In the field of heavy-section casting, the production of large ductile cast iron components presents significant challenges due to prolonged solidification times and the inherent tendency for shrinkage defects. As a researcher focused on material forming technologies, I embarked on a detailed investigation into the casting process of a critical machine tool stand made from ductile cast iron (grade QT400-18). The initial manufacturing approach, which employed traditional riser-based feeding, resulted in subsurface shrinkage porosity at critical junctions, leading to part rejection. This article documents my comprehensive analysis, numerical simulation efforts, and the development of an optimized riser-free casting process that successfully produced sound castings while enhancing process efficiency and yield.
The machine tool stand is a structurally complex component with internal cavities and significant variation in wall thickness. Its overall dimensions are 2230 mm in length, 1750 mm in width, and 550 mm in height, with a weight of approximately 2483 kg. The wall thickness ranges from a minimum of 20 mm to a maximum of 80 mm, particularly around the top-mounted bearing support platforms which house thick shaft bosses. This disparity classifies it as a medium-to-large, intricate casting. The functional requirements demand high internal soundness and surface finish on machined faces, with no allowances for defects like shrinkage cavities, porosity, or sand inclusions.
The original casting methodology utilized furan resin no-bake sand for manual molding. The gating system was a multi-ingate step design to ensure balanced filling. The chemical composition of the molten ductile cast iron was meticulously controlled, as summarized in Table 1. The pouring temperature was maintained between 1310°C and 1330°C. To address the anticipated shrinkage in the thick bearing support sections, two large risers were placed atop these regions. However, practical production revealed minor but unacceptable shrinkage porosity at the junctions between the risers and the casting body, a clear indication of inadequate feeding.
| Element | Base Iron Range | After Nodularization Range |
|---|---|---|
| Carbon (C) | 3.6 – 4.0 | 3.5 – 3.9 |
| Silicon (Si) | 1.3 – 1.6 | 2.2 – 2.5 |
| Manganese (Mn) | ≤ 0.5 | ≤ 0.45 |
| Sulfur (S) | ≤ 0.03 | ≤ 0.02 |
| Phosphorus (P) | ≤ 0.04 | ≤ 0.04 |
| Residual Magnesium (Mg) | – | ≤ 0.05 |
To diagnose this issue, I employed AnyCasting simulation software to analyze the solidification sequence. The temperature field and solidification time results for the original riser-equipped design confirmed that the last areas to solidify were indeed the regions beneath the risers in the bearing support platforms. The simulation predicted a solidification sequence where the risers themselves froze before the critical hot spots in the casting, rendering them ineffective for feeding. This phenomenon can be partially described by Chvorinov’s rule for solidification time:
$$ t = k \left( \frac{V}{A} \right)^2 $$
where \( t \) is the solidification time, \( V \) is the volume of the casting section, \( A \) is its surface area, and \( k \) is the mold constant. For the thick bearing boss, the high \( V/A \) ratio leads to a longer local solidification time \( t \). In a糊状凝固 (mushy freezing) alloy like ductile cast iron, interdendritic feeding is difficult, and a riser must remain liquid longer than this hot spot to be effective. The simulation showed this was not achieved. The fraction solid evolution over time, \( f_s(t) \), in these regions indicated a prolonged pasty stage where shrinkage porosity could nucleate.
The fundamental issue stems from the solidification characteristics of ductile cast iron. During cooling, graphite precipitation from the austenite-graphite eutectic leads to a significant volumetric expansion, often referred to as graphite expansion. The net volume change, \( \Delta V_{net} \), during solidification is a balance between the contraction of the liquid metal and austenite (\( \Delta V_{shrinkage} \)) and the expansion from graphite nucleation (\( \Delta V_{graphite} \)):
$$ \Delta V_{net} = \Delta V_{shrinkage} – \Delta V_{graphite} $$
For successful feeding without risers, the process must harness \( \Delta V_{graphite} \) to compensate for \( \Delta V_{shrinkage} \). This requires a rigid mold cavity to contain the expansion pressure and a gating system that seals off early, forcing the expansion to act inwardly on the still-solidifying metal. The original process with large risers provided a “soft” escape for this expansion pressure, negating its feeding benefit and even drawing liquid back into the riser, creating shrinkage at its root.
Based on this analysis, I formulated an optimized riser-free casting process. The core principle was to eliminate the traditional risers and create conditions for controlled, pressurized solidification using the mold’s rigidity and early gating freeze-off. The key modifications are detailed in Table 2.
| Process Element | Original Process | Optimized Riser-Free Process | Purpose |
|---|---|---|---|
| Top Features | Two large cylindrical risers | Replaced with flat, wedge-shaped vent holes | Allow gas escape, freeze quickly, seal cavity. |
| Ingate Design | Standard thickness | Reduced to a flat rectangle, 18 mm thick | Ensure ingates solidify before casting, creating a sealed system. |
| Cooling Aids | None at bearing bosses | External or internal chills placed adjacent to bearing boss regions | Accelerate local solidification, refine microstructure, balance overall cooling. |
| Bearing Bore | Cored and cast | Option A: Cast with external chills. Option B: Not cast; left as solid for machining with chills. | Improve local soundness and mechanical properties. |
| Gating System | Step gating with multiple ingates | Retained, but with thinner ingates; sprue and lower ingates use ceramic tubes. | Maintain balanced fill while enabling early freeze-off. |
The new gating system design maintained the multi-ingate step configuration but with crucially thinner cross-sections. The ingate area calculation was revised to ensure rapid freezing. The flow rate \( Q \) during filling is governed by:
$$ Q = A_{ingate} \cdot v $$
where \( A_{ingate} \) is the total ingate cross-sectional area and \( v \) is the flow velocity. By reducing the ingate thickness, the cross-sectional area was decreased, but the number of ingates was maintained to keep \( Q \) sufficient for non-turbulent filling within the desired time. The primary goal was to manipulate the freezing time of the ingates relative to the casting. Using a simplified thermal model, the time for an ingate of thickness \( d \) to solidify can be approximated by:
$$ t_{ingate} \propto d^2 $$
By reducing \( d \), \( t_{ingate} \) decreases quadratically, ensuring it solidifies early in the overall casting solidification cycle, thereby isolating the mold cavity.
