As a researcher in the field of materials processing, I have devoted extensive efforts to investigating the repair of surface defects in gray cast iron components. The motivation for this work stems from the fact that gray cast iron, due to its simple production process and low cost, is widely used in machinery manufacturing, metallurgy, mining, petrochemical and other industrial sectors. However, during the casting process, a sand foundry defect such as porosity, slag inclusion, or shrinkage is almost inevitable. Furthermore, subsequent machining operations may introduce over-tolerance, and in-service conditions often cause wear or scoring. These sand foundry defect types not only degrade the performance of cast iron parts but also lead to significant economic losses. Repairing these defects is therefore of great economic and social value.
Among the various surface engineering techniques available, plasma arc spray welding stands out because of its high degree of automation, simple operation, and high production efficiency. For this study, I focused on the repair of gray cast iron (HT250) surface defects using plasma spray welding with Ni25 self-fluxing alloy powder. For comparison, I also employed manual arc cold welding using two typical cast iron electrodes, Z116 (high vanadium type) and Z308 (pure nickel type). The objective was to evaluate the feasibility, microstructural characteristics, and mechanical properties of the resulting repair layers. Throughout this research, I paid special attention to the influence of the original sand foundry defect geometry and the thermal cycle on the solidification behavior and the formation of hard, brittle phases such as white iron in the semi-melted zone.
Experimental Materials and Methods
The base material used in all experiments was HT250 gray cast iron, machined into plates with dimensions of 60 mm × 40 mm × 10 mm. The chemical composition of HT250 is listed in Table 1.
| C | Si | Mn | P | S |
|---|---|---|---|---|
| 3.1–3.3 | 1.6–1.8 | 0.7–0.9 | <0.25 | <0.12 |
Three types of repair materials were selected: Ni25 nickel-based self-fluxing alloy powder (particle size 50–153 mesh, apparent density 4.8 g/cm³, flowability 15 s/50 g, melting point approximately 1300 °C), Z308 nickel-based cast iron electrode (diameter 3.2 mm), and Z116 high-vanadium cast iron electrode (diameter 3.2 mm). The Ni25 powder was supplied in the form of nearly spherical particles, which ensured good flowability during plasma spray welding. The chemical composition of Ni25 is given in Table 2.
| C | Cr | B | Si | Fe | Ni |
|---|---|---|---|---|---|
| 0.05 | 5.06 | 1.58 | 3.04 | 3.2 | balance |
Before welding, the HT250 plates were thoroughly cleaned to remove oil and rust. Plasma spray welding was carried out using a GAP 2001 DC system, while manual arc welding was performed with an inverter DC welding machine. For the plasma spray welding process, I used high-purity argon as both the plasma gas and the shielding gas. The powder was fed into the arc column through a carrier gas flow. After welding, the specimens were sectioned using wire electrical discharge machining into 10 mm × 10 mm × 15 mm blocks for metallographic observation and mechanical testing.
Optimization of Plasma Spray Welding Parameters
Since the plasma spray welding process involves multiple interrelated parameters, I adopted an orthogonal experimental design to determine the optimal parameter combination. The microhardness of the deposited layer was chosen as the evaluation index, with the target of matching the hardness of the HT250 substrate (measured average 238.5 HV). Three factors were investigated: transferred arc current (A), plasma gas flow rate (L/min), and powder feed gas flow rate (L/min). Table 3 shows the levels of each factor.
| Level | Current (A) | Plasma gas (L/min) | Powder gas (L/min) |
|---|---|---|---|
| 1 | 45 | 1.2 | 2.0 |
| 2 | 55 | 1.5 | 2.5 |
| 3 | 65 | 1.8 | 3.0 |
The orthogonal array L9(3³) was employed, and the microhardness results are summarized in Table 4.
| Specimen | A (Current) | B (Plasma gas) | C (Powder gas) | Hardness (HV) |
|---|---|---|---|---|
| 1 | 45 | 1.2 | 2.0 | 255.7 |
| 2 | 45 | 1.5 | 2.5 | 250.3 |
| 3 | 45 | 1.8 | 3.0 | 262.7 |
| 4 | 55 | 1.2 | 2.5 | 247.6 |
| 5 | 55 | 1.5 | 3.0 | 235.6 |
| 6 | 55 | 1.8 | 2.0 | 240.5 |
| 7 | 65 | 1.2 | 3.0 | 273.2 |
| 8 | 65 | 1.5 | 2.0 | 255.6 |
| 9 | 65 | 1.8 | 2.5 | 264.2 |
From the range analysis, I calculated the average hardness for each level of each factor, as shown in Table 5.
