GTAW Surfacing of UNS N10003 on 316H: Parameter Optimization and Dilution Control
Literature Overview and Core Finding
This paper by Yang Fei, Li Chaowen, and colleagues from Shanghai University of Science and Technology and the Shanghai Institute of Applied Physics addresses a critical issue in the construction of molten salt reactors: how to apply a corrosion-resistant nickel-based alloy (UNS N10003) onto a structural stainless steel (316H) using a cost-effective welding process. The authors used tungsten inert gas (GTAW) welding to deposit UNS N10003 onto 316H stainless steel, optimized the welding parameters, and characterized the resulting microstructure and hardness.
The key findings are:
- Dilution decreases with increasing wire feed speed, while deposit height increases.
- When the wire feed speed is constant, changing the current produces a deposit height variation of 0.2-0.5 mm.
- The interface can be divided into three zones: convective mixing zone (WM), non-convective mixing zone (UZ), and heat-affected base metal zone (BM).
- The deposit is primarily austenitic, with Mo-rich M2C carbides in the WM zone and δ-ferrite in the UZ zone.
- The hardness of the 316H base metal is (160±10) HV, while the WM zone is (202±11) HV.
Process Window and Parameter Optimization
The Dilution Problem
In dissimilar metal welding, dilution is the enemy. The base metal (316H) contains less nickel and molybdenum than the filler metal (UNS N10003). If dilution is high, the deposit will have a lower nickel content, reducing its corrosion resistance. The authors' finding that dilution decreases with increasing wire feed speed is critical.
| Parameter | Effect on Dilution | Effect on Deposit Height |
|---|---|---|
| Wire feed speed ↑ | Dilution ↓ | Deposit height ↑ |
| Current ↑ | Dilution ↑ (implied) | Deposit height varies 0.2-0.5 mm |
Practical Implications
From my experience, the wire feed speed is the primary control for dilution in GTAW surfacing. A higher wire feed speed means more filler metal is deposited per unit of arc energy, which reduces the proportion of base metal in the deposit. This is a simple but effective way to control dilution.
However, there is a limit. If the wire feed speed is too high, the filler metal may not melt completely, leading to lack of fusion. The authors' optimization of parameters is essential to find the sweet spot.
Microstructural Zones and Dilution Effects
The Three-Zone Interface Model
The authors' identification of three distinct zones at the interface is a valuable contribution. This model helps explain the microstructural evolution during dissimilar metal welding.
| Zone | Description | Microstructure |
|---|---|---|
| WM (convective mixing zone) | Fully mixed weld metal | Austenitic, with Mo-rich M2C carbides |
| UZ (non-convective mixing zone) | Partially mixed zone | Austenitic with δ-ferrite |
| BM (heat-affected base metal zone) | Base metal affected by heat | 316H austenitic structure |
Why the UZ Zone Matters
The UZ zone is the most critical for corrosion resistance. This zone has a composition that is between the base metal and the filler metal. It may have insufficient nickel for the intended service environment. In molten salt service, this could lead to localized corrosion at the interface.
Carbide Precipitation
The Mo-rich M2C carbides in the WM zone are a concern. Molybdenum is added to nickel-based alloys for corrosion resistance, but when it forms carbides, it is no longer available for solid solution strengthening. This can reduce the corrosion resistance of the deposit. The authors found that these carbides are small and distributed on dendrite boundaries, which explains the higher hardness of the WM zone.
Engineering Practice and Challenges
Molten Salt Reactor Applications
The motivation for this work is the construction of molten salt reactors, which use a fluoride salt as both coolant and fuel carrier. The salt is extremely corrosive, and the structural material must have excellent corrosion resistance. UNS N10003 (also known as Hastelloy N) was developed specifically for this application.
Practical Considerations
- Preheat and interpass temperature: For 316H base metal, preheating is generally not required. However, the interpass temperature should be controlled to prevent excessive heat buildup.
- Post-weld heat treatment: This is a critical question. The authors do not discuss PWHT, but for a dissimilar metal joint, PWHT can cause carbide precipitation and reduce corrosion resistance. In many cases, PWHT is avoided for this type of joint.
- Multi-pass welding: The authors note that this work provides a theoretical basis for multi-pass, multi-layer welding. In practice, I would expect to use 2-3 layers to achieve the required deposit thickness and composition.
Key Questions and Reflections
What the Paper Does Not Tell Us
The paper does not provide corrosion test data. The authors optimized parameters and characterized the microstructure, but they did not test the corrosion resistance of the deposit. This is a significant gap. The whole point of using UNS N10003 is corrosion resistance, and the paper does not demonstrate that the deposit actually has the required corrosion resistance.
My Practical Judgment
The parameter optimization is sound, and the microstructural characterization is thorough. However, I would want to see corrosion test results before specifying this process for a critical application. The dilution control is the key issue, and the authors have demonstrated that wire feed speed is an effective control. But dilution is not the only factor. The carbide precipitation in the WM zone could be a problem, and the δ-ferrite in the UZ zone could be a site for selective corrosion.
Summary
This paper provides a solid foundation for the GTAW surfacing of UNS N10003 onto 316H stainless steel. The parameter optimization is practical and the microstructural characterization is thorough. The three-zone interface model is a valuable contribution to understanding dissimilar metal welding. However, the paper lacks corrosion test data, which is the critical performance criterion for this application. I would recommend that the authors extend this work to include corrosion testing, particularly in molten salt environments, before the process is applied to a real reactor component. The dilution control strategy is sound, but the microstructure needs to be validated against the service environment.
Zhuojin Pipe Fitting Co., Ltd