Effect of Microstructural Evolution on Plasticity of GH4065A Superalloy Cast Ingot During Homogenization Hot Treatment

Abstract

Improved plasticity in superalloy castings minimizes processing defects, reduces stress concentration, and enhances mechanical performance. To obtain the microstructure–plasticity relationship, GH4065A ingots were homogenized at 1140–1200 °C for 5–80 h. Microstructural analysis tracked the evolution of dendritic crystals and precipitates (including η phase, carbides, and borides). Tensile tests were conducted to assess plasticity in terms of elongation and reduction in area. Results show that increasing temperature accelerated dendritic dissolution. While 1140 °C was ineffective for short-term dendrite elimination, temperatures of 1160–1200 °C achieved near-complete dissolution within 30–60 h. Precipitates evolution was also observed: the η phase dissolved preferentially, while the sizes of carbides and borides gradually decreased, especially at 1200 °C. Electron probe microanalysis confirmed Nb as the most segregated element. With higher temperatures, Nb diffused from microsegregated zones toward homogeneity. Plasticity improved notably when the Nb segregation coefficient was ~1.5 but decreased at ~1. The optimal homogenization parameters were determined as 1180 °C for 15–60 h. This study provides key processing guidelines for GH4065A ingots, supporting enhanced service performance and operational safety of related components.

1. Introduction

The manufacture of engine components from wrought superalloys requires cogging [1]. However, the as-cast microstructure has low ductility. This not only limits the permissible deformation per heating cycle but also increases cracking susceptibility. To address this, heat treatments aimed at microstructural optimization are employed [2,3,4]. These treatments are tailored to the ingot’s grain and precipitate characteristics to enhance its plasticity [5,6]. Therefore, this research investigates microstructural evolution and plasticity changes during the homogenization of GH4065A ingots.

Generally, the Ni-based superalloy is composed of multiple metallic elements (e.g., Ni, Co, Cr, Al, and Ti) and non-metallic elements (e.g., B and C), featuring a typical microstructure consisting of a face-centered cubic (FCC) γ matrix and L1₂-structured γ′ precipitates [7,8,9,10]. The production of ingots typically involves a multi-step melting process: primary melting by vacuum induction melting (VIM), followed by secondary and tertiary refining using either vacuum arc remelting (VAR) or electroslag remelting (ESR) [11,12,13]. However, during the melting process, the distribution of elements can be influenced by various factors, including charge material preparation, furnace charging, melting, refining, and alloying procedures, which often result in microsegregation of both metallic and non-metallic elements [14,15]. When excessive segregation occurs, carbides and borides may precipitate, but topologically close-packed (TCP) phases, especially those with the most prominent characteristics, are the key precipitates to focus on [16,17,18,19,20]. These precipitates generally exhibit significant lattice mismatch and differences in shear and elastic modulus compared to the γ matrix and γ′ phases, promoting the formation of crystallographic defects and dislocation pile-ups at their interfaces [21,22,23]. During subsequent deformation of the ingot, these segregation-induced precipitates can severely restrict plastic deformation.

Numerous studies have investigated the relationship between homogenization heat treatment and improved plasticity in nickel-based superalloy ingots. Yang et al. suggested that homogenization temperature influences mechanical properties by regulating the balance among elemental diffusion and phase evolution. Increasing heat-treatment temperature can eliminate inter-dendritic brittle phases, but it also promotes pore coarsening, leading to non-monotonic changes in crack propagation resistance and elongation. Specifically, low temperatures result in residual brittle phases that reduce plasticity, while high temperatures cause a significant increase in porosity and the formation of large voids, inducing stress concentration and crack initiation, which similarly impair elongation [5]. Wu et al. observed severe elemental segregation in the as-cast microstructure of superalloys, which adversely affects plasticity. Specifically, Nb and Mo were found to segregate to inter-dendritic regions. The segregation index (δ), a measure of the extent of elemental segregation, is defined as the ratio of the element’s concentration in the inter-dendritic regions to that in the dendritic core regions. When δ > 1, it indicates that the element is enriched in the inter-dendritic regions compared to the dendritic core regions; when δ < 1, it means the element is depleted in the inter-dendritic regions relative to the dendritic core regions. In this study, the δ for Nb and Mo, which is less than 1, decreases rapidly within the first 10 h at 1160–1260 °C, followed by a slower decrease. [24]. Ma et al. reported significant segregation in the ingot, which was alleviated after homogenization, leading to a more uniform elemental distribution. Elements such as Ni, Co, Al, Fe, Cr, and W (with segregation coefficient > 1) tend to concentrate in dendritic cores, whereas Ti, Nb, and Mo (with segregation coefficient < 1) segregate to inter-dendritic regions. Clearly, elemental segregation is a common phenomenon in superalloy ingots and has a substantial impact on alloy plasticity [25]. The alloy (GH4065A) in this work is a new-generation wrought superalloy. To date, no systematic study has been conducted on the homogenization heat treatment of the ingot. Therefore, this research provides valuable engineering insights for the manufacturing of structural components.

