In recent years, the aerospace industry has increasingly adopted high-strength ZL205A aluminum alloy for structural components, particularly large axisymmetric bodies such as rocket engine casings and missile shells. However, the occurrence of segregation defects in these castings has become a serious concern, leading to costly rejections and potential safety hazards. This research focuses on understanding the origin and characteristics of segregation defects in large ZL205A axisymmetric castings and evaluates the effectiveness of optimized casting techniques and grain refinement in mitigating these defects. The study employs scanning electron microscopy (SEM), energy dispersive spectroscopy (EDS), differential scanning calorimetry (DSC), and finite element method (FEM) simulations. Key findings indicate that segregation defects mainly manifest as fish-scale patterns located at the junction between the casting body and slit gates, with coarse Al₂Cu eutectic phases distributed along grain boundaries. These defects reduce tensile strength and elongation by approximately 11% and 58%, respectively. Low-pressure casting, improved sand core collapsibility, and optimal temperature field control significantly reduce segregation by lowering temperature gradients and solidification shrinkage resistance. Additionally, a multistep addition of Al-5Ti-B grain refiners refines the microstructure from 185 μm to 91 μm average grain size, enhancing fluidity by 33.7% and improving hot-cracking resistance. The refined grains increase intergranular bridging areas, delaying grain separation and minimizing segregation formation. The results provide practical guidance for producing sound ZL205A castings with reduced sand foundry defect risks.
1. Introduction
ZL205A is an Al-Cu based high-strength casting alloy developed in China, with nominal composition of 4.6–5.3 wt% Cu, 0.3–0.5 wt% Mn, 0.15–0.35 wt% Ti, 0.05–0.20 wt% Zr, 0.05–0.25 wt% Cd, and trace B and V. Its T6 heat treatment delivers an ultimate tensile strength exceeding 490 MPa, which surpasses many wrought aluminum alloys and makes it attractive for replacing forged components in aerospace structures. The application of ZL205A has expanded to large thin-walled axisymmetric components such as solid rocket motor casings, missile adapters, and aircraft structural parts. However, with increased production, a persistent problem has emerged: segregation defects detected by X-ray inspection. These defects appear as white streaks or patches on radiographs, causing significant property degradation and frequent rejection of castings. Despite extensive research, the exact mechanisms and effective process countermeasures have not been fully established, especially for large sand castings. This paper addresses this gap by systematically investigating the formation mechanism of segregation defects in large ZL205A axisymmetric castings and evaluating two mitigation strategies: (1) optimization of the casting process (low-pressure casting and sand core design) to control solidification thermal gradients and stress, and (2) optimization of grain refinement (multi-step Al-Ti-B addition) to enhance alloy performance. The ultimate goal is to reduce the occurrence of these sand foundry defects and provide reliable engineering solutions.
2. Materials and Experimental Procedures
2.1 Material Preparation
All ZL205A alloy melts were prepared in a 500 kg resistance furnace using high-purity aluminum (99.99%), and master alloys of Al-50Cu, Al-10Mn, Al-4Ti, Al-4V, Al-2Zr, Al-4Ti-B, and metallic cadmium. The nominal chemical composition is listed in Table 1. The melting procedure involved sequential charging of Al-Mn, Al-Ti, Al-Zr, Al-V master alloys and return scrap, followed by melting and superheating to 690–710 °C. Al-Cu master alloy and Cd (wrapped in aluminum foil) were then added and stirred. For the grain-refinement optimization, Al-4Ti-B wire (KBM, φ10 mm) was added either as a single addition after degassing (conventional) or in two steps: 2/3 of the total amount added during the melting stage, and the remaining 1/3 added 30 minutes before pouring. Degassing was performed with C₂Cl₆ tablets (0.3 wt%) followed by rotary argon lance treatment for 20 minutes. The melt was then held at 710–720 °C for 15 minutes before pouring.
