In the production of heavy-duty marine diesel engines, the cylinder block represents one of the most critical and challenging castings. The stringent quality requirements, particularly the necessity for ultrasonic testing in key areas such as cylinder bores and bearing saddles, demand an exceptionally sound casting, free from major discontinuities. A specific series of these engine blocks, with cast weights reaching up to 20 tons in material QT400-15, presented a recurring and significant quality issue: the formation of slag inclusions, also commonly referred to as black dross or slag spots. These defects appeared in various locations on the castings, severely impacting the mechanical properties, especially elongation and impact toughness, and jeopardizing the batch production stability.
Root Cause Analysis of Slag Inclusions
Slag inclusions in ductile iron are typically categorized into two types: primary and secondary slag inclusions. Primary slag inclusions originate from oxides, sulfides, and other non-metallic compounds formed during melting or during the post-inoculation treatment process. These are carried into the mold cavity with the metal stream. Secondary slag inclusions form due to re-oxidation of the metal during the mold filling stage. They are frequently found on the upper surfaces of the casting, the lower surfaces of cores, or in isolated pockets where metal flow is stagnant. The presence of these defects indicates a breakdown in the control of metal quality, fluid dynamics, or temperature during the casting process.
To pinpoint the exact cause of the persistent slag inclusions in these large cylinder blocks, numerical simulation using MAGMA software was employed to analyze the original gating and feeding design. The simulation focused on the filling pattern, velocity field, and temperature distribution.
The original gating system was a single-side ingress design. Simulation results revealed several critical issues contributing directly to the formation of slag inclusions:
- Excessive Flow Velocity and Turbulence: The initial metal entry was characterized by a very high instantaneous velocity, causing severe splashing and air entrainment within the gating system, particularly near the ingates. The ingates were positioned directly opposite core prints at the bottom of the mold, further exacerbating turbulent flow and splashing against the sand core. This turbulence drastically increases the surface area of iron exposed to air, promoting re-oxidation and the creation of secondary slag inclusions. The kinetic energy of the stream can also break up existing slag films, dispersing them throughout the liquid.
- Poor Temperature Distribution and Cold Zones: The temperature field simulation showed a significant thermal gradient across the casting. The side opposite the single ingate was markedly colder than the side where the metal entered. Low-temperature regions are problematic because the viscosity of the molten iron increases as temperature drops, described by the relationship for dynamic viscosity ($\eta$):
$$ \eta \propto e^{(E_a / RT)} $$
where $E_a$ is the activation energy, $R$ is the gas constant, and $T$ is the absolute temperature. Higher viscosity severely impedes the buoyant rise of slag particles, trapping them within the casting. Furthermore, the original ingate was not located at the lowest point of the casting cavity in its pouring orientation. This created a “dead zone” in the lower areas opposite the ingate where early-arriving, cooler metal—potentially carrying the initial slag from the ladle—could stagnate, allowing slag inclusions to settle and become entrenched. - Pouring Basin and Ladle Practice: The simulation also indicated severe splashing in the pouring basin during mid-fill. In practice, the large casting volume required two ladles. The introduction of the second ladle’s stream into the basin likely caused additional turbulence, disturbing the slag layer that should have floated to the basin’s surface and dragging it into the sprue.
The key parameters contributing to slag inclusions are summarized in the table below:
| Factor | Original Process Issue | Impact on Slag Formation |
|---|---|---|
| Flow Dynamics | Single-side ingress; high velocity; ingate facing core. | Severe turbulence & re-oxidation (secondary slag inclusions). |
| Thermal Profile | Large thermal gradient; cold zones; dead metal areas. | Increased viscosity traps primary & secondary slag inclusions. |
| System Design | Closed gating system; poor slag trapping in basin. | Primary slag inclusions carried into cavity. |
Process Optimization and Corrective Actions
Based on the numerical simulation findings, a comprehensive set of corrective actions was implemented to tackle the root causes of the slag inclusions.
