In engineering applications, the failure of components due to fatigue under cyclic contact stress is a primary cause of equipment downtime and diminished processing quality. Rolling contact fatigue (RCF) is a critical factor in such failures. For mechanical parts like rolling bearings, gears, and wheel-rail contacts, various forms of surface damage lead to performance degradation. Typical damage from RCF under dry or boundary-lubricated conditions includes pitting, spalling, cracking, and “crushed” damage, often evolving into cracks propagating through the rolling element. As noted in prior studies, cyclic contact stresses induce localized plastic strain, and the accumulation of this plastic strain leads to issues like stress concentration or fatigue cracking, making it a principal cause of crack initiation in RCF.
Gray iron casting has garnered significant attention due to its low cost and excellent properties, including low melting point, good fluidity, castability, excellent machinability, and respectable wear resistance. However, its fatigue wear resistance (FWR) is often considered poor in contact fatigue environments due to the presence of low-hardness, cavity-like graphite flakes, which act as inherent defects. Researchers have employed various surface modification techniques to enhance the FWR of gray iron casting, such as physical vapor deposition (PVD), chemical vapor deposition (CVD), thermal spraying (TS), and plasma transferred arc (PTA) welding, with some success. Nevertheless, these methods have limitations: coating thickness is often limited to less than 100 µm, restricting use in high-stress scenarios; material mismatch at the coating-substrate interface risks crack formation; and compositional differences lead to chemical discontinuity. Consequently, failures often initiate within or at the interface of these coatings, sometimes leading to more severe damage than in untreated materials.
Inspired by bionic principles observed in nature, where alternating hard and soft phases on a surface can improve performance, recent work has focused on laser surface modification. Laser remelting (LR) creates localized hardened zones, or “units,” on the material surface without adding external material. These units act as hard phases to resist deformation, while the softer untreated matrix acts as a buffer to release strain, reducing stress accumulation and improving properties like wear and fatigue resistance. However, for gray iron casting, the reinforcement mechanism in LR is primarily the dissolution and re-precipitation of carbon, leading to phase transformations and carbide formation. The inherent inhomogeneous distribution of graphite in the matrix results in non-uniform microstructures after LR—areas near graphite form high-carbon martensite and carbides, while areas away from graphite form low-carbon martensite, leaving behind untransformed ferrite. This microstructural inconsistency limits the degree of performance enhancement.
To address this and achieve a more homogeneous and superior hardened zone, our study proposes and investigates a novel “laser carburizing” (LC) process. This technique involves pre-placing a layer of amorphous carbon powder on the surface of the gray iron casting prior to laser irradiation. The process aims to introduce additional carbon into the melt pool, forcing a more uniform distribution and enabling the formation of a higher volume fraction of hardening phases. The core objective of this experimental work is to systematically evaluate the improvement in fatigue wear resistance of gray iron casting achieved through laser carburizing, compare its efficacy against conventional laser remelting, and elucidate the underlying mechanisms driving this enhancement. We specifically investigate the influence of input laser energy density on the microstructure, mechanical properties, and resultant FWR.
1. Materials and Experimental Methodology
1.1 Base Material and Sample Preparation
The substrate material was a commercial-grade gray iron casting (equivalent to HT150 according to GB 5612-85). Its nominal chemical composition is provided in Table 1. The as-received microstructure, shown in the referenced image, consists of type-A flake graphite (G) embedded in a matrix of predominantly pearlite (P) with approximately 5% ferrite (F). The average graphite flake width was about 4 µm. The base material had a Vickers microhardness of approximately 260 HV0.2.
