In the production of ductile iron castings, particularly for critical components like crankshafts, the presence of slag inclusions remains a significant and persistent challenge. These slag inclusions, often referred to as black slag defects, are unique to ductile iron and frequently lead to the rejection of cast parts due to their detrimental effects on mechanical properties, fatigue strength, and wear resistance. This article presents a comprehensive study conducted from my perspective as a foundry engineer and researcher, focusing on the metallographic and electron probe microanalysis (EPMA) of slag inclusions in diesel engine crankshafts. The aim is to elucidate the nature, origin, and formation mechanisms of these defects and to develop effective countermeasures. The investigation leverages advanced analytical techniques, including optical microscopy and electron probe microanalysis, to provide a detailed characterization of slag inclusions. Throughout this work, the term ‘slag inclusions’ will be repeatedly emphasized to underscore its central role in ductile iron quality control.
The formation of slag inclusions is intrinsically linked to the complex physicochemical reactions occurring during the spheroidization and inoculation treatments of molten iron. In conventional production, slag inclusions predominantly manifest on the upper surfaces of castings or the lower surfaces of cores, appearing as non-lustrous, dark patches in fracture sections. Their presence not only compromises the structural integrity of components like crankshafts but also leads to substantial economic losses due to high scrap rates. My investigation began with a critical review of our existing production process for ductile iron crankshafts, which historically employed a double-tapping treatment method using rare-earth magnesium (RE-Mg) alloys via the sandwich (or冲入法) process. Despite the use of fluxing agents like cryolite and covering with charcoal ash, slag inclusions were not eliminated, prompting a systematic re-evaluation of the treatment methodology.
The core of the process modification involved shifting from a double-tapping to a single-tapping treatment method. The revised procedure is as follows: Prior to tapping, approximately 0.3-0.5% (by weight of the molten iron) of limestone powder is added to the ladle as a replacement for cryolite powder. The molten iron is then tapped in a single pour onto the RE-Mg alloy in the ladle. After the completion of the vigorous alloy reaction, the melt is stirred thoroughly to promote agglomeration and flotation of reaction products. The slag is subsequently skimmed off. Finally, inoculating ferrosilicon is added to the surface of the molten iron, followed by another stirring and skimming operation before covering with charcoal ash and pouring. This adjustment significantly increased the treatment temperature and enhanced the turbulent mixing, thereby improving the buoyancy and removal efficiency of the nascent slag. The result was a near-complete elimination of macroscopic slag inclusion defects on the crankshaft castings, as visually confirmed. This practical improvement served as the foundation for the subsequent detailed microanalysis to understand the fundamental reasons behind its success.
To delve into the microstructural characteristics of the slag inclusions, samples were extracted from crankshafts produced by both the old (double-tap) and new (single-tap) processes. Metallographic examination under an optical microscope revealed two primary morphologies of slag inclusions: blocky/stringer-type and discrete dot-type. The blocky inclusions were typically located just beneath the casting’s upper surface (cope side) and are generally classified as primary slag, formed during the spheroidization and inoculation reactions. In contrast, the dot-type inclusions were often found distributed beneath the blocky ones, representing secondary slag formed during the period between final slag removal and the complete solidification of the casting.
The chemical composition of these slag inclusions was rigorously investigated using a JCXA-733 electron probe microanalyzer. Both wavelength-dispersive spectroscopy (WDS) for point analysis and element distribution mapping (area scanning) were performed. The primary elements of interest were oxygen (O), sulfur (S), silicon (Si), and magnesium (Mg). Point analysis data from various regions—normal matrix, dot-type slag inclusion zones, and blocky slag inclusion zones—are summarized in the table below. The intensity counts for O and S X-rays provide a semi-quantitative comparison of their relative concentrations.
