In the field of metallic materials, nodular cast iron, or ductile iron, holds a significant position due to its excellent combination of mechanical properties and castability. Its microstructure, characterized by spherical graphite nodules embedded within a metallic matrix (typically ferrite, pearlite, or a mixture thereof), is the cornerstone of its performance. The pursuit of enhancing the properties of nodular cast iron, particularly in the as-cast state to avoid costly heat treatments, is a continuous endeavor. Traditionally, this involves balancing the matrix phases through alloying with elements like Cu, Ni, or Mo. However, these methods often lead to increased cost or a trade-off between strength and ductility. This has prompted the exploration of novel approaches, such as the introduction of ceramic particles to modify the microstructure intrinsically.
Among various ceramics, titanium carbide (TiC) stands out for its exceptional properties: high melting point (approximately 3067°C), high hardness (28-35 GPa), high Young’s modulus, and excellent chemical stability. Its application as a reinforcement phase in ferrous alloys like steels and white cast irons has shown promise in refining grains and improving wear resistance. However, its systematic study within the context of nodular cast iron is notably scarce. The production of nodular cast iron involves unique processes like spheroidization and inoculation, which introduce numerous heterogeneous nuclei, and its solidification mode is often a “mushy” or divorced eutectic. These characteristics pose distinct challenges and opportunities when introducing exogenous particles like TiC.
My investigation focuses on the effect of adding a small quantity of TiC particles, via an ex-situ addition method, on the microstructure and mechanical properties of as-cast nodular cast iron. The ex-situ method allows for control over particle size, morphology, and addition amount, enabling a clearer analysis of cause and effect. The primary challenge with ex-situ addition is ensuring uniform distribution and good interfacial bonding within the iron melt. To address this, I employed a high-energy ball milling process to modify the TiC powder prior to its introduction into the melt.
The experimental work began with the pre-treatment of the TiC powder. Commercial TiC powder (purity ≥99.9%, diameter 3-5 µm) was mixed with iron powder (purity ≥99.5%, average diameter 30 µm) in a specific ratio. This mixture was subjected to high-energy ball milling in a planetary mill for 24 hours at a speed of 100 rpm, using a ball-to-powder weight ratio of 10:1. This process is crucial for several reasons. First, it fragments the larger iron particles. More importantly, it induces repeated cold welding and fracturing, effectively coating or embedding the finer TiC particles within the ductile iron particles. This composite powder has a higher effective density than pure TiC, reducing buoyancy in the molten iron, and the iron shell improves wettability and compatibility with the nodular cast iron matrix.
The melting of the nodular cast iron was conducted in a medium-frequency induction furnace, using high-purity pig iron as the base material, with ferromanganese and ferrosilicon used for composition adjustment. The target superheat temperature was 1550°C. The treatment was performed using a sandwich method in a pouring ladle. For the TiC-containing sample, 0.1 wt.% of the ball-milled TiC-Fe composite powder was placed at the bottom of the ladle. This was covered by 1.2 wt.% rare-earth magnesium ferrosilicon alloy as the spheroidizing agent, which was then topped with 0.5 wt.% 75-ferrosilicon as the inoculant. The base iron was poured onto this stack, initiating a vigorous spheroidizing reaction that also helped disperse the introduced particles. The melt was then cast into resin-bonded sand molds to produce Y-block test castings. A reference sample without TiC addition was prepared under identical conditions for comparison.
The chemical composition of the final castings, determined by optical emission spectrometry, is presented in Table 1. The key difference is the titanium content. The reference sample (Sample A) contains a background level of 0.006% Ti, while the sample with added TiC powder (Sample B) shows a Ti content of 0.022%. Based on stoichiometry, this corresponds to an effective TiC addition of approximately 0.02 wt.%, indicating a yield of about 20% considering the initial 0.1 wt.% addition. Losses are attributed to slag entrapment and turbulence during the spheroidizing reaction.
| Sample | C | Si | Mn | P | S | Ti | Mg | Ce |
|---|---|---|---|---|---|---|---|---|
| A (Base) | 3.03 | 2.40 | 0.52 | 0.006 | 0.008 | 0.006 | 0.041 | 0.029 |
| B (With TiC) | 3.03 | 2.37 | 0.51 | 0.006 | 0.008 | 0.022 | 0.042 | 0.027 |
Microstructural analysis was performed on polished and etched samples using optical microscopy (OM) and scanning electron microscopy (SEM) coupled with energy dispersive spectroscopy (EDS). The most striking initial observation was the significant improvement in the graphite morphology in Sample B. Compared to the base nodular cast iron, the TiC-containing sample exhibited a higher number of graphite nodules, which were also more spherical and slightly smaller in average diameter. Quantitative image analysis was conducted according to relevant standards, and the statistical data is summarized in Table 2.
