The production of high-integrity, heavy-section ductile iron castings, such as gear housings, presents significant metallurgical and foundry challenges. Among these, slag inclusion defects are particularly prevalent and detrimental, acting as stress concentrators that severely compromise mechanical properties, especially fatigue strength and elongation. This article details a first-person perspective on the holistic production process for a QT400-18RT grade gear housing, analyzes the root causes of slag inclusion formation, and presents validated countermeasures. The goal is to establish a stable, high-quality production methodology that consistently meets stringent technical requirements.
The specified gear housing, with a major wall thickness of 100-108 mm and a weight exceeding 2000 kg, demands exceptional material properties. The key specifications include a ferritic matrix (QT400-18RT), low silicon content (<2.0%), and acceptance via attached test blocks requiring tensile strength ≥370 MPa, yield strength ≥240 MPa, and elongation ≥12%. Furthermore, a minimum nodularity of 90% and the absence of defects like slag inclusion, cold shuts, and shrinkage verified by ultrasonic testing are mandatory.
Foundry Process Design and Parameters
A robust molding process is the foundation for dimensional accuracy and surface quality. We employ a furan resin sand system, ensuring a mold/core strength (σ_b) greater than 0.98 MPa. Rigid flasks and adequate mold wall thickness are crucial to prevent mold wall movement, thereby minimizing risks of shrinkage and penetration defects. The gating system is designed as an open type to ensure calm filling and minimize turbulence, which is a primary vector for slag inclusion entrapment. Ceramic tubes are used for the sprue and ingates to resist erosion. Key process parameters are summarized below:
| Component | Specification | Quantity | Purpose/Note |
|---|---|---|---|
| Sprue | Ø90 mm ceramic tube | 1 | Primary metal entry |
| Runner | 70/90mm x 130mm cross-section | – | Distributes metal |
| Ingates | Ø30 mm ceramic tubes | 10 | Calm entry into cavity |
| Chills (Various) | e.g., Ø90x120mm, 140x160x100mm | ~120 pieces total | Directional solidification, microstructure control |
| Feeder (Wedge) | Neck: 100x20x20mm; Top: 120x120mm; H: 500mm | 1 | Feed metal to compensate for shrinkage |
The strategic placement of chills, both external and internal (direct-contact), is critical for achieving directional solidification from the thick sections towards the feeder, preventing internal shrinkage defects that can sometimes be confused with slag inclusion clusters during inspection.
Melting, Composition, and Treatment Philosophy
The chemistry and treatment processes are meticulously controlled to achieve the high-elongation ferritic matrix while minimizing factors that promote slag inclusion.
Charge Material and Base Chemistry
To limit carbide-forming and trace elements, the charge consists of 20-30% high-purity pig iron balanced with 70-80% selected steel scrap. No returns are used to avoid the accumulation of trace elements from repeated treatments. The target base iron composition is set to provide a favorable foundation for subsequent spheroidization and inoculation.
| Element | Target Range (wt.%) | Rationale |
|---|---|---|
| Carbon (C) | 3.8 – 3.9 | Upper range for good fluidity and graphitization potential. |
| Silicon (Si) | 0.9 – 1.0 | Low base silicon allows for strong, controlled inoculation. |
| Manganese (Mn) | 0.2 – 0.3 | Kept low to avoid pearlite stabilization and embrittlement. |
| Phosphorus (P) | ≤ 0.05 | Minimized to prevent phosphide eutectic formation. |
| Sulfur (S) | ≤ 0.02 | Low initial sulfur reduces MgS/MgO slag formation during treatment. |
Spheroidization and Inoculation Treatment
Treatment is performed in a 5-ton medium-frequency induction furnace using a trench-type ladle and the sandwich method. The spheroidizer (Mg-Fe-Si alloy with controlled rare earths) is placed in the well, covered by a primary inoculant (Ca-Ba-FeSi), and then sealed with a covering agent. The reaction is controlled to be brisk but not violent.
Inoculation is the key to achieving a high nodule count and roundness, which directly influences mechanical properties. A multi-stage inoculation process is employed to combat fade and ensure effectiveness:
$$ \text{Total Inoculation} = I_{primary} + I_{stream} + I_{late} $$
Where:
$$ I_{primary} = 0.3-0.5\% \text{ (ladle bottom)}, \quad I_{stream} = 0.4-0.5\% \text{ (during tapping)}, \quad I_{late} = 0.1-0.15\% \text{ (during pouring)} $$
The late inoculation uses a fine-grade (0.2-0.7 mm) sulfur-oxygen inoculant added via a dedicated feeder into the pouring stream, ensuring intimate mixing just before the metal enters the mold.
Pouring Practice
The principle is “quick and quiet.” The pouring basin must be kept full to create a positive pressure head and prevent vortex formation that can draw surface oxides (slag inclusion precursors) into the sprue. The late inoculant is added synchronously with the metal stream into this full basin.
Initial Results and the Emergence of Slag Inclusion Defects
Initial production runs yielded promising results on attached test blocks (A1, A2). Chemical analysis and mechanical properties met all specifications.
| Sample ID | C | Si | Mn | P | S | Mg | RE | Rm (MPa) | Rp0.2 (MPa) | A (%) | Nodularity (%) |
|---|---|---|---|---|---|---|---|---|---|---|---|
| A1 | 3.57 | 1.98 | 0.22 | 0.022 | 0.010 | 0.045 | 0.0059 | 383 | 246 | 25.5 | 95 |
| A2 | 3.58 | 2.00 | 0.23 | 0.023 | 0.009 | 0.047 | 0.0055 | 379 | 241 | 26.0 | 95 |
However, ultrasonic testing of the casting from sample A1 indicated discontinuities. Metallographic analysis of the flagged area confirmed the presence of macroscopic slag inclusion defects, primarily located on the upper surfaces and coping regions of the casting.