| Cu | Mn | Ti | Zr | Cd | B | V | Al |
|---|---|---|---|---|---|---|---|
| 4.6–5.3 | 0.3–0.5 | 0.15–0.35 | 0.05–0.20 | 0.05–0.25 | 0.0005–0.06 | 0.05–0.30 | Balance |
2.2 Casting of Large Axisymmetric Components
The investigated component is a large axisymmetric shell with a diameter of 1400 mm, height of 500 mm, and wall thickness of 30 mm. Both gravity pouring and low-pressure casting were employed. The gating system was designed using vertical slit gates fed by a central sprue with a “米” (cross) shaped runner. The slit gate parameters were optimized using three-factor, three-level orthogonal experiments combined with ProCAST simulations. The final optimized parameters were: slit width δ = 25 mm, slit length b = 35 mm, and runner diameter d = 80 mm. The sand molds and cores were prepared using 70/140 mesh silica sand bonded with phenolic urethane resin (CP5140/CP5235) at 10 mL/kg sand. For advanced collapsibility, the core was made hollow with a wall thickness of 100 mm and divided into four segments with 6–8 mm gaps between them. The coating used was ZA-3 proprietary wash. For low-pressure casting, the pressure profile was: fill pressure 25 kPa, increase to 40 kPa during mold filling, and hold at 45 kPa for 300 seconds. The pouring temperature was maintained at 715 °C for all experiments.
The photograph above illustrates a modern automatic pouring line that can be employed for reproducible low-pressure casting of large aluminum components. Automation is critical in minimizing manual variations that can lead to sand foundry defect formation, including segregation.
2.3 Heat Treatment and Sample Preparation
After casting, test blocks were subjected to T5 heat treatment: solution treatment at 538 ± 5 °C for 10–18 hours followed by water quenching, and aging at 155 ± 5 °C for 8–10 hours. Mechanical test specimens were machined from the casting body at locations identified as defect-free or containing segregation defects, based on X-ray radiography. Two types of tensile specimens were used: round specimens with diameter 5 mm (Φ5) and flat specimens with thickness 6 mm (flat 6). The gage length for all specimens was 25 mm. Tensile tests were conducted at room temperature using an Instron 5582 machine at a strain rate of 1 mm/min. Yield strength (R₀.₂), ultimate tensile strength (Rm), and elongation (A) were measured. Microhardness measurements were performed on selected specimens using a Struers Duramin-2 tester with a 100 g load.
2.4 Microstructural Characterization
Optical microscopy (OM) was carried out on a Leica DM2500M microscope. SEM and EDS analyses were performed using a JEM-7001F field emission scanning electron microscope equipped with an EDAX energy dispersive spectrometer. For DSC analysis, a NETZSCH STA 409C instrument was used. Specimens of 3 mm diameter and 1 mm thickness were heated from room temperature to 700 °C at a rate of 10 °C/min in a protective argon atmosphere. Fracture surfaces of tensile specimens were examined by SEM in both secondary electron (SE) and backscattered electron (BSE) modes to assess the fracture mode and the presence of segregation phases.
3. Characteristics and Mechanisms of Segregation Defects
3.1 Macroscopic Features
X-ray inspection of production castings revealed that segregation defects are predominantly localized at the bottom region of slit gates, specifically at the junction between the slit runner and the casting body. The typical radiographic appearance consists of white, fish-scale-like patterns, as shown in the schematic description common to this defect type. This morphology arises from the accumulation of high-density Cu-rich phases, which attenuate X-rays more strongly than the surrounding aluminum matrix. Statistical analysis of over 30 rejected castings indicated that the defects are always associated with the last solidified regions corresponding to local hot spots.
3.2 Microstructural Characteristics
Figure 3 (not shown here, but described) presents optical and SEM micrographs of the segregation area. The microstructure consists of primary α-Al dendrites surrounded by a network of coarse white phases. EDS analysis was performed on both white and gray regions. The white regions were found to contain approximately 48.9 wt% Al and 51.1 wt% Cu, corresponding to an atomic ratio of Al:Cu = 2:1, which matches the stoichiometry of Al₂Cu (θ phase). The gray regions contained about 93.5 wt% Al, 0.4 wt% Mn, and 6.1 wt% Cu, typical of α-Al with some eutectic constituents. In the defect-free areas, the Al₂Cu phase was present in much smaller, isolated particles, whereas in the segregation regions, the phase formed continuous networks along grain boundaries. DSC analysis of the segregated region exhibited a distinct endothermic peak starting at 544.6 °C, corresponding to the melting of the eutectic Al₂Cu phase, whereas the defect-free sample showed no such peak (main solution peak starting at 640.8 °C). This confirms that the segregated region contains excessive low-melting-point eutectic compounds that cannot be fully dissolved during T5 heat treatment.
