In our foundry operations, during experimental trials of advanced processes such as pneumatic desulfurization and in-mold spheroidization for producing ductile iron crankshafts, we encountered a perplexing issue: the appearance of unusual sand casting defects on machined surfaces. These sand casting defects, manifesting as white patches and black spots, were particularly prevalent in the upper sections of main bearing and connecting rod journal areas, leading to significant quality concerns and reduced yield. This comprehensive analysis delves into the nature, origin, and mitigation strategies for these sand casting defects, employing a multi-faceted experimental approach and theoretical reasoning to elucidate the underlying mechanisms.
Macroscopically, the white patches, with linear dimensions ranging approximately from 2 to 10 mm, presented as areas of white, powdery material that could be easily dislodged with a probing needle, leaving behind distinct pits or cavities. The black spots, conversely, appeared as dense clusters of micro-shrinkage porosity often associated with dark, slag-like inclusions. The recurrent and batch-wise occurrence of these sand casting defects posed a serious threat to the mechanical integrity of the crankshafts, especially impacting fatigue and impact resistance—critical properties for engine components. To systematically address these sand casting defects, we initiated a rigorous investigation program encompassing macro-examination, chemical analysis, and microstructural characterization.
The initial step involved hot acid etching to reveal the sub-surface nature of these sand casting defects. Specimens containing the defects were immersed in a 50% hydrochloric acid aqueous solution maintained at 70±5°C for 30 minutes. After etching, rinsing, alkali neutralization, and drying, the specimens were examined. For white patch areas, the white material was partially or wholly dissolved, confirming its chemical reactivity and leaving behind a network of pores and cavities. This behavior suggested the presence of a basic oxide susceptible to acid attack. The reaction can be conceptually represented as:
$$ \text{MO} + 2\text{H}^+ \rightarrow \text{M}^{2+} + \text{H}_2\text{O} $$
where MO represents a metallic oxide. For black spot areas, the etching process tended to enlarge the apparent area of shrinkage and more clearly delineate the slag inclusions, indicating their relative stability in the acid medium. This differential response provided the first clue that these two types of sand casting defects, while potentially related, had distinct chemical compositions.
| Defect Type | Visual Change Post-Etching | Inferred Chemical Nature |
|---|---|---|
| White Patches (White Spots) | White powder dissolves, revealing underlying porosity and pits. | Contains acid-soluble compounds, likely basic oxides. |
| Black Spots | Shrinkage porosity zones become more pronounced; slag inclusions remain. | Contains relatively acid-stable compounds, likely complex oxides/silicates. |
To precisely identify the elements concentrated within these sand casting defects, we employed spectroscopic analysis (optical emission spectroscopy). The analysis focused on comparing the spectral lines from the defect zones with those from sound areas of the same casting. The results were striking: in white patch regions, the spectral lines for magnesium (Mg) were exceptionally intense and bright, indicating a severe localized enrichment of magnesium. Quantitative estimation suggested magnesium levels could reach up to 1% in these zones, far exceeding the nominal bulk composition. In contrast, areas away from the white patches showed uniform and expected low levels of magnesium. For black spot regions, the Mg lines were notably weaker, but lines for rare earth elements (e.g., Cerium – Ce) were prominent. Other elements like silicon (Si), manganese (Mn), and iron (Fe) showed no significant variation. This led to the preliminary inference that the white powder was predominantly magnesium oxide (MgO), while the black spots consisted of slag rich in rare earth (RE) oxides, possibly along with other complex compounds. The formation of these oxides during the spheroidization treatment can be described by:
$$ 2\text{Mg} + \text{O}_2 \rightarrow 2\text{MgO} $$
$$ 4\text{RE} + 3\text{O}_2 \rightarrow 2\text{RE}_2\text{O}_3 $$
These reactions are exothermic and occur at the interface between the treating alloy and the molten iron, generating what is termed “primary slag.” The entrapment of this slag is a fundamental cause of the observed sand casting defects.
