In the realm of nuclear power plant safety, emergency diesel generator sets serve as a critical line of defense, ensuring operational stability during power failures. Among these systems, the water circulation pump—a key auxiliary component—relies on a robust pump casing to facilitate heat exchange and cooling. The casing must withstand corrosive media, high-pressure loads, and rapid thermal cycles, making its manufacturing process a focal point of engineering. This article delves into the application of precision investment casting for producing austenitic-ferritic stainless steel pump casings, specifically tailored to meet the stringent requirements of nuclear standards like RCCM-2007. Through iterative process design, simulation-assisted optimization, and empirical validation, we explore how precision investment casting can achieve high-integrity components with complex geometries and superior material properties.
The pump casing under study is fabricated from M3402 (Z3NCD19-10M) austenitic-ferritic stainless steel, a material specified in the RCCM-2007 code for its balanced mechanical strength and corrosion resistance. With a nominal weight of 12 kg and dimensions of approximately 300 mm × 140 mm × 240 mm, the casing features intricate fluid channels, varying wall thicknesses from 3 mm to 22 mm, and multiple reinforcing ribs. Such complexity necessitates a casting method that ensures dimensional accuracy, minimal defects, and consistent microstructure. Precision investment casting, also known as lost-wax casting, is ideal for this purpose, as it allows for the production of near-net-shape parts with fine surface details and tight tolerances. Our research focuses on optimizing this precision investment casting process to address challenges like shrinkage porosity, composition control, and performance validation.
To begin, we outline the initial process design for the precision investment casting of the pump casing. The material composition of M3402 stainless steel is critical, with strict limits on sulfur and phosphorus (S ≤ 0.015 wt%, P ≤ 0.030 wt%) to enhance ductility and reduce hot tearing. The chemical composition targets are summarized in Table 1, derived from RCCM-2007 specifications and adjusted for casting feasibility. We employ medium-frequency induction melting using pre-alloyed 316L scraps and master alloys, followed by deoxidation and degassing to achieve the desired chemistry. The precision investment casting process starts with wax pattern fabrication, where inject parameters are optimized to replicate the complex geometry. Key parameters include an injection temperature of 58–60°C, pressure of 0.45–0.50 MPa, and a dwell time of 40–45 s, as detailed in Table 2.
| Element | C | Si | Mn | P | S | Cr | Ni | Mo | Cu |
|---|---|---|---|---|---|---|---|---|---|
| Specification | ≤0.04 | ≤1.50 | ≤1.50 | ≤0.030 | ≤0.015 | 18–21 | 9–12 | 2.25–2.75 | ≤1 |
| Achieved | 0.03 | 0.91 | 1.13 | 0.02 | 0.004 | 18.5 | 10.6 | 2.28 | 0.91 |
The gating system design is pivotal in precision investment casting to ensure proper filling, feeding, and defect minimization. Initially, a horizontal gating with top and side gates was used, but it led to shrinkage porosity in thick sections like the neck area. This defect arose due to inadequate feeding during solidification, where isolated liquid pools formed in thermal centers. To address this, we optimized the system to a vertical gating layout, enhancing directional solidification and feeding efficiency. The gating dimensions were calculated based on proportional methods. For instance, the ingate cross-section was determined using the hot-spot diameter \(D_c\) (25 mm for critical sections) and the relation: $$D_g = k \cdot D_c$$ where \(D_g\) is the equivalent ingate diameter, and \(k\) is a factor between 0.6 and 1.0. With \(k = 0.8\), we obtained \(D_g = 20\) mm. The sprue diameter \(D_s\) was then derived as: $$D_s \geq (1.1 \sim 1.2) D_g$$ resulting in \(D_s = 22\) mm for a square-section sprue. These calculations guided the final gating design, featuring multiple tapered ingates and a central sprue to promote progressive solidification.
Shell building in precision investment casting involves successive coatings of ceramic materials to form a robust mold. We used a silica-based process with zircon flour for the face coats and molochite for backup layers. The shell consisted of 10.5 layers: two face coats with 80–120 mesh zircon sand, two transition coats, six backup coats, and a seal coat. Drying parameters were controlled at 24 ± 2°C and 65 ± 5% relative humidity, with drying times ranging from 6–8 hours for face coats to over 24 hours for seal coats, as summarized in Table 2. This meticulous shell-making ensures mold strength and permeability, critical for preventing defects like mold cracking or gas entrapment during pouring.
