In modern foundry engineering, the application of numerical simulation software to model the filling and solidification processes of castings has become an indispensable tool. This technology allows engineers to directly observe these critical phases on a computer. More importantly, it enables the effective prediction of potential defects—such as shrinkage porosity and cavities—including their probable size, location, and the timing of their formation during the solidification sequence. This predictive capability is paramount during the initial casting process design stage, allowing for the optimization of gating and feeding systems before any metal is poured, thereby ensuring final casting quality and significantly reducing development time and cost.
This article details a practical application where I utilized this approach for a critical steel component. The subject was a structural shell casting known as a PIVOT housing. The primary challenge was that this component was designed to withstand substantial pressure loads, imposing a stringent technical requirement: the final casting must be completely free from shrinkage defects. To achieve this, I employed a workflow integrating three-dimensional modeling with advanced solidification simulation to design, analyze, and optimize the casting process.
The geometry of the steel shell casting was relatively complex, featuring several intersecting sections that created pronounced thermal hot spots. Its key dimensions were approximately 500 mm x 530 mm x 130 mm, with a final weight of about 60 kg. The specified material was equivalent to Chinese standard ZG30, a medium-carbon cast steel. The fundamental casting parameters were established first: a linear shrinkage allowance of 2% was applied, a machining allowance of 4 mm was added to all relevant surfaces, and a horizontal parting plane was selected for the mold. The initial process was designed for single-cavity molds using hand-molded sodium silicate-bonded sand.
Initial Casting Process Design
The first step was to create a complete digital model. I used Pro/ENGINEER software to build an accurate 3D solid model of the shell casting itself. Subsequently, the initial gating and feeding system was designed around this model based on established foundry engineering principles and empirical rules.
The gating system was designed as a traditional pressurized system to promote rapid and tranquil filling. The dimensions for the sprue, runner, and ingates were calculated using standard choke area formulas derived from handbooks. The key parameters for the initial setup are summarized in the table below.
| Component | Cross-Sectional Shape | Dimensions (mm) | Calculated Area (mm²) |
|---|---|---|---|
| Sprue (Top) | Circular | φ40 | ~1257 |
| Runner | Trapezoidal | Top: 30, Bottom: 25, Height: 20 | ~550 |
| Ingates (x2) | Rectangular | 30 x 10 | 300 (each) |
Feeding design is critical for steel shell castings due to their high shrinkage volume and tendency for skin-forming (pasty) solidification. The material, ZG30, has a significant freezing range, leading to mushy zone formation. The solidification time, $t_s$, for a section can be approximated by Chvorinov’s Rule:
$$
t_s = B \left( \frac{V}{A} \right)^n
$$
where $V$ is the volume, $A$ is the surface area, $B$ is the mold constant, and $n$ is an exponent typically close to 2. The modulus method was used to size the risers. The modulus $M$ of a section is defined as its volume-to-cooling-surface-area ratio:
$$
M = \frac{V}{A}
$$
A riser must have a greater modulus than the section it feeds to solidify last. The main thermal junctions in the shell casting were identified. The initial strategy involved placing two side risers (φ120 mm x 200 mm) to feed the thick sections furthest from the gating system, while relying partially on the gating system itself to provide some feeding to nearer hot spots—a technique sometimes called “gating risering.”
The chemical composition specification for the ZG30 steel is provided below.
| Element | C | Si | Mn | P | S |
|---|---|---|---|---|---|
| Content | 0.27 – 0.35 | 0.17 – 0.37 | 0.50 – 0.80 | ≤ 0.03 | ≤ 0.03 |
The melting was planned for a 100 kg medium-frequency coreless induction furnace. Key thermal parameters were set: a tap temperature range of 1580°C – 1600°C and a pouring temperature range of 1540°C – 1560°C. The target pouring time was set at 6 seconds per mold based on prior experience.
Solidification Simulation of the Initial Design
To virtually test this initial design, I employed numerical simulation. The Pro/ENGINEER models of the casting, gating, and risering system were exported in the STL (stereolithography) file format, which is a standard for describing 3D surface geometry. These files were then imported into the InteCAST (also known as HuaZhu CAE) simulation software system. This system is a Finite Difference Method (FDM) based code specifically designed for simulating casting processes.
The simulation workflow consists of three core modules:
- Pre-processing: The STL files are assembled, and the computational domain is discretized into a finite difference mesh. A mesh size of 3 mm was selected, resulting in a model with approximately 5.4 million cells. Material properties (for ZG30 steel and the sand mold) and boundary conditions (interfacial heat transfer coefficients) were assigned from the software’s database.
- Computational Processing: For this initial analysis, focusing on defect prediction, only the solidification thermal field was calculated, ignoring fluid flow. The governing energy equation for transient heat conduction 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, $T$ is temperature, $t$ is time, $k$ is thermal conductivity, $L$ is latent heat of fusion, and $f_s$ is the solid fraction. The software solves this equation numerically, accounting for the release of latent heat. - Post-processing: After the simulation run is complete, results are visualized. The most critical output for this study was the map of predicted shrinkage defects, generated using a widely-accepted porosity criterion model based on the local thermal gradient $G$ and solidification rate $R$. Porosity is predicted in regions where the $G/\sqrt{R}$ value falls below a critical threshold specific to the alloy.
