Recent developments in nickel base material welding considering the influence of shielding gas on the hot cracking resistance

Th. Kannengiesser, M. Wolf, H. Schobbert

Federal Institute for Materials Research and Testing (BAM)

Unter den Eichen 87

D-12200 Berlin, Germany

ABSTRACT

Nickel base alloys are frequently applied in safety-relevant fields such as chemical plant construction or power plant engineering. Particularly on account of their cubic face-centered solidification characteristic, these materials are frequently susceptible to metallurgy-specific hot cracking during fusion welding. A classification of these materials according to their hot cracking resistance in the MVT-Test (Modified Varestraint Transvarestraint Test) is carried out by the example of a series of base and welding filler materials. Moreover, it has been proven by the MVT-Test that the hot cracking resistance of nickel base alloy (Alloy 602 CA) can be improved by selecting appropriate shielding gases.

.

1 Introduction

Nickel base alloys are characterized by their largely inert behaviour towards chemically aggressive media. Furthermore, some of these alloys, e.g. Alloy 602 CA (NiCr25FeAlY, M-No. 2.4633, Nicrofer 6025 HT) can be applied as construction materials for service temperatures up to 1200°C. Owing to these properties, nickel base alloys are among others predestined for use in chemically, mechanically and/or thermally highly stressed components and constructions, e.g. in specific fields of chemical plant construction, power plant engineering or offshore technology [1]. Nearly all nickel base alloys can be welded with conventional arc welding processes such as TIG-, plasma, MIG/MAG- and MMA-welding. In addition to this, an increasing number of nickel base materials are currently being qualified for further welding processes. At present, investigations are underway for MAG-tandem welding, combining two MAG-torches in a common welding process, using Alloy 625 (Nicrofer 6020 hMo), Alloy C4 (Nicrofer 6616 hMo), Alloy 59 (Nicrofer 5923 hMo) and Alloy B2 (Nicrofer 6928).

Based on the current state-of-the-art, this study was aimed at classifying nickel base alloys with respect to their hot cracking resistance during welding depending on practice-relevant external influencing factors using an externally-loaded hot cracking test.

2 Shielding gases for nickel base material welding

During welding of nickel base alloys, vital importance has to be attached to the selection of an appropriate shielding gas. Generally, pure argon can be used in TIG-welding, but with the high-temperature Alloy 602 CA (NiCr25FeAlY), for example, the application of this shielding gas almost inevitably leads to the formation of hot cracks. On the part of industry, extensive investigations have therefore been initiated in order to improve the weldability of this material via the shielding gas influence on the hot cracking resistance. The results from these research efforts have proven that this material is reliably weldable with the shielding gas argon + 1% to 3% nitrogen [2]. Investigations undertaken on the part of the Federal Institute for Materials Research and Testing (BAM) with the MVT-Test using Alloy 602 CA (NiCr25FeAlY) are presented in this paper.

For MIG-welding of most nickel base alloys, pure argon can be used as shielding gas, which, however, results in an unstable arc behaviour. In addition to this, there might be a lack of melt flowing towards the base material.

For the nickel base materials, special shielding gases have been developed to which, apart from helium and/or hydrogen, only a very small amount of carbon dioxide has been added. Already with a carbon dioxide content of only 0.05% in the shielding gas, high arc stability is reached. Additions of hydrogen and helium yield an improved flow behaviour of the melt owing to higher heat input. This is the reason why 3-component- and 4-component-shielding gases, respectively, containing argon, carbon dioxide, hydrogen and/or helium are usually applied for MAG-welding of nickel base alloys. For MAG-welding of nickel base Alloy 602 CA (NiCr25FeAlY), a special 4-component-shielding gas with nitrogen portions is used.

3 Experimental procedure and materials

The hot cracking behaviour of materials during welding is determined in a complex manner by numerous influencing factors. In order to investigate the effect of individual parameters on hot crack formation, hot cracking tests under external loading need to be applied. These tests allow defined deformation of material specimens during welding causing mechanical loading of the solidified weld pool area. The weld pool area of real components experiences similar loading due to superposed thermomechanical shrinkage and strain motions throughout the entire component or structure, respectively. Strains in the weld pool area and the resulting strain rates, respectively, constitute a basic influencing factor for hot crack formation. The described hot cracking experiments were conducted using the Modified Varestraint Transvarestraint-Test (MVT-Test) introduced in the Federal Institute for Materials Research and Testing (BAM) as a reference procedure [3], Figure 1.

