This study focuses on the development and optimization of a sand casting process for a critical high-pressure steam chamber component, a type of sand casting part commonly used in steam turbine applications. The component is subjected to extreme service conditions, including temperatures up to approximately 565°C and pressures around 8.28 MPa, demanding exceptionally high internal quality and integrity from the final sand casting parts. The material specified is the alloy cast steel ZG15Cr2Mo1. Traditional trial-and-error methods for such complex sand casting parts are costly and time-consuming. Therefore, leveraging numerical simulation technology is crucial for designing an efficient and reliable casting process. This work details the systematic approach, from initial process design based on theoretical calculations to iterative optimization using ProCAST simulation software, to eliminate shrinkage defects in this large-scale sand casting part.
1. Component Characteristics and Requirements
The steam chamber is a pressure-containing sand casting part with a complex internal cavity for rotor assembly. Its geometry, after accounting for machining allowances, features significant variations in wall thickness. The main dimensions are 1648 mm × 620 mm × 1077 mm, with a casting volume of 0.251 m³ and a calculated weight of 1957.8 kg. The wall thickness analysis confirms it meets the minimum wall thickness requirements for steel sand casting parts, with a maximum thickness of 153 mm and a minimum of 30 mm. The alloy composition is critical for its high-temperature performance, as detailed in Table 1.
| C | Mn | Si | Cr | Mo | S | P |
|---|---|---|---|---|---|---|
| ≤0.18 | 0.40–0.70 | ≤0.60 | 2.00–2.75 | 0.90–1.20 | ≤0.030 | ≤0.030 |
Note: For every 0.01% reduction in C content from the upper limit, an increase of 0.04% in Mn is permitted, with a maximum Mn content of 1.2%.
The key physical parameters for process design include a liquidus temperature of 1501°C, a density of 7.8 g/cm³, and a constrained linear shrinkage rate of 1.8%, necessitating an effective feeding system design for these sand casting parts.
2. Foundry Process Scheme Design
The design of the sand casting process for these large sand casting parts involves several critical decisions regarding molding, gating, and feeding.
2.1 Selection of Pouring Position and Parting Line
Two primary options were considered for positioning this sand casting part in the mold: vertical (standing upright) and horizontal (lying on its side). A horizontal pouring position was selected. This choice offers several advantages for producing such sand casting parts: it allows the parting line to be placed at the largest cross-section, simplifying pattern withdrawal; it positions the major thick sections (like the base) in the upper parts of the mold cavity, facilitating the placement of feeders (risers) for effective feeding; and it eliminates the need for complex coring support (chaplets), thereby streamlining the overall process for these sand casting parts.
2.2 Mold and Core Design
Given the high pouring temperature (exceeding 1580°C) required for steel sand casting parts, a high-refractoriness sand system is essential. A phenolic resin-bonded sand with high silica content (SiO₂ ≥ 97%) was selected for both mold and core making. An alumina-based alcohol coating was specified to be applied via spraying to enhance the surface finish of the sand casting parts. The internal geometry of the steam chamber necessitates three sand cores: two main cores (Core 1 and 2) to form the left and right sections of the internal cavity, and a smaller core (Core 3) to form the external shell passages, allowing them to be cast integrally in these sand casting parts.
2.3 Process Parameters
For small-batch production of these sand casting parts, manual molding with self-setting resin sand is employed. The dimensional tolerance grade is set to CT13 per GB/T 6414-2017, and the weight tolerance grade is MT12 per GB/T 11351-2017, with a tolerance value of ±8%. The constrained linear shrinkage is set at 1.8% for accurate pattern sizing.
3. Theoretical Calculation for Gating and Feeding System
Initial system design for these heavy sand casting parts relies on empirical formulas to ensure complete filling and adequate feeding.
The total poured weight \(G_L\), including a 7% allowance for mold dilation, is calculated as:
$$G_L = (1 + 0.07) \times G_C = 2094.846 \text{ kg}$$
where \(G_C = 1957.8 \text{ kg}\) is the net casting weight.
