Joule effect, produced when the current is transferred through
~
powder

the

conducting

bath, keeps the slag bath temperature constant. The welding current is fed over one or more endless wire
ignition with arc powder fusion
slag

electrodes which melt in the highly heated slag bath. Molten pool and slag bath which both form the weld pool are, sideways retained by the groove faces and, in general, by

slag molten pool weld metal

start of welding
br-er6-13e.cdr

welding

end of welding
ISF 2002

Process Phases During ES Welding

Figure 6.13
2005

6. Narrow Gap Welding, Electrogas- and Electroslag Welding

84

water-cooled copper shoes which are, with the complete welding unit, and in relation with the welding speed, moved progressively upwards. To avoid the inevitable welding defects at the beginning of the welding process (insufficient penetration, inclusion of unmolten powder) and at the end of the welding (shrinkholes, slag inclusions), run-up and run-off plates are used. The electroslag welding process can be divided into four process phases, Figure 6.13. At the beginning of the welding process, in the so-called ignition phase, the arc is ignited for a short period and liquefies the non-conductive welding flux powder into conductive slag. The arc is extinguished as the electrical conductivity of the arc length exceeds that of the conductive slag. When the desired slag bath level is reached, the lower ignition parameters (current and voltage) are, during the so-called Data-Increase-Phase, raised to the values of the stationary welding process. This occurs on the run-up plate. The subsequent actual welding process starts, the process phase. At the end of the weld, the switch-off phase is initiated in the run-off plate. The solidifying slag bath is located on the run-off plate which is subsequently removed. The electroslag welding with consumable feed wire (channel-slot welding) proved to be very useful for shorter welds. The copper sliding shoes are replaced by fixed Cu cooling bars and the welding nozzle by a steel tube, Figure 6.14. The length of the sheathed steel tube, in general, corresponds with the weld seam length (mainly shorter than 2.500 mm) and the steel tube melts during welding in the ascending slag bath. Dependent on the plate thickness, welding can be carried out with one single or with several wire and strip electrodes. A feature of this process variation is the handiness
workpiece drive motor welding cable run-off plate workpiece wire or strip electrode

Electroslag fusing nozzle process (channel welding) position: vertical plate thickness: 15 mm materials: unalloyed, lowalloy and highalloy steels welding consumables:

of the welding device and the easier welding area preparation. Also curved

= ~

fusing feed nozzle workpiece cable workpiece run-up plate copper shoes workpiece

seams can be welded with a bent consumable electrode. As the groove width can be significantly rewhen comparing

wire electrodes: 2.5 - 4 mm strip electrodes: 60 x 0.5 mm plate electrodes: 80 x60 up to 10 x 120 mm fusing feed nozzle: 10 - 15 mm welding powder: slag formation with high electrical conductivity

copper shoes
br-er 6-14e.cdr

duced

Electroslag Welding with Fusing Wire Feed Nozzle

Figure 6.14
2005

6. Narrow Gap Welding, Electrogas- and Electroslag Welding

85

with conventional processes, and a strip shaped filler material with a consumable guide piece is used, this welding process is rightly placed under the group of narrow gap welding techniques. Likewise in electrogas welding, the electroslag welding process is also characterised by a large molten pool with a simultaneously - low heating and cooling rate. Due to the low cooling rate good degassing and thus almost porefree hardening of the slag bath is possible. Disadvantageous, however, is the formation of a coarse-grain structure. There are, however, possibilities to improve the weld properties, Figtechnological measures
post weld heat treatment decrease of peak temperature and dwell times at high temperatures increase of welding speed reduction of energy per unit length continuous normalisation furnace normalisation increase of deposit rate application of several wire electrodes, metal powder addition decrease of groove volume V, double-V butt joints, multi-pass technique reduction of S-, P-, H2-, N2and O2 - contents and other unfavourable trace elements

metallurgical measures
increase of purity application of suitable base and filler metals addition of suitable alloy and micro-alloy elements (nucleus formation)

ure 6.15. To avoid postweld heat treatment the electroslag welding process with local continuous normalisation has been developed for plate thicknesses of up to approx. 60 mm, Figure 6.16. The welding temperature in the weld region drops below the Ar1temperature and is subsequently heated to the normalising temperature
1 2 3 4 5 6 7 8 9 10