Rectangular blocks (120 mm × 15 mm × 5 mm) were sectioned from cast plates. All surfaces were ground and polished sequentially with finer grit SiC papers to a mirror-like finish to minimize the influence of machining marks prior to laser treatment and testing. For the laser carburizing process, high-purity (98%) amorphous carbon powder was used. A sodium silicate binder (water glass) was mixed with the carbon powder to form a slurry, which was pre-placed uniformly on the sample surface to a controlled thickness before laser processing to prevent powder spatter.
| C | Si | Mn | P | S | Fe |
|---|---|---|---|---|---|
| 3.0–3.2 | 1.5–1.8 | 0.7–0.9 | < 0.15 | < 0.12 | Bal. |
1.2 Laser Surface Treatment
A pulsed Nd:YAG laser system (wavelength 1.064 µm, max power 300 W) was used for surface treatment under an argon shielding atmosphere. The laser head was fixed vertically, and the focal length was adjusted to vary the laser spot size and, consequently, the energy density on the surface. The primary processing parameters were kept constant: pulse duration of 5.0 ms, pulse frequency of 9 Hz, and single-pulse energy of 5.37 J. The laser-treated track, referred to as a “unit,” was created by translating the sample at a speed of 0.3 mm/s. Multiple parallel unit tracks were fabricated with a center-to-center spacing of 5 mm along the sample’s length.
By changing the focal length, different laser spot diameters and thus different energy densities were achieved. The key parameter defining the laser input is the energy density (Ed), calculated based on the spot area. A series of samples were prepared, as detailed in Table 2. This includes laser remelted (LM) samples (without carbon pre-placement) and laser carburized (LC) samples (with 0.2 mm carbon paste pre-placement) at three distinct energy densities. Untreated parent material (PM) samples were also prepared for baseline comparison.
| Sample ID | Spot Dia. (mm) | Carbon Layer (mm) | Energy Density (J/mm²) | Process |
|---|---|---|---|---|
| PM | – | – | – | Parent Material |
| LM-H | 0.8 | 0 | 10.70 | Laser Remelting |
| LC-H | 0.8 | 0.2 | 10.70 | Laser Carburizing |
| LM-M | 1.0 | 0 | 6.84 | Laser Remelting |
| LC-M | 1.0 | 0.2 | 6.84 | Laser Carburizing |
| LM-L | 1.2 | 0 | 4.75 | Laser Remelting |
| LC-L | 1.2 | 0.2 | 4.75 | Laser Carburizing |
1.3 Microstructural and Mechanical Characterization
Cross-sections perpendicular to the laser tracks were prepared using standard metallographic techniques. The geometry (width, depth, area) of the laser-modified units was measured using optical microscopy (OM). Microstructural analysis was performed using scanning electron microscopy (SEM). Phase identification within the units was carried out using X-ray diffraction (XRD) with Cu Kα radiation. The microhardness profiles across the units and the matrix were measured using a Vickers microhardness tester with a 0.2 kg load and a 10 s dwell time. Tensile and compressive properties of the treated surface layers were inferred from tests on small specimens extracted to include the modified zones, using a material testing system (MTS).
1.4 Fatigue Wear Testing and Analysis
Rolling contact fatigue (RCF) tests were conducted on a purpose-built test rig. The contact was lubricated with SAE 46 mechanical oil to establish an elastohydrodynamic lubrication film. A loading mechanism applied a cyclic load via ten cylindrical rollers (made of GCr15 bearing steel, 63 HRC) rolling back and forth over the sample surface. The normal load per roller was 80 N, and the equivalent rolling speed was 7 cm/s. Samples were cleaned and weighed before testing. The tests were run for a total duration of 70 hours, with the samples being weighed every 10 hours to determine mass loss. The mass loss (ML) and average mass loss rate (MLR) during the steady-state wear period were used to evaluate and compare the fatigue wear resistance. Worn surface morphologies were examined using OM, SEM, and 3D surface profilometry to identify failure modes.
To understand the contact stress distribution, a simplified finite element analysis (FEA) model was constructed. A 2D plane-strain model simulated a single roller in pure rolling contact with the treated surface under an 8 N load. The roller was modeled as a rigid body. The material properties assigned to the matrix and the laser-treated unit (approximated as a homogeneous region with higher modulus and hardness) are listed in Table 3. The maximum contact (Hertzian) and subsurface shear stresses were analyzed.