| Analysis Location | Magnesium (Mg) Signal | Rare Earth (e.g., Ce) Signal | Other Elements (Si, Mn, Fe) | Primary Inference |
|---|---|---|---|---|
| White Patch Zone | Very Strong, Intense lines | Normal to Weak | Normal, Uniform | Severe Mg enrichment; defect is MgO. |
| Black Spot Zone | Weak | Strong | Normal, Uniform | RE element enrichment; defect is RE-oxide slag. |
| Sound Area (Remote from Defect) | Normal, Low | Normal, Low | Normal, Uniform | Baseline composition. |
Metallographic examination provided microstructural confirmation and further details about these sand casting defects. Specimens were prepared by standard polishing techniques and observed under optical and scanning electron microscopes. In regions corresponding to white patches (after partial removal of the powder during preparation), the pits were found to contain residual particles with distinct octahedral morphology and sharp edges, characteristic of crystalline MgO. These particles were often surrounded by or mixed with darker gray phases identified as complex oxide slags. The matrix adjacent to these pits frequently exhibited numerous fine, silver-gray inclusions, which energy-dispersive X-ray spectroscopy (EDS) identified as magnesium sulfides (MgS) and other sulfide compounds. A notable feature was the distribution of these sulfides, often encircling graphite nodules or existing as discrete particles in the ferrite matrix. Furthermore, clear evidence of surface oxidation and decarburization was observed in layers adjacent to the defect cavities, indicating exposure to air or oxidizing conditions during or after solidification. The microstructure near these sand casting defects often showed anomalies in graphite morphology, including regions of graphite flotation with oversized, exploded, or dendritic graphite shapes.
For black spot defects, the metallographic view was dominated by interconnected shrinkage porosity networks. At the edges of these porous zones and within the cavities, a网状分布 (network-like distribution) of grayish inclusions was prevalent. EDS analysis confirmed these as complex oxides containing rare earths, along with occasional titanium nitrides (TiN). The surrounding matrix consistently exhibited severe graphite flotation, characterized by a high concentration of large, often deformed graphite nodules, including flower-like, dendritic, flake-like, and very large spherical forms. Within this floated graphite zone, clusters of夹杂物 (inclusions) were identified as composite compounds containing magnesium, rare earths, sulfur, and oxygen—essentially, complex oxy-sulfides. The co-occurrence of shrinkage porosity, slag inclusions, and abnormal graphite strongly suggests a common root cause related to melt cleanliness and solidification dynamics, all contributing to these sand casting defects.
| Defect Type | Key Microstructural Features | Identified Inclusions/Phases | Associated Graphite Anomalies |
|---|---|---|---|
| White Patches | Pits/cavities with residual angular particles; surrounding sulfide inclusions; decarburized layers. | MgO (crystalline), MgS, FeO/MgO complex oxides. | Localized graphite distortion; some flotation. |
| Black Spots | Extensive shrinkage porosity networks; slag networks at pore edges. | Rare Earth Oxides (RE2O3), TiN, Mg-RE-O-S complexes. | Severe graphite flotation zone with oversized, deformed nodules. |
Synthesizing all analytical evidence, we concluded that both white patch and black spot defects are direct consequences of entrapped primary slag formed during the in-mold spheroidization treatment—a classic yet severe manifestation of sand casting defects. Our process involved prior pneumatic desulfurization, which successfully reduced sulfur to very low levels (“double-zero” grade). While this minimized sulfide formation, it had an unintended consequence: the extremely low impurity content resulted in the formation of a very fine, powdery, and dispersed primary slag during the subsequent magnesium/rare-earth treatment. The physics of particle settling in molten iron is governed by Stokes’ law:
$$ v = \frac{2}{9} \frac{(\rho_s – \rho_{Fe}) g r^2}{\eta} $$
Here, $v$ is the terminal settling velocity of a slag particle, $\rho_s$ and $\rho_{Fe}$ are the densities of the slag and iron melt respectively, $g$ is gravitational acceleration, $r$ is the effective particle radius, and $\eta$ is the dynamic viscosity of the molten iron. For the fine, powdery slag particles (with $r$ potentially in the micrometer range), the settling velocity $v$ is exceedingly small. Coupled with the short processing time between spheroidization and the completion of pouring and solidification, these fine slag particles had insufficient time to float up and coalesce into a removable slag layer at the melt surface. Consequently, they remained suspended in the molten iron and were swept into the mold cavity. During solidification, these particles were pushed by advancing dendrites into interdendritic regions, particularly accumulating at the upper surfaces of the casting (due to buoyancy) and in stagnant zones near cores. Upon machining, these subsurface clusters of inclusions were exposed. If the cluster consisted mainly of brittle MgO powder, it would be machined away, leaving a pit that appeared white due to the reflective powder—the “white patch” sand casting defect. If the cluster was a mixture of more coherent slag (RE oxides) and associated shrinkage porosity from the volume deficit of the inclusions, it would appear as a dark, porous area—the “black spot” sand casting defect. The surrounding graphite flotation is a secondary effect, caused by the local enrichment of carbon in the final liquid due to the exclusion of solutes (like Mg and RE) into the slag, altering the eutectic solidification conditions.