2. Experiments

2.1. Materials and Methods

The experimental material was a GH4065A superalloy ingot with dendritic crystals and multiple precipitated phases. Its nominal chemical composition (wt.%) was shown in

Table 1. Nominal chemical composition of the Ni-based superalloy (wt.%).

Elements	C	Cr	Co	W	Mo	Al	Ti	Nb	B	Fe	Zr	Ni
wt.%	0.011	16.0	13.0	4.0	4.0	2.1	3.7	0.7	0.015	1.0	0.05	Bal.

. The ingot was produced by vacuum induction melting, electroslag remelting, and vacuum consumable remelting. In this section, an explanation of how solidification calculations were carried out is provided. The phase diagram from room temperature to 1400 °C for the alloy, calculated based on its chemical composition using JMatPro software, serves as the basis for these calculations.

The microstructure of the superalloy ingot before homogenization heat treatment and typical locations (center, 1/2R, and edge) were examined (line-intercept method) for their grain distribution and precipitated phases, as shown in Figure 1. Clearly, dendritic crystals with lengths and widths of approximately 300 to 400 μm were equiaxially distributed. Although slight differences were observed among the dendrites at the center, the position at half of the radius (1/2R), and edge locations, these variations were considered acceptable. Thus, higher-magnification characterization further revealed the presence of intragranular/inter- γ′ phases as well as η phases and carbides.

2.2. Microstructural Characterization

To elucidate the microstructure of the ingot and the effects of homogenization heat treatment, this study combined thermophysical analysis with experimental characterization. Differential scanning calorimetry (DSC, STA 449C, NETZSCH, Saarbrücken, Germany) analysis was performed using a simultaneous thermal analyzer to determine the phase transformation temperatures of the alloy. Based on these results, homogenization heat treatments were conducted at various conditions (1140–1200 °C for 5–80 h). The evolution of dendritic structures under the coupled influence of temperature and time as well as the dissolution processes of various precipitates were characterized using scanning electron microscopy (SEM, AMBER, TESCAN, Brno, Czech Republic). The degree of elemental segregation was quantified using electron probe microanalysis (EPMA, JXA-iSP100, JEOL, Tokyo, Japan). Measurements were taken at a minimum of 10 points within both the dendritic core and inter-dendritic regions for each condition, and the average composition for each region was calculated. An effective segregation ratio, R, was then determined using the following equation: R = C_{inter-dendritic}/C_{dendritic core}, where C represents the average concentration of a given element; R > 1 indicates enrichment in the inter-dendritic regions (behaviorally analogous to k < 1), while R < 1 indicates enrichment in the dendritic cores (analogous to k > 1). In this work, the term “segregation coefficient” (k) refers to this experimentally derived ratio R. Additionally, tensile tests were carried out to evaluate the influence of microstructural evolution on plasticity, determined by measuring the elongation and reduction of area.

3. Results and Discussion

3.1. Solidification Calculations

Based on the chemical composition, the phase diagram from room temperature to 1400 °C was calculated using JMatPro software, as shown in Figure 2. The phase diagram suggests that the theoretical solidification sequence for the ingot follows: Liquid (L) → L + γ (1348 °C) → L + γ + MC (1270 °C) → L + γ + MC + η (1190 °C). Notably, in addition to the γ matrix with face-centered structure and γ′ with L1₂ structure, topologically close-packed (TCP) phases (η: 1155–1190 °C, σ: 1155–1175 °C), carbides (MC: 1155–1270 °C, M₂₃C₆: 850–915 °C), and borides (M₃B₂: 1155–1170 °C) can also precipitate. Based on the aforementioned results and thermal–physical property analysis, this research conducted a homogenization heat treatment on the ingot at temperatures ranging from 1140 to 1200 °C for durations spanning 5 to 80 h.

