In the pursuit of advancing foundry technology, my research has focused on the implementation of innovative processes such as pneumatic desulfurization and mold-side nodularization treatments for ductile iron castings. During the course of this experimentation, a specific and recurring challenge emerged: the appearance of unusual white and black patch defects on the cast components, which are not commonly observed in conventional casting processes. These sand casting defects presented a significant obstacle, impacting both product quality and yield, necessitating a thorough investigation into their nature and origin.
The defects manifested predominantly on machined surfaces of ductile iron crankshafts, specifically on the upper-drag sections of the main bearing journals and connecting rod journals. The so-called “white patches” and “black patches” appeared in batches, severely compromising the integrity of the parts. Understanding these sand casting defects became paramount, requiring a multi-faceted analytical approach involving macro-examination, hot acid etching, spectroscopic analysis, and metallography.
Macroscopic and Microstructural Characterization of Defects
Initial macroscopic inspection provided the first clues. The white patches, with dimensions ranging from several millimeters to over a centimeter, were composed of a white, powdery substance. When probed with a needle, this powder would dislodge, leaving behind a pit or cavity. The black patches, in contrast, were identified as areas of concentrated shrinkage porosity and slag inclusions. To delve deeper, samples containing these sand casting defects were subjected to a series of tests.
The hot acid etching test proved particularly revealing. Using a 50% hydrochloric acid solution heated to 70±5°C for an immersion time of 30 minutes, followed by neutralization and drying, the underlying structures were exposed. For white patch samples, the white material was dissolved by the acid, but in its place, a significant amount of subsurface porosity and cavities was revealed. For black patch samples, the extent of the visible porosity and slag areas increased, confirming their inherent weakness.
Spectroscopic analysis (optical emission spectroscopy) was conducted directly on the defect sites. The results were striking:
- White Patches: The spectrum from the white powder showed an exceptionally bright magnesium (Mg) line, indicating a high concentration or enrichment of magnesium at these locations. Analysis away from the defect showed a more uniform, lower magnesium content.
- Black Patches: The spectrum from the black patch areas showed weaker and less intense magnesium lines. The analysis pointed towards a composition rich in rare earth (RE) elements, suggesting the presence of rare earth oxide/sulfide slag complexes.
Based on this, a preliminary inference was made: the white powder was likely magnesium oxide (MgO), while the black patches were slag, primarily composed of rare earth oxides and sulfides.
Metallographic examination under a polished state provided definitive microstructural evidence. Within the pits left by the dissolved white powder, residual particles with an octahedral shape and sharp edges were observed, consistent with the crystalline structure of MgO. These pits and their immediate surroundings were also riddled with dark gray oxide slag and silvery-gray sulfide inclusions, often surrounding graphite nodules. A notable finding was the presence of obvious decarburization on both sides of oxide films near these defects. Furthermore, the areas surrounding both types of sand casting defects consistently exhibited signs of graphite flotation, characterized by exploded, dendritic, thick-flake, and large, irregularly shaped graphite formations. The matrix in these zones was often ferritic due to the decarburization effect of the oxidizing slag.
For the black patches, the edges of the shrinkage porosity zones revealed a network of grayish rare earth oxides and titanium nitride inclusions. The surrounding areas, again, showed severe graphite flotation and the presence of complex magnesium sulfide-based inclusions.
The key characteristics of these two distinct yet related sand casting defects are summarized in the table below.
| Defect Feature | White Patch (MgO Inclusions) | Black Patch (RE Slag & Porosity) |
|---|---|---|
| Macroscopic Appearance | White, powdery surface deposit. | Dark, dense area of porosity/slag. |
| Primary Composition | Magnesium Oxide (MgO) crystals. | Rare Earth (Ce, La) Oxides/Sulfides, TiN. |
| Associated Microstructures | Pits, oxide films, decarburization, graphite flotation. | Shrinkage microporosity network, graphite flotation, complex sulfides. |
| Mechanism of Exposure | Machining removes surface layer, exposing brittle MgO clusters which may fall out. | Machining cuts into subsurface slag clusters and interconnected shrinkage pores. |
Root Cause Analysis: The Role of Mold-Side Treatment Slag
The conclusive finding from the integrated analysis was that both the white and black patch sand casting defects originate from primary slag formed during the mold-side nodularization treatment. This process, while advantageous for certain production layouts, introduces specific kinetic challenges. The incoming iron had undergone efficient pneumatic desulfurization, achieving very low sulfur levels (“double-zero” grade). While beneficial for nodularization efficiency, this also reduces the population of endogenous sulfide particles that can act as nucleation sites for slag agglomeration. Furthermore, the timeline from treatment to solidification in the mold is exceptionally short.