The propensity for slag entrapment and the severity of these sand casting defects can be modeled as a function of process parameters. Let $C_{slag}$ be the concentration of slag particles in the melt before pouring, $t_{float}$ the available time for flotation, and $v_{avg}$ the average settling/flotation velocity. The fraction of slag entrapped, $F_{entrap}$, can be approximated by:
$$ F_{entrap} \propto C_{slag} \cdot \exp\left(-\frac{t_{float} \cdot v_{avg}}{H}\right) $$
where $H$ is a characteristic length (e.g., melt depth). In our short-cycle process, $t_{float}$ was minimal, and $v_{avg}$ was low due to fine particle size, leading to a high $F_{entrap}$ and thus prominent sand casting defects.
Beyond slag-related issues, we also investigated broader compositional factors that could exacerbate brittleness and other sand casting defects in ductile iron. Elements like phosphorus (P), silicon (Si), manganese (Mn), and chromium (Cr) promote the formation of hard, brittle phases at grain boundaries. Phosphorus, in particular, forms a phosphide eutectic network that severely embrittles the matrix. The combined embrittling effect can be semi-quantitatively expressed as an impairment factor $I$ on toughness:
$$ I = k_P[P] + k_{Si}[Si] + k_{Mn}[Mn] + k_{Cr}[Cr] + k_{slag} \cdot A_{slag} $$
where $[X]$ denotes weight percent of element X, $k_X$ are empirical coefficients representing their potency, $k_{slag}$ is a factor for slag inclusions, and $A_{slag}$ is a measure of slag content or area fraction. High values of $I$ correlate with low impact energy and increased susceptibility to other sand casting defects like cracking. Electron probe micro-analysis confirmed significant segregation of chromium (up to 1.5%) in phosphide complexes at grain boundaries, amplifying their brittleness. This element segregation, stable against normal annealing, underscores that preventing such sand casting defects requires strict compositional control from the outset.
| Factor | Typical Range Investigated | Effect on Microstructure | Contribution to Sand Casting Defects / Embrittlement |
|---|---|---|---|
| Phosphorus (P) | 0.05% – 0.10% | Forms continuous phosphide eutectic at grain boundaries. | High: Major cause of intergranular brittleness. |
| Silicon (Si) | 2.2% – 3.0% | Solid solution strengthens ferrite; promotes graphite formation. | Moderate: Can increase hardness and reduce ductility if excessive. |
| Manganese (Mn) | 0.3% – 0.8% | Segregates to grain boundaries; stabilizes carbides. | Moderate: Enhances P segregation; promotes carbide formation. |
| Chromium (Cr) | 0.05% – 0.25% | Strong carbide former; segregates with P. | High: Forms hard Cr-containing phosphides/carbides. |
| Residual Magnesium (Mg) | 0.03% – 0.06% | Essential for spheroidization; promotes dross formation. | High (if excessive): Increases slag/MgO formation, causing white patches. |
| Rare Earth (RE) Residual | 0.01% – 0.03% | Aids spheroidization; forms RE oxides/sulfides. | High (if excessive): Increases slag volume, causing black spots. |
| Slag Entrapment | N/A (Process-dependent) | Introduces inclusions, pores, and graphite abnormalities. | Very High: Direct cause of white/black spot defects and stress concentrators. |
To eliminate the specific white and black spot sand casting defects, we devised a targeted solution focusing on modifying the physical state and behavior of the primary slag. The core strategy was to transform the fine, powdery, and dispersed slag into a cohesive, viscous liquid phase that could readily coalesce and be separated from the molten iron. This was achieved by introducing a small but critical addition of specific fluxing agents into the reaction chamber (or well) of the in-mold spheroidization system. The additives included fluorspar (calcium fluoride, CaF₂), sodium fluorosilicate (Na₂SiF₆), and glass powder or glass fibers. Their synergistic functions are multi-fold and can be described through their effects on slag properties:
1. Fluorspar (CaF₂): It acts as a powerful flux, lowering the melting point and viscosity of the silicate-based slag. The calcium ions can also react with oxides, forming more complex but fluid calcium aluminosilicates. The reaction facilitates slag fluidity:
$$ \text{SiO}_2 (s) + \text{CaF}_2 (s) \xrightarrow{\text{heat}} \text{CaSiO}_3 (l) + \text{SiF}_4 (g) $$
The gaseous SiF₄ may also promote stirring. The overall effect is to promote the coalescence of fine particles into larger droplets.