| Region | Elastic Modulus (GPa) | Poisson’s Ratio | Density (g/cm³) |
|---|---|---|---|
| Matrix (PM) | 130 | 0.23 | 7.4 |
| Laser-treated Unit | 150 | 0.25 | 7.8 |
2. Characteristics of Laser-Treated Units
2.1 Geometry and Formation
The rapid melting and subsequent ultra-fast solidification during laser processing resulted in units with a distinct, near-parabolic cross-sectional profile. The geometrical dimensions of the units are summarized in Table 4. A clear trend was observed: with increasing laser spot size (and thus decreasing energy density), the unit width increased slightly, but the depth decreased significantly, leading to a substantial reduction in the unit’s cross-sectional area. This is attributed to the lower energy density distributing heat over a wider but shallower region, resulting in less melting penetration.
Comparing LC and LM units processed at identical energy densities reveals that LC units were consistently deeper and had a larger cross-sectional area than their LM counterparts, despite having a slightly smaller width. This is fundamentally linked to the enhanced laser energy absorption. The effective heat input ($Q_{eff}$) considering absorption can be expressed as:
$$ Q_{eff} = \eta \frac{P}{V} $$
where $\eta$ is the absorption coefficient, $P$ is laser power, and $V$ is scanning speed. The pristine surface of gray iron casting has a relatively low absorption (η ~20%) for the Nd:YAG laser wavelength. The pre-placed carbon layer, being dark and amorphous, dramatically increases this absorption coefficient (η ~90%). Consequently, for the same nominal laser parameters, the LC process delivers significantly more effective energy into the material, leading to a larger melt pool and deeper penetration. The slightly narrower width of LC units is attributed to the lower thermal conductivity of the carbon layer, which somewhat constrains lateral heat diffusion.
| Sample ID | Width (mm) | Depth (mm) | Cross-sectional Area (mm²) |
|---|---|---|---|
| LM-H | 1.26 | 0.84 | 0.71 |
| LC-H | 1.03 | 1.10 | 0.76 |
| LM-M | 1.43 | 0.24 | 0.23 |
| LC-M | 1.25 | 0.50 | 0.42 |
| LM-L | 1.52 | 0.10 | 0.10 |
| LC-L | 1.37 | 0.15 | 0.14 |
2.2 Phase Composition and Microstructure
XRD analysis confirmed the phase transformations induced by laser processing. The parent gray iron casting showed peaks for ferrite (α-Fe) and graphite. After both LM and LC treatment, the graphite peaks disappeared, and new peaks corresponding to martensite (α’-Fe), retained austenite (γ-Fe), and cementite (Fe3C) appeared. While the phase constituents were similar for LM and LC units, the diffraction peaks for the LC units were consistently broader. The full width at half maximum (FWHM) for the primary martensite (110) peak is compared in Table 5. Broader peaks indicate finer crystallite size and/or higher lattice strain, suggesting that the laser carburizing process promoted further microstructural refinement.
| Sample ID | 2θ (°) | FWHM (°) |
|---|---|---|
| LM-H | 44.660 | 0.286 |
| LC-H | 44.600 | 0.315 |
| LM-M | 44.684 | 0.307 |
| LC-M | 44.599 | 0.326 |
| LM-L | 44.634 | 0.315 |
| LC-L | 44.626 | 0.338 |
SEM micrographs revealed striking microstructural differences. In the LM units, the microstructure was non-uniform, featuring a mixture of martensite, retained austenite, and, notably, patches of untransformed pro-eutectoid ferrite. This inhomogeneity stems from the local variations in carbon availability from the dissolving graphite flakes during the brief melting period. Areas distant from graphite sources could not achieve sufficient carbon concentration to fully transform to austenite, leaving behind soft ferrite upon cooling.
In contrast, the microstructure of LC units was significantly more homogeneous and refined. The added carbon source ensured a uniformly high carbon concentration throughout the melt pool. This promoted the formation of a predominantly ledeburitic structure—a eutectic mixture of austenite (which subsequently transformed to martensite) and cementite—arranged in a fine, interconnected network. Within the cells of this network, either fine martensite or fine lamellar pearlite was observed, depending on the local cooling rate. The volume fraction of the hard ledeburite phase was substantially higher in LC units (estimated 60-70%) compared to LM units (40-45%). Furthermore, as shown in the micrographs, the scale of the microstructure (e.g., ledeburite spacing, martensite lath size) decreased with increasing energy density for both processes, due to higher cooling rates. However, at any given energy density, the LC microstructure was invariably finer and more homogeneous than the LM one.
