The production of large steel castings, often characterized by single-unit or low-volume batches, presents a significant economic challenge. The substantial upfront cost associated with each unique mold makes achieving a 100% yield rate paramount for controlling overall production expenses. Traditional trial-and-error methods for process design are not only time-consuming and costly but also carry a high risk of failure for such critical components. In our foundry, specializing in castings ranging from 20 to 100 tons, the adoption of Computer-Aided Engineering (CAE) simulation has become an indispensable strategy. This technology allows for the virtual prototyping of the casting process, enabling the prediction and mitigation of potential casting defects before any metal is poured. This article details our application of Flow Science’s FLOW-3D software in the development of a large “cold type” steel casting, demonstrating its critical role in optimizing gating, risering, and the use of chills to ensure soundness.
The component in question is a cylindrical steel casting (ZG230-450) with a length of 3655 mm, a maximum outer diameter of 1410 mm, and an inner bore of 415 mm. Its most challenging feature is a substantial wall thickness of approximately 500 mm, with a total mass of around 23 tons. This combination of large size and heavy section dramatically increases the susceptibility to shrinkage-related casting defects, such as macro-porosity and micro-porosity (shrinkage cavities and sponginess), which form during solidification due to inadequate feeding.
| Parameter | Value |
|---|---|
| Material | ZG230-450 (Cast Steel) |
| Length | 3655 mm |
| Max. Outer Diameter | 1410 mm |
| Inner Diameter | 415 mm |
| Max. Wall Thickness | ~500 mm |
| Approx. Weight | 23,000 kg |
| Primary Defect Risk | Shrinkage Porosity |
The initial casting process was designed for a horizontal pouring orientation. An open gating system was employed to minimize flow resistance, featuring a single downsprue with a cross-sectional area of 80 cm², branching into three levels of horizontal runners (each 80 cm²). The first-level runner fed four ingates, while the second and third levels each fed two ingates, all with an area of 80 cm². Two open-top risers were placed at the highest points of the casting to provide a reservoir of molten metal to compensate for solidification shrinkage. The primary goal was to achieve directional solidification, progressing from the casting extremities toward these risers.
Numerical Simulation: Filling and Solidification Analysis
Numerical analysis using FLOW-3D, which employs the Volume of Fluid (VOF) method for accurate free-surface tracking, was conducted to validate and optimize this initial design. The simulation models key physical phenomena, including fluid flow, heat transfer, solidification, and the formation of casting defects.
Filling Process and Oxide Formation
The filling sequence was analyzed with an initial pour velocity of 2 m/s at the downsprue. The simulation confirmed that the open gating system promoted a calm and orderly fill. The metal front rose steadily from the bottom of the mold cavity towards the top, completing the fill in approximately 214 seconds. A critical aspect analyzed was the potential for oxide film entrapment, a common casting defect arising from turbulent surface folding. The streamlined filling pattern allowed oxides to float freely to the metal surface. The final simulation frame showed these oxides concentrated exclusively within the open risers, effectively removing them from the final casting body and eliminating this potential source of weakness.
Initial Solidification and Defect Prediction (Without Chills)
Following the fill, the solidification phase was simulated. The results indicated a problematic thermal profile. The thinner sections at the ends and near the bore solidified first, as expected. However, the massive, uniform wall section in the central part of the casting displayed very little thermal gradient. This led to what is essentially a simultaneous solidification pattern across the bulk of the casting wall. According to solidification theory, a significant volume of isolated liquid pockets will form in the final stages, leading to internal shrinkage porosity or micro-shrinkage because feed metal cannot reach them. The famous Chvorinov’s rule, while simplified, hints at this issue for sections of similar modulus:
$$ t_s = B \left( \frac{V}{A} \right)^n $$
where \( t_s \) is solidification time, \( V \) is volume, \( A \) is surface area, \( B \) is a mold constant, and \( n \) is an exponent (typically ~2). Sections with similar \( V/A \) ratios solidify at similar times, preventing directional feeding.
The simulation output graphically confirmed this. As shown in the defect prediction plot, severe and widespread micro-porosity was present throughout the lower half of the casting wall. Furthermore, the macroscopic shrinkage cavity (pipe) was correctly located within the risers, but it was excessively deep, threatening the integrity of the casting top. This validated the need for a riser feed but highlighted the unacceptable internal casting defect.
| Defect Type | Location (Without Chills) | Severity | Cause |
|---|---|---|---|
| Macro-shrinkage (Pipe) | Open Risers | High (Very deep) | Expected, but requires controlled depth. |
| Micro-porosity / Sponginess | Throughout lower casting wall | Critical (Widespread) | Simultaneous solidification; lack of thermal gradient. |
| Oxide Inclusions | None in casting body | None | Calm filling enabled flotation to risers. |
Process Optimization: Implementing Chill Casting
To eliminate the pervasive micro-porosity casting defect, the solidification sequence needed to be forcibly altered from simultaneous to directional. The strategy was to accelerate cooling at the casting bottom to create a strong vertical thermal gradient, establishing a bottom-to-top solidification front. This was achieved by placing external chills along the bottom quarter of the casting’s outer circumference.