| Sample | Nodularity (%) | Graphite Count (N/mm²) | Graphite Area Fraction (%) | Ferrite Content (Vol.%) |
|---|---|---|---|---|
| A (Base) | 70.2 | 42.5 | 9.30 | 61.5 |
| B (With TiC) | 79.3 | 59.7 | 11.39 | 66.0 |
The data reveals compelling changes: the graphite nodule count increased by 40.5%, the nodularity improved by 9.1 percentage points (nearly one full quality grade), the graphite area fraction increased, and the ferrite content in the matrix rose by 4.5%. This multifaceted improvement can be explained by the role of TiC particles as highly effective heterogeneous nucleation sites for graphite. The potency of a substrate as a nucleant for a solidifying phase is often assessed using Bramfitt’s two-dimensional lattice misfit theory. The disregistry $\delta$ between the nucleating solid and the substrate is calculated as:
$$ \delta_{(hkl)_n}^{(hkl)_s} = \frac{1}{3} \sum_{i=1}^{3} \frac{|d_{[uvw]_s}^i \cos \theta – d_{[uvw]_n}^i|}{d_{[uvw]_n}^i} \times 100\% $$
where $(hkl)_s$ and $(hkl)_n$ are low-index planes of the substrate and nucleus, $[uvw]_s$ and $[uvw]_n$ are low-index directions in those planes, $d$ is the atomic spacing along a direction, and $\theta$ is the angle between the two directions. A disregistry value $\delta < 12\%$ is generally considered favorable for heterogeneous nucleation. For the TiC (111) plane and the graphite (0001) basal plane, the calculated disregistry is approximately 8.33%, confirming that TiC is an excellent candidate for nucleating graphite. Therefore, the introduced TiC particles provided additional nucleation sites, increasing the graphite count. A higher number of nodules reduces the diffusion distance for carbon atoms during the eutectic and eutectoid reactions, which can accelerate the growth of ferrite around the nodules, leading to a higher final ferrite content. Furthermore, the increased nucleation rate releases more latent heat, potentially reducing the undercooling ($\Delta T$) at the solidification front. Since graphite grows via a lateral growth mechanism (e.g., screw dislocation), a lower $\Delta T$ can lead to a slower growth rate, which favors the maintenance of a spherical shape and reduces the probability of graphite degeneracy, thereby improving nodularity.
The influence of TiC particles extended beyond graphite and into the metallic matrix itself. SEM-EDS analysis confirmed the presence of Ti-rich particles, approximately 3-5 µm in size, distributed within the matrix. Their interaction with the matrix during solid-state transformation was profound. I observed several key phenomena:
1. Formation of “Pearlite Colony” Boundaries: In some regions, TiC particles were found aligned along what appeared to be the boundaries of a pearlite colony. These particles seemed to have been pushed and accumulated at the boundaries of former austenite dendrites during solidification. During the subsequent eutectoid transformation, these particles acted as barriers, leading to the formation of a region with a distinct pearlitic structure, sometimes with a core of granular cementite. This suggests that TiC particles can influence the local transformation kinetics and morphology.
2. Modification of Cementite Morphology: TiC particles located at pearlite colony boundaries or within ferrite grains were often associated with altered cementite morphology. Instead of the typical continuous lamellae, the surrounding cementite appeared as short rods or even discrete particles. This can be attributed to the fact that TiC itself can act as a heterogeneous nucleant for cementite. The disregistry between TiC (001) and cementite (Fe$_3$C) (001) is about 9.54%, which is within the favorable range. When cementite nucleates on a TiC particle, its growth pattern can be disrupted, leading to a divorced or granular form rather than well-formed lamellae.
3. Refinement of Ferrite Grains: TiC particles located at ferrite grain boundaries were observed. These particles, present before the $\gamma \rightarrow \alpha$ transformation, pin the migrating austenite/ferrite interface during transformation, hindering grain boundary motion. This Zener pinning effect leads to a refinement of the final ferrite grain size. Measurement using the linear intercept method according to ASTM standards confirmed this. The ferrite grain size number, G, increased from 8.17 in Sample A to 8.62 in Sample B. The relationship between grain size number and the number of grains per square millimeter, $n_A$, is given by:
$$ n_A = 2^{G-1} $$
This corresponds to an increase of several hundred grains per square millimeter in the TiC-modified nodular cast iron.