| Sample Description | O Kα Intensity (counts) | S Kα Intensity (counts) |
|---|---|---|
| Crankshaft (Single-tap process) – Normal Matrix | 1250 | 580 |
| Crankshaft (Double-tap process) – Normal Matrix | 1400 | 650 |
| Double-tap process – Dot-type Slag Inclusion Zone | 4500 | 5200 |
| Double-tap process – Blocky Slag Inclusion Zone | 6800 | 3100 |
The data reveals a striking enrichment of oxygen and sulfur within the slag inclusion regions. Specifically, in dot-type slag inclusions, the oxygen intensity is approximately 3.2 times higher, and sulfur is about 8.0 times higher than in the normal matrix of the double-tap processed sample. For blocky slag inclusions, oxygen is about 4.9 times higher, and sulfur is about 4.8 times higher. This unequivocally demonstrates that the slag inclusions are composed of oxides and sulfides. The single-tap process sample shows lower baseline levels of oxygen and sulfur in its normal matrix, indicating a more effective purification of the melt prior to pouring, which correlates with the observed reduction in slag inclusions.
Element distribution maps (X-ray area scans) provide a visual representation of the localization of these elements. In the blocky slag inclusions, the maps for magnesium and oxygen show strong co-localization, with their distribution images almost perfectly overlapping. This indicates that the blocky slag inclusions are primarily magnesium-rich oxides, likely compounds such as MgO or complex magnesium silicates. The formation of these oxides can be described by reactions such as:
$$ 2Mg_{(in\ Fe)} + O_2 \rightarrow 2MgO_{(s)} $$
$$ 2Mg_{(in\ Fe)} + SiO_2_{(slag)} \rightarrow 2MgO_{(s)} + Si_{(in\ Fe)} $$
For the dot-type slag inclusions, the distribution maps for sulfur, oxygen, and silicon all show enrichment, with sulfur being the most pronounced. This suggests that the dot-type slag inclusions are predominantly rare-earth oxy-sulfides or sulfides. The presence of rare-earth elements (from the spheroidizing alloy) facilitates the formation of stable sulfides and oxy-sulfides. The chemical affinity can be represented by the free energy of formation for such compounds. For instance, the stability of cerium oxysulfide (Ce2O2S) is high, and its formation can be simplified as:
$$ 2Ce_{(in\ Fe)} + 2O + S \rightarrow Ce_2O_2S_{(s)} $$
The overall propensity for slag inclusion formation is governed by thermodynamic and kinetic factors. The activity of surface-active elements like oxygen and sulfur in the molten iron is critical. The equilibrium between dissolved elements and their compounds can be expressed using law of mass action. For example, the solubility product for magnesium oxide formation in iron melt is a function of temperature and composition:
$$ K_{MgO} = a_{Mg}^2 \cdot a_O $$
where \(a_i\) represents the activity of element i. When the ionic product exceeds the solubility product, precipitation of MgO occurs, leading to the formation of primary slag inclusions. Similarly, for sulfide formation involving rare-earth elements (RE):
$$ K_{RE_xS_y} = a_{RE}^x \cdot a_S^y $$
The process modification to a single-tap method with limestone flux positively impacts these equilibria. Limestone (CaCO3) decomposes at high temperature to form CaO and CO2. The CaO acts as a basic flux, reacting with acidic oxides like SiO2 and Al2O3 present in the slag, forming lower-melting-point calcium silicates and aluminates that are more fluid and easier to remove. This can be represented as:
$$ CaO_{(s)} + SiO_2_{(s)} \rightarrow CaSiO_{3(l)} $$
The increased fluidity and basicity of the slag enhance the absorption and removal of both oxide and sulfide particles from the melt. Furthermore, the higher treatment temperature and vigorous stirring in the single-tap process increase the rate of slag particle coalescence and flotation according to Stokes’ law:
$$ v = \frac{2}{9} \frac{(\rho_{metal} – \rho_{slag}) g r^2}{\eta} $$
where \(v\) is the flotation velocity, \(\rho\) are densities, \(g\) is gravity, \(r\) is the particle radius, and \(\eta\) is the melt viscosity. By promoting the growth of slag particles (increasing \(r\)) and potentially reducing viscosity \(\eta\) through temperature increase, the removal rate is significantly accelerated, reducing the population of slag inclusions entrapped during solidification.