2.3 Hardness and Mechanical Properties
The microhardness of the treated units, measured at their centers, is plotted against energy density. Two key trends are evident: First, for both LM and LC processes, the hardness of the unit increases with increasing energy density. This correlates with the observed microstructural refinement and potentially a higher martensite content at higher cooling rates. Second, and more importantly, at every energy level, the LC unit exhibits a significantly higher hardness than the corresponding LM unit. The LC-H sample achieved the peak hardness. This enhancement is a direct consequence of the laser carburizing process: the homogeneous, fine-grained ledeburitic structure with a high density of hard cementite and fine martensite/pearlite offers superior resistance to plastic deformation.
This was corroborated by bulk mechanical tests. Samples containing the laser-treated surface layers showed markedly improved resistance to both tensile and compressive deformation compared to the untreated gray iron casting. The laser carburized samples displayed higher yield strength, ultimate tensile strength, and fracture elongation in tension, as well as higher compressive strength. The degree of improvement was positively correlated with the laser energy density used during carburizing. This enhanced global mechanical performance underscores the effectiveness of the LC units in strengthening the material.
3. Fatigue Wear Performance and Mechanisms
3.1 Wear Test Results
The fatigue wear resistance was evaluated based on mass loss during RCF testing. The mass loss versus time curves for all samples typically showed an initial running-in period with a higher wear rate, followed by a transition to a steady-state period with a relatively constant wear rate. The transition to steady-state occurred earlier for all laser-treated samples compared to the parent material, indicating a quicker surface stabilization. The steady-state mass loss rate (MLR) and the total mass loss after 70 hours are the critical metrics for comparison.
The results, summarized visually in wear plots, clearly demonstrate the superiority of the laser-treated surfaces. The untreated gray iron casting (PM) exhibited the highest total mass loss and average MLR. All laser treatments reduced the wear. The ranking of performance, from best to worst, was: LC-H > LC-M ≈ LM-H > LM-M > LC-L > LM-L > PM. This ranking leads to several important conclusions:
- Laser surface treatment significantly improves the fatigue wear resistance of gray iron casting.
- Laser carburizing (LC) is consistently more effective than laser remelting (LM) at the same energy density.
- The improvement from both processes is strongly dependent on energy density, with higher energy density leading to better wear resistance. The LC-H sample showed the most dramatic improvement.
3.2 Contact Stress Distribution and Failure Modes
The finite element analysis provided crucial insight into why the non-smooth, laser-treated surface performs better. For a smooth surface (PM), the maximum contact pressure follows a classic Hertzian distribution. For a surface with alternating hard units and a softer matrix, the stress field is redistributed. The FEA results showed that the maximum contact stress is concentrated on the hard laser-treated units, which have a higher elastic modulus. Conversely, the contact stress on the softer matrix regions between the units is lower than the stress on a fully smooth surface of the parent material.
This stress redistribution is beneficial. The high-stress cycles are now borne by the units, which are specifically engineered (through grain refinement, carbide formation, and graphite elimination) to have high fatigue strength and resistance to crack initiation. The softer matrix regions, now operating under reduced stress, are less likely to initiate cracks from their inherent graphite flakes. This mimics the bionic principle of load sharing between hard and soft phases.
Examination of the worn surfaces confirmed this mechanism. The untreated gray iron casting surface showed severe damage characterized by large spalls and deep cracks, often initiating from and connecting graphite flakes. On laser-treated samples, the matrix areas showed less severe pitting compared to the PM. Most significantly, the laser-treated units themselves showed only very mild, shallow pitting. The primary failure mode in the units was micro-pitting, whereas the untreated matrix areas and the pure PM sample exhibited macro-pitting and spalling. This direct observation proves that the units successfully resisted severe fatigue damage.