The culmination of these microstructural modifications—improved graphite characteristics, increased ferrite content, refined ferrite grains, and altered pearlite morphology—was expected to significantly impact the mechanical properties. I conducted tensile tests at room temperature and Charpy V-notch impact tests at both room temperature and -20°C. The results are summarized in Table 3.
| Sample | Yield Strength (MPa) | Tensile Strength (MPa) | Total Elongation (%) | Impact Toughness, RT (J/cm²) | Impact Toughness, -20°C (J/cm²) |
|---|---|---|---|---|---|
| A (Base) | ~285 | ~450 | 14.1 | 8.36 | 7.65 |
| B (With TiC) | ~280 | ~445 | 16.8 | 10.96 | 9.32 |
The most notable improvements are in ductility and toughness. The total elongation increased by 19.1%, room temperature impact toughness by 31.1%, and sub-zero impact toughness by 21.8%, while the yield and tensile strength remained virtually unchanged. This is a highly desirable outcome, breaking the traditional strength-ductility trade-off for this class of material.
Fractographic analysis via SEM provided insights into the mechanisms behind this enhanced toughness. The fracture surface of the base nodular cast iron (Sample A) showed a mixed mode of failure. There were regions of dimpled rupture around graphite nodules, indicative of microvoid coalescence, but also significant areas of cleavage fracture characterized by flat facets and river patterns. This cleavage was more pronounced in the impact fracture surfaces, especially at -20°C.
In contrast, the fracture surface of the TiC-modified nodular cast iron (Sample B) exhibited a more ductile morphology. The dimples were more numerous, finer, and more uniformly distributed. The area fraction of cleavage facets was substantially reduced. The presence of fine TiC particles contributed to this in two key ways. First, they themselves can initiate very fine microvoids during plastic deformation. The growth and coalescence of these fine voids require more energy and distribute strain more evenly, postponing the onset of catastrophic crack propagation. Second, the refinement of the ferrite grain size, as described by the Hall-Petch relationship for cleavage strength, increases the resistance to crack initiation and propagation. The Hall-Petch relationship is often expressed for yield strength, but a similar principle applies to fracture stress:
$$ \sigma_f = \sigma_0 + k_f d^{-1/2} $$
where $\sigma_f$ is the fracture stress, $\sigma_0$ and $k_f$ are material constants, and $d$ is the grain diameter. A smaller grain size $d$ leads to a higher $\sigma_f$. Furthermore, the change in cementite morphology from continuous lamellae to short rods or particles reduces stress concentration sites and makes crack propagation through the pearlitic regions more difficult. The improved graphite nodularity and increased nodule count also play a role by ensuring a more uniform stress distribution and providing more sites for blunting of advancing cracks.
The reason the tensile strength did not increase concurrently can be understood as a balance of opposing factors. On one hand, graphite improvement and ferrite grain refinement are strengthening mechanisms. On the other hand, the reduction in the volume fraction of the stronger pearlite phase and the presence of stress-concentrating irregular particles (from powder agglomerates) can lower the strength. In this case, these factors balanced out, leaving the strength nearly unchanged while allowing the ductility-enhancing mechanisms to dominate.
In conclusion, my investigation demonstrates that the exogenous addition of a very small quantity (effective 0.02 wt.%) of TiC particles, after suitable pre-treatment via high-energy ball milling, is a highly effective method for enhancing the microstructure and mechanical properties of as-cast nodular cast iron. The TiC particles act as powerful heterogeneous nucleation sites for graphite, leading to a significant increase in nodule count, improved nodularity, and a refined graphite structure. This, in turn, promotes the formation of more ferrite. Additionally, the particles interact with the matrix, refining ferrite grains, modifying pearlite/carbide morphology, and creating unique microstructural features. The synergistic effect of these modifications results in a remarkable increase in ductility (19.1% higher elongation) and toughness (over 30% higher impact energy) without compromising the tensile strength. This work opens a new pathway for producing high-toughness, as-cast nodular cast iron through microstructural engineering with ceramic particles, potentially reducing reliance on expensive alloying elements or post-casting heat treatments.