For many years, our foundry specialized in producing large, high-power marine diesel engines, with ductile iron crankshafts serving as their core components. We possessed extensive experience in manufacturing these critical parts. However, our production journey was significantly hampered by a persistent and severe quality issue: a high scrap rate primarily caused by slag inclusion defects, often referred to as “black spots.” At its peak, the scrap rate for crankshafts reached alarmingly high levels, with slag inclusion defects accounting for a dominant portion of this total waste. This situation severely impacted production schedules and economic performance. In response, a dedicated quality task force was established to conduct a scientific and rigorous investigation into the causes and mechanisms behind the formation of these slag inclusion defects. Through systematic analysis and the implementation of targeted corrective measures, we achieved remarkable improvements. The scrap rate declined steadily over several years, eventually falling to a minimal level. Notably, for a continuous period thereafter, not a single crankshaft was scrapped due to slag inclusion defects, fundamentally transforming production reliability, reducing costs, and enhancing overall economic and social benefits.
The manifestation of the slag inclusion defect followed a distinct pattern. These defects were predominantly located on the upper surfaces of the crankpin and main journals, as well as on the cope-side shoulder faces relative to the parting plane. Their characteristic shape was crescent-like. The severity of the slag inclusion defect was most pronounced in areas farthest from the ingate. Typically, these flaws were only revealed during machining. If the slag was contained within the machining allowance, it could be removed. However, less severe inclusions, invisible to the naked eye, would become clearly evident as pronounced indications under magnetic particle inspection. In the most serious cases, large, dull, cloud-like patches were visibly apparent on the machined surface, areas which oxidized first upon exposure to air.
To understand the nature of the slag inclusion defect, comparative metallographic and chemical analyses were conducted. Samples were taken from both a defective region on a main journal (far from the ingate) and a sound region of the same casting. Microscopic examination of the sound area revealed a standard microstructure of spherical graphite within a matrix of pearlite and ferrite. In contrast, the defective area showed blocky and curved, ribbon-like slag inclusions distributed within the matrix. Under dark-field illumination, these inclusions appeared gray-black surrounded by a bright border, confirming their primary identity as oxidized slag particles. Interestingly, the graphite nodules adjacent to these slag particles remained spherical. Chemical analysis of the two regions provided critical quantitative insight, as summarized in the table below.
| Element | Slag Inclusion Zone (wt.%) | Sound Zone (wt.%) |
|---|---|---|
| C | 3.65 | 3.68 |
| Si | 2.85 | 2.55 |
| Mn | 0.65 | 0.55 |
| P | 0.045 | 0.035 |
| S | 0.018 | 0.012 |
| Mg | 0.045 | 0.040 |
| RE (Rare Earth) | 0.035 | 0.025 |
The data clearly indicates elevated levels of Si, Mn, P, S, Mg, and Rare Earth (RE) in the slag inclusion defect region. A substantial portion of these elements is believed to exist in compound forms. The mechanism for the formation of this slag inclusion defect is a complex sequence of events during melting, treatment, and pouring. During the spheroidization treatment, elements like Mg and RE from the inoculant react vigorously with oxygen and sulfur present in the base iron, forming compounds such as MgO, RExOy, MgS, and RExSy. The density ($\rho$) of these compounds relative to the molten ductile iron ($\rho_{iron} \approx 7.1 \, g/cm^3$) is crucial:
$$ \rho_{MgO} \approx 3.6 \, g/cm^3, \quad \rho_{RE_2O_3} \approx 7.3 \, g/cm^3, \quad \rho_{MgS} \approx 2.8 \, g/cm^3 $$
Lighter compounds like MgO and MgS can float up readily. However, dense RE oxides have a density very close to that of the iron melt, making their natural buoyancy-driven removal difficult. Consequently, a significant amount remains suspended. This constitutes the primary slag formation. During the pouring and mold filling process, fresh metal surfaces are continuously exposed to air, leading to secondary oxidation and the formation of oxide films. Turbulence and splashing within the mold cavity fragment and entrain these films. While these films can float, they also tend to adsorb and carry the suspended dense RE oxide particles, as well as other non-metallic inclusions and free graphite, toward the upper surfaces of the casting. The final slag inclusion defect is thus an agglomerate of primary and secondary oxidation products. The rate of inclusion flotation can be approximated by Stokes’ law for small spheres:
$$ v = \frac{2 g r^2 (\rho_{melt} – \rho_{inclusion})}{9 \eta} $$
where $v$ is the terminal velocity, $g$ is gravity, $r$ is the inclusion radius, and $\eta$ is the melt viscosity. This shows that flotation is severely hampered for small particles ($v \propto r^2$) and for inclusions with density close to the melt.
The thermal conditions during solidification play a decisive role in determining the final location and severity of the slag inclusion defect. The mold and extensive use of chills rapidly extract heat, creating a “stagnant” mushy zone at the casting surface. The depth of this zone ($d$) is inversely related to the pouring temperature ($T_p$) and the cooling rate ($\dot{T}$):
$$ d \propto \frac{1}{(T_p – T_{liquidus})} \quad \text{and} \quad \dot{T} = \frac{k (T_{melt} – T_{mold})}{\rho C_p \delta^2} $$
where $k$ is thermal conductivity, $\rho$ is density, $C_p$ is specific heat, and $\delta$ is the solidified shell thickness. A higher pouring temperature results in a thinner mushy zone, allowing more time and a clearer path for slag particles to float to the surface before being trapped. Conversely, a low pouring temperature deepens this zone, capturing inclusions within the casting wall and leading to a more severe subsurface slag inclusion defect.
