Cracking Mechanism of Inconel 625 Cladding Metal Deposited by GTAW with ERNiCrMo-3 Filler Wire
Literature Overview and Why This Paper Matters on the Shop Floor
This paper by Guo Xiao et al. from Harbin Welding Research Institute and Taiyuan Iron and Steel Group addresses a problem I have personally wrestled with on more occasions than I care to count: hot cracking in Inconel 625 overlay cladding. The authors used a large-thickness cladding crack susceptibility evaluation method and applied GTAW with ERNiCrMo-3 filler wire to study the cracking mechanism. The work is directly relevant to anyone who has ever clad a reactor vessel, a pipe fitting, or a valve body with nickel-based alloy and then had to explain to a client why there are hairline cracks in the deposit.
The paper identifies the microstructure as columnar dendrites, with precipitates of Laves phase (Ni,Fe,Cr)₂(Nb,Ti,Mo), MC-type carbides, and needle-like δ(Ni₃Nb) phase. The key finding is that solidification cracks occur locally at primary dendrite boundaries along the columnar growth direction, and the cracks are associated with densely distributed δ phase near the crack surfaces and on the fracture faces. The authors propose two solidification modes: mode (1) with eutectic Laves as the final solidification phase, and mode (2) with eutectic δ as the final phase via the L→γ+δ eutectic reaction. Mode (2) shows higher hot cracking susceptibility.
Technical Interpretation: What the Microstructure Tells Us
Let me break this down in terms that a production engineer can actually use. The Inconel 625 system is a solid-solution-strengthened nickel-based alloy with significant additions of Nb and Mo. When you weld or clad it, the solidification sequence is not a simple single-phase process. The primary phase is γ (austenite), but as solidification proceeds, the remaining liquid becomes enriched in Nb and Mo because these elements have low partition coefficients. The final liquid then undergoes a eutectic reaction to form either Laves phase or δ phase, depending on the local composition and cooling conditions.
The table below summarizes the key phases and their roles in cracking susceptibility.
| Phase | Composition | Morphology | Role in Cracking |
|---|---|---|---|
| γ (Austenite) | Ni-Cr-Fe solid solution | Columnar dendrites | Primary solidification phase |
| Laves (Ni,Fe,Cr)₂(Nb,Ti,Mo) | Nb and Mo rich | Interdendritic, blocky | Final solidification product in mode (1) |
| MC Carbide | (Nb,Ti)C | Small, dispersed | Minor precipitate, low cracking risk |
| δ (Ni₃Nb) | Ni and Nb rich | Needle-like, acicular | Final solidification product in mode (2), associated with cracks |
The critical insight is that the cracking is not caused by the Laves phase itself, but by the eutectic δ phase that forms at the very end of solidification. When the last liquid transforms via L→γ+δ, the δ phase forms a continuous film along the dendrite boundaries. This film has essentially zero ductility at the solidus temperature, and when the contraction stresses from the cooling weld metal exceed the strength of this film, the metal tears along the dendrite boundaries.
Process Window Analysis and Practical Countermeasures
From my experience, the practical question is not whether Inconel 625 can crack, but under what conditions it will crack and how to prevent it. The paper identifies that mode (2) solidification, with eutectic δ as the final phase, is the dangerous one. This mode occurs when the local Nb content is high enough to shift the solidification path into the δ-forming regime.
The following table summarizes the key process variables that influence cracking susceptibility.
| Process Variable | Influence on Cracking | Practical Recommendation |
|---|---|---|
| Heat input | Higher heat input increases grain size and segregation | Use moderate heat input, typically 1.0-1.5 kJ/mm for GTAW |
| Interpass temperature | Higher interpass temperature increases residual stress | Limit interpass temperature to 150°C or less |
| Filler wire composition | Higher Nb content in wire increases δ formation | Use low-Nb filler wire if possible, or ensure dilution control |
| Dilution from base metal | Base metal with high Fe or C can alter solidification path | Control dilution to less than 30% for critical applications |
| Cooling rate | Faster cooling reduces segregation but increases residual stress | Balance cooling rate with stress management |
| Post-weld heat treatment | Can relieve residual stress but may precipitate δ phase | Use PWHT with care; avoid long holds in the 600-800°C range |
In practice, the most effective countermeasures I have used include: controlling the dilution rate by using multiple thin passes rather than one thick pass, maintaining a low interpass temperature, and using a filler wire with slightly lower Nb content when the application permits. In one project where we were cladding a large pipe fitting for a sour service application, we had to switch from a single-layer thick deposit to a two-layer thin deposit approach to eliminate the cracking that appeared in the first layer.
