Transient Liquid Phase Diffusion Bonding of Heat-Resistant Steel Pipes with TP304 and T91 Substrates

Literature Overview

This paper by Chen Sijie, Li Xingeng, and Jing Xiaotian (2005), published in Welding Technology, investigates the transient liquid phase (TLP) diffusion bonding of heat-resistant steel pipes using three different interlayer foils: BNi2, Fe78Si9B13, and FeNiCrSiB. The substrates are TP304 stainless steel and T91 ferritic-martensitic steel, both of which are widely used in high-temperature power plant applications. The study examines the microstructure, mechanical properties, and elemental distribution of the bonded joints to determine optimal process parameters.

Core Technical Content

TLP bonding is a solid-state joining technique that involves a brief liquid phase formation followed by diffusion-driven solidification, resulting in a joint with properties approaching those of the base material. The process is particularly attractive for joining dissimilar metals, such as austenitic stainless steel (TP304) and ferritic-martensitic steel (T91), where conventional welding often produces brittle intermetallic phases.

The microstructural analysis reveals that the BNi2 and FeNiCrSiB interlayers produce narrow, diffusion-controlled bond lines with minimal intermetallic compound formation. The elemental distribution maps show a gradual transition from the base metal composition to the interlayer composition, indicating effective diffusion during the holding period. In contrast, the Fe78Si9B13 interlayer produces excessive intermetallic phases that embrittle the joint.

Process and Standards Analysis

The bonding process was conducted under argon atmosphere to prevent oxidation. The key process parameters—temperature, time, and interlayer thickness—were optimized through a series of trials. The optimal window identified is approximately 1150°C for 120 minutes with an interlayer thickness of 50-100 μm. This temperature is below the solidus temperature of both substrates, ensuring that only the interlayer melts during the initial stage of the process.

From a standards perspective, TLP bonding is not covered by conventional welding codes such as ASME Section IX or EN ISO 15614. The qualification of TLP joints for pressure vessel or piping applications requires a custom qualification procedure based on the specific process parameters, substrate materials, and service conditions. The paper's approach of characterizing the joint microstructure and mechanical properties provides the basis for such a qualification procedure.

The use of BNi2 as an interlayer for TP304 is notable because nickel-based interlayers are commonly used in TLP bonding of nickel alloys. The successful application to austenitic stainless steel demonstrates the versatility of this approach. For T91, the FeNiCrSiB interlayer is more appropriate because it provides better compositional compatibility with the ferritic-martensitic matrix.

Integration with Engineering Practice

In power plant applications, the dissimilar metal joint between TP304 and T91 is a common challenge. TP304 is used for its excellent corrosion resistance in high-temperature aqueous environments, while T91 is preferred for its superior creep strength at elevated temperatures. Conventional welding of these materials produces a heterogeneous weld zone with significant residual stresses and potential for intergranular cracking.

TLP bonding offers an alternative to fusion welding for joining these materials. The joint produced by TLP bonding has a more homogeneous microstructure and lower residual stresses compared to a fusion weld. However, the process requires precise control of the interlayer thickness and bonding parameters, which can be challenging for large-diameter pipes. The paper's results suggest that TLP bonding is feasible for heat-resistant steel pipes, but further development is needed to scale the process to industrial dimensions.

Key Questions and Reflections

A significant practical concern is the cost of the interlayer foils. BNi2 and FeNiCrSiB are expensive materials, and the need for precise thickness control adds to the process complexity. For high-volume applications, the cost advantage of conventional welding may outweigh the performance benefits of TLP bonding.

Another concern is the effect of thermal cycling on the TLP joint. Power plant components are subjected to repeated start-up and shutdown cycles, which induce thermal stresses. The paper does not address the fatigue performance of the TLP joint under cyclic loading. Future work should evaluate the fatigue resistance of the joint under realistic thermal cycling conditions.

The elemental distribution maps show that the bond line width is approximately 100-200 μm after bonding. This is relatively narrow, which is favorable for maintaining the mechanical properties of the joint. However, the bond line may be susceptible to corrosion attack if the service environment is aggressive. Engineers should consider the corrosion resistance of the bond line in the context of the specific service conditions.

Study Insights and Implications

This paper demonstrates that TLP bonding is a viable technique for joining heat-resistant steel pipes, producing joints with mechanical properties equal to or exceeding those of the base material. The selection of the appropriate interlayer is critical: BNi2 and FeNiCrSiB are effective for TP304, while FeNiCrSiB is the preferred choice for T91. The Fe78Si9B13 interlayer is unsuitable for both substrates due to excessive intermetallic formation.

The process parameters identified—1150°C for 120 minutes under argon atmosphere—provide a practical starting point for engineers developing TLP bonding procedures for heat-resistant steel pipes. However, the qualification of this process for code applications requires additional testing, including fatigue, creep, and corrosion resistance evaluations. The work represents a meaningful contribution to the joining of dissimilar high-temperature alloys and opens avenues for further research on the long-term performance of TLP joints in service.