A more detailed comparative analysis of the two processes is presented in the following table, summarizing key parameters and outcomes related to slag inclusion formation:
| Process Parameter / Outcome | Double-Tap Treatment | Single-Tap Treatment with Limestone |
|---|---|---|
| Treatment Temperature | Lower (due to heat loss between taps) | Higher (single continuous tap) |
| Melt Turbulence & Mixing | Moderate | High (enhanced by single vigorous pour) |
| Primary Slag (Oxide) Formation | High; poor removal efficiency | Controlled; efficient flotation and removal |
| Secondary Slag (Sulfide/Oxy-sulfide) Formation | High; due to residual O and S | Low; effective deoxidation and desulfurization |
| Final Slag Viscosity & Fluidity | Higher (with cryolite) | Lower (with basic CaO-based slag) |
| Observed Slag Inclusion Defect Rate | High | Negligible |
| Dominant Slag Inclusion Type in Casting | Both blocky (Mg-O) and dot (RE-S-O) | Rare, minor dot-type if any |
The prevention strategies derived from this analysis are multi-faceted. First and foremost, minimizing the sources of oxygen and sulfur is paramount. This involves using high-purity charge materials, pre-melt desulfurization where possible, and ensuring dry, preheated ladles and molds to avoid moisture-derived oxygen. Second, the treatment process must be optimized to maximize slag separation. The single-tap method with a basic flux like limestone proves highly effective. The timing and technique of slag skimming are also critical; it must be performed after sufficient holding time for slag agglomeration but before temperature drop increases viscosity. Third, controlling the solidification dynamics can help. A directional solidification pattern that pushes remaining liquid slag inclusions towards risers or non-critical areas can mitigate their harmful effects in the final casting. The use of chills or controlled cooling can assist in this regard.
To further quantify the relationship between process variables and slag inclusion severity, a theoretical model can be considered. The number density of slag inclusions entrapped, \(N_{slag}\), can be expressed as a function of initial impurity concentration, removal efficiency, and solidification front velocity:
$$ N_{slag} \propto \frac{[O]_{initial} \cdot [S]_{initial} \cdot (1 – \epsilon_{removal})}{v_{solidification}} $$
where \([O]_{initial}\) and \([S]_{initial}\) are the initial concentrations of oxygen and sulfur, \(\epsilon_{removal}\) is the fractional efficiency of slag removal prior to pouring (which is enhanced by the single-tap process), and \(v_{solidification}\) is the average solidification front velocity. Lower initial concentrations, higher removal efficiency, and faster solidification (which can trap fewer particles but may also lead to finer distribution if entrapped) all contribute to reducing \(N_{slag}\).
In conclusion, the comprehensive investigation combining practical process improvement with advanced microanalytical techniques has yielded profound insights into the problem of slag inclusions in ductile iron crankshafts. The slag inclusions are definitively identified as heterogeneous aggregates rich in oxygen and sulfur, manifesting as two distinct types: primary blocky inclusions dominated by magnesium oxides and secondary dot-type inclusions dominated by rare-earth oxy-sulfides. The successful mitigation was achieved by adopting a single-tap treatment process utilizing limestone as a basic flux. This approach elevated the treatment temperature, enhanced melt purification through improved slag flotation and removal dynamics, and ultimately reduced the activity of oxygen and sulfur in the molten metal. The persistent challenge of slag inclusions can thus be systematically addressed through a holistic strategy encompassing charge material control, optimized treatment metallurgy, and careful foundry practice. This work underscores the importance of fundamental understanding in solving practical foundry defects and provides a replicable framework for quality enhancement in ductile iron casting production.