3.2. Effect of Temperature on Microstructural Evolution

The phase evolution in superalloys is fundamentally linked to changes in their thermal properties. The specific heat capacity and corresponding DSC curves for the superalloy ingot at high temperatures are presented in Figure 3, where the heat flow differential directly responds to these transformations. It can be seen that the specific heat capacity curve exhibited distinct transitions at critical temperatures. Below approximately 1080 °C, it increased gradually, then decreased rapidly near 1100 °C. A subsequent sharp increase followed by a decrease was observed between 1140 °C and 1160 °C. Finally, a sharp increase in specific heat capacity began around 1300 °C, indicating a significant phase transformation. Correspondingly, the DSC curve confirmed this trend: an endothermic event was recorded between 1138 °C and 1164 °C, which evolved up to ~1180 °C, while another distinct thermal reaction initiated between 1288 °C and 1295 °C. By comparing with the equilibrium phase diagram calculated via JMatPro, the first endothermic event (~1136–1180 °C) can be attributed primarily to the dissolution of TCP phases and M₃B₂. The second, higher-temperature event (~1288–1295 °C) likely corresponds to the dissolution of more stable phases, such as primary MC-type carbides.

Based on the experimental results and considering a solidus temperature of about 1110 °C, temperatures around 1300 °C are deemed unsuitable for homogenization due to the risk of severe damage to the γ matrix and γ′ phases. In contrast, significant microstructural evolution was observed within the 1100–1200 °C range, with 1140 °C identified as the onset temperature for these changes. Notably, according to JMatPro calculations, the γ′ phase is expected to dissolve completely by approximately 1100 °C, which is consistent with the DSC signal ending around 1140–1160 °C. Consequently, the microstructural evolution observed above ~1140 °C is attributed primarily to the dissolution of precipitates (e.g., carbide, borides, and TCP phases). Hence, the optimal homogenization heat treatment should be conducted between 1140 and 1200 °C.

3.3. Dissolution Process of Dendrites and Precipitation Phases

Temperature is the decisive factor governing microstructural evolution rates. To this end, a representative heat treatment at 1140 °C for 80 h was conducted to preliminarily assess the microstructure’s sensitivity to temperature. The characteristics of dendritic grain at the center, 1/2R, and edge positions are shown in Figure 4. Specifically, we utilized the line-intercept method to measure the grain sizes at these positions and then calculated their average values for further analysis. Evidently, the dendritic grains at the center and 1/2R positions underwent significant dissolution, whereas those at the edge only partially dissolved. Subsequently, higher-magnification images were obtained to reveal the characteristics of precipitates after the 80 h homogenization treatment at 1140 °C, as shown in Figure 5. Carbides, borides, and η phases persisted in the ingot; however, the intragranular γ′ phases were noticeably dissolved at this temperature. Based on these findings, it is indicated that 1140 °C is insufficient to rapidly alter the grain structure or dissolve the precipitates.

The microstructural homogenization process was further studied by increasing aging temperatures. The corresponding evolution of dendritic grains in the central position after heat treatment at 1160, 1180, and 1200 °C for 5, 15, 30, 60, and 80 h, respectively, is shown in Figure 6. Clearly, after aging at 1160 °C for 30 h, the dendritic structures were significantly reduced and were almost completely eliminated after 80 h. At 1180 °C, the dendrites were markedly reduced following a 30 h treatment and were fully eliminated after 60 h. Notably, the size of the precipitates was considerably smaller compared to those formed at 1160 °C. When the temperature was increased to 1200 °C, the timeline for dendritic reduction and elimination was nearly identical to that at 1180 °C; however, the precipitate size was significantly smaller. Consequently, it is reasonable to postulate that the substantial grain coarsening observed after solution treatment at 1200 °C is primarily because of a reduced Zener pinning pressure. This reduction is driven mainly by the decreased volume fraction of grain-boundary pinning precipitates due to their partial dissolution.