During the vigorous reaction of the nodularizing alloy with the molten iron, primary slag is instantaneously generated. This slag consists of the oxidation products of highly reactive elements like Magnesium and Rare Earths (e.g., MgO, Ce2O3, La2O3), as well as their sulfides (e.g., MgS, CexSy). The thermodynamic driving force for their formation is very high. The reactions can be simplified as:
$$ \text{[Mg]}_{in\,Fe} + \frac{1}{2}\text{O}_2 \ (or\ [O]) \rightarrow \text{MgO}_{(s)} $$
$$ 2\text{[Ce]}_{in\,Fe} + 3[O] \rightarrow \text{Ce}_2\text{O}_{3(s)} $$
$$ \text{[Mg]}_{in\,Fe} + [S] \rightarrow \text{MgS}_{(s)} $$
In a conventional ladle treatment, there is a holding period allowing buoyant forces to float these slag particles to the surface for removal. In the constrained, rapid mold-side process, this critical separation time is virtually eliminated. The slag particles, which are fine and dispersed, become entrapped within the advancing solidification front. They are carried into the mold cavity by the flowing metal and ultimately imprisoned, primarily in the upper surfaces of the casting or in dead zones near cores, where flow velocity diminishes.
The physical state and morphology of the slag are crucial. The white patch defect is directly linked to solid, particulate MgO. Magnesium has a very high affinity for oxygen, and the oxide formed is a solid with a high melting point (~2852°C). It forms as discrete, fine crystals or clusters that are extremely difficult to coagulate. During machining, these brittle clusters at or just below the surface are exposed and can be plucked out, leaving a pit and revealing the white powder.
The black patch defect is associated with more complex, often lower-melting-point slag systems involving rare earth oxides and sulfides. These slags may be partially liquid or form sticky aggregates. Their presence disrupts the orderly feeding of molten metal during the final stages of solidification, leading to the formation of localized shrinkage porosity (micro-shrinkage) intertwined with the slag particles. This creates a continuous weak zone. Graphite flotation in these areas is promoted because the slag particles and the associated oxide films push the eutectic solidification front, creating conditions where graphite nodules grow excessively and irregularly in the last-to-freeze regions rich in carbon and slag.
Theoretical Framework and Influencing Factors
The formation of these sand casting defects can be modeled by considering the interplay of fluid dynamics, particle kinetics, and solidification science. The fundamental issue is the failure to separate a dispersed secondary phase (slag) from the liquid metal before it solidifies.
1. Slag Particle Dynamics: The fate of a slag particle depends on its Stokes settling velocity versus the local solidification rate. The terminal velocity (vt) for a small spherical particle in a melt can be approximated by Stokes’ law:
$$ v_t = \frac{2 (\rho_{metal} – \rho_{slag}) g r^2}{9 \eta} $$
Where:
$\rho_{metal}$ and $\rho_{slag}$ are the densities of the molten iron and slag particle, respectively,
$g$ is gravitational acceleration,
$r$ is the radius of the slag particle,
$\eta$ is the dynamic viscosity of the molten iron.
For typical values ($\rho_{metal} \approx 7000\ kg/m^3$, $\rho_{MgO} \approx 3600\ kg/m^3$, $\eta \approx 0.005\ Pa\cdot s$), the velocity is heavily dependent on particle size. For a fine MgO particle with r = 10 µm, vt is on the order of $10^{-4}$ m/s. In a mold-side process where the time available for flotation before solidification front advancement (tsolid) might be only 30-60 seconds, the maximum possible travel distance is a few millimeters. This is utterly insufficient for removal from the melt.