| Parameter | K1/3 | K2/3 | K3/3 | Range R |
|---|---|---|---|---|
| A | 256.2 | 241.2 | 264.3 | 23.1 |
| B | 258.8 | 247.2 | 255.8 | 11.6 |
| C | 250.6 | 254.0 | 257.1 | 6.5 |
The range values clearly show that the transferred arc current has the most significant influence on the hardness of the spray-welded layer, followed by the plasma gas flow rate, while the powder feed gas rate has the least effect. The optimal parameter combination that produced a hardness closest to the substrate (238.5 HV) was found to be A2B2C1, i.e., a current of 55 A, a plasma gas flow of 1.5 L/min, and a powder gas flow of 2.0 L/min. The corresponding voltage was 25 V, the shielding gas flow was 12 L/min, and the nozzle-to-workpiece distance was 10 mm. These optimal parameters were used for all subsequent plasma spray welding experiments.
Microstructural Characterization
After metallographic preparation, the specimens were etched with appropriate reagents. The Ni25 spray-welded specimen was etched with a solution of 30 ml nitric acid and 2 ml hydrofluoric acid for 10 s. The Z116 weld was etched with 5% nital for 10 s, while the Z308 weld was etched with a mixture of 50% nitric acid and 50% acetic acid for 5 minutes. Microstructures were observed using optical microscopy and scanning electron microscopy (SEM). The chemical composition and elemental distribution across the fusion zone were analyzed by energy dispersive spectroscopy (EDS), and phase identification was performed by X-ray diffraction (XRD).
Ni25 Plasma Spray-Welded Layer
The Ni25 spray-welded layer exhibited a clear interface with the substrate, indicating sound metallurgical bonding. The microstructure of the spray-welded layer consisted mainly of dendrites, which grew in a direction perpendicular to the fusion line. This dendritic morphology is characteristic of rapid solidification with a large degree of constitutional supercooling in the molten pool. The XRD pattern revealed that the main phases present were Fe₀.₆₄Ni₀.₃₆, FeNi₃, Ni, FeCr₀.₂₉Ni₀.₁₆C₀.₀₆, and Ni₃Si. The presence of these phases is consistent with the nickel-rich composition of the Ni25 powder and the interdiffusion of elements such as Fe and Cr between the coating and the substrate. The EDS line scan across the interface showed a smooth transition of Ni, Fe, and Cr, confirming mutual diffusion. Notably, no white iron structure was observed in the semi-melted zone of the Ni25 spray-welded layer, which is a significant advantage for subsequent machining.
Z308 and Z116 Overlay Welds
For comparison, I also prepared overlay welds using Z308 and Z116 electrodes. The microstructure of the Z308 weld consisted of an austenitic matrix with graphite flakes and intergranular Ni-Si eutectic constituents. The XRD analysis confirmed the presence of Fe₀.₆₄Ni₀.₃₆, FeNi₃, C, and Ni₃Si. The Z308 weld exhibited a narrow white iron zone at the fusion boundary, approximately 50 μm wide, which can be attributed to the strong graphitizing effect of nickel.
The Z116 weld, on the other hand, showed a microstructure composed mainly of ferrite with dispersed Fe₃Si, Fe₂SiV, and FeV compounds. The XRD patterns confirmed these phases. The semi-melted zone of the Z116 weld contained a much wider white iron layer, around 200 μm in width, with a hardness as high as 720.9 HV. This wide hard zone is detrimental to machining and reduces the overall quality of the repair.