3.3 Effect on Mechanical Properties
Table 2 summarizes the average tensile properties obtained from Φ5 and flat 6 specimens, both for defect-free and segregated locations. For the defect-free material, the average UTS, yield strength, and elongation were 434.5 MPa, 340 MPa, and 9.82% for Φ5 specimens, and 435.3 MPa, 340 MPa, and 9.18% for flat specimens, respectively. For segregated specimens, the average values were significantly lower: UTS 386.2 MPa, yield strength 340 MPa, and elongation 3.86%. Thus, the segregation defect reduces tensile strength by about 11% and elongation by about 58%, while yield strength remains essentially unchanged. Microhardness measurements showed that the segregated regions had hardness values of 140–160 HV compared to 162–164 HV in defect-free areas, indicating a softer eutectic network.
| Specimen condition | Specimen type | Rm (MPa) | R₀.₂ (MPa) | A (%) |
|---|---|---|---|---|
| Defect-free | Φ5 round | 434.5 ± 14 | 337.8 ± 17 | 9.82 ± 2.1 |
| Defect-free | flat 6 | 435.3 ± 10 | 340 ± 12 | 9.18 ± 1.7 |
| Segregated | flat 6 | 386.2 ± 16 | 340 ± 16 | 3.86 ± 0.5 |
3.4 Fracture Behavior
Fracture surfaces of defect-free tensile specimens showed a typical ductile rupture characterized by fine dimples, indicating microvoid coalescence. In contrast, specimens with segregation defects exhibited brittle intergranular fracture with large areas of smooth, bright facets. The BSE imaging revealed that these facets were covered by a continuous network of Al₂Cu phase. This brittle phase acts as a crack initiation site and provides an easy path for crack propagation along grain boundaries, leading to premature failure. These observations confirm that the segregation defect is detrimental to the structural integrity of the casting.
3.5 Mechanism of Segregation Formation
ZL205A solidifies over a wide range of about 89 °C, and its solidification mode is predominantly mushy. During solidification, the primary α-Al dendrites form first, rejecting copper into the remaining liquid. Since the partition coefficient of Cu is less than unity, the intergranular liquid becomes enriched in Cu. In a sound casting, this enriched liquid solidifies as fine eutectic particles uniformly distributed. However, when the casting is subjected to thermal stresses or inadequate feeding, localized hot spots remain liquid longer than surrounding solid. These hot spots, often at the junction between the casting and the slit gate, may be simultaneously subjected to three stress components: (1) thermal stress due to differential cooling rates between the inner and outer walls; (2) tensile stress caused by solidification contraction of the casting body; and (3) compressive stress from the sand core resisting contraction. When the combined stress exceeds the strength of the partially solidified dendrite network, cracks or gaps form in the mushy zone, and the Cu-rich liquid is drawn into these gaps by capillary and shrinkage forces. This process results in large pools of Al₂Cu eutectic forming continuous films along the grain boundaries, which correspond to the fish-scale patterns observed on radiographs. This mechanism is supported by the FEM simulation results showing that the last solidified region is indeed at the casting-slit gate interface.
4. Influence of Casting Process on Segregation Defects
4.1 Optimization of Gating System
The gating system plays a crucial role in controlling the thermal field and feeding behavior. Three key parameters (slit width δ, slit length b, and runner diameter d) were varied in three levels using an orthogonal L9 design. ProCAST simulations were used to evaluate each design based on mold filling stability and solidification temperature gradients. The optimal combination, as confirmed by experimental casting, was δ = 25 mm, b = 35 mm, and d = 80 mm. This design promotes a smooth upward filling with minimal turbulence and avoids premature freezing in the gate. Using this configuration, the filling time was reduced by 20% compared to the original design, and the occurrence of shrinkage-related defects decreased notably.