The morphology was characteristic of dross or slag films entrapped during mold filling or early solidification, distinctly different from gas or shrinkage pores.
Root Cause Analysis of Slag Inclusion Formation
Slag inclusion in ductile iron can be classified as primary (formed during treatment/tapping) or secondary (formed during pouring and mold filling). Given the controlled residuals of Mg and RE (see table above), which were within low and acceptable ranges, the focus shifted to process dynamics. The relationship between pouring temperature (Tpour) and slag inclusion propensity is particularly critical and can be modeled conceptually. The likelihood of defect formation Lslag is a function of multiple variables:
$$ L_{slag} = f(T_{pour}, \eta(T), v_{flow}, [O,S]_{surf}, t_{float}) $$
Where η(T) is the temperature-dependent viscosity, vflow is the flow velocity, [O,S]surf is the surface oxide/sulfide film stability, and tfloat is the time available for slag particles to float out. An analysis of the initial process identified three major contributing factors:
| Factor | Initial Condition | Mechanism Promoting Slag Inclusion |
|---|---|---|
| Pouring Temperature | 1300 – 1340°C (Last casting often below 1300°C) | Low temperature increases metal viscosity (η↑), drastically reducing tfloat. Slag particles cannot ascend to the cope surface before solidification. |
| Gating System & Practice | Open system, but manual pouring with inconsistent basin fullness. | An unfilled basin creates vortices, drawing existing surface slag films into the sprue. Turbulence in runners fragments oxides, making them harder to trap. |
| Chill Quality and Management | Uncontrolled reuse, surface oxidation/rust. | Oxidized chill surfaces introduce exogenous oxides. Moisture or decomposition products from coatings can generate gas and promote local re-oxidation, leading to subsurface slag inclusion or slag-associated gas holes. |
The interplay of low pouring temperature and suboptimal gating practice was deemed the primary driver for the observed defects. The viscosity effect is dominant; as temperature decreases, the Stokes’ law velocity for a slag particle to rise becomes impractically slow:
$$ v_{rise} = \frac{2 g r^2 ( \rho_{Fe} – \rho_{slag} )}{9 \eta(T)} $$
Where r is the particle radius, g is gravity, and ρ are densities. For a given particle size, vrise is inversely proportional to viscosity η, which increases exponentially as temperature approaches the liquidus.
Implemented Countermeasures and Validation
Based on the root cause analysis, the following integrated corrections were implemented:
1. Optimized Pouring Temperature Regime
The target pouring temperature was raised to 1360-1380°C for all castings in the pour. This required adjusting ladle pre-heating and sequencing to ensure the last casting still met the minimum threshold. The higher temperature significantly reduces viscosity, increasing vrise and tfloat, allowing more time for intrinsic and reaction slags to coalesce and rise to the metal surface in the pouring basin or feeder.
2. Enhanced Gating System with Filtration
While retaining the open system for calm filling, a ceramic foam filter (150x150x32 mm) was inserted at the junction of the runner and ingates. This acts as a mechanical barrier to trap any macroscopic slag that passed the basin. Additionally, a slag trap was added at the end of the runner. The practice of maintaining a full pouring basin was strictly enforced and monitored.
3. Strict Chill Management Protocol
A chill management system was instituted:
- Lifecycle Limit: Chills were retired after a maximum of 5 uses.
- Surface Preparation: The working face of chills was machined or ground smooth before each use to remove all rust, scale, and burned-on sand.
- Pre-heating: Chills were preheated to around 120-150°C to eliminate moisture and reduce thermal shock, but kept below the oxidation threshold.
This eliminated exogenous sources of oxidation and gas that could lead to localized slag inclusion defects.
Results and Conclusion
The implementation of these coordinated measures—elevated pouring temperature, gating filtration, and rigorous chill control—was evaluated over a subsequent production run of 18 gear housing castings. The results were definitive:
- Ultrasonic Testing: 100% pass rate with no indications characteristic of slag inclusion defects.
- Dimensional and Visual Inspection: All castings were sound, with clean upper surfaces free from dross scabs or pits.
- Consistent Quality: The mechanical properties and microstructure from attached test blocks remained consistently within specification, confirming that the improvements did not adversely affect the metallurgy.
In conclusion, the stable production of heavy-section ductile iron castings like gear housings requires a systems-engineering approach. While proper chemistry and nodularizing treatment are fundamental, the control of dynamic process factors is paramount in defeating slag inclusion defects. The key learnings are:
- Pouring temperature is a critical lever; it must be sufficiently high to ensure low metal viscosity and adequate slag floatation time, with a practical minimum threshold around 1360°C for such castings.
- Even a well-designed gating system benefits from ceramic filtration as a fail-safe to trap residual slag, and operational discipline in maintaining a full pouring basin is non-negotiable.
- Auxiliary tooling like chills must be treated as consumable process-critical items with strict quality and lifecycle controls to prevent them from becoming defect initiation sites.
By systematically addressing the triad of temperature, turbulence, and exogenous contamination, the incidence of slag inclusion defects can be reduced to near zero, enabling reliable and economical production of high-performance ductile iron components.