Figure 1. The Modified Varestraint Transvarestraint – Test during testing

In this externally loaded hot cracking test, a bead-on-plate weld is deposited by fully mechanical TIG-welding on the surface of a material specimen with the standard size of 100 mm x 490 mm x 10 mm. When the arc passes the specimen centre, the specimen is bent at constant rate around a die by means of a plunger with known radius. Due to this deformation, the specimen is subjected to strain of defined level and rate. By variation of the bending radius and of the bending direction (Varestraint – Transvarestraint), different types of external loading can be realized. Since the bending process takes place within a short period of time during welding, hot cracks originate in a restricted test zone of the MVT-specimen. As a result of the stress-strain distribution occurring during bending, the hot cracks are almost exclusively formed at the specimen surface. Therefore, metallographic preparation can be dispensed with in compliance with the standardized test evaluation method specified in DVS-Guideline 1004-2.

Both base materials and welding filler materials can be investigated with the MVT-Test. MVT-specimens for welding filler material investigation are cut out from a weld with the corresponding weld metal. The weld metal with a minimum width of 12 mm may be located transversely or longitudinally in relation to the geometry of the MVT-specimen, but must always be in its centre. During the MVT-Test, the primary weld metal is remolten by the TIG-torch and mechanically loaded during solidification.

The hot cracking resistance of a material is quantified in the MVT-test procedure by adding up the lengths of all cracks detected on the specimen surface at 25-fold light microscopic magnification. The total crack length measured after a test is plotted in an MVT-diagram versus the total strain. Based on the three diagram sectors defined as “hot crack-proof”, “increasing hot cracking susceptibility” and “hot crack hazard”, the hot cracking resistance of a material is assessed according to the sector membership of the plotted values for the respective set of parameters. Since the strain rates applied in the MVT-Test rise with increasing total strain, a relationship between hot crack formation and the respective critical deformation rate can also be seen from the MVT-diagram.

Table 1. Survey of the examined materials tabulated according to their application purpose

Table 2. Chemical composition of the examined base materials

Composition (wt-%) Material			Ni	Mo			Al	Nb		C
			Cr	Fe	Ti			Co	Si
NiCr25FeAlY			R				1.8-2.4			0.15-0.25
			22-24	8-11	0.1-0.2
NiCr23Co12Mo			R	8-10			0.6-1.5			0.05-0.1
			21-23	2	0.2-0.6			10-13	0.025
NiCr22Mo9Nb			R	8-10				3.2-3.8
			21-23	3
NiMo16Cr16Ti			R	14-17						0.009
			14.5-17.5	3	0.7			2	0.05
NiCr23Mo16Al			R	15-16.5			0.1-0.4
			22-24	1.5				0.3	0.1
NiMo28			R	26-30						0.01
			0.4-1	1.5-2				1	0.08
Ni36			35-37							0.03
			0.2	R					0.2
X10NiCrAlTi32-21			30-34				0.15-0.60			0.12
			19-23	R	0.15-0.60				1.00

Table 3. Designation and chemical composition of the examined welding filler materials

Composition (wt-%) Welding filler material	Ni	Mo	max. C	Nb	max. Y
	Cr	Fe	Co	Al	max. Zr
SG-NiCr25FeAlY FM 602 CA M-No. 2.4649 Nicrofer S 6025	60		0.25		0.12
	25	10		2	0.10
SG-NiCr22Co12Mo FM 617 M-No. 2.4627 Nicrofer S 5520	55	10	0.10
	20	1.5	11
SG-NiCr23Mo16 FM 59 M-No. 2.4607 Nicrofer S 5923	59	16	0.01
	23	1
SG-NiMo27 FM B2 M-No. 2.4615 Nimofer S 6928	69	28	0.01
	1	2
Ni36 FM 36 M-No. 1.3912 Pernifer S 6436	36		0.03
		64

This contribution presents investigations using the MVT-Test aimed at quantifying the hot cracking resistance of six different nickel base alloys, one high-alloyed steel and one iron-nickel alloy. It covers examinations of the hot cracking behaviour depending on the shielding gas applied to base material specimens of the high-temperature material Alloy 602 CA (NiCr25FeAlY, M-No. 2.4633) and of the wet corrosion material Alloy C4 (NiMo16Cr16Ti, M-No. 2.4610). The designations and chemical compositions of the investigated base materials are listed in Table 1 and Table 2. Weld metal investigations carried out using welding filler material for four of the six nickel base alloys and for the iron-nickel alloy are also discussed. The designations and the chemical compositions of the welding filler materials applied in the experiments are compiled in Table 3.

