In our foundry’s production of ductile iron castings, we have long grappled with various casting defects, most notably the pervasive issue of slag inclusion defect. This defect, often appearing as non-metallic inclusions in the subsurface or interior of castings, severely impacts machinability, pressure tightness, and mechanical properties. Literature and our practical experience consistently pointed to a primary culprit: excessive residual magnesium content in the treated iron. Following the adoption of the pressure magnesium addition method, while magnesium utilization improved significantly, the associated rise in residual magnesium levels made defects like the slag inclusion defect, along with subcutaneous blowholes, shrinkage porosity, and dispersed shrinkage, increasingly pronounced. This paper details our systematic investigation and successful implementation of methods to reduce the total magnesium addition and precisely control the residual magnesium content, thereby effectively suppressing the formation of slag inclusion defect.
1. The Fundamental Relationship: Residual Magnesium and Slag Formation
The mechanism linking magnesium to slag inclusion defect is primarily chemical and thermodynamic. Magnesium is a highly reactive element. When added to iron for graphite spheroidization, it first reacts with sulfur and oxygen present in the melt:
$$Mg + S \rightarrow MgS$$
$$2Mg + O_2 \rightarrow 2MgO$$
Both magnesium sulfide (MgS) and magnesium oxide (MgO) are solid, non-wettable compounds at iron casting temperatures. If the residual magnesium content after treatment is excessively high, it increases the chemical potential for the formation of these compounds throughout the holding and pouring stages. These compounds agglomerate, often combining with silicates from slag or mold reactions, to form complex, coarse inclusions. The strong surface tension between these inclusions and the liquid iron prevents their flotation and removal, leading to their entrapment within the casting matrix as the dreaded slag inclusion defect. Therefore, the core principle of our approach was to determine the minimum residual magnesium level required for reliable nodularization and rigorously control the process to achieve it, avoiding unnecessary excess.
2. Principles for Determining Magnesium Addition
Traditionally, the magnesium addition amount (\(Mg_{added}\)) was determined empirically based on several factors:
$$Mg_{added} = f([S]_{initial}, [Mg]_{residual-target}, Mg_{losses})$$
Where:
\([S]_{initial}\) = Initial sulfur content of the base iron.
\([Mg]_{residual-target}\) = Target residual magnesium needed for spheroidization (typically > 0.03%).
\(Mg_{losses}\) = Magnesium lost to oxidation during treatment, transfer, and pouring.
Due to the difficulty in accurately predicting \(Mg_{losses}\), a common practice was to add a large safety margin of magnesium. This over-addition inevitably led to high residual levels and exacerbated the slag inclusion defect. We revised this principle: the addition must be based strictly on achieving complete graphite spheroidization, with the target residual magnesium kept as low as feasibly possible within the effective range (e.g., 0.03%-0.05%). The key was to maximize magnesium utilization (\(\eta_{Mg}\)) to achieve this low residual from a reduced addition.
3. Our Production Methodology for Reducing Magnesium Addition
Our standard production setup uses an acid cupola for melting. The charge consists of pig iron, ductile iron returns, ferromanganese (FeMn), ferrosilicon (FeSi), magnesium ingot (Mg > 99%), and a silicon-calcium alloy (SiCa). The pivotal equipment is our proprietary “bottle-type” pressure addition ladle, as shown in the schematic. Its design, particularly the depth of the immersion bell and the headspace volume, is critical for efficiency.
The process sequence is as follows:
- Charge Make-up and Base Iron Chemistry: We aim for a near-eutectic, high-carbon (≈3.8%), low-silicon (≈1.2%) base iron composition. This promotes higher magnesium absorption and better nodularization.
- Pressure Magnesium Treatment: This is the most critical step for controlling residual magnesium. The base iron is tapped into the pre-heated pressure ladle containing the magnesium ingot within a sealed bell. The lid is secured, and the treatment occurs under a controlled pressure atmosphere.
- Post-Inoculation: After treatment, the slag is skimmed. Ferrosilicon inoculant is then added to the ladle to promote graphite formation and avoid chilling. We found that vigorous stirring after inoculation is essential for uniform silicon recovery (~90%).
