ZHUOJIN-LOGOZhuojin Pipe Fitting Co., Ltd
Zhuojin Pipe Fitting Co., Ltd
STEEL PIPE · FITTING · WELDING  |  NOTES FROM A 30-YEAR INDUSTRY VETERAN

Liquid Metal Penetration Cracking in 316L Stainless Steel During TIG Brazing with Copper-Based Filler: A Metallurgical Autopsy

Background: The ITER Thermal Anchor Application

The thermal anchor described in this paper is not an exotic laboratory curiosity; it is a critical component in the ITER magnet support system, tasked with intercepting heat flow from the support structure to the superconducting coils. The anchor is internally cooled by liquid helium, and it must simultaneously withstand cryogenic temperatures, strong electromagnetic forces, and repeated thermal cycling. The joint between the anchor and the stainless steel support must therefore deliver high strength, adequate toughness, and excellent thermal conductivity. TIG brazing with a copper-based filler was selected because copper offers outstanding thermal conductivity and the brazing process produces a sound, leak-tight joint. On paper, this is a sound engineering choice. In practice, it failed catastrophically during engineering testing, with cracks propagating through the 316L austenitic stainless steel tube and causing helium leakage. I have seen this class of failure before in dissimilar-metal joints where a low-melting-point filler metal is used in contact with a high-strength austenitic matrix, and the root cause is almost never a single factor.

The Failure Phenomenon: What Actually Happened

The authors report that during TIG brazing, the copper-based filler metal penetrated the grain boundaries of the 316L stainless steel and propagated along them, resulting in intergranular cracking and leakage. This is the classic signature of liquid metal embrittlement (LME) or liquid metal penetration, a phenomenon where a liquid metal wets and penetrates the grain boundaries of a solid metal, causing severe loss of ductility and premature fracture. The mechanism is not simple dissolution; it involves a combination of wetting, grain boundary diffusion, and stress-assisted crack propagation. The key observation is that the penetration occurred during the welding process itself, not during subsequent service, which immediately points to the thermal cycle and the associated residual stress state as primary contributors. In my experience, when a copper-based filler is used to braze austenitic stainless steel, the copper can penetrate the grain boundaries at temperatures above approximately 400°C, and the depth of penetration is strongly dependent on the time at temperature and the magnitude of tensile stress. The authors correctly identify that both diffusion penetration and welding-induced stress are contributing factors, and I would add that the combination is far more dangerous than either alone.

Metallurgical Investigation: Grain Boundary Penetration and Stress Interaction

The investigation employed optical metallography, scanning electron microscopy (SEM), and energy-dispersive X-ray spectroscopy (EDS) to characterize the joint specimens made with different filler metals and brazing methods. The results are instructive and align with what I have seen in similar failures. The metallographic sections show clear evidence of copper-rich phases along the austenitic grain boundaries, with crack-like features extending from the brazing interface into the base metal. The SEM images reveal a continuous network of grain boundary penetration, and the EDS spectra confirm the presence of copper at the crack tips. This is not a simple diffusion layer; it is a path of least resistance that follows the grain boundaries, which are inherently weaker and more chemically active than the grain interiors. The critical insight is that the penetration is not merely a diffusion phenomenon; it is driven by the stress state at the joint. When the joint is under tensile stress during cooling, the liquid copper is drawn into the grain boundaries, and as the joint solidifies, the copper acts as a wedge, propagating the crack further. This is a classic stress-assisted liquid metal penetration mechanism, and it explains why the failure occurred during welding rather than in service.

Analysis Method Key Findings Interpretation
Optical Metallography Copper-rich phases along grain boundaries; crack-like features at interface Grain boundary penetration is the dominant failure mode
SEM Continuous grain boundary network of penetration Penetration follows the path of least resistance
EDS Copper confirmed at crack tips and along grain boundaries Copper is the active penetrating species
Fracture Surface Intergranular, brittle appearance Loss of ductility due to liquid metal embrittlement

Root Cause Analysis: Why Copper Filler Penetrates 316L

The root cause of this failure can be traced to the fundamental metallurgical incompatibility between copper-based filler metals and austenitic stainless steel at elevated temperatures. Copper has a limited but non-negligible solubility in austenitic stainless steel, and at brazing temperatures, the copper can diffuse along grain boundaries, which are high-energy, high-diffusivity paths. The driving force for penetration is the reduction in surface energy when the liquid copper wets the grain boundaries, combined with the tensile stress that opens the boundaries and allows the liquid to advance. In the ITER thermal anchor application, the joint geometry and the restraint of the surrounding structure create a significant tensile stress state during cooling, which accelerates the penetration. The authors note that the welding stress is a critical factor, and I would emphasize that the stress state is not just a consequence of the thermal cycle but also of the joint design and the stiffness of the surrounding components. In my experience, the most effective countermeasure is to reduce the time at temperature and to minimize the tensile stress at the joint, either through joint design, filler selection, or post-weld stress relief. The paper's conclusion that both factors are important is correct, but the practical lesson is that you cannot fix this problem by addressing only one of them.

Practical Countermeasures and Lessons for Brazing Operations

For engineers working on similar dissimilar-metal joints, the lessons from this failure are clear and actionable. First, the filler metal selection must be based on a thorough assessment of the base metal compatibility, not just the thermal conductivity or the ease of brazing. Copper-based fillers are attractive for thermal applications, but they carry a significant risk of LME when used with austenitic stainless steel. Alternative fillers, such as silver-based or nickel-based alloys, may offer a better balance of thermal conductivity and compatibility, although they may require different brazing parameters. Second, the joint design must be optimized to minimize stress concentration and to allow for some compliance during the thermal cycle. A rigid joint with a large thermal mass will always be more susceptible to stress-assisted penetration. Third, the brazing process parameters must be controlled to minimize the time at temperature, particularly in the range where copper is liquid and the grain boundaries are susceptible. This means using the shortest possible brazing cycle and ensuring rapid cooling after the filler has solidified. Finally, post-weld inspection should include not only visual and leak testing but also metallographic examination of the joint interface, particularly if the joint is subjected to cyclic loading or cryogenic service. I have seen too many cases where a joint passes a leak test but fails after a few thermal cycles because the grain boundary penetration was present but not yet through-wall. The ITER experience is a stark reminder that in critical cryogenic applications, you cannot afford to take shortcuts in joint qualification.

Concluding Reflections

This paper is a valuable contribution to the understanding of liquid metal penetration in austenitic stainless steel brazed joints, and it highlights a failure mode that is often overlooked in design reviews. The key takeaway for me is that the combination of diffusion penetration and welding stress is a lethal combination, and that the solution requires a holistic approach that addresses both the metallurgical compatibility and the mechanical stress state. In my thirty years in this industry, I have learned that the most dangerous failures are rarely the result of a single root cause; they are almost always the result of a chain of events that individually seem manageable but collectively produce a catastrophic outcome. The ITER thermal anchor failure is a textbook example of this principle, and the lessons learned here should be applied to any brazed joint where a low-melting-point filler is used in contact with a high-strength austenitic matrix. I would recommend that any engineer working on similar applications review this paper carefully and incorporate the findings into their joint qualification procedure, because the cost of failure in a cryogenic system is far higher than the cost of a more rigorous qualification.