Figure 2. MVT-Diagram – Investigations on base material specimens; heat input = 7.5 kJ/cm; the test gas (shielding gas in the MVT-Test) is indicated in brackets.

4 Results

4.1 Base materials in the MVT-Test

Figure 2 shows the experimental results of the base material specimens made from Alloy 800 H, Alloy 59, Alloy 36, Alloy 625, Alloy 617 and Alloy B2. The selected heat input was 7.5 kJ/cm for all tests. From a direct comparison between these materials it was found that the high-alloyed steel Alloy 800 H has the lowest hot cracking resistance, since its total crack lengths are at the three strain levels of 1%, 2% and 4% in Sector 2 (increasing hot cracking susceptibility), and hence at all applied strain levels, exceeded the total crack lengths of all the other base material specimens tested. In the nickel base Alloy B2 and the iron-nickel Alloy 36, no hot cracking was detected at any strain level in the MVT-Test. These two alloys exhibit maximum hot cracking resistance. The total crack lengths of nickel base Alloy 59, Alloy 625 and Alloy 617, except Alloy 59 at the strain level of 4%, keep within Sector 1 “hot crack-proof” for all strain levels. This is to say that for Alloy B2, Alloy 36, Alloy 617, Alloy 625, and conditionally for Alloy 59, the risk of hot cracking must be regarded as low.

Figure 3. MVT-Diagram – Investigations on base material specimens for determining the influence of shielding gases on the hot cracking resistance of Alloy 602 CA.

For Alloy 602 CA, MVT-base material tests were conducted at total strain levels of 1% and 4% using seven different shielding gases in order to quantify the hot cracking resistance depending on different shielding gas compositions. Figure 3 shows the influence of these shielding gases on the MVT-result. With pure argon as shielding gas, total crack lengths were measured which at the strain levels of 1% and 4% occur in Sector 2 “increasing hot cracking susceptibility” near the borderline of Sector 3 “hot crack hazard”. When argon + 2% H₂ was used as shielding gas, a significant reduction of the hot cracking resistance was found. The measured total crack lengths are all within Sector 3 “hot crack hazard”. However, with nitrogen portions added to the shielding gas, a substantial increase in hot cracking resistance was noticed. For all investigated nitrogen-containing shielding gases (Ar + 1% N₂, Ar + 3% N₂, Ar + 5% N₂ + 5%He, Ar + 5% N₂ + 5% He + 0.5% CO₂, Ar + 5% N₂ + 5% He + 0.05% CO₂), the measured total crack lengths kept within Sector 1 which means “hot crack-proof”. Thus, the improving effect of nitrogen-containing shielding gas on the hot cracking resistance of Alloy 602 CA has been quantified by the MVT-Test.

Further investigations were carried out to answer the question of whether nitrogen is bound in the weld metal of this material during the welding process and, beyond it, of whether its effect on the hot cracking resistance of the weld metal is persistent. For that purpose, TIG-remelt welds were produced on the MVT-base material specimens using shielding gases with defined nitrogen portions. The specimens were subsequently examined in the MVT-Test with pure argon. The measured (short) total crack lengths were nearly identical to those of the reference specimens tested in the MVT-Test with the respective nitrogen-containing shielding gases. These results demonstrate that the nitrogen is firmly included in the weld metal and does not even lose its effect by a second overweld under pure argon. The questions of how the nitrogen is present in the melt and of whether it is dissolved in the crystal lattice or precipitates as a phase cannot be answered clearly on the basis of these experiments. But its influence is obviously not due to an effect from the gas phase on surface stresses.

For nickel base Alloy C4, MVT-tests were performed to determine the influence of shielding gas on the hot cracking resistance using the shielding gases pure argon, Ar + 5% N₂, Ar + 50% He and Ar + 2% H₂ and a heat input of 7.5 kJ/cm. When the shielding gases Ar, Ar + 50% He and Ar + 2% H₂ were applied, no hot cracking was found up to the maximum strain level of 4%. Only with the shielding gas Ar + 5% N₂, a very short total crack length of 1.67 mm was measured at the strain level of 4%. This material has therefore to be classified as “hot crack-proof” or “supremely hot crack-proof”, respectively, with any of the shielding gases examined in the MVT-Test. A dependence of this material on the applied shielding gas mixture cannot be ascertained definitely.