Subsequently, characterization of the precipitates was also performed to determine the influence of aging temperature and time, as shown in Figure 7. At 1160 °C and up to 60 h, microstructures contained η phase, carbides, and borides. After 80 h, the η phase dissolved, though carbides and borides persisted at grain boundaries. At 1180 °C, the η phase was eliminated within 30 h, leaving carbides as the dominant phase; notches forming between carbides after 60 h suggest incipient decomposition. At 1200 °C, η phase dissolution completed after merely 15 h. Both carbides and borides remained but were significantly finer. Extending aging to 30 h resulted in spherical borides and decomposed, reduced carbides, with continued dissolution upon further aging. This pronounced acceleration of precipitate dissolution at 1200 °C implicates a substantial weakening of grain boundary pinning.

This result identifies that while both 1180 °C and 1200 °C are beneficial for dendritic refinement and precipitate dissolution, 1160 °C shows limited efficacy in modifying carbides. Considering that 1200 °C may cause excessive grain coarsening and plasticity loss due to rapid precipitate dissolution, a homogenization temperature ~1180 °C is proposed, offering an optimal balance between microstructural homogenization and grain size control.

3.4. Plasticity Enhancement Associated with Elemental Diffusion and Homogenization

3.4.1. Elongation and Reduction of Area Associated with Microstructure Evolution

To evaluate the effect of grain and precipitate evolution on the plasticity, homogeneous heat treatments were performed at 1160, 1180, and 1200 °C for durations of 5, 15, 30, 60, and 80 h. Subsequently, tensile tests were conducted on specimens taken from the axial and chordwise directions, respectively, and the corresponding elongation and reduction in area were measured (see Figure 8). After aging at 1160 °C for 5 to 80 h, the reduction in area for specimens taken from the axial and chordwise directions exhibited nearly-identical values, while the elongation exhibited significant differences. Compared to the initial specimens, both axial and chordwise specimens exhibited marked improvements in plasticity after just 5 h of aging at 1180 °C. Extending the duration to 15 h further enhanced elongation, with sustained high plasticity maintained up to 80 h. However, at 1200 °C, persistent differences in plasticity were noted, likely due to excessive dissolution of precipitates and grain growth induced by the elevated temperature. Consequently, aging at 1180 °C for 15 to 60 h represents the optimal heat treatment parameters for this ingot.

3.4.2. Effect of Temperature on Elemental Diffusion and Homogenization

During homogenization heat treatment, microstructural evolution involves elemental diffusion and redistribution, thereby reducing microsegregation [26,27]. To quantify this process, the effective segregation coefficients (k) for key elements were statistically analyzed via EPMA after aging at 1180 °C for 0 to 80 h (Figure 9). A coefficient k > 1 signifies that the element is enriched in the inter-dendritic regions, whereas k < 1 indicates enrichment in the dendritic cores (thermodynamic k > 1). After aging, Nb exhibited the most significant decrease in its coefficient (towards 1), followed by Ti and W, while other elements remained largely unchanged. This indicates that Nb, with an initial k >> 1, had the most severe inter-dendritic segregation in the as-cast ingot, and its diffusion during homogenization was the most pronounced. The strong inter-dendritic segregation of Nb and Ti is a key driver for the formation of MC-type carbides in these regions. Conversely, the core-enriched W contributes primarily to solid solution strengthening in the dendritic regions.

Furthermore, to determine the influence of temperature on Nb diffusion, the effective diffusion coefficient of Nb was calculated using Equations (1)–(3) [28,29]. The values of D₀^eff and Q^eff can be influenced by the contents of an individual element and are defined by Equations (2) and (3). The parameters required for calculations can be found in references [30,31,32]. The calculated D^eff of Nb at 1140, 1160, 1180, and 1200 °C was determined to be 2.76 × 10⁻¹⁶ m²·s⁻¹, 3.79 × 10⁻¹⁶ m²·s⁻¹, 6.43 × 10⁻¹⁶ m²·s⁻¹, and 1.08 × 10⁻¹⁵ m²·s⁻¹, respectively. It can be observed that when the temperature exceeds 1160 °C, the diffusion coefficient approximately doubles with every 20 °C increase. Consequently, at 1200 °C, the elemental diffusion rate may become excessively high, which can affect grain size and indirectly alter the grain structure. Taking all factors into consideration, processing at 1180 °C ensures elemental homogeneity while enabling a well-optimized microstructure.