To systematically reduce the addition, we fixed all other melting and treatment parameters (tap temperature, headspace volume, treatment time) within narrow windows. We then varied the magnesium addition in steps from 0.20% down to 0.14% of the iron weight and measured the resulting residual magnesium and graphite quality. The results from a representative production batch are summarized in Table 1.
| Heat No. | Mg Added (%) | Initial S (%) | Final S (%) | Residual Mg (%) | Mg Utilization, η (%) | Graphite Nodularity |
|---|---|---|---|---|---|---|
| A-105 | 0.20 | 0.030 | 0.008 | 0.068 | 42.6 | Excellent (≥95%) |
| A-108 | 0.18 | 0.028 | 0.007 | 0.056 | 44.8 | Excellent |
| A-112 | 0.16 | 0.029 | 0.007 | 0.047 | 46.5 | Excellent |
| A-115 | 0.14 | 0.031 | 0.008 | 0.036 | 44.2 | Good (≈90%) |
The data shows a clear, controllable relationship. By reducing the addition to 0.14-0.16%, we successfully maintained residual magnesium in the 0.036-0.047% range, which proved sufficient for good nodularization while being low enough to mitigate the slag inclusion defect.
4. Key Factors for High Magnesium Utilization and Low Residual Mg
Maximizing magnesium utilization (\(\eta_{Mg}\)) is essential for reducing the added amount. It is calculated as:
$$\eta_{Mg} = \frac{[Mg]_{residual} \times 100 + 0.76 \times ([S]_{initial} – [S]_{final})}{Mg_{added}} \times 100\%$$
The factor 0.76 represents the stoichiometric mass of magnesium required to remove 1% of sulfur (\(Mg/S\) atomic mass ratio ≈ 24.3/32.1 ≈ 0.76). Our focus was on optimizing the variables in this equation.
4.1 Ladle Design and Operation
The “bottle-type” pressure ladle, with a high height-to-diameter ratio for the immersion bell, ensures deep plunging and prolonged contact time between magnesium vapor and iron, increasing absorption. A tight seal minimizes oxidation losses. Using a cast copper bell instead of steel prevents premature melting and allows for a more controlled release of magnesium.
4.2 Controlling Base Iron Sulfur
A significant portion of added magnesium is consumed by desulfurization. The magnesium consumed for desulfurization (\(Mg_{for-S}\)) is approximately:
$$Mg_{for-S} \approx 0.76 \times \Delta S$$
where \(\Delta S = [S]_{initial} – [S]_{final}\). Therefore, a lower initial sulfur directly reduces the magnesium required. While using a basic lining or low-sulfur charge is ideal, we focused on cupola operation. Intensifying melting by using a hot-blast system increased tap temperature and reduced coke consumption, which indirectly lowered sulfur pickup. Comparative data is shown in Table 2.
| Cupola Condition | Blast Temperature (°C) | Coke Ratio (%) | Avg. Initial Sulfur, [S]initial (%) |
|---|---|---|---|
| Before Hot-Blast Retrofit | Ambient | 14 | 0.032 – 0.035 |
| After Hot-Blast Retrofit | 250 – 300 | 11 | 0.028 – 0.031 |
4.3 The Role of Silicon-Calcium (SiCa) Addition
We introduced a small, consistent post-inoculation addition of SiCa alloy (0.1-0.2%). Calcium acts as a surface-active element, modifying the morphology of the oxides and sulfides formed. Instead of forming coarse MgO/MgS clusters prone to causing slag inclusion defect, calcium promotes the formation of finer, more globular, and potentially less harmful complex oxy-sulfides (e.g., Ca, Mg, Al silicates). This modification improves their flotation and removal from the melt before pouring, directly reducing the incidence of slag inclusion defect.
5. Experimental Verification: Reduction of Slag Inclusion Defects
To conclusively verify the effect of lower residual magnesium and SiCa addition on the slag inclusion defect, we conducted dedicated production-scale tests. We poured standard test blocks (Y-blocks or step blocks) under different treatment conditions and performed thorough macro- and micro-examination.
