6. Narrow Gap Welding, Electrogas - and Electroslag Welding

6.

Narrow Gap Welding,

Electrogas - and

Electroslag Welding

6. Narrow Gap Welding, Electrogas- and Electroslag Welding 77

2005

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, the process

characteristics mentioned

in the upper part of Figure

6.1 are frequently con-

nected with the definition

for narrow gap welding. In

spite of these “definition

difficulties” all questions

about the universally valid

advantages and disadvan-

tages of the narrow gap

welding method can be

clearly answered.

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 sev-

eral important process characteristics. Narrow gap TIG welding with cold or hot wire addi-

tion is mainly applied as

an orbital process method

or for the joining of high-

alloy as well as non-

ferrous metals. This

method is, however, hardly

applied in the practice.

The other processes are

more widely spread and

are explained in detail in

the following.

© ISF 2002

Narrow Gap Welding

br-er6-01e.cdr

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

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

- 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

Disadvantages

Figure 6.1

Figure 6.2

© ISF 2006

Survey of Narrow Gap Welding TechniquesBased on Conventional Technologies

br-er6-02e.cdr

process withstraightened

wire electrode(1P/L, 2P/L, 3P/L)

process withoscillating

wire electrode(1P/L)

process withtwin electrode(1P/L, 2P/L)

process withlengthwisepositioned

strip electrode(2P/L)

process withlinearly oscillating

filler wire

process withstripshaped

filler andfusing feed

electrogasprocess with

linearlyoscillating

wire electrode

electrogasprocess with

bent,longitudinally

positionedstrip electrode

MIG/MAG-processes

(1P/L,2P/L,3P/L)

process with hotwire addition(1P/L, 2P/L)

process with coldwire addition(1P/L, 2P/L)

submerged arcnarrow gap welding

electroslag narrowgap welding

gas-shielded metal arcnarrow gap welding

tungsten innertgas-shielded

narrow gap welding

flat position vertical up position all welding positions


2005

In Figure 6.3, a systematic subdivision of the

various GMA narrow gap technologies is

shown. In accordance with this, the fundamen-

tal distinguishing feature of the methods is

whether the process is carried out with or with-

out 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 accor-

dance with the different types of arc move-

ment.

In the following, some of the GMA narrow gap

technologies are explained:

Using the turning tube method, Figure 6.4, side

wall fusion is achieved by the turning of the contact tube; the contact tip angles are set in de-

grees 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 oscil-

lation frequencies can be realised - independent of the filler wire properties. In contrast, when

the corrugated wire

method with mechanical

oscillator is applied, arc

oscillation is produced by

the plastic, wavy deforma-

tion of the wire electrode.

The deformation is ob-

tained by a continuously

swinging oscillator which is

fixed above the wire feed

rollers. Amplitude and fre-

quency of the wave motion

Figure 6.3

Figure 6.4

Principle of GMANarrow Gap Welding

br-er 6-04e.cdr

corrugated wire method with mech. oscillator

1 - wire reel2 - drive rollers3 - wire mechanism for wire guidance4 - inert gas shroud5 - wire guide tube and shielding gas tube6 - contact tip

1 - wire reel2 -

3 - 4 - inert gas shroud5 - wire feed nozzle and shielding gas tube6 - contact tip

mechanical oscillator for wire deformation

drive rollers

12 -

14

8 -

10

1 1

22

33

4 4

5 5

6 6

© ISF 2006

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

br-er6-03e.cdr

D

A

C

B

long-wire method(1 P/L, 2 P/L)

thick-wire method(1 P/L, 2 P/L)

twin-wire method(1 P/L)

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

twisted wire method(1 P/L)

coiled-wire method(1 P/L)

rotation method(1 P/L)

corrugated wire methodwith mechanical oscillator

(1 P/L)

corrugated wire methodwith oscillating rollers

(1 P/L)

corrugated wire methodwith contour roll (1 P/L)

zigzag wire method(1 P/L)

wire loop method(1 P/L)

GMA narrow gap weldingno wire-deformation

GMA narrow gap weldingwire-deformation

explanation:P/L:Pass/Layer

A: method without forced arc movementB: method with rotating arc movementC: method with oscillating arc movementD: method with two or more filler wires


2005

can be varied over the total amplitude of oscil-

lation and the speed of the mechanical oscil-

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 pro-

duced 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

can be noticed. Thanks to the correct setting

of the oscillation parameters and the precise,

centred torch manipulation no sidewall fusion

defects occurred, in spite of the low sidewall

penetration depth. A further advantage of the weave-bead technique is the high crystal re-

structuring rate in the weld metal and in the basemetal adjacent to the fusion line – an advan-

tage that gains good toughness properties.