4.2 Comparison of Gravity Pouring vs. Low-Pressure Casting
FEM thermal simulations were performed for both gravity and low-pressure casting. The results are summarized in Table 3. In gravity casting, the maximum temperature difference within the casting during solidification was as high as 324 °C, with a horizontal gradient of 312 °C and vertical gradient of 186 °C. In low-pressure casting, these values were reduced to 126 °C, 108 °C, and 64 °C, respectively. The reduction is achieved because low-pressure casting allows controlled filling from bottom to top, maintaining a favorable vertical temperature gradient. This promotes directional solidification from the top downward, ensuring that the last liquid remains in the gating system rather than in the casting body. As a result, the risk of suction of Cu-rich liquid into the casting is markedly reduced.
| Process | Peak ΔT (casting) | Peak ΔT (horizontal section) | Peak ΔT (vertical direction) |
|---|---|---|---|
| Gravity casting | 324 °C | 312 °C | 186 °C |
| Low-pressure casting | 126 °C | 108 °C | 64 °C |
Experimental calibration castings were produced under identical conditions with both processes. Three batches of castings produced by gravity pouring exhibited a total of 125 segregation defects (including linear, flake, and fish-scale types). In contrast, three batches produced by low-pressure casting had only 23 defects. SEM observations revealed that in gravity-cast specimens, the grain boundary Al₂Cu network was coarse and continuous, while in low-pressure cast specimens the network was much finer and more broken, indicating a lower segregation severity. Therefore, low-pressure casting is a highly effective method for suppressing this sand foundry defect.
4.3 Improvement of Sand Core Collapsibility
Another critical factor is the resistance provided by the sand core during solidification contraction. In the original core design, a solid core creates high compressive stresses on the hot spot region. To improve collapsibility, the core was redesigned as a hollow shell with a thickness of 100 mm and divided into four separate segments with 6–8 mm gaps. During solidification, the segments can move slightly inward as the casting shrinks, releasing the accumulated stress. This reduces the likelihood of hot tearing and subsequent liquid filling of cracks. Experimental verification with both gravity and low-pressure casting showed that the defect count was reduced from 125 to 63 for gravity casting and from 23 to 10 for low-pressure casting when the improved core was used. This confirms the beneficial effect of core collapsibility on minimizing sand foundry defect, particularly segregation.
5. Influence of Grain Refinement on Segregation Defects
5.1 Optimization of Al-5Ti-B Addition
The traditional grain refinement method adds all Al-5Ti-B at once during the melting stage. However, due to the low solubility of Ti in aluminum (0.14 wt%), excess Ti tends to precipitate as coarse Al₃Ti particles that settle to the bottom of the crucible, resulting in a loss of nucleating sites and poor refinement. To overcome this, a two-step addition procedure was developed: 2/3 of the Al-4Ti-B was added after superheating, and the remaining 1/3 was added 30 minutes before pouring. This multistep approach maintains a higher concentration of active TiAl₃ particles in the melt, increasing the nucleation density and slowing the settling process.
5.2 Microstructure and Property Improvements
Evaluation of cylindrical cast samples showed that the conventional single-step addition yielded an average grain size of 185 μm, whereas the optimized two-step addition produced a fine equiaxed grain structure with an average size of 91 μm – a 50.8% reduction. The corresponding as-cast mechanical properties improved significantly: ultimate tensile strength increased from 195 MPa to 219 MPa, yield strength from 83 MPa to 94 MPa, and elongation from 8% to 12%. This enhancement is attributed to Hall-Petch strengthening, which can be expressed by:
$$\sigma_s = \sigma_0 + K d^{-1/2}$$
where σ₀ and K are material constants, and d is the average grain diameter. A finer grain size leads to a higher yield strength. In addition, the refined microstructure distributes the brittle Al₂Cu phase more evenly along many small grain boundaries, reducing the continuity of the brittle network.