5.1 Test Methodology
Test blocks were poured from heats treated with varying magnesium additions, with and without SiCa addition. Key process parameters were recorded. The cast blocks were sectioned and examined for macro-slag inclusions on fractured surfaces and polished sections. Microstructure, chemical analysis, and mechanical properties were evaluated.
5.2 Results and Analysis
The results from a series of test casts are consolidated in Table 3. The “Slag Index” is a qualitative rating from 1 (severe, numerous large inclusions) to 5 (virtually free of macro inclusions).
| Cast ID | Mg Added (%) | Residual Mg (%) | SiCa Added? | Slag Index (1-5) | Tensile Strength (MPa) | Elongation (%) | Graphite Nodularity (%) |
|---|---|---|---|---|---|---|---|
| S-1 (Reference) | 0.20 | 0.065 | No | 2 (Poor) | 720 | 4.5 | 98 |
| S-2 | 0.16 | 0.048 | No | 3 (Moderate) | 740 | 6.0 | 96 |
| S-3 | 0.16 | 0.046 | Yes (0.15%) | 4 (Good) | 750 | 7.5 | 97 |
| S-4 | 0.14 | 0.035 | Yes (0.15%) | 5 (Excellent) | 735 | 8.0 | 90 |
The correlation is striking. Cast S-1, with high residual Mg (0.065%), showed a pronounced slag inclusion defect. Simply reducing the addition to 0.16% (S-2) improved the situation. The combination of reduced magnesium addition (0.16% or 0.14%) and SiCa addition (S-3, S-4) resulted in excellent surface quality and minimal slag inclusion defect. Importantly, the mechanical properties remained excellent, and graphite nodularity was maintained at acceptable levels (>90%) even at the lowest addition of 0.14%.
The microstructural analysis confirmed that the inclusions present in high-residual-Mg samples were predominantly coarse MgO/MgS clusters. In samples treated with lower Mg and SiCa, the inclusions were fewer, smaller, and often of a more complex, modified chemistry.
6. Comprehensive Process Control Summary
Our successful strategy to minimize the slag inclusion defect rests on an integrated control of the entire process, summarized by the following equation and principles:
The target residual magnesium is the cornerstone:
$$[Mg]_{residual-target} \approx 0.04\%$$
To achieve this from a minimized addition:
$$Mg_{added} = \frac{ [Mg]_{residual-target} + 0.76 \times \Delta S }{\eta_{Mg}}$$
Therefore, the operational goals are:
- Maximize Utilization (\(\eta_{Mg}\)): Achieve >45% via optimal pressure ladle design (deep bell, correct headspace), perfect sealing, and consistent, rapid treatment.
- Minimize Sulfur Pickup (\(\Delta S\)): Use a stable, low-sulfur charge and optimized, intensified cupola melting (e.g., with hot blast) to keep \([S]_{initial}\) consistently below 0.032%.
- Employ Inclusion Modifiers: Add a small, controlled amount of SiCa (0.1-0.2%) post-treatment to modify inclusion morphology, aiding flotation and reducing the severity of any entrained slag inclusion defect.
- Strict Process Discipline: Fix all other parameters (tap temperature, treatment timing, inoculation practice) to reduce variability, making the relationship between \(Mg_{added}\) and \([Mg]_{residual}\) predictable and controllable.
7. Conclusion
Through systematic investigation and production trials, we have established a reliable and practical methodology for significantly reducing the slag inclusion defect in ductile iron castings. The core of this methodology is the deliberate reduction of total magnesium addition coupled with precise process control to maintain a low but sufficient residual magnesium level (0.035-0.048%). This is made feasible by maximizing magnesium utilization through equipment and operational optimization, and by controlling base iron sulfur. The supplemental use of silicon-calcium alloy further enhances the cleanliness of the iron by modifying residual inclusions. This integrated approach has allowed us to consistently produce high-quality ductile iron castings with excellent mechanical properties and a dramatic reduction in defect-related scrap, proving that controlling residual magnesium is the most effective lever for mitigating the slag inclusion defect.