In the manufacturing of critical components for rail transit vehicles, such as bogies, frames, and motor housings, steel casting plays a pivotal role due to its excellent isotropic properties, high strength, and good toughness. Among various steel casting materials, G20Mn5, a low-carbon low-alloy cast steel per European standard EN10293, is widely used for its superior comprehensive mechanical properties and weldability. However, defects like shrinkage porosity or cracks can occur during casting or service, necessitating repair welding to ensure structural integrity and reduce costs. This study investigates the microstructural evolution and mechanical performance of G20Mn5 steel casting joints repaired via MAG (Metal Active Gas) welding under five different process conditions, aiming to identify the optimal repair strategy for industrial applications.
Defects in steel casting components, if within acceptable limits, can be remedied through welding repairs, but the process must preserve or enhance the base material’s properties. The repair of steel casting involves considerations such as preheating, post-weld heat treatment (PWHT), and cooling rates to avoid detrimental phases like martensite or coarse Widmanstätten structures, which could compromise toughness and fatigue resistance. This research employs a first-person perspective to detail experimental procedures, results, and analyses, emphasizing the importance of process optimization for steel casting in demanding environments like rail transit.
| Material | C | Si | Mn | P | S | Cr | Cu | Ni | V |
|---|---|---|---|---|---|---|---|---|---|
| G20Mn5 Steel Casting | 0.17-0.23 | ≤0.60 | 1.00-1.60 | ≤0.020 | ≤0.020 | ≤0.30 | ≤0.30 | ≤0.80 | ≤0.03 |
| 4Si1 Welding Wire | 0.06-0.14 | 0.80-1.20 | 1.60-1.90 | ≤0.025 | ≤0.025 | ≤0.15 | ≤0.15 | ≤0.15 | ≤0.03 |
The base material, G20Mn5 steel casting, was procured as plates measuring 500 mm × 150 mm × 12 mm, simulating typical sections in bogie components. For repair welding, the AG 46 4 M21 4Si1 wire (diameter 1.2 mm) was selected based on its matching mechanical properties with the steel casting, ensuring high-strength matching joints. The chemical compositions and mechanical properties are summarized in Tables 1 and 2, highlighting the compatibility essential for reliable steel casting repairs.
| Material | Yield Strength (MPa) | Tensile Strength (MPa) | Elongation (%) | Impact Energy (J) at Test Temperature |
|---|---|---|---|---|
| G20Mn5 Steel Casting | ≥300 | ≥480 | ≥20 | ≥27 at -20°C |
| 4Si1 Welding Wire | ≥460 | 530-680 | ≥20 | ≥47 at -40°C |
Five distinct repair welding processes were designed, varying in preheating, post-weld insulation, and heat treatment conditions, as detailed in Table 3. All processes used MAG welding with a shielding gas mixture of 80% Ar and 20% CO2 at a flow rate of 18 L/min. The welding parameters, including current, voltage, and speed for each pass (root, filler, and cover), were kept constant to isolate the effects of thermal management. For steel casting repairs, thermal cycles critically influence microstructural transformations; thus, processes ranged from no preheating with slow cooling to preheating at 120-150°C followed by natural cooling or annealing.
| Process ID | Preheating | Post-weld Treatment | Heat Treatment |
|---|---|---|---|
| Process 1 | 120-150°C | Natural Cooling | Annealing |
| Process 2 | 120-150°C | Natural Cooling | None |
| Process 3 | 120-150°C | Insulation Cover for Slow Cooling | None |
| Process 4 | None | Natural Cooling | Annealing |
| Process 5 | None | Insulation Cover for Slow Cooling | None |
The welding parameters for each pass are consistent across processes: root pass at 155-170 A and 17-19 V, filler passes at 270-310 A and 28-32 V, and cover pass at 280-310 A and 29-32 V, with travel speeds adjusted between 3-9 mm/s. Such parameters ensure adequate penetration and fusion in steel casting repairs, minimizing defects like lack of fusion or excessive heat input.
Microstructural analysis was conducted on cross-sections of repaired joints using optical microscopy after polishing and etching with 5% nital. Mechanical testing included tensile tests per GB/T 2651-2023, hardness traverses per GB/T 2654-2008, Charpy impact tests per GB/T 2650-2008 at room temperature, and fatigue tests per GB/T 3075-2021 under axial tension-tension loading with stress ratio R=0. Fatigue limits were defined at 107 cycles. All tests aimed to evaluate whether the repaired steel casting joints meet or exceed base material requirements.
The microstructure of steel casting and welded joints is pivotal in determining mechanical performance. For G20Mn5 steel casting, the base material typically consists of ferrite and pearlite, but welding introduces thermal cycles that alter phase distributions. In all five repair processes, the weld zone (WZ) exhibited similar microstructures: proeutectoid ferrite along columnar grain boundaries, acicular ferrite, granular bainite, and minor pearlite, with no martensite observed. This indicates that the cooling rates, even without preheating, were sufficient to avoid hard phases detrimental to steel casting integrity.
In the heat-affected zone (HAZ), particularly the coarse-grained region (CGHAZ), microstructures varied. Processes 2, 3, and 5