Analysis of Layered Wide-Gap Brazing of Ni-Based Superalloys: Microstructural and Process Evolution

Abstract

Conventional Wide-gap brazing (WGB) has been widely utilized in the repair of hot-section turbine componentry, especially for difficult-to-weld Ni-based superalloys due to its ability to repair defects upwards of 250 µm. However, for larger repairs, its intermixed nature leads to the formation of deleterious intermetallics and residual porosity, rendering the repair ineffective. Recently, a layered variant of WGB, layered wide-gap (L-WGB) has been adopted, allowing for repairs in the mm range to be realized. While the microstructural and process evolution of conventional WGB has been explored in detail, much is still unknown about the L-WGB variant. Thus, this study aims to gain a deeper understanding of the microstructural and process evolution of L-WGB through the implementation of differential scanning calorimetry, electron microscopy, X-Ray microscopy, and computational thermodynamic calculations, and simulations using the MAR-M247/BNi-9 system.

First, initial thermal characterization of the braze alloy and the layered brazements revealed differences in melting behavior, suggesting the presence of a two-stage infiltration sequence. Thus, thermal analysis, 3-D X-Ray microscopy, and computational thermodynamic simulations were employed to investigate the mechanisms in which the two-stage infiltration sequence manifests. To begin with, it was found that significant diffusional homogenization between braze alloy and MAR-M247 particles, occurred prior to liquation. As a result, only about 15% of the braze melted during Stage 1 of layered wide-gap brazing as compared to approximately 49% when only the melting of pure BNi-9 was considered for the same temperature interval. Stage 1 melting began at the ternary eutectic transformation temperature, 1053°C, with the complete melting of Ni3B and partial melting of γ-Ni. Upon liquation, the braze liquid readily infiltrated the MAR-M247 skeleton via capillary action. As a result of rapid boron diffusion into the MAR-M247 skeleton from the infiltrating braze, extensive base metal dissolution and eventual diffusional solidification occurred, with Stage 1 infiltration terminating at 1073°C.

Stage 2 melting and infiltration began at 1102°C with the re-melting of the partially infiltrated braze as well as melting of remaining un-infiltrated braze atop the layered brazement, with near complete infiltration (~98%) attained by 1150°C (546 s). Further, it was found that braze infiltration kinetics varied between Stages 1 and 2, with 58% and 40% of the MAR-M247 skeleton infiltrated after 120 s and 228 s, respectively. Rapid infiltration kinetics during Stage 1 of braze infiltration were attributed to strong capillary forces and increased liquid volume from MAR-M247 dissolution. Conversely, Stage 2 infiltration was sluggish, likely due to diminished MAR-M247 dissolution kinetics and evolving pore structure. Moreover, reduced dissolution kinetics likely limited liquid generation and may have enabled particle re-arrangement, further altering the pore connectivity.

Next, using similar experimental techniques, the influence of brazing time and temperature on the microstructural evolution of L-WGB was investigated. Once liquefied, rapid braze infiltration into the MAR-M247 skeleton occurred via capillary action. During infiltration, partial and complete dissolution of the MAR-M247 skeleton occurred, which led to diffusional solidification at 1068°C. Upon further and complete infiltration, it was found that rapid densification was achieved prior to isothermal brazing temperatures. The post-braze microstructure contained γ -Ni matrix grains, precipitated Cr,W,Mo-rich MxBy borides, athermal solidification products along matrix grain boundaries and triple junctions, as well as internal porosity. It was found that brazing temperature dictated the athermal solidification products with binary eutectic (CrB + γ -Ni) at 1150°C and ternary eutectic (CrB + γ -Ni + Ni3B) at 1180°C and 1205°C. These findings agreed with Scheil-Gulliver predictions. Brazing time influenced the compositional homogeneity of the braze liquid, altering solidification behavior. This resulted in higher and lower solidification ranges for shorter and longer brazing times, respectively. Further, it was found that liquid fraction within the brazement increased with both brazing temperature and time, suggesting a persistent liquid phase. This finding was accompanied by an increase in volume fraction of athermally solidified intermetallics, consistent with an increase in liquid phase with increased brazing time and temperature. Lastly, γ -Ni grain growth occurred, although heterogeneity between the upper and lower regions of the brazement was observed. The upper region displayed larger grains on average when compared to the lower region. This was attributed to early breakdown of the diffusion affected zone, and subsequent boride migration during liquid infiltration. These borides may have hindered γ -Ni grain growth via a grain boundary pinning mechanism.

Finally, 3-D X-ray microscopy was utilized to evaluate the effect of brazing time and temperature on residual porosity and densification, post-infiltration. All samples possessed high densification, with residual porosity and lateral shrinkage values of 1.09-1.25% and 9.67-11.67%, respectively for brazements brazed at 1180°C and 1205°C with 0- and 60-minute isothermal holds. High degrees of densification were attributed to compressive forces early on from the infiltrating braze. Additionally, no significant difference in residual porosity was observed for the brazing conditions selected. All brazements contained fine to moderately fine porosity (~ 5.77-20 µm), uniformly distributed throughout. Larger pores were also present within the brazements, with low aspect ratios pores found in the upper regions of the brazements for brazements brazed at 1180°C and 1205°C + 0-minute holds, and moderate aspect ratio pores in the lower regions of the brazement when brazed at 1205°C + 60-minute holds. It was found that brazing temperature and time influenced the pore characteristics. With increased brazing temperature and time, there was a reduction in the overall pore number density and increased average diameter, suggesting pore coalescence and coarsening. Additionally, large pores exhibited a morphological shift with increased brazing time and temperature, i.e. from low to high aspect ratio. This was likely driven by an increase in liquid fraction, promoting pore filling and reshaping.