The simulation results for the initial design were revealing and concerning. The defect prediction map clearly showed two distinct regions of predicted shrinkage porosity within the body of the shell casting. Crucially, these defects were located in the thick sections adjacent to the gating system. The analysis of the solidification sequence—observing the progressive isolation of liquid pockets—provided the explanation.
The simulation animated the solidification front progression. It showed that while the two dedicated side risers effectively created directional solidification toward themselves in their respective regions, the strategy of using the gating system for feeding failed. The thermal analysis revealed that the gates and runner sections solidified and lost their thermal connectivity to the casting’s hot spots before those hot spots had fully solidified. This premature isolation created isolated liquid pools within the shell casting that could not be fed, leading to the predicted microporosity. The feeding distance from the gates was insufficient for the geometric and thermal characteristics of these particular sections of the steel shell casting.
Process Optimization Based on Simulation Insights
The simulation provided a clear diagnostic: the feeding system was inadequate. Relying on the gating channels for significant feeding was unreliable for this geometry. The solution was to implement a fully riser-fed approach for all major hot spots. The concept of riser effective feeding distance became the guiding principle. The effective feeding distance $L_f$ for a riser can be estimated empirically, often as a multiple of the casting section thickness $T$:
$$
L_f = k \cdot T
$$
where $k$ is a factor dependent on alloy, casting geometry, and cooling conditions. For steel in sand molds, $k$ typically ranges from 4 to 6. Applying this to the shell casting sections indicated that the two original risers were insufficient to cover the entire casting volume.
Consequently, the process was redesigned. The new layout incorporated a total of four cylindrical risers. To maintain a favorable yield (weight of sound casting / total poured weight), the riser size could be reduced compared to the initial large ones, as their feeding responsibility was now shared and their effective range was fully utilized. The new risers were sized at φ100 mm x 150 mm. The gating system was modified to be purely for filling, not for feeding, and was repositioned to efficiently deliver metal to the cavity without creating new thermal imbalances. The computational mesh for this optimized design contained approximately 544,000 cells.
| Parameter | Initial Design | Optimized Design |
|---|---|---|
| Number of Riser | 2 | 4 |
| Riser Dimension | φ120 mm x 200 mm | φ100 mm x 150 mm |
| Primary Feeding Source | 2 Risers + Gating System | 4 Risers |
| Predicted Defects in Casting | Yes (2 zones) | No |
| Approximate Yield | ~65% | ~63% |
Simulation and Production Validation of the Optimized Design
The optimized design was subjected to the same rigorous numerical simulation. The results were markedly different. The solidification sequence animation now showed a clear, progressive solidification front moving from the extremities of the casting and from thin sections toward the four risers. The liquid pockets remained interconnected with the riser’s liquid reservoir until the final stages of solidification. The defect prediction algorithm confirmed the success: no shrinkage porosity or cavities were predicted within the critical sections of the shell casting. The only predicted shrinkage was contained safely within the riser bodies themselves, which are removed during finishing.
Encouraged by the simulation results, the optimized design was released for production. Molds were prepared using the specified sodium silicate sand process. A 100 kg heat of ZG30 steel was melted, tapped, and poured within the prescribed temperature ranges. After shakeout, heat treatment (annealing to relieve stresses and homogenize the structure), and cleaning, the castings were visually inspected and then sectioned for rigorous internal quality checks.
The production validation was successful. Macroscopic examination of the sectioned hot spots showed sound, dense metal with no evidence of macro-shrinkage or significant microporosity clusters. The castings proceeded to machining, where they were fully validated to meet all dimensional and serviceability requirements. The integration of 3D modeling and numerical simulation enabled a right-first-time approach for this demanding steel shell casting, eliminating the need for multiple, costly, and time-consuming physical trial casts.
Conclusion
This project underscores the transformative impact of numerical simulation in foundry practice, particularly for critical components like pressure-bearing steel shell castings. The key takeaways are:
- Predictive Power: Simulation accurately identified a flawed feeding strategy in the initial design, predicting specific defect locations that aligned with known geometric and thermal risks.
- Mechanistic Understanding: It provided more than just a yes/no answer; the visualization of the solidification sequence offered a clear physical explanation—premature isolation of liquid pools—guiding the direction of the redesign.
- Effective Optimization: The simulation served as a virtual testbed, allowing for the rapid evaluation of the four-riser design and confirming its efficacy before any metal was poured. The principle of ensuring directional solidification toward adequately sized and positioned risers was conclusively validated.
- Economic and Temporal Efficiency: The entire optimization cycle was completed digitally, dramatically shortening the process development lead time, reducing material and energy waste from physical trials, and ensuring a faster route to a reliable, high-quality production process for the shell casting.
The methodology demonstrated—from 3D CAD modeling through to virtual solidification analysis and defect prediction—forms a robust framework for modern casting process design. It moves the industry from reliance on empirical trial-and-error towards a science-based, predictive engineering discipline. For complex steel shell castings where internal soundness is non-negotiable, such simulation-led optimization is not just beneficial but essential for achieving quality, performance, and economic goals.