5.3 Effect on Fluidity and Hot Cracking Resistance
Fluidity was evaluated using a spiral test. The conventional refined alloy produced a flow length of 184 mm, while the optimized refined alloy produced a flow length of 246 mm, an increase of 33.7%. The better fluidity improves the feeding ability and reduces the formation of shrinkage discontinuities. Hot-cracking tests were conducted using a ring-shaped test mold with various core diameters. The conventional refined alloy exhibited cracking at ring widths of 27.5 mm and 25 mm, while the optimized alloy did not crack even at 22.5 mm. This improved resistance to hot tearing is essential because it prevents the formation of cracks that later become filled with Cu-rich liquid, thus eliminating a major source of segregation defects.
5.4 Influence of Grain Refinement on Casting Trials
Keeping the optimized low-pressure casting process and improved core design fixed, three batches of large axisymmetric castings were poured using the optimized grain refinement procedure. X-ray inspection revealed a total of only 16 segregation defects in those three batches, compared to 23 defects in the batches prepared with conventional grain refinement (with the same casting process). When combined with the low-pressure process and improved core collapsibility, the overall reduction from the original gravity casting with conventional refinement was from 125 defects to 10 defects per three batches. Thus, the combination of optimized casting process and grain refinement offers a powerful solution to suppress sand foundry defect, especially macro-segregation, in ZL205A large castings.
6. Discussion
From the above experimental results, it is evident that segregation defects in ZL205A large axisymmetric castings arise from a complex interplay of thermal, mechanical, and metallurgical factors. The primary cause is the cumulative stress in the mushy zone during the late stage of solidification. When the local stress exceeds the cohesion of the partially solidified dendrites, intergranular fissures form, and Cu-rich eutectic liquid is drawn into them. Low-pressure casting reduces the peak thermal gradients and promotes a more uniform cooling pattern, thus reducing thermal stress. The redesigned sand core with high collapsibility allows the core to yield under the contraction force, relieving mechanical stress. Grain refinement increases the number of grain boundaries and the total contact area between grains. During the late solidification stage, intergranular bridging (often called “dendrite bridges”) forms, which increases the cohesion between grains. According to the grain bridge model, the critical stress for grain separation is proportional to the effective bridging area. For a constant volume fraction of solid, the total bridge area increases as grain size decreases, as approximated by:
$$A_{\text{bridge}} \propto \left(\frac{1}{d}\right) \cdot \text{constant}$$
Thus, finer grains yield a larger cohesive area, delaying the onset of separation and raising the critical threshold for hot cracking. This mechanism explains the observed improvement in hot-cracking resistance and the corresponding reduction in segregation defects.
Additionally, the refined alloy has better fluidity, which enables easier feeding of the intergranular spaces and reduces the likelihood of shrinkage-induced voids. The combination of low-pressure casting, optimized gating, improved core collapsibility, and optimized grain refinement provides a comprehensive framework for producing sound ZL205A large axisymmetric castings with minimal segregation. These findings are directly applicable to industrial production, where sand foundry defect reduction is a critical quality target.
7. Conclusions
The following conclusions can be drawn from this investigation:
- Segregation defects in ZL205A large axisymmetric castings appear as fish-scale patterns located predominantly at the junction between the casting body and slit gates, resulting from the accumulation of coarse Al₂Cu eutectic along grain boundaries.
- The segregation defect reduces tensile strength by about 11% and elongation by about 58%, while yield strength remains unchanged. Fracture mode changes from ductile microvoid coalescence to brittle intergranular failure.
- Low-pressure casting reduces the peak temperature difference in the casting from 324 °C to 126 °C, and combined with a hollow, segmented sand core, it reduces the number of segregation defects by more than 90% compared to conventional gravity casting.
- A two-step addition of Al-4Ti-B grain refiner reduces the average grain size from 185 μm to 91 μm, improves fluidity by 33.7%, enhances as-cast tensile strength by 11%, and significantly improves hot-cracking resistance.
- The optimized casting process and grain refinement together provide an effective approach to mitigate the sand foundry defect of segregation in large ZL205A axisymmetric castings, ensuring reliable performance in aerospace applications.