2. Solidification Front Engulfment: The critical velocity for a particle to be pushed by an advancing solid/liquid interface rather than engulfed is described by models that balance interfacial forces. For non-metallic particles like oxides with poor wettability by iron, there is a critical solidification velocity below which they are pushed. However, in ductile iron, the solidification mode is eutectic, with graphite nodules growing in a quasi-solid manner. The local solidification rate can be high, and the complex morphology of the solidification front readily entraps particles, especially if they are agglomerated. The presence of slag particles themselves can pin the solidification front, leading to cellular or dendritic growth around them and creating the observed micro-porosity.
3. Key Process Factors Aggravating the Defects: Several factors intrinsic to the process exacerbated these sand casting defects:
| Process Factor | Effect on Slag Formation & Behavior | Link to Defect Severity |
|---|---|---|
| Ultra-Low Sulfur Iron | Reduces endogenous MnS/CeS particles that can act as nucleation sites for slag coalescence. | Leads to finer, more dispersed, and more stable slag particles that are harder to float and remove. |
| Short Reaction-to-Solidification Time | Dramatically reduces the time available for Stokesian flotation and slag agglomeration. | Virtually guarantees entrapment of a significant population of primary slag particles. |
| High Residual Magnesium/Rare Earths | Increases the thermodynamic potential for oxidation and slag formation during pouring and mold filling. | Increases the volume fraction of primary slag, raising the probability of defect formation. |
| Turbulent Mold Filling | Can break up any forming slag skins or aggregates, re-dispersing fine particles. | Promotes even distribution of fine slag throughout the casting, rather than concentration in risers. |
Solution Strategy: Slag Modification and Control
The analysis clearly indicated that eliminating these sand casting defects required intercepting the root cause: the entrapment of fine, dispersed primary slag. The solution strategy shifted from trying to prevent slag formation (thermodynamically impossible) to actively modifying its physical properties and creating a mechanism for its separation within the very short process window.
The core concept was to alter the slag’s morphology from a fine, solid powder (like MgO) or a sticky paste into a fluid, cohesive liquid phase that could coalesce easily. Furthermore, a means to actively concentrate and trap this coalesced slag before the metal entered the mold cavity was necessary. This was achieved by introducing specific fluxing agents into the reaction chamber (nodularizing chamber) of the mold-side treatment system.
The additives included:
- Fluorspar (CaF2): A powerful flux that significantly lowers the melting point and viscosity of silicate and aluminate slags. By incorporating CaF2, the various oxide and sulfide products are fused into a single, low-melting-point, fluid slag phase.
- Sodium Fluosilicate (Na2SiF6): Acts similarly, providing fluoride ions to flux the slag and sodium which can form low-melting-point compounds.
- Glass Powder or Glass Fibers: These provide a readily fusible silica network (SiO2) that readily dissolves other oxides, forming a fluid glassy slag. It also provides a bulk material to help coagulate fine particles.
The mechanisms and outcomes of this intervention are summarized below:
| Action | Mechanism | Result |
|---|---|---|
| Slag Fluidity & Coalescence | Fluxes (CaF2, Na2SiF6, Glass) lower slag liquidus temperature and interfacial tension. | Fine MgO/RE-oxide particles are wetted and absorbed into a single, low-viscosity liquid slag droplet. Coalescence rate increases dramatically ($\propto r$, not $r^2$). |
| Slag-Metal Separation | Large, fluid slag droplets have a much higher Stokes velocity. $$ v_t \propto r^2 $$ For a droplet with r=1mm, vt is ~10,000 times faster than for a 10µm particle. | Within the short residence time in the treatment chamber, these large droplets can separate from the metal stream. |
| Active Slag Trapping | Design of the treatment chamber (e.g., a whirl gate or centrifugal separator) uses rotational flow to concentrate the denser metal to the periphery and gather the lighter, coalesced slag at the center or behind a baffle. | The purified metal is extracted from the outer flow path to feed the casting, while the concentrated slag mass is retained in a dedicated “slag pocket” within the chamber itself. |
The effectiveness of this engineered approach was decisive. Subsequent production of crankshafts using the modified mold-side treatment process, incorporating the designed flux additions and separation principles, completely eliminated the occurrence of both white and black patch sand casting defects. The mechanical properties, particularly impact toughness and fatigue strength—which are severely degraded by such stress-concentrating defects—were restored to their expected high levels, resulting in a substantial improvement in product quality and casting yield. This case underscores that a deep, mechanism-based understanding of sand casting defects is essential for developing targeted and effective solutions, especially when implementing advanced foundry processes.