The role of chills is critical in this ductile cast iron process. Chills act as heat sinks, extracting heat rapidly from specific areas. The heat extraction rate \( \dot{q}_{chill} \) can be modeled using:
$$ \dot{q}_{chill} = h_{interface} \cdot A_{chill} \cdot (T_{melt} – T_{chill}) $$
where \( h_{interface} \) is the interfacial heat transfer coefficient, \( A_{chill} \) is the contact area, \( T_{melt} \) is the metal temperature, and \( T_{chill} \) is the initial chill temperature. Placing external chills on the sides of the bearing support platforms dramatically increased the local effective surface area \( A \), reducing the local \( V/A \) ratio and solidification time \( t \) according to Chvorinov’s rule. This served two purposes: it refined the graphite nodule structure in this critical high-stress area, and more importantly, it helped synchronize the solidification of this hot spot with the rest of the casting wall, promoting directional solidification towards the now-sealed internal cavities.
Numerical simulations of the optimized design vividly illustrated the improvement. The solidification time field showed a much more uniform progression of the solidus isotherm. The wedge-shaped vents and thin ingates were confirmed to freeze early, depicted in the simulation as solidified zones while the main casting body was still in a mushy state. A comparative analysis of the solidification parameters is shown in Table 3.
| Parameter / Region | Original Design (with Risers) | Optimized Design (Riser-free with Chills) |
|---|---|---|
| Last Point to Solidify | Under riser neck, at bearing boss junction. | Dispersed within internal cavity walls; no isolated hot spot. |
| Solidification Time Gradient | High gradient between riser and boss. | More uniform gradient; chills create steeper local gradient at bosses. |
| Feeding Path Status | Riser path closed before casting hot spot solidified. | Ingates/vent paths closed early; cavity sealed for graphite expansion. |
| Predicted Shrinkage | High probability at riser root (Confirmed in practice). | Very low probability (Confirmed in practice). |
The practical implementation validated the simulation results. Several castings were produced using the riser-free methodology with varying chill configurations. In all cases, the castings were sound upon visual and non-destructive inspection. The bearing support areas, whether cast with a cored hole and external chills or left solid for machining with internal chills, exhibited dense, porosity-free structures. The graphite expansion was successfully contained by the high-strength resin sand mold and the locked steel flask, creating the necessary internal pressure for feeding. The process yield, defined as the weight of the usable casting divided by the total weight of metal poured, showed a marked improvement due to the elimination of riser metal. The yield improvement \( \Delta Y \) can be expressed as:
$$ \Delta Y = \frac{W_{casting}}{W_{poured, new}} – \frac{W_{casting}}{W_{poured, old}} \approx \frac{W_{risers}}{W_{poured, old}} $$
where \( W \) denotes weight. Since \( W_{poured, new} \) is less than \( W_{poured, old} \) by approximately the weight of the risers, the yield increased significantly.
Further metallurgical analysis of the optimized ductile cast iron castings revealed the benefits of the controlled cooling. The use of chills promoted a finer graphite nodule count and smaller nodule size in the chilled zones, enhancing the mechanical properties locally. The overall microstructure was uniform, without the degenerate graphite or carbides sometimes associated with slow cooling in heavy sections. The success of this riser-free approach hinges on several interdependent factors, which I have summarized into a governing equation set for feasibility assessment:
1. Mold Rigidity Condition: The mold wall movement \( \delta \) must be negligible.
$$ \delta = \frac{P_{expansion} \cdot V_{cavity}}{E_{mold} \cdot A_{wall}} \approx 0 $$
where \( P_{expansion} \) is the pressure from graphite expansion, \( E_{mold} \) is the effective modulus of the mold assembly, and \( A_{wall} \) is the restraining area.
2. Gating Freeze-Off Condition: The solidification time of the ingate \( t_{ingate} \) must be less than the time for the casting skin to form a seal \( t_{skin} \).
$$ t_{ingate} < t_{skin} $$
3. Expansion Compensation Condition: The available graphite expansion volume \( V_{graphite} \) must suffice to compensate for the total shrinkage volume \( V_{shrinkage} \) within the sealed cavity.
$$ V_{graphite} \geq V_{shrinkage} = \beta \cdot V_{casting} $$
where \( \beta \) is the volumetric shrinkage coefficient for the specific ductile cast iron composition.
This case study demonstrates that for complex, heavy-section ductile cast iron castings, a shift from riser-dependent feeding to a controlled, riser-free process is not only viable but advantageous. The key lies in engineering the solidification sequence through intelligent gating design, strategic use of chills, and leveraging the natural graphite expansion within a rigid mold system. This approach simplifies molding operations, reduces cleaning and finishing labor, and most importantly, boosts the casting yield, contributing to more sustainable and cost-effective manufacturing of large ductile cast iron components. The principles established here—emphasizing mold rigidity, early flow path freeze-off, and thermal management with chills—provide a robust framework for optimizing similar casting processes for ductile cast iron across various industries.