The effect of chills can be conceptualized through the heat transfer equation at the chill-casting interface:
$$ q = h (T_{cast} – T_{chill}) $$
where \( q \) is the heat flux, \( h \) is the interfacial heat transfer coefficient, \( T_{cast} \) is the casting surface temperature, and \( T_{chill} \) is the chill temperature. The high thermal conductivity of the steel chills (compared to sand) and intimate contact create a high effective \( h \), dramatically increasing \( q \) and extracting heat from the casting bottom rapidly.
Simulation of Optimized Solidification (With Chills)
A new simulation was run incorporating the chills. The results demonstrated a complete transformation of the solidification pattern. A pronounced cooling wave initiated from the chilled bottom and the two ends, progressing uniformly upward towards the risers. This established a clear “directionally solidified” zone, satisfying the fundamental requirement for soundness: the creation of a continuous liquid feed path to the riser until the entire casting is solid.
The defect prediction results were markedly improved. The extensive internal micro-porosity was completely eliminated. Only a minimal amount of surface shrinkage was predicted on the lower part of the inner bore, which would be safely removed by subsequent machining. The shrinkage pipe remained in the riser, now functioning as intended.
To quantitatively illustrate the chills’ impact, the temperature histories at three points in the mid-length of the casting were compared: Point A (near outer bottom), Point B (mid-wall), and Point C (near inner top).
| Condition | Point A Cooling Rate | Point B Cooling Rate | Point C Cooling Rate | Max Temp Difference (A-C) at 7.5h |
|---|---|---|---|---|
| Without Chills | Low | Low | Low | < 50 °C |
| With Chills | Very High | High | Low | > 400 °C |
The temperature profiles clearly show that without chills, all points cool at a similar, slow rate with minimal gradient. With chills, Point A cools drastically faster, Point B at a moderate rate, and Point C remains hot much longer. This creates and maintains the essential vertical thermal gradient (\( \frac{dT}{dy} \)) needed for directional solidification, mathematically ensuring that the solidus isotherm moves progressively upward, preventing the isolation of liquid pools and the consequent casting defect formation.
The final solidification sequence can be summarized by monitoring the solid fraction \( f_s \) over time. The condition for sound casting is that at any time \( t \), the gradient of the solid fraction ensures a liquid path:
$$ \nabla f_s \cdot \vec{g} > 0 $$
where \( \vec{g} \) is the direction opposite to gravity (pointing towards the riser). The simulation with chills confirmed this condition was met, while the initial design violated it, leading to the casting defects.
Conclusion and CAE Value Proposition
This case study underscores the transformative power of CAE simulation in the realm of large steel casting production. The virtual analysis provided unambiguous insights that would have been impossible or prohibitively expensive to obtain physically.
1. Gating System Validation: The simulation confirmed that the multi-level open gating system promoted non-turbulent filling, successfully avoiding the casting defect of oxide inclusion entrapment within the casting body.
2. Critical Defect Prediction: The initial process simulation accurately predicted a catastrophic, widespread casting defect in the form of internal micro-porosity due to unfavorable simultaneous solidification, identifying a major flaw before tooling was even created.
3. Effective Optimization: CAE enabled the rapid design and validation of a corrective measure—the strategic placement of bottom chills. The simulation quantitatively demonstrated how the chills established the necessary thermal gradient, converting the solidification mode to directional and completely eliminating the internal shrinkage casting defect.
4. Risk Mitigation and Cost Savings: By identifying and solving the problem digitally, the foundry avoided the tremendous cost of a physical trial involving over 20 tons of steel, along with associated mold materials, energy, labor, and potential schedule delays. The guarantee of a sound casting first time dramatically reduces the economic risk inherent in single-piece production.
In conclusion, the integration of FLOW-3D CAE technology is no longer merely an advanced tool but a fundamental necessity for the competitive and reliable production of large-scale steel castings. It shifts the paradigm from reactive defect correction to proactive casting defect prevention, ensuring yield, quality, and economic viability in this demanding sector of metalcasting.