- Gating System Redesign: The single-side ingress was changed to a symmetrical, two-side ingress system. The cross-sectional areas of the main runner and ingates were significantly increased to transform the system from a pressurized (choked) to a non-pressurized (open) design. The governing equation for flow rate $Q$ is:
$$ Q = A \cdot v $$
where $A$ is the cross-sectional area and $v$ is the flow velocity. By increasing $A$, the metal entrance velocity $v$ is reduced for a constant $Q$, leading to laminar, quiescent filling. Furthermore, the ingate positions were offset to avoid direct impingement on core prints. - Optimization of Pouring Parameters: The pouring temperature was raised from the original 1330-1340°C to 1360-1370°C. This has a dual benefit: it lowers the metal’s viscosity (as per the equation above), enhancing slag floatation, and it extends the fluidity, allowing better filling of thin sections. The pouring practice was also refined using a “two-stage stopper” technique to better control the initial pour rate and minimize initial turbulence.
- Enhanced Process Controls: Strict protocols were enforced for thoroughly drying ladles and pouring basins to prevent gas evolution from residual moisture. The quality and compaction of molds and cores were rigorously checked to minimize erosion. Careful management of the two-ladle pour was instituted to minimize disturbance in the pouring basin during the switchover.
The new design was subjected to numerical simulation to verify its effectiveness. The results confirmed a dramatic improvement:
- Flow Field: Metal velocity was substantially reduced. Filling became calm and progressive, with minimal splashing or air entrainment. The symmetrical fill eliminated the previous cold dead zone, ensuring early metal could circulate and allow inclusions to float upwards.
- Temperature Field: The thermal gradient was significantly reduced, with a much more uniform temperature distribution across both sides of the block at equivalent heights. This homogeneity prevents local cold spots where slag inclusions could be trapped.
The comparative effects of the changes are outlined below:
| Aspect | Original Process | Optimized Process | Mechanism for Reducing Slag Inclusions |
|---|---|---|---|
| Ingress Design | Single-side | Symmetrical two-side | Eliminates dead zones, promotes uniform fill & temperature. |
| System Type | Pressurized | Non-pressurized (Open) | Lowers entry velocity, reduces turbulence & re-oxidation. |
| Ingate Position | Facing core print | Offset from obstacles | Prevents direct impingement and associated splashing. |
| Pouring Temp. | 1330-1340°C | 1360-1370°C | Reduces viscosity ($\eta$), enhances slag floatation velocity ($v_s$). |
The terminal velocity of a spherical slag particle rising in molten iron can be estimated using Stokes’ law:
$$ v_s = \frac{2}{9} \cdot \frac{(\rho_m – \rho_i) \cdot g \cdot r^2}{\eta} $$
where:
- $v_s$ = settling (or rising) velocity
- $\rho_m$ = density of molten iron
- $\rho_i$ = density of the inclusion (slag)
- $g$ = acceleration due to gravity
- $r$ = radius of the inclusion particle
- $\eta$ = dynamic viscosity of the molten iron
This formula clearly shows that increasing the temperature (which decreases $\eta$) directly increases the floatation speed ($v_s$) of slag particles, giving them more time to escape to the slag catcher or riser before the metal solidifies.
Production Validation and Conclusion
The optimized process was implemented in a production trial of 20 cylinder blocks. The results were conclusive: the occurrence of slag inclusion defects was effectively controlled, achieving a 100% qualification rate for the trial batch. The castings successfully passed all subsequent machining, assembly, and engine testing procedures, confirming the internal soundness achieved.
In conclusion, the systematic application of numerical simulation was instrumental in diagnosing the complex, multi-faceted origins of the slag inclusion problem in these large ductile iron castings. It moved the investigation from empirical guesswork to a physics-based analysis of fluid flow and heat transfer. The key to solving the issue lay not in a single change, but in a holistic optimization targeting fluid dynamics (reducing turbulence), thermal management (ensuring uniform temperature), and process control (optimizing temperature and pour practice). This integrated approach, guided by simulation, transformed a problematic casting into one capable of stable, high-quality batch production, demonstrating the critical role of virtual prototyping in modern foundry engineering for eliminating defects like slag inclusions.