In my work on the design and simulation of water glass sand casting for water valves, I focused on addressing common casting defects such as shrinkage cavities, porosity, and gas holes. The water valve is a critical component in fluid control systems; its quality directly influences the stable operation of the entire system. Water glass sand casting is widely adopted due to its adaptability and low cost. I aimed to improve casting quality while reducing production costs through careful process design and numerical simulation.
1. Part Structure and Technical Requirements
The water valve I studied has an overall dimension of 480 mm × 448 mm × 250 mm. It consists of four channels and a hollow internal flow passage. The valve features multiple connecting faces and sealing surfaces, which impose high requirements on surface quality and dimensional accuracy. Additionally, the sealing performance at the joints must be guaranteed. I selected ductile iron as the casting material because it offers high strength and hardness compared to aluminum alloys, along with low cost and good corrosion resistance. The cast product must be free of defects such as porosity, slag inclusions, and shrinkage porosity. After casting, a stress-relief heat treatment (aging) is applied to the casting.
2. Molding Method and Sand Material Selection
Considering the size, structure, and small-batch production characteristics of the water valve, I chose water glass sand casting. This method is suitable for parts with relatively simple geometries, offers flexible and simple operation, and is cost-effective for small batches or single-piece production. The molding sand is a mixture of water glass and quartz sand in a specific proportion. A hardener is added and stirred rapidly, then the mixture is poured into the mold and compacted. The water glass undergoes chemical reaction to harden, forming the desired mold and core. The water glass sand casting process exhibits strong adaptability, simple procedure, and good heat resistance.
3. Molding Scheme and Gating System Design
To reduce cost and improve productivity, I adopted a two-cavity mold (one pattern producing two castings). By analyzing the valve structure, I found it is symmetrical about the horizontal plane through its center. Therefore, I chose the central horizontal plane as the parting surface, dividing the mold into a cope and a drag. The internal cavity of the valve is formed by sand cores. The molding layout is shown in the schematic representation. I placed the runner at the parting surface for easy mold construction. To ensure rapid and stable filling, I designed four ingates located at four different positions of the valve, converging into a single sprue. Analyzing the solidification sequence, I predicted that the last solidifying region would be at the top end of the largest channel outlet. That location is prone to shrinkage defects due to lack of feed metal, so I added a riser at that position to provide feeding. The gating system design includes a sprue, four runners, and four ingates, with one riser.
The main parameters of the gating system are summarized in Table 1.
| Parameter | Value |
|---|---|
| Sprue height (mm) | 200 |
| Sprue top diameter (mm) | 40 |
| Sprue bottom diameter (mm) | 30 |
| Runner cross-section (mm²) | 500 |
| Ingate cross-section (mm²) per ingate | 120 |
| Riser diameter (mm) | 60 |
| Riser height (mm) | 100 |
4. Filling and Solidification Simulation Analysis
I used AnyCasting software to simulate the filling and solidification processes. The simulation aimed to verify that the design avoids turbulence and ensures proper feeding, thereby minimizing sand casting defects. The filling simulation results showed that the molten metal flows from the sprue into four runners and then enters the mold from the bottom of the casting, gradually rising to the middle and finally to the top and the riser. The flow remained smooth without turbulence. The total filling time was approximately 5.2 s. The filling velocity and pressure can be described by Bernoulli’s equation adapted for liquid metal flow in sand molds:
$$
v = \sqrt{2g h – \frac{2 \Delta P}{\rho}}
$$
where \(v\) is the velocity of the metal front, \(g\) the gravitational acceleration, \(h\) the metallostatic head, \(\Delta P\) the pressure loss due to friction and mold permeability, and \(\rho\) the density of the molten metal. In my simulation, the average filling velocity at the ingates was about 0.8 m/s, which is within the recommended range to avoid sand erosion and air entrapment.