3.3 Mechanisms of Fatigue Wear Improvement
The enhancement in fatigue wear resistance of the gray iron casting via laser carburizing can be attributed to a multi-faceted mechanism operating at different stages of the fatigue process:
1. Elimination of Crack Initiation Sites: The flake graphite in the base gray iron casting acts as pre-existing micro-cracks and stress concentrators. The laser melting process completely dissolves this graphite within the unit volume. By removing these intrinsic defects, the primary and easiest path for fatigue crack initiation is eradicated, significantly increasing the number of cycles required for crack nucleation.
2. Increased Difficulty of Crack Propagation: Should a micro-crack initiate (e.g., from a pore or inclusion), its propagation is hindered by the refined and hardened microstructure of the LC unit.
– Grain Refinement: The extremely fast solidification creates a fine-grained structure. Finer grains mean more grain boundaries, which act as barriers to crack propagation, deflecting its path and consuming more energy.
– Increased Hard Phase Content: The high volume fraction of hard, brittle cementite in the ledeburite network and the martensite matrix increases the material’s strength and hardness. While this can reduce toughness, in the context of high-cycle contact fatigue where stresses are primarily elastic, the increased resistance to plastic flow (ratcheting) is paramount. The accumulation of plastic strain per cycle ($\Delta \varepsilon_r$) is a driving force for failure. According to models for ratcheting-induced failure, the fatigue life ($N_f$) is related to a critical strain ($\varepsilon_c$) and the ratcheting strain per cycle:
$$ N_f = \frac{\varepsilon_c}{\Delta \varepsilon_r} $$
The LC unit’s higher yield strength directly reduces $\Delta \varepsilon_r$, thereby increasing $N_f$.
– Absence of Graphite: The absence of graphite flakes means there are no weak interfaces or ready-made paths for cracks to link up and grow easily, unlike in the base material where cracks readily propagate from one graphite flake to another.
3. Beneficial Stress Redistribution: As shown by FEA, the presence of hard units on the surface alters the contact stress field, shielding the softer, more vulnerable matrix from experiencing the peak stresses. This systemic effect protects the bulk of the material, confining the most demanding stress conditions to the regions best equipped to handle them.
The laser carburizing process synergistically combines all these effects more effectively than laser remelting. By introducing excess carbon, it ensures a homogeneous, high-carbon melt that yields a uniform, fine dispersion of hard carbides and a fully martensitic (or fine pearlitic) matrix without soft ferrite islands. This microstructural superiority translates directly to higher hardness, better resistance to plastic ratcheting, and consequently, superior fatigue wear resistance for the gray iron casting.
4. Conclusion
This study demonstrates that localized laser carburizing is a highly effective surface engineering strategy for significantly enhancing the rolling contact fatigue wear resistance of gray iron casting. The key findings are:
- The laser carburizing process, involving the pre-placement of a carbon layer, generates modified surface units with a more homogeneous, refined, and ledeburite-rich microstructure compared to conventional laser remelting. This is due to increased laser energy absorption and a higher, more uniform carbon concentration in the melt pool.
- The microhardness and bulk mechanical properties (tensile/compressive strength) of the laser carburized layers are superior to those of laser remelted layers and increase with the applied laser energy density.
- In rolling contact fatigue tests, laser carburized samples exhibited the lowest mass loss and wear rate. The improvement was more pronounced than for laser remelted samples and scaled positively with energy density.
- The enhancement mechanism is multifaceted: it involves the elimination of graphite-based crack initiation sites, the creation of a refined microstructure that impedes crack propagation, and a beneficial redistribution of contact stresses that shields the softer matrix. The laser carburized units themselves resist severe damage, exhibiting only mild pitting, while protecting the overall surface.
This laser carburizing technique presents a practical and efficient method for selectively hardening critical surfaces on gray iron casting components subjected to cyclic contact loads, potentially extending their service life in demanding engineering applications such as bearing guides, engine components, and machinery beds. The process is efficient, can be performed in ambient conditions, and offers precise control over the treated regions, making it a promising alternative to conventional coating-based surface treatments for gray iron casting.