In my experience with sand-lined metal mold casting processes for producing automotive brake drums, I have encountered a significant challenge related to nitrogen porosity defects in gray iron castings. This defect typically manifests as irregular, dense, and continuous holes on the outer circumference of the brake drum, especially at the upper corner of the large outer circle. After machining 1–2 mm, these defects become visible, and in severe cases, obvious cavities appear on the rough casting. This issue not only affects the aesthetic quality but also compromises the mechanical properties, such as wear resistance and thermal conductivity, which are critical for brake drum performance. As gray iron casting is widely used in automotive components due to its excellent castability and strength, addressing nitrogen porosity is essential for ensuring product reliability and longevity.
The sand-lined metal mold casting process offers advantages over traditional methods by leveraging rapid cooling to produce dense microstructure and high mechanical properties. However, the fast cooling rate can also trap gases, leading to defects like nitrogen porosity. In this article, I will delve into the analysis of these defects, drawing from my firsthand observations and experiments. I will cover the entire casting process, from melting and molding to pouring, and then focus on the defect analysis using techniques like EDS spectroscopy. Additionally, I will present improvement measures implemented in our production line, supported by tables and formulas to summarize key data. Throughout, I will emphasize the importance of gray iron casting in industrial applications and how optimizing it can enhance product quality.
Let me begin by describing the casting process we employed. The production used medium-frequency induction furnaces for melting, with a charge composition of 60–70% scrap steel and the remainder as returns, including gates, risers, machining chips, and rejects. Carbon raiser was added using graphite-based materials with fixed carbon content above 98%, sulfur content ≤ 0.5%, ash ≤ 2%, volatile matter ≤ 1%, and particle size of 1–5 mm. The charging sequence involved placing 50% of the returns (including chips) at the furnace bottom, adding 70% of the carbon raiser, then loading scrap steel while distributing the remaining 30% carbon raiser in two batches. This approach ensured a high-carbon melt pool formation upon heating, aiding in scrap steel melting by lowering its melting point through surface carburization. The molten iron was heated to 1,500–1,530°C, held at 1,530–1,550°C for 5 minutes for high-temperature settling with repeated slag removal. The tapping temperature ranged from 1,490 to 1,530°C, and 75# ferrosilicon was used for inoculation during tapping. The minimum pouring temperature for the last casting was not below 1,360°C, and the entire ladle pouring time did not exceed 15 minutes.
For molding, we used reclaimed coated sand with new sand additions from Inner Mongolia Tongliao sand, with a grain size of 70–140 mesh. The new sand addition rate was initially 5–20%, with the rest being reclaimed sand. The coated sand underwent thermal regeneration, including processes like crushing, screening, magnetic separation, thermal regeneration, micro-powder removal, cooling, storage, and inspection. The coated sand properties included room-temperature tensile strength ≥ 2.5 MPa, room-temperature bending strength ≥ 6 MPa, hot tensile strength ≥ 1.3 MPa, hot bending strength ≥ 3.8 MPa, loss on ignition ≤ 2.6%, gas evolution ≤ 20 mL, and melting point of 95–110°C. The gating system was designed as a top-pour system with a central sprue, using a funnel-shaped pouring cup with a pressure angle of at least 40° relative to the top plane of the casting. A 15 PPI ceramic filter was placed in the pouring cup to filter the molten iron, and a sprue well with a radius of at least R50 was set below to buffer the flow and reduce turbulence. The gating system was semi-closed, with the ratio of total choke area to total runner area to total ingate area as 1:4.17:2.04. Depending on the product structure, 4 or 5 ingates were used, each with a cross-sectional area not less than 240 mm². The pouring time per mold did not exceed 25 seconds, and stress relief grooves were added on both sides of each ingate to facilitate removal without damaging the casting body. The runner width matched the ingate width, and its thickness was set at 1.25 times the flange wall thickness to ensure smooth metal flow and slag trapping.
To analyze the nitrogen porosity defects, I conducted EDS spectroscopy on the defective areas. The scan revealed irregular defects perpendicular to the casting surface, penetrating 2–5 mm deep. The defect composition primarily included carbon, nitrogen, iron, oxygen, and silicon, with significant segregation of elements. The table below summarizes the elemental content from EDS analysis:
| Element | Mass Fraction (%) |
|---|---|
| Nitrogen (N) | 49.5 |
| Carbon (C) | 28.58 |
| Iron (Fe) | 12.71 |
| Oxygen (O) | 8.87 |
| Silicon (Si) | 0.34 |
From this data, carbon was enriched inside the defects, while nitrogen was distributed uniformly around the defect area. The presence of continuous or discontinuous graphite films within the defects, adjacent to decarburized matrix structures, indicated typical nitrogen-induced porosity. This is a common issue in gray iron casting when nitrogen content exceeds solubility limits during solidification.