Two narrow-gap welding

variations with a rotating

arc movement are shown in

Figure 6.6. When the rota-

tion method is applied, the

arc movement is produced

by an eccentrically protrud-

ing wire electrode (1.2 mm)

from a contact tube nozzle

which is rotating with fre-

quencies between 100 and

150 Hz. When the wave

© ISF 2002br-er6-05e.cdr

plate thickness:gap preparation:

elctrode diameter:amperage:pulse frequency:arc voltage:welding speed:wire oscillation:oscillation width:shielding gas:primery gas flow:secondary gas flow:number of passes:

300 mmsquare-butt joint, 9 mmflame cut1.2 mm260 A120 HZ30 V22 cm/min80 min4 mm80% Ar/ 20% Co

25 l/min50 l/minapprox. 70

-1

2

Figure 6.5

Figure 6.6

Principle ofGMA Narrow Gap Welding

br-er 6-06e.cdr

rotation method spiral wire method

1 - wire reel2 - drive rollers3 - mechanism for nozzle rotation4 - inert gas shroud5 - shielding gas nozzle6 - wire guiding tube

1 - wire reel2 - wire mechanism for wire deformation3 - drive rollers4 - wire feed nozzle and shielding gas supply5 - contact piece

13 -

14

1 1

2

2

3

3

4 4

5

6 5

9 -

12


2005

wire method is used, the 1.2 mm solid wire is being spiralwise deformed. This happens be-

fore 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

GMA narrow gap welding

methods which are oper-

ated without forced arc

movement, where a reli-

able sidewall fusion is ob-

tained either by the wire

deflection through constant

deformation (tandem wire

method) or by forced wire

deflection with the contact

tip (twin-wire method). In

both cases, two wire elec-

trodes with thicknesses

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 in-

cline 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 multi-

pass 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

Principle of GMANarrow Gap Welding

br-er 6-07e.cdr

tandem method

1 - wire reel2 - deflection rollers3 - drive rollers4 - inert gas shroud5 - shielding gas nozzle6 - wire feed nozzle and contact tip

1

2

3

4

5

6

9 -

12

350

twin-wire method

1 - wire reel2 - drive rollers3 - inert gas shroud4 - wire feed nozzle and shielding gas supply5 - contact tips

1

2

3

4

5

15 -

18

Figure 6.7


2005

and 25 mm are used. The gap width is, de-

pending on the number of passes per layer,

between 20 and 25 mm. SA twin-wire weld-

ing is, in general, carried out using two elec-

trodes (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 to-

wards 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

shapes in relation to plate thickness for SA

welding (DIN 8551 part 4) with those for GMA

welding (EN 29692) and the unstandardised,

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 mate-

rial 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.

The practical application of SA narrow gap

welding for the production of a flanged calotte

joint for a reactor pressure vessel cover is de-

picted in Figure 6.10. The inner diameter of the

pressure vessel is more than 5,000 mm with

br-er6-09e.cdr

Comparison of theWeld Groove Shape

8

8

s

10°

s

16

7°

8°

6

3 s s

3

10

double-U butt weldSA-DU weld preparation

(8UP DIN 8551)

square-edge butt weldSA-SE weld preparation

(3UP DIN 8551)

double-U butt weldGMA-DU weld preparation

(Indexno. 2.7.7 DIN EN 29692)

narrow gap weldGMA-NG weld preparation

(not standardised)

Figure 6.9

br-er6-08e.cdr

Submerged ArcNarrow Gap Welding

strip electrode

twin-wire electrode

α

s

xa

so

h

h

w

w

f

p

H

a

z

vw

s

v weld speed

a electrode deflectionH stick out z distance torch - flank

s gap widthh bead heightw bead widthp penetration depth

w

SO stick out

s gap widtha electrode deflectionx distance of strip tip to flank

twisting angle

h bead hightw bead width

α

Figure 6.8


2005

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.

Electrogas welding (EG) is characterised by a vertical groove which is bound by two water-

cooled copper shoes. In the groove, a filler wire electrode which is fed through a copper noz-

zle, 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.

© ISF 2002br-er6-10e_sw.cdr

Figure 6.10

br-er6-11e.cdr

Electrogas Welding

+-

weld advance

workpiecewire guide

electrode

shielding gas

arc

weld pool

Cu-shoe

weld metal water

designation:position:plate thickness:

materials:gap width:electrodes:

amperage:voltage:weld speed:shielding gas:

gas-shielded metal arc welding (GMAW acc. DIN 1910 T.4)vertical (width deviations of up to 45°)6 - 30 mm square-butt joint or V weld seam 30 mm double-V weld seamunalloyed, lowalloy and highalloy steels8 - 20 mmonly 1 (flux-cored wire, for slag formation betweencopper shoe and weld surface) Ø 1.6 - 3.2 mm350 - 650 A28 - 45 V2 - 12 m/hunalloyed and lowalloy steelsCO or mixed gas (Ar 60% and 40% Co )highalloy steels: argon or helium

2 2

Figure 6.11


2005

The application of electrogas welding for low-

alloyed 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 expo-

sure to temperatures of more than 1500°C and

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 weld-

adjacent regions above critical temperatures

and thus the brittleness effects are significantly

reduced.