4.2 Weld metals in the MVT-Test

The assessment of the hot cracking resistance of welding filler materials is, as a general rule, more difficult than that of base materials. For example, welding filler materials are often not available as full materials, from which MVT-specimens could be prepared. Furthermore, molten base material and welding filler material intermingle and produce a melt which may exhibit a significantly different hot cracking behaviour from that of the base material or of non-remolten filler material separately.

This is the reason why the influence of the welding process, of the welding parameters and of the base material should be appropriately considered in the assessment of the hot cracking resistance of a welding filler material. The drawback of externally loaded hot cracking tests conducted immediately after welding a filling run is that they do not allow quantification of the stressing in the weld pool area with adequate exactness. Dual-step MVT-testing is therefore carried out for weld metal investigations. In a first step, so-called MVT-weld metal specimens consisting of base material and of the weld metal to be examined are cut out of a test piece. In the second step, this weld metal is remolten by a TIG-arc in the MVT-Test and mechanically loaded in accordance with the specified test procedure.

Figure 4. MVT-Diagram – Investigations on weld metal specimens of Alloy 59 + FM 59 with variation of the welding parameters; MVT-test parameters: heat input 7.5 kJ/cm, test gas (shielding gas for the MVT-Test) Ar 4.8.

The diagram in Figure 4 shows the influence of the welding procedure and of the applied welding parameters, respectively, during welding of the filling runs on the result of the MVT-Test. The examined welding filler material was FM 59 (M-No. 2.4607) in combination with the similar base material Alloy 59 (M-No. 2.4605). The filling runs were produced using three different welding procedures (pulsed MAG- and MIG-welding, each with a high and a low heat input, as well as SA-welding). It can be seen from the MVT-diagram that the measured total crack lengths of all specimens under 1% test strain and of the SA-welded specimen under 4% test strain, except the specimen pulsed MAG-welded with the high heat input, are within Sector 1 “hot crack-proof”. In sum, the SA-welded weld metal specimens exhibit a significantly lower hot cracking resistance compared to the pulsed MIG- and the pulsed MAG-welded specimens, whereas the comparison between the pulsed MAG- and the pulsed MIG-welded specimen showed only a minor difference in hot cracking resistance.

Figure 5. MVT-Diagram – Investigations on nickel base weld metal specimens; similar base material MVT-heat input = 7.5 kJ/cm; the test gas is indicated in brackets.

The MVT-diagram in Figure 5 shows results of investigations of the welding filler materials FM 602 CA, FM 36, FM 617 and FM B2. From the weld metal specimens of welding filler material FM 36, two specimen series were examined. Specimen series 1 was manually TIG-welded with the shielding gas Ar + 2% H₂. For specimen series 2, pulsed MAG-welding was applied with the shielding gas Ar + 2.5% CO₂. It should be noted at this point that MAG-welding of Alloy 36 with the filler material FM 36 is currently being qualified and is the subject of further research. Two specimen series with the welding filler material FM 617 were also examined. Specimen series 1 was pulsed MAG-welded with a helium content of 15 percent by volume in the shielding gas, specimen series 2 was also pulsed MAG-welded with a helium content of 30 percent by volume in the shielding gas. In the weld metal specimens of welding filler material FM 36 (specimen series 1 and specimen series 2) as well as FM B2, no hot cracks occurred at any of the tested total strain levels. This means that these welding filler materials must be categorized as “supremely hot crack-proof” for the investigated parameters. For the weld metal specimens of welding filler material FM 617 (specimen series 1 and specimen series 2) as well as FM 602 CA, the measured crack lengths are in Sector 1. These welding filler materials must therefore assessed as “hot crack-proof” for the investigated parameters. For the weld metal specimens of welding filler material FM 617, a marginal dependence of the hot cracking resistance on the applied shielding gas was observed. For specimen series 2 with a helium content of 30 percent by volume in the shielding gas, the total crack length at a strain level of 4% is significantly greater that the corresponding crack length of specimen series 1.

5 Discussion of hot crack formation in conjunction with the weld pool geometry

A distinctive feature in the investigations concerning the influence of nitrogen in the shielding gas on hot cracking under an external load from the Varestraint variant of the MVT-Test (bending in longitudinal direction to the weld) is the formation of so-called craters at the end of the weld in the material Alloy 602 CA, Figure 6.