| Layer Type | Coating Material | Stucco Sand | Drying Time (h) | Temperature (°C) | Humidity (%) |
|---|---|---|---|---|---|
| Face Coat (2 layers) | Silica sol + Zircon flour | 80–120 mesh Zircon | 6–8 | 24 ± 2 | 65 ± 5 |
| Transition Coat (2 layers) | Silica sol + Molochite flour | 30–60 mesh Molochite | ≥10 | 24 ± 2 | 65 ± 5 |
| Backup Coat (6 layers) | Silica sol + Molochite flour | 16–30 mesh Molochite | ≥12 | 24 ± 2 | 65 ± 5 |
| Seal Coat | Backup coating slurry | None | ≥24 | 24 ± 2 | 65 ± 5 |
Following process optimization, we employed ProCAST simulation software to validate the precision investment casting design. ProCAST utilizes finite element analysis to model fluid flow, heat transfer, and solidification, predicting defects such as shrinkage porosity and cold shuts. The simulation accounted for material properties of M3402 steel, including thermal conductivity and latent heat, with a pouring temperature of 1640 ± 10°C. The governing equations for fluid flow and heat transfer are based on Navier-Stokes and energy conservation principles. For instance, the heat transfer equation during solidification is: $$\rho c_p \frac{\partial T}{\partial t} = \nabla \cdot (k \nabla T) + \rho L \frac{\partial f_s}{\partial t}$$ where \(\rho\) is density, \(c_p\) is specific heat, \(k\) is thermal conductivity, \(T\) is temperature, \(L\) is latent heat, and \(f_s\) is solid fraction. The simulation results, shown in Figure 6 (not referenced per guidelines), indicated that the optimized vertical gating promoted directional solidification, with the last-to-freeze regions shifted to the sprue, thereby eliminating shrinkage in the casing body. The Niyama criterion, expressed as: $$N_y = \frac{G}{\sqrt{\dot{T}}}$$ where \(G\) is temperature gradient and \(\dot{T}\) is cooling rate, was used to assess porosity risk. Values above a threshold (typically 1 °C1/2·mm-1) indicated sound regions, confirming the efficacy of our precision investment casting approach.
The optimized precision investment casting process was then implemented in trial production. After shell baking at 1050 ± 10°C for 1.5 hours, melting and pouring were conducted under controlled atmosphere. Post-casting, the casings underwent solution heat treatment at 1090 ± 10°C for 2–2.5 hours, followed by water quenching to dissolve secondary phases and homogenize the microstructure. We conducted extensive evaluations to assess quality. Dimensional accuracy was verified using 3D scanning with a Hexagon RA7535SI flexible arm, comparing point cloud data to the CAD model. The results, processed in Geomagic software, showed deviations within ±0.5 mm for internal channels, meeting CT7 tolerance per GB6414 standards. Surface roughness averaged Ra 12.5 μm, suitable for nuclear applications.
Non-destructive testing via X-ray radiography (XGY-4510 system) revealed no internal defects like porosity, cracks, or inclusions in the optimized castings, a marked improvement over initial trials. Microstructural analysis was performed using a GX71 inverted metalloscope. The M3402 material exhibited a dual-phase structure of austenite and ferrite, with ferrite content measured at approximately 13.5% via magnetic induction methods, within the specified 12–25% range per RCCM-2007. This dual-phase morphology enhances strength and stress corrosion cracking resistance, crucial for seawater service. Mechanical properties were evaluated according to RCCM-2007 requirements, with tensile and impact tests summarized in Table 3. All values surpassed minimum specifications, demonstrating the robustness of the precision investment casting process.
| Test Type | Temperature | Property | Specified Minimum | Measured Value |
|---|---|---|---|---|
| Tensile | Room Temperature | Yield Strength (Rp0.2, MPa) | ≥210 | 340 |
| Room Temperature | Tensile Strength (Rm, MPa) | ≥480 | 553 | |
| Room Temperature | Elongation (A, %) | 35 | 52 | |
| Tensile | 350°C | Yield Strength (Rp0.2, MPa) | ≥140 | 175 |
| 350°C | Tensile Strength (Rm, MPa) | ≥390 | 406 | |
| Charpy V-Notch Impact | Room Temperature | Average Energy (J) | 80 | 160 |
Further analysis involved solidification modeling to refine the precision investment casting process. Using ProCAST, we simulated temperature fields and solidification sequences. The fraction solid evolution over time, \(f_s(t)\), can be described by: $$f_s = 1 – \exp\left(-\int_0^t \frac{dt}{\tau}\right)$$ where \(\tau\) is a time constant dependent on cooling conditions. The simulation predicted that thermal gradients in the optimized gating exceeded 10°C/mm in critical sections, ensuring rapid feeding and reducing microporosity. Additionally, we calculated the solidification time \(t_s\) for thick sections using Chvorinov’s rule: $$t_s = C \left(\frac{V}{A}\right)^n$$ where \(V\) is volume, \(A\) is surface area, \(C\) is a mold constant, and \(n\) is an exponent (~2 for sand molds). For the 22 mm thick flange, \(t_s\) was estimated at 120 seconds, aligning with simulation outputs. These insights guide future designs for similar precision investment casting projects.
The success of this precision investment casting study hinges on several factors. First, material composition control is paramount; low sulfur and phosphorus levels minimize hot tearing, while balanced chromium and nickel promote phase stability. Second, gating optimization via vertical design enhances feeding, as evidenced by the elimination of shrinkage defects. Third, simulation tools like ProCAST provide a cost-effective means to predict and mitigate defects before physical trials, reducing development time. Lastly, rigorous post-casting treatments and inspections ensure compliance with nuclear standards. This holistic approach underscores the versatility of precision investment casting for high-performance components.
In conclusion, our research demonstrates that precision investment casting is a viable method for manufacturing nuclear-grade stainless steel pump casings with complex geometries and stringent requirements. Through iterative design, simulation validation, and empirical testing, we optimized the process to produce defect-free M3402 casings with excellent dimensional accuracy, microstructure, and mechanical properties. The precision investment casting technique, supported by advanced modeling, offers a reliable pathway for fabricating critical safety components in nuclear and other high-integrity industries. Future work could explore additive manufacturing for wax patterns or enhanced simulation algorithms to further refine the precision investment casting process for even more challenging applications.