The original gating system design was identified as a major contributor to the slag inclusion defect. It featured a single, concentrated entry point at the riser end of the long crankshaft casting. This design caused severe turbulence, waterfall effects, and metal splashing as the metal flowed through the complex geometry of crank throws and wide shoulders, leading to excessive oxidation. Furthermore, it created a pronounced thermal gradient, with the metal at the far end being significantly cooler and more oxidized, explaining the severe slag inclusion defect in that region. The entire mold acted as a heated runner, preventing simultaneous and balanced solidification. The key corrective action was to redesign the gating system into a dispersed, multiple-ingate configuration. This minimized temperature drop, promoted quiescent mold filling, reduced thermal gradients, and enabled more uniform solidification, thereby drastically reducing the potential for slag inclusion defect formation.
A multi-pronged approach was implemented to address the root causes of the slag inclusion defect. The table below summarizes the key influencing factors and the corresponding control measures.
| Key Factor | Target/Principle | Implemented Action |
|---|---|---|
| Molten Metal Temperature | Maximize superheat to lower viscosity, promote inclusion flotation, and delay surface solidification. | Computer-controlled cupola operation, strict charge sequencing (high-quality materials in early melts), use of block CaC2, optimized blast air control, and thorough slagging. Target tap temperature > 1450°C. |
| Carbon Equivalent (CE) | Target near-eutectic composition for maximum fluidity and longest liquid residence time. | Control CE = %C + ⅓ %Si to approximately 4.3-4.5. The liquidus temperature $T_L$ is minimized at the eutectic, maximizing superheat $\Delta T = T_{pour} – T_L$. |
| Inoculant & Spheroidizer | Minimize addition to reduce slag generation and temperature drop. | Use minimum amount required for effective nodularization and to avoid chill. Employ multi-component alloys (e.g., Si-Ca, RE-Ca-Si) for better deoxidation, forming low-melting-point, low-density slags. |
| Covering & Fluxing Agents | Aggregate, dissolve, and remove slag; protect metal surface from re-oxidation. | Strategic use of sodium fluoroaluminate (Cryolite, Na3AlF6): 0.1% added to treatment ladle, 0.15% stirred in after treatment, and a final cover. Its decomposition gases help float inclusions. Expanded perlite powder used as an excellent covering agent to aggregate slag and form a protective barrier. |
| Process Timing | Minimize time for re-oxidation and slag re-dissolution before pouring. | Complete all treatment, skimming, and covering operations within 10 minutes, followed by a 5-minute stillness period for final inclusion flotation. |
| Pouring Temperature | Maintain temperature high enough to prevent premature solidification trapping slag. | Strictly control pouring temperature in the range of 1320°C – 1350°C. Below 1300°C, the risk of slag inclusion defect increases sharply. |
| Mold & Core Gas Venting | Prevent back-pressure and gas-driven turbulence that increases oxidation. | Ensure core moisture < 1.5%. All core vents connected to exterior. Multiple venting paths including main parting line vents and small risers on heavy sections. |
The thermodynamic driving force for the formation of oxide inclusions can be described by the Gibbs free energy of formation. For the key reaction:
$$ 2[Mg] + [O] \rightarrow MgO_{(s)} $$
The equilibrium constant $K$ and the resulting dissolved oxygen level are functions of temperature and magnesium content:
$$ K_{MgO} = \frac{a_{MgO}}{[a_{Mg}]^2 [a_O]} \approx \frac{1}{[\%Mg]^2 [\%O]} $$
A low residual magnesium content, while sufficient for nodularization, is beneficial in reducing the driving force for MgO formation, thereby mitigating the source of primary slag inclusion defect. Similarly, controlling sulfur content is critical because MgS formation competes with MgO and can act as nucleation sites for larger slag agglomerates. The relationship is complex but underscores the need for low base sulfur levels (target < 0.015%).
In conclusion, the slag inclusion defect in large-section ductile iron crankshafts is a multifaceted problem involving both primary (treatment-related) and secondary (pouring-related) oxidation mechanisms. The primary inclusions consist of oxides and sulfides of magnesium, rare earths, and silicon. Our systematic analysis confirmed that the dominant factor governing the severity of the slag inclusion defect is the temperature of the molten metal throughout the process chain. Secondary but significant factors include the base iron sulfur content and the residual rare earth level. The residual magnesium content, while important for microstructure, has a less pronounced direct effect on slag formation severity. The successful eradication of this defect hinged on a holistic strategy: a radical redesign of the gating system to promote tranquil filling, a relentless focus on achieving and maintaining the highest possible metal temperature, precise control of composition and treatment additions, and the disciplined application of effective fluxing and covering technologies. This comprehensive approach, treating the process as an integrated system, proved essential for achieving stable, high-quality production free from the costly and persistent slag inclusion defect.