2. Sodium Fluorosilicate (Na₂SiF₆): Upon heating, it decomposes and provides both fluoride and silicate ions, further enhancing slag fluidity and acting as a scavenger for oxides. It helps in forming a slag with a lower surface tension, improving wettability and aggregation.
3. Glass Powder/Fibers: These primarily serve as physical nucleation sites for the fine slag particles. The molten glass provides a sticky surface to which the fine MgO and RE oxide particles can adhere, effectively increasing their apparent size $r$ in Stokes’ law. According to the equation, even a modest increase in $r$ significantly increases the settling velocity $v$ (since $v \propto r^2$), dramatically improving flotation efficiency.
The modified process can be modeled as shifting the slag particle size distribution from a fine mode (mean radius $r_1$) to a coarse mode (mean radius $r_2$), where $r_2 \gg r_1$. The time required for 95% removal, $t_{95}$, scales inversely with the square of the radius:
$$ t_{95} \propto \frac{1}{r^2} $$
Thus, by increasing the effective slag agglomerate size, the required flotation time is drastically reduced, making removal feasible within the existing process window.
Operationally, these additives were placed in the spheroidization chamber. During the treatment, the exothermic reactions and melt turbulence ensured mixing. The resulting slag became a viscous, coherent mass rather than a powder. Furthermore, the design of our treatment system incorporated a centrifugal mixing element. The combined effect of increased slag droplet size and centrifugal force caused the dense, viscous slag to collect centrally in the mixing chamber or against dedicated baffles, effectively acting as a slag trap. The cleansed iron then flowed into the mold cavity. After implementing this optimized practice, subsequent batches of crankshaft castings were thoroughly inspected. The results were conclusive: the incidence of both white patch and black spot sand casting defects dropped to virtually zero. Mechanical testing confirmed the recovery of impact and fatigue properties, validating the effectiveness of the intervention.
| Modification Component | Mechanism of Action | Parameter Affected | Quantitative Benefit (Conceptual) |
|---|---|---|---|
| Addition of Fluxes (CaF₂, Na₂SiF₆) | Lowers slag melting point & viscosity; promotes chemical coalescence. | Slag viscosity (ηslag), Interfacial energy (γ). | Increases coagulation rate constant $k$ in $dn/dt = -k n^2$. |
| Addition of Glass Nucleants | Provides nucleation sites for slag particle adhesion, increasing agglomerate size. | Effective slag particle radius (r). | Increases settling velocity $v$ by factor of $(r_2/r_1)^2$. |
| Centrifugal Separation in Chamber | Provides additional force to drive dense slag to collection zones. | Effective acceleration on slag (a). | Separation force $F = (\rho_s – \rho_{Fe}) \cdot V \cdot a$. |
| Baffles/Slag Traps | Physically intercept and retain coalesced slag. | Slag removal efficiency (ηremoval). | $η_{removal} \rightarrow ~1$ for agglomerated slag. |
In conclusion, our investigation systematically unraveled the nature of the white patch and black spot sand casting defects in ductile iron crankshafts produced via in-mold spheroidization. These defects are intrinsically linked to the entrapment of primary slag—a mixture of magnesium oxides, rare earth oxides, and their sulfides—formed during the treatment stage. The ultra-low sulfur content from prior desulfurization and the short process cycle exacerbated the problem by favoring the formation of a fine, non-coalescing slag that remained suspended in the melt. Through combined macro- and micro-analytical techniques, we positively identified the chemical and microstructural signatures of these sand casting defects. The solution emerged from a fundamental understanding of colloid and fluid dynamics in molten metals: by modifying slag properties to promote aggregation and employing mechanical means to enhance its separation, we successfully eradicated these detrimental sand casting defects. This case study highlights that combating complex sand casting defects often requires a dual approach of rigorous root-cause analysis followed by innovative process engineering grounded in physicochemical principles. Continuous vigilance over melt treatment parameters, slag dynamics, and final composition remains paramount to consistently producing high-integrity castings free from such sand casting defects.