$$D_{\mathit{eff}}=D_0^{eff}\mathit{exp}\{-{\displaystyle \frac{Q_{eff}}{RT}}\}$$

(1)

$$Q_{\mathit{eff}}=Q_{Ni}+{{\displaystyle \sum }}_m{x}_m{Q}_{m,Ni}$$

(2)

$$D_0^{eff}=1/({{\displaystyle \sum }}_m\frac{x_m}{D_0^{m,Ni}})$$

(3)

where the D₀^eff and Q^eff represent the pre-exponential coefficient and activation energy, respectively, while T denotes the temperature, and R is the ideal gas constant (8.314 J/mol·K). Q_Ni represents the activation energy for self-diffusion of pure nickel, Q_m,Ni is the activation energy for inter-diffusion of solute m in nickel, D₀^m,Ni is the pre-exponential factor for solute m in nickel, and x_m is the atomic concentration of element m.

The distribution of Nb was characterized by EPMA after aging at 1180 °C for 5, 15, 30, 60, and 80 h, with the results shown in Figure 10. After 15 h of aging, the segregation of Nb in the ingot was significantly reduced. When the aging time was further increased to 30 and 60 h, segregation persisted only in localized areas. Ultimately, after 80 h, the segregation was virtually eliminated. To elucidate the intrinsic relationship between plasticity and Nb homogenization, the reduction in area and the Nb segregation coefficient were determined based on experimental data and characterization results for aging temperatures from 1140 to 1200 °C and times from 0 to 80 h, as shown in Figure 11. A comparative analysis revealed that when the Nb segregation coefficient approaches 1.5, the plasticity is markedly improved. Conversely, as the coefficient continues to decrease and nears 1, the plasticity begins to deteriorate, although it remains substantially higher than that of the original ingot. Thus, the combined effects of sustained temperature and duration effectively reduced its segregation, which contributes to the improvement of the plasticity. As stated above, this phenomenon shows a strong correlation with the alloy’s plasticity evolution. It is crucial to emphasize that while the elimination of Nb segregation (i.e., a segregation coefficient approaching 1) through homogenization generally enhances overall plasticity by creating a more uniform matrix, achieving an overly uniform state can have a dual effect on Nb-containing MC-type carbides (e.g., NbC). First, it removes the strong chemical driving force for their localized formation. Second, prolonged high-temperature exposure can lead to their coarsening and a reduction in their number density due to Ostwald ripening and partial dissolution. An unfavorable size and distribution of these carbides can then act as crack initiation sites, thereby deteriorating ductility. This suggests that an optimal balance exists during heat treatment between segregation reduction and precipitate morphology control.

4. Conclusions

Homogenization heat treatment at 1140–1200 °C for 5–80 h significantly improved the plasticity of the superalloy ingot. This work systematically examined the grain structure and precipitate characteristics after aging and correlated the microstructural evolution with plasticity indicators. It was determined that Nb is the primary segregating element, and its homogenization contributes markedly to the enhancement of plasticity. The main conclusions are summarized as follows:

• The as-cast superalloy ingot exhibits a microstructure composed of dendritic crystals and various precipitates, including carbides, borides, and η phases;
• Temperature defines the evolution of grains and precipitates. An increase to 1160 °C significantly modifies the microstructure, while at 1200 °C, precipitates undergo rapid and extensive dissolution, leading to pronounced changes in carbide size and volume fraction;
• The homogenization heat-treatment revealed Nb exhibited the most pronounced segregation tendency, followed by Ti and W, with negligible changes in other elements. This confirms that Nb segregation is the most severe in the ingot and has a detrimental effect on the plasticity;
• The aging window of 15 to 60 h at 1180 °C was identified as optimal for grain and precipitates. It enables effective elemental homogenization (Nb) and controlled precipitation, which consequently leads to a remarkable improvement in the plasticity.