Using a bottom-pour ladle with one nozzle of 70 mm diameter (pouring rate \(v_{\text{ladle}} \approx 120 \text{ kg/s}\)), the theoretical pouring time \(t\) is:
$$t = \frac{G_L}{N \cdot n \cdot v_{\text{ladle}}} = \frac{2094.846}{1 \times 1 \times 120} \approx 17.46 \text{ s}$$
A practical pouring time of \(t = 18 \text{ s}\) is adopted for simulation.
The metal rise velocity \(v_L\) within the mold cavity is verified to ensure it is sufficient to prevent cold shuts and excessive heating of the sand mold for these sand casting parts:
$$v_L = \frac{h_C}{t} = \frac{620 \text{ mm}}{18 \text{ s}} \approx 35.5 \text{ mm/s}$$
This meets the required rise speed of >35 mm/s for complex steel sand casting parts.
An unpressurized (choke-at-the-bottom) gating system is designed. The cross-sectional area ratios for the various elements are set as:
$$\Sigma A_{\text{nozzle}} : \Sigma A_{\text{sprue}} : \Sigma A_{\text{runner}} : \Sigma A_{\text{ingate}} = 1.0 : (1.8 \sim 2.0) : (1.8 \sim 2.0) : (2.0 \sim 2.5)$$
Three ingates are placed along the parting plane to promote uniform temperature distribution during the filling of these sand casting parts.
The minimum required metallostatic pressure head \(h_M\) is checked to ensure proper filling of the furthest points in these extensive sand casting parts:
$$h_M \geq L \cdot \tan \alpha$$
Where \(L = 790 \text{ mm}\) is the longest flow distance from the sprue, and \(\alpha = 6^\circ\) is the recommended pressure angle for the wall thickness. With a sprue height yielding \(h_M = 83 \text{ mm}\), the condition is satisfied as \(83 \geq 790 \cdot \tan(6^\circ) \approx 83\), confirming the gating design is theoretically sound for these sand casting parts.
4. Initial Numerical Simulation and Defect Analysis
The initial process design, including the gating system without specific feeders, was modeled and simulated in ProCAST. The key simulation parameters were: Pouring Temperature = 1600°C, Pouring Time = 18 s, Heat Transfer Coefficient (metal-sand) \(h = 750 \text{ W/(m²·K)}\). The mesh of the system, including molds and cores, was carefully generated to ensure computational accuracy for these sand casting parts.
4.1 Filling Analysis
The simulation of the filling stage showed a stable, progressive filling of the mold cavity. No signs of misruns or cold shuts were observed, indicating that the designed gating system is capable of completely filling these large sand casting parts without major turbulence-related issues.
4.2 Solidification and Shrinkage Prediction
The solidification simulation revealed significant shrinkage porosity defects, which are critical quality concerns for pressure-tight sand casting parts. The major defects were predicted in two locations:
- Massive shrinkage cavities in the thick sections of the steam chamber base.
- Subsurface shrinkage porosity in the transition regions between the top bosses and the internal partition walls.
These areas act as thermal centers (hot spots) because their geometry leads to delayed solidification, creating isolated liquid pools that cannot be fed effectively after the surrounding thinner sections solidify. This is a typical challenge in sand casting parts with non-uniform wall thickness.
5. Process Optimization Based on Simulation Results
To address the shrinkage defects predicted in the initial simulation, the process was systematically optimized. The strategy combines the use of feeders (risers) to provide liquid metal补给 and chills to promote directional solidification towards those feeders in these sand casting parts.
5.1 Feeder (Riser) Design and Placement
Feeders were added to the top of the identified hot spots to act as reservoirs of molten metal. The feeder neck is designed to solidify after the connected hot spot in the sand casting part. Four main feeders were positioned using modulus calculations:
- Two feeders over the thick base sections.
- Two feeders over the top boss regions.