The formation mechanism of nitrogen porosity in gray iron casting relates to the solubility of nitrogen in iron. Nitrogen dissolves interstitially in molten iron, and as temperature decreases during solidification, its solubility drops. This can be expressed by the following formula for nitrogen solubility in iron:
$$ S_N = k \cdot e^{-\frac{\Delta H}{RT}} $$
where \( S_N \) is the nitrogen solubility, \( k \) is a constant, \( \Delta H \) is the heat of dissolution, \( R \) is the gas constant, and \( T \) is the temperature in Kelvin. In rapid cooling processes like sand-lined metal mold casting, the surface solidifies quickly, trapping nitrogen gas that evolves from the melt. When the nitrogen content in the molten iron exceeds about 120 ppm, porosity becomes likely. Our measurements showed an actual nitrogen content of 71.7 ppm in the melt, which was below this threshold, suggesting that the defect source was elsewhere.
Upon further investigation, I identified three primary sources of nitrogen in gray iron casting: scrap steel, carbon raiser, and coated sand. Scrap steel, especially common carbon steel and high-manganese steel, can have nitrogen contents ranging from 40–120 ppm. Carbon raisers vary widely; for instance, coal-based carbon raisers contain 0.2–0.7% nitrogen, while graphite-based ones have less than 300 ppm. Coated sand, however, proved to be a significant contributor due to the use of hexamethylenetetramine (urotropine) as a curing agent. In coated sand formulations, resin addition is 2–2.5% of the base sand mass, with urotropine at 15% of the resin mass. Since urotropine contains 40% nitrogen by mass, the coated sand can have a nitrogen content of 0.12–0.15%, far exceeding the 120 ppm limit. During heating and pouring, urotropine decomposes at temperatures above 230°C, releasing ammonia gas that can infiltrate the molten iron, leading to porosity in gray iron casting.
To quantify the nitrogen input from coated sand, I derived a formula based on the sand composition. Let \( m_s \) be the mass of base sand, \( r \) be the resin addition rate (2–2.5%), \( u \) be the urotropine fraction in resin (15%), and \( n \) be the nitrogen fraction in urotropine (40%). The nitrogen content in coated sand, \( N_{sand} \), can be calculated as:
$$ N_{sand} = m_s \cdot r \cdot u \cdot n $$
For typical values, this yields \( N_{sand} \) around 0.12–0.15%, or 1,200–1,500 ppm, which is substantially higher than the melt’s tolerance. This analysis underscored the need for improvements in coated sand management.
Based on these findings, I implemented several measures to mitigate nitrogen porosity defects in gray iron casting. First, regarding coated sand quality, we switched to a new supplier for raw materials, ensuring higher-quality phenolic resin and urotropine. We increased the new sand addition rate from 5–20% to 20–40% in the sand mix, reducing the recycled sand content and its residual nitrogen. Additionally, we separated the management of reclaimed sand from shell molding lines and iron mold lines, as their performance requirements and sand grain sizes differ, preventing cross-contamination. The table below summarizes these improvements:
| Improvement Area | Before | After |
|---|---|---|
| Coated Sand Supplier | Standard supplier | High-quality supplier |
| New Sand Addition Rate | 5–20% | 20–40% |
| Reclaimed Sand Management | Mixed use | Separated by production line |
Second, in production process controls, we made several adjustments. We added more vent plugs to the molds, particularly at bottom and corner areas, to facilitate gas escape during sand coating and pouring. This enhanced venting is crucial for gray iron casting to release gases from coated sand decomposition. We also raised the mold and iron mold temperatures: the mold temperature was set to 230–250°C, and the iron mold temperature to 240–280°C, accelerating urotropine decomposition and gas evolution before pouring. To further aid gas expulsion, we increased the number of vent channels on the iron molds from 8 to 20. Moreover, we extended the time between mold closing and pouring from 10 minutes to at least 20 minutes, allowing more time for gases to dissipate. Finally, we slightly increased the pouring temperature to reduce molten iron viscosity, promoting gas bubble flotation and reducing defect formation. The relationship between pouring temperature and gas solubility can be described by:
$$ \eta = A \cdot e^{\frac{B}{T}} $$
where \( \eta \) is the viscosity, \( A \) and \( B \) are constants, and \( T \) is the temperature. Higher temperatures lower viscosity, making it easier for gases to escape in gray iron casting.
These improvements were systematically applied, and we monitored the outcomes over an extended production run. The results were highly positive: after implementing these changes, we continuously produced over 100,000 brake drum castings without any recurrence of nitrogen porosity defects. This success highlights the effectiveness of targeted interventions in gray iron casting processes. To encapsulate the key parameters, I have compiled a table of optimized process conditions:
| Process Parameter | Optimized Value |
|---|---|
| Mold Temperature | 230–250°C |
| Iron Mold Temperature | 240–280°C |
| Vent Channels per Mold | 20 |
| Time Before Pouring | ≥20 minutes |
| Pouring Temperature | Increased by 10–20°C |
| New Sand Addition Rate | 20–40% |
In conclusion, my analysis of nitrogen porosity defects in gray iron casting revealed that coated sand was the primary nitrogen source, not the molten iron. By focusing on sand quality and process adjustments, we achieved a defect-free production. This experience underscores the importance of holistic process control in gray iron casting, where material inputs and thermal management play critical roles. Future work could explore advanced sand formulations with lower nitrogen content or alternative curing agents to further enhance gray iron casting quality. As gray iron casting continues to be vital for automotive and industrial applications, such improvements contribute to sustainable manufacturing and product performance. I hope this detailed account provides valuable insights for practitioners in the field of gray iron casting.