C-content limits Mn, Si, Mo, Cr, Ni, Cu, Nb, V, Zr, Ti

br-er 6-15e.cdr

Possibilities to Improve Weld Seam Properties

Figure 6.15

temperature C
1. filler wire 2. copper shoes 3. slag pool 4. metal pool 5. water cooling 6. slag layer 7. weld seam 8. distance plate 9. postheating torch 10. side plate 11. heat treated zone

2 2000 1500 7 8 9 11 900 500 10 950

(>Ac3). The specially designed torches follow the copper shoes along the weld seam. By reason of the residual heat in the workpiece the necessary temperature can be

br-er 6-16e.cdr

ES Welding with Local Continuous Normalisation

reached in a short time.

Figure 6.16
2005

6. Narrow Gap Welding, Electrogas- and Electroslag Welding

86

In order to circumvent an expensive postheat weld treatment which is often unrealistic for use on-site, the electroslag high-speed welding process with multilayer technique has been developed. Similar to electrogas welding, the weld cross-section is reduced and, by application of a twin-wire electrode in tandem arrangement and addition of metal powder, the weld speed is increased, as in contrast to the conventional technique. In the heat affected zones toughness values are determined which correspond with those of the unaffected base metal. The slag bath and the molten pool of the first layer are retained by a sliding shoe, Figure 6.17. The weld preparation is a double-V butt weld with a gap of approx. 15 mm, so the carried along sliding shoe seals the slag and the metal bath. Plate preparation is, as in conventional electroslag welding, exclusively done by flame cutting. Thus, the advantage of easier weld preparation can be maintained.

12 11 1 2 3 4 9 5 6 7 8 4
1 magnetic screening 2 metal powder addition 3 tandem electrode 4 water cooling 5 copper shoe (water cooled) 6 slag pool 7 molten pool 8 solidified slag 9 welding powder addition 10 weld seam

1 2 3 4 9 5 6 7 8 4
1 magnetic screening 2 metal powder supply 3 three-wire electrode 4 water cooling 5 copper shoe (water cooled) 6 slag pool 7 molten pool 8 solidified slag 9 welding powder supply 10 weld seam 11 first pass 12 second pass
ISF 2002

10

10

br-er6-17e.cdr

ES-welding in 2 passes with sliding shoe

ISF 2002

br-er6-18e.cdr

ES-welding of the outer passes

Figure 6.17

Figure 6.18

For larger plate thicknesses (70 to 100 mm), the three passes layer technique has been developed. When welding the first pass with a double-V-groove preparation (root width: 20 to 30 mm; gap width: approx. 15 mm) two sliding shoes which are adjusted to the weld

2005

6. Narrow Gap Welding, Electrogas- and Electroslag Welding

87

groove are used. The first layer is welded using the conventional technique, with one wire electrode without metal powder addition. When welding the outer passes flat Cu shoes are again used, Figure 6.18. Three wire electrodes, arranged in a triangular formation, are used. Thus, one electrode is positioned close to the root and on the plate outer sides two electrodes in parallel arrangement are fed into the bath. The single as well as the parallel wire electrodes are fed with different metal powder quantities which as outcome permit a welding speed 5 times higher than the speed of the single layer conventional technique and also leads to strong increase of toughness in all zones of the welded joint. If wall thicknesses of more than 100 mm are to be welded, several twin-wire electrodes with metal powder addition have to be used to reach deposition rates of approx. 200 kg/h. The electroslag welding process is limited by the possible crack formation in the centre of the weld metal. Reasons for this are the concentration of elements such as sulphur and phosphor in the weld centre as well as too fast a cooling of the molten pool in the proximity of the weld seam edges.

2005