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 the conducting

bath, keeps the slag bath

temperature constant. The

welding current is fed over

one or more endless wire

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

br-er6-12e.cdr

Electroslag Welding

1

2

3

4

5

6

7

8

9

1. base metal2. welding boom3. filler metal4. slag pool5. metal pool

6. copper shoe7. water cooling8. weld seam9. Run-up plate

designation:position:plate thickness:gap width:materials:electrodes:

amperage:voltage:welding speed:slag hight:

resistance fusion weldingvertical (and deviation of up to 45°) 30 mm (up to 2,000 mm)24 - 28 mmunalloyed, lowalloy and highalloy steels1 or more solid or cored wires Ø 2.0 - 3.2 mmplate thickness range per electrode: fixed 30 - 50 mmoscillated: up to 150 mm550 - 800 A per electrode35 - 52 V0.5 - 2 m/h30 - 50 mm

Figure 6.12

© ISF 2002

Process Phases During ES Welding

br-er6-13e.cdr

~

ignition with arc powder fusion

start of welding welding end of welding

powder

slag

weld metal

molten pool

slag

Figure 6.13


2005

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 conduc-

tive 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 subse-

quently 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

of the welding device and

the easier welding area

preparation. Also curved

seams can be welded with

a bent consumable elec-

trode. As the groove width

can be significantly re-

duced when comparing Electroslag Welding withFusing Wire Feed Nozzle

br-er 6-14e.cdr

=~

drive motorwire or stripelectrode

run-off plate

welding cable

workpiece workpiece

fusingfeed nozzle

workpiece cable run-up plate

copper shoes

workpieceworkpiece

copper shoes

Electroslag fusingnozzle process (channel welding)

position: verticalplate thickness: 15 mmmaterials: unalloyed, lowalloy and highalloy steels

welding consumables:

wire electrodes: Ø 2.5 - 4 mmstrip electrodes: 60 x 0.5 mmplate electrodes: 80 x60 up to 10 x 120 mmfusingfeed nozzle: Ø 10 - 15 mm welding powder: slag formation with high electrical conductivity

Figure 6.14


2005

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 tech-

niques.

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 cool-

ing 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, Fig-

ure 6.15.

To avoid postweld heat

treatment the electroslag

welding process with lo-

cal continuous normali-

sation has been devel-

oped for plate thicknesses

of up to approx. 60 mm,

Figure 6.16. The welding

temperature in the weld

region drops below the Ar1-

temperature and is subse-

quently heated to the nor-

malising temperature

(>Ac3). The specially de-

signed torches follow the

copper shoes along the

weld seam. By reason of

the residual heat in the

workpiece the necessary

temperature can be

reached in a short time.

ES Welding with LocalContinuous Normalisation

br-er 6-16e.cdr

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

temperature °C

2000

1500

900

500

950

1

2

3

4

5

6

7

8

9

10

2

7

8

9

11

10

Figure 6.16

Possibilities to ImproveWeld Seam Properties

br-er 6-15e.cdr

technological measures metallurgical measures

increase of purity

application of suitablebase and filler metals

addition of suitable alloyand micro-alloy elements(nucleus formation)

reduction ofS-, P-, H -, N -and O - contentsand otherunfavourabletrace elements

2 2

2

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

post weld heattreatment

decrease of peak temperatureand dwell times at hightemperatures

increase of welding speed

reduction of energy perunit length

continuousnormalisationfurnacenormalisation

increase of deposit rate

decrease of groove volume

application ofseveral wireelectrodes,metal powderaddition

V, double-V buttjoints, multi-passtechnique

Figure 6.15


2005

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 applica-

tion 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, Fig-

ure 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 con-

ventional electroslag welding, exclusively done by flame cutting. Thus, the advantage of eas-

ier weld preparation can be maintained.

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

© ISF 2002br-er6-17e.cdr

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

ES-welding in 2 passes with sliding shoe

1

2

3

4

9

5

6

7

8

4

10

Figure 6.17

© ISF 2002br-er6-18e.cdrES-welding of the outer passes

1

11

2

3

4

9

5

6

7

8

4

10

12

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

Figure 6.18


2005

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 elec-

trodes, 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 phos-

phor in the weld centre as well as too fast a cooling of the molten pool in the proximity of the

weld seam edges.

Documents

6. Narrow Gap Welding, Electrogas - and Electroslag Welding