White Iron Problem in Cast Iron Repair
One of the most critical issues in welding cast iron is the formation of white iron in the semi-melted zone. White iron is an extremely hard and brittle phase that forms when carbon in the liquid iron precipitates as cementite (Fe₃C) rather than graphite. This occurs due to rapid cooling and insufficient graphitizing elements. The presence of a sand foundry defect often complicates the situation because the defect area may have a different local cooling rate or chemical heterogeneity, promoting even more white iron formation. In my experiments, I compared the semi-melted zones of the three repair methods. Table 6 summarizes the approximate white iron widths observed.
| Repair method | Width of white iron zone (μm) |
|---|---|
| Plasma spray welding (Ni25) | 0 |
| Z308 electrode | ≈50 |
| Z116 electrode | ≈200 |
The plasma spray welding process uses a highly concentrated arc as the heat source, which produces a high energy density and rapid heating and cooling rates. This narrows the distance between the liquidus temperature at the fusion boundary and the solidus temperature of the gray cast iron, effectively reducing the width of the semi-melted zone. In addition, the Ni-based alloy powder contains elements such as Ni and Si that promote graphitization, further suppressing white iron formation. Since a sand foundry defect often is located on the surface or slightly below it, the localized thermal conditions during repair are critical. The absence of white iron in the Ni25 plasma-sprayed layer means that the repaired surface can be easily machined without tool damage or excessive wear.
In contrast, manual arc welding with Z308 and Z116 produces a broader thermal gradient, leading to a wider semi-melted zone. The Z308 electrode, being nickel-based, provides strong graphitization and limits the white iron width to about 50 μm. The Z116 electrode, although it contains vanadium to tie up carbon, still results in a 200 μm white iron zone due to the lower graphitizing power and higher cooling rate. This zone not only makes machining difficult but also reduces the reliability of the repair, especially when the original defect was a deep or geometrically complex sand foundry defect that requires multiple weld passes.
Mechanical Properties of the Repaired Layers
To evaluate the overall quality of the repaired layers, I performed microhardness measurements, tensile tests, and crack resistance assessments.
Microhardness
Vickers microhardness measurements were taken along a line perpendicular to the fusion line, starting from the weld metal, across the semi-melted zone, and into the substrate. The applied load was 500 g with a dwell time of 10 s. Figure 1 shows the hardness profiles for the three types of repairs.
| Repair method | Weld metal (HV) | Semi-melted zone (HV) | Substrate (HV) |
|---|---|---|---|
| Ni25 plasma | 230–240 | ≈290 | ≈238 |
| Z308 | 170–190 | ≈456 (narrow peak) | ≈238 |
| Z116 | 270–280 | ≈720 (wide peak) | ≈238 |
The Ni25 spray-welded layer exhibited an average hardness of about 235.6 HV, which is remarkably close to the base metal hardness of 238.5 HV. This matching hardness ensures uniform machining behavior and avoids the abrupt transitions that cause tool breakage. The Z308 weld metal was softer than the base metal, while the Z116 weld metal was harder. However, the Z308 semi-melted zone had a sharp hardness peak of 456.6 HV, and the Z116 zone reached an extreme peak of 720.9 HV with a wider distribution. Such high local hardness is unacceptable for applications requiring machining of the repaired area.
Tensile Strength
Tensile test specimens were fabricated from HT250 according to GB/T 9439-88, with a V-shaped notch machined at the center. The notch was filled using the optimized plasma spray welding process. After welding and surface grinding, the specimens were tested at room temperature. The results are presented in Table 8.
| Specimen | Ultimate tensile strength (MPa) | Fracture location |
|---|---|---|
| 1 | 212.67 | Weld center |
| 2 | 183.34 | Weld center |
| 3 | 196.53 | Weld center |
| 4 | 200.39 | Weld center |
| Average | 198.23 | – |
All fracture locations were at the weld center, indicating that the bonding strength between the plasma-sprayed layer and the substrate exceeded the tensile strength of the weld metal itself. The average tensile strength of 198.23 MPa is acceptable for many cast iron repair applications and confirms that the metallurgical bond achieved is strong and reliable.
Crack Resistance
I evaluated the crack sensitivity of the three repair methods using the rigid circular disc test. In this test, a weld bead is deposited along a diameter of the disc, and the disc provides high restraint. The occurrence of cracks in the weld bead is used as a measure of its resistance to cracking. The results are summarized in Table 9.
| Repair method | Weld appearance | Crack observed? |
|---|---|---|
| Ni25 plasma | Good, smooth | No crack |
| Z308 | Good, uniform | No crack |
| Z116 | Irregular, poor | Two transverse cracks |
The Ni25 plasma-sprayed weld and the Z308 weld both showed excellent crack resistance, with no cracks detected after the restraint test. The Z116 weld, however, exhibited two distinct transverse cracks, indicating a higher susceptibility to cracking. This poor crack resistance can be attributed to the presence of the wide white iron zone and the lower ductility of the ferritic matrix with hard vanadium carbide precipitates.