Figure 6. Influence of the shielding gas on the shape of the solidification front at otherwise the same welding parameters

Craters at the end of a weld caused by the abrupt cutoff of the heat source reflect to a certain extent the shape of the weld pool, i.e. the shape of the solidification front at the surface. Pure argon engenders a flat solidification front. The nitrogen addition, by contrast, entails a markedly steeper front. The weld depth with the use of the nitrogen-containing shielding gas is greater (~ 3.6 mm) than that with the use of pure argon at the same welding speed (~ 2.3 mm). The nitrogen addition involves an increase in the welding voltage of around 10% at the same welding current and a slight diminution in the weld width. The power increase of 10% seems to be too small to explain this effect. Experiments have already been described in Section 4.1, in which a weld produced with nitrogen-containing shielding gas was subsequently subjected to an MVT-Test using pure argon. The crater at the end of the second weld produced with pure argon has the same steep solidification front as the weld obtained with nitrogen addition. Consequently, arc constriction as a result of the nitrogen addition is out of the question. It is more reasonable to deduce changed flow conditions in the weld pool due to changed surface stresses as a result of dissolved nitrogen in the melt, where applicable. Another possibility could be an increase in solidification temperature caused by the nitrogen addition. Investigations aimed at a conclusive interpretation are ongoing. In the following, potential implications of this effect for the formation of hot cracks are first of all discussed.

Figure 7 shows an instantaneous photograph from high-speed video recordings of hot crack formation at the solidification front in the MVT-Test. The crack originates from the solidification front, where it is connected with the melt, and grows with the progressing solidification front into the newly formed weld metal without being completely closed by the melt, which can be proven by the crack identification in the weld metal after cooling. It is therefore essential to establish the cause of the melt deficit which finally involves material separation and hot cracking, respectively. It should also be emphasized in this context that the crack appearances in the MVT-specimens relating to Figure 6 coincide very well with the craters at the end of the welds insofar as the directions of the crack origins are perpendicular to the respective solidification front. The crack paths take their course in the spaces between cells (dendrites, grains) which, except for deviations forced by epitaxy, grow perpendicularly to the solidification front, i.e. longitudinally to the temperature gradient.

Figure 7. Hot crack formation at the solidification front; material: Alloy 602 CA.

6 Conclusions

Considering the described test parameters, the MVT-base material investigations can be summarized as follows:

Alloy 36 (shielding gas pure argon) and Alloy B2 (shielding gas argon + 2%H₂) are “supremely hot crack-proof”.

Alloy 617 (shielding gas Ar + 2%H₂) and Alloy 625 (shielding gas pure argon) are “hot crack-proof”.

Alloy 59 (shielding gas pure argon) is “conditionally hot crack-proof”.

Alloy 800 H (shielding gas pure argon) has to be classified as being subject to an “increasing hot crack hazard”.

It has been proven that the hot cracking resistance of Alloy 602 CA is significantly influenced by the shielding gas composition.

Nitrogen additions in the shielding gas increase the hot cracking resistance and hydrogen additions in the shielding gas decrease the hot cracking resistance as compared to pure argon.

Considering the described welding and test parameters, the MVT-weld metal investigations can be summarized as follows:

Concerning welding filler material FM 59, the welds performed with the pulsed MIG- and the pulsed MAG-welding process have to be classified as “hot crack-proof” and “conditionally hot crack-proof”, respectively. The produced SA-weld is classified as “conditionally hot crack-proof”.

The results of the weld metal investigations for the welding filler materials FM 36 and FM B2 showed that these materials are “supremely hot crack-proof”. The results of the weld metal investigations for the welding filler materials FM 602 CA and FM 617 revealed that these materials have to be classified as “hot crack-proof”.

6 Literature

[1] Heubner, U., u.a.

Nickelwerkstoffe und hochlegierte Sonderstähle.

Kontakt & Studium, Bd. 153, Expert Verlag, Ehningen 1993

[2] Zinke, M., A. Hübner

Auswirkungen stickstoffhaltiger Schutzgase beim WIG - Schweißen hoch warmfester Nickelbasiswerkstoffe.

Der Praktiker 9 / 2002, S. 308-312. Fachzeitschrift des DVS, 54. Jahrgang, Deutscher Verlag für Schweißtechnik, Düsseldorf 2002

[3] Wilken, K., H. Heuser, Th. Hoffmann

Heißrissverhalten von Chrom - Nickel - Stählen und Nickelbasislegierungen. DVS-Berichte Bd. 183, S. 29-35. Deutscher Verlag für Schweißtechnik, Düsseldorf 1997