After filling, the solidification simulation was performed. The solidification sequence is depicted in the simulation output: the three port regions and the outermost surfaces solidify first because they are in contact with numerous sand cores. Thicker sections and regions with more molten metal solidify later. For example, the junctions between two adjacent ports of the two castings, and the intersection of runners and sprue, solidify later. The riser and runners provide feeding to the late-solidifying regions. The solidification time was computed using the heat conduction equation:
$$
\rho c_p \frac{\partial T}{\partial t} = \nabla \cdot (k \nabla T) + Q
$$
where \(T\) is temperature, \(t\) time, \(\rho\) density, \(c_p\) specific heat, \(k\) thermal conductivity, and \(Q\) the latent heat source term. The simulation parameters used are listed in Table 2.
| Parameter | Value |
|---|---|
| Pouring temperature (°C) | 1330 |
| Initial mold temperature (°C) | 25 |
| Thermal conductivity of mold (W/m·K) | 0.8 |
| Specific heat of mold (J/kg·K) | 1000 |
| Density of mold (kg/m³) | 1600 |
| Latent heat of ductile iron (kJ/kg) | 250 |
| Solidus temperature (°C) | 1150 |
| Liquidus temperature (°C) | 1260 |
The total solidification time was approximately 480 seconds. The simulation results indicated that the riser remained liquid until the casting was mostly solidified, ensuring effective feeding.
5. Defect Prediction and Analysis
A critical part of my work was the prediction of sand casting defects. The AnyCasting software uses the Niyama criterion to predict shrinkage porosity. The Niyama number is defined as:
$$
N_y = \frac{G}{\sqrt{R}}
$$
where \(G\) is the temperature gradient and \(R\) is the cooling rate. When \(N_y\) falls below a threshold (typically 1.0 for ductile iron), the region is susceptible to micro-shrinkage. In my simulation, I calculated the Niyama distribution and identified areas with low values. The defect probability map is shown in the simulation output. The results revealed that the intersection of runners and sprue had a high risk of shrinkage porosity. However, the water valve castings themselves showed no significant defect probability. The riser, as the last filling region, also showed no defects due to adequate feeding. This indicated that my gating and riser design successfully minimized sand casting defects such as shrinkage cavities and micro-porosity. I also checked for gas porosity and sand inclusion by analyzing the filling pattern; the smooth flow without turbulence reduces the formation of such defects. In conclusion, the simulation confirmed that the designed process can produce sound castings free of major sand casting defects.
The image above illustrates common sand casting defects that I aimed to avoid. My simulation results showed that the proposed process effectively mitigates these defects.
6. Melting and Pouring Control
I used a medium-frequency induction furnace for melting. The melting process requires strict control of temperature and time to ensure uniform alloy composition. Slag removal and degassing operations were performed to prevent inclusions and gas holes. Before pouring, the sand mold was preheated to reduce thermal stress between the casting and the mold. The pouring temperature was controlled at 1330 °C for ductile iron, which allows complete filling and promotes directional solidification. The pouring velocity was kept steady to avoid air entrapment. The chemical composition of the ductile iron is given in Table 3.
| C | Si | Mn | P | S | Mg | Fe |
|---|---|---|---|---|---|---|
| 3.6-3.8 | 2.4-2.8 | 0.3-0.5 | ≤0.05 | ≤0.02 | 0.03-0.06 | Balance |
The melt treatment included nodularization using magnesium alloy and inoculation for graphite formation. The pouring process used a bottom-pour ladle to minimize slag carryover.
7. Results and Discussion
Based on the simulation and process control, trial production was carried out. The cast water valves were inspected for defects. Non-destructive testing (ultrasonic and visual inspection) showed no significant shrinkage, porosity, or sand inclusion. The microstructure was examined and confirmed to have the desired nodular graphite structure. The mechanical properties met the required specifications. The success of the trial validated that the designed water glass sand casting process, with optimized gating and risering, effectively suppresses sand casting defects. A comparison of defect probability before and after optimization is shown in Table 4.
| Defect type | Original design | Optimized design (this work) |
|---|---|---|
| Shrinkage porosity (volume %) | 2.5% | 0.3% |
| Gas porosity (area %) | 1.8% | 0.1% |
| Sand inclusion (count per casting) | 3 | 0 |
The reduction in sand casting defects demonstrates the effectiveness of the simulation-guided process design.
8. Conclusion
In this work, I systematically designed a water glass sand casting process for a water valve. By carefully selecting sand materials, parting surface, molding layout, and gating/riser system, I avoided common sand casting defects such as shrinkage, porosity, and slag inclusions. AnyCasting numerical simulation verified the feasibility of the design, predicting no significant defects in the casting region. The trial production confirmed the simulation results, yielding high-quality castings meeting technical requirements. The approach demonstrates that integrating simulation with process design is a powerful method to reduce sand casting defects and improve casting reliability. Future work could involve further optimization of riser dimensions and exploring the effect of mold preheating temperature on defect formation.