Discussion
From the above results, it is clear that plasma spray welding with Ni25 powder offers distinct advantages over manual arc welding with Z116 and Z308 electrodes for repairing surface defects in gray cast iron. The key benefits include:
- No white iron formation: The concentrated heat input of the plasma arc minimizes the semi-melted zone, preventing the formation of brittle white iron. This is particularly important when the original sand foundry defect has created a thin section or a sharp edge that would otherwise cool rapidly and promote cementite formation.
- Hardness matching: The Ni25 spray-welded layer has a hardness close to that of the HT250 substrate, enabling easy machining and uniform wear behavior.
- High bond strength: The metallurgical bonding resulting from the interdiffusion of alloying elements provides high tensile strength, as evidenced by the consistent fracture locations at the weld center rather than at the interface.
- Good crack resistance: The absence of hard brittle phases and the high ductility of the nickel-based alloy contribute to superior crack resistance under high restraint conditions.
In contrast, the Z116 electrode, despite being specifically designed for cast iron welding, suffers from a wide white iron zone and poor crack resistance. The Z308 electrode provides better results than Z116 due to its higher nickel content, but it still cannot match the performance of the plasma-sprayed Ni25 layer. Therefore, for repairing critical components where a sand foundry defect would otherwise render the part unusable, plasma spray welding is the preferred technique.
It should also be noted that the presence of a sand foundry defect not only affects the mechanical integrity of the casting but also influences the repair process. Defects such as shrinkage cavities or gas porosities can trap oxides or moisture, which may cause porosity in the weld deposit. Plasma spray welding, with its cleaning action and the use of inert shielding gases, effectively mitigates these issues. The high temperature of the plasma arc also helps in degassing and floating out the impurities, resulting in a sound deposit.
Conclusions
Based on my experimental investigations, the following conclusions can be drawn:
- The optimal plasma spray welding parameters for Ni25 powder on HT250 were determined through orthogonal experiments: transferred arc current of 55 A, plasma gas flow of 1.5 L/min, powder feed gas flow of 2.0 L/min, arc voltage of 25 V, shielding gas flow of 12 L/min, and nozzle-to-workpiece distance of 10 mm. The current has the strongest effect on layer hardness, followed by plasma gas flow and powder gas flow.
- The Ni25 spray-welded layer forms a metallurgical bond with the substrate, with clear evidence of elemental diffusion. The microstructure consists mainly of austenite dendrites with small amounts of secondary phases such as Fe₀.₆₄Ni₀.₃₆, FeNi₃, Ni, FeCr₀.₂₉Ni₀.₁₆C₀.₀₆, and Ni₃Si.
- The Z116 electrode produces a ferritic matrix with dispersed vanadium-rich compounds, while the Z308 electrode produces an austenitic matrix with graphite precipitation. Both arc-welded deposits contain white iron in the semi-melted zone, with the Z116 weld having a much wider white iron zone (≈200 μm) than the Z308 weld (≈50 μm).
- The Ni25 plasma-sprayed layer exhibits no white iron, making it the most suitable method for repairing HT250 defects. Its average microhardness is 235.6 HV, which is very close to the substrate hardness of 238.5 HV.
- Tensile tests revealed that all plasma-sprayed specimens fractured at the weld center, with an average ultimate tensile strength of 198.23 MPa, confirming high bond strength.
- Crack resistance tests showed that Ni25 plasma-sprayed and Z308 weld deposits have excellent resistance to cracking, while the Z116 weld deposit is prone to transverse cracking due to its wide brittle white iron zone.
In summary, plasma spray welding with Ni25 powder is a highly effective method for repairing surface defects of gray cast iron. It not only eliminates the harmful white iron structure that is commonly associated with arc welding processes but also provides a repaired layer with matched hardness, high bond strength, and excellent crack resistance. This technique is particularly valuable for salvaging castings that would otherwise be scrapped due to a severe sand foundry defect. Future work should explore the application of plasma spray welding to other cast iron grades and to defects that are larger or located in more complex geometries.