The size of each feeder was determined based on the modulus (Volume/Surface Area ratio) of the feeding region it serves, ensuring it remains liquid long enough to feed the shrinkage in these specific zones of the sand casting parts.
5.2 Chill Design and Placement
External chills (made of cast iron or steel) were introduced to accelerate cooling in strategic areas. Their purpose is to modify the solidification sequence, ensuring that the regions between the feeders and the chills solidify directionally. Chills were placed:
- On the internal sand core surfaces adjacent to the thick base sections.
- Against the sand molds forming the internal partition walls.
This arrangement encourages solidification to start from the chilled areas and progress towards the feeders, effectively channeling shrinkage porosity out of the critical sand casting parts and into the feeders.
5.3 Simulation of the Optimized Process
The modified casting system, incorporating feeders and chills, was simulated under the same conditions. The results demonstrated a marked improvement:
- The large shrinkage cavities in the base were completely eliminated, with the shrinkage now concentrated within the feeders.
- The overall shrinkage porosity in the casting body was significantly reduced.
- The total shrinkage porosity percentage was calculated to be 2.71%, primarily located in the feeders, which are later removed from the final sand casting parts.
The simulation confirmed that the optimized design successfully redirected the shrinkage defects, greatly enhancing the internal soundness of the high-pressure steam chamber sand casting parts. However, the simulation also indicated that the accelerated cooling from the chills on the partition walls could slightly deepen the micro-porosity gradient in the adjacent boss regions, suggesting a potential area for further fine-tuning of chill size or the use of insulating sleeves on feeder necks.
6. Discussion on Process Design for Complex Sand Casting Parts
This case study highlights a systematic methodology for developing robust processes for high-integrity sand casting parts. The iterative cycle of design-simulation-redesign is powerful. Key factors influencing the quality of such sand casting parts include:
- Thermal Modulus Control: The balance between section moduli of the casting, feeders, and feeder necks is paramount for effective feeding. This can be expressed by ensuring the feeder modulus \(M_f\) is greater than the casting modulus \(M_c\) at the junction: \(M_f > 1.2 \times M_c\) is a common rule of thumb.
- Solidification Gradient Management: The use of chills (Q) and heating (insulation) allows the foundry engineer to manipulate the solidification time \(t_s\) across the sand casting part. The objective is to create a positive temperature gradient \(\frac{dT}{dx}\) pointing towards the feeder. The local solidification time can be approximated by Chvorinov’s rule: \(t_s = B \cdot \left( \frac{V}{A} \right)^n\), where \(V/A\) is the local modulus, and \(B\) and \(n\) are constants dependent on the mold material and metal properties. Chills effectively reduce the local value of \(B\), decreasing \(t_s\).
- Feed Path Efficiency: The distance over which a feeder can effectively feed is limited in sand casting parts. This “feeding distance” depends on the alloy’s solidification characteristics and the presence of chills or cooling fins. Ensuring all points in a sand casting part are within the effective feeding range of a riser or chill is crucial.
For future optimization of these specific sand casting parts, variables such as feeder size and insulation, chill material and dimensions, and pouring temperature could be explored through Design of Experiments (DOE) coupled with simulation to find the global optimum that minimizes both macro- and micro-shrinkage.
7. Conclusion
This research successfully demonstrates the application of numerical simulation in designing and optimizing the sand casting process for a large, high-integrity ZG15Cr2Mo1 steel steam chamber. The initial process design, derived from theoretical calculations, was found to produce significant shrinkage cavities in the thick sections of the sand casting part. Through an iterative simulation-guided approach, an optimized process was developed incorporating strategically placed feeders and chills. The final simulation results confirmed that this optimized scheme effectively eliminated major shrinkage defects, relocating the porosity to the sacrificial feeders. The total predicted shrinkage porosity was reduced to 2.71%, markedly improving the expected internal soundness of the final sand casting parts. This work underscores the critical role of simulation technology in the cost-effective and reliable development of complex sand casting parts for demanding industrial applications, reducing the need for physical trials and accelerating time-to-market for high-quality sand casting parts.