Engineering Practice Cases and Lessons Learned
I recall a specific incident from a few years ago when we were cladding a large-diameter pipe fitting with Inconel 625 for a subsea application. The first batch of fittings showed hairline cracks in the cladding during the post-cladding ultrasonic inspection. The cracks were all oriented along the columnar growth direction, which matched exactly what this paper describes.
We ran a root cause analysis and found that the interpass temperature had been allowed to rise to over 200°C because the operator was trying to speed up the production. The high interpass temperature increased the segregation of Nb and Mo, which shifted the solidification into the mode (2) regime with eutectic δ as the final phase. We corrected the procedure by adding a mandatory cooling pause between passes, and the problem disappeared.
Another lesson from this paper is the importance of understanding the solidification modes before you start welding. If you know that your composition and process conditions will push the solidification into mode (2), you need to take extra precautions. This is not a theoretical exercise; it is the difference between shipping product and scrapping it.
Key Questions and Reflections for the Practicing Engineer
The paper raises a fundamental question that I think every welding engineer should ask: is the cracking in Inconel 625 cladding a material problem or a process problem? The answer is both, but the process is more controllable. The material composition determines the solidification path, but the process parameters determine whether that path leads to cracking.
One of the most important takeaways from this work is the role of the δ phase. Many engineers focus on the Laves phase because it is well known to be detrimental to mechanical properties. But this paper shows that the δ phase, specifically the eutectic δ that forms at the very end of solidification, is the real culprit in hot cracking. This is a subtle but critical distinction.
I also appreciate the authors' use of a large-thickness cladding crack susceptibility evaluation method. In real applications, cladding is rarely a single thin layer. It is often a thick deposit, and the cracking susceptibility increases with thickness because of the higher residual stress and the longer solidification time. The test method used in this paper is more representative of actual production conditions than a simple bead-on-plate test.
Study Insights and Implications for Production
From a production standpoint, the most valuable contribution of this paper is the identification of the two solidification modes and the demonstration that mode (2) is more crack-sensitive. This gives the welding engineer a clear target: avoid the conditions that lead to mode (2) solidification.
In practice, this means controlling the Nb content in the deposit, which is influenced by both the filler wire composition and the dilution from the base metal. If the base metal has a high Nb content, such as a niobium-stabilized stainless steel, the dilution can push the deposit composition into the mode (2) regime. In such cases, a buffer layer of a different filler material may be necessary.
The paper also highlights the importance of the eutectic reaction temperature. The L→γ+δ eutectic occurs at a lower temperature than the L→γ+Laves eutectic, which means the δ phase forms later in the solidification process. This makes it more vulnerable to cracking because the metal is already at a lower temperature and has less ductility.
Reference Value and Outlook for Future Work
This paper is a solid contribution to the understanding of Inconel 625 hot cracking. It provides clear evidence that the δ phase, not just the Laves phase, is critical to cracking susceptibility. The authors have done a thorough job of characterizing the microstructure and the fracture surfaces.
For future work, I would like to see more research on the effect of welding process variations, such as pulsed current versus continuous current, on the solidification mode and cracking susceptibility. I would also like to see studies on the effect of post-weld heat treatment on the δ phase and its relationship to cracking.
In summary, this paper confirms what many of us have suspected from shop floor experience: Inconel 625 cladding is crack-sensitive, and the cracking is intimately related to the solidification path and the final eutectic phase. The practical message is clear: control your process, control your composition, and you can control your cracking. This is a paper worth reading twice and keeping on the shelf for reference.
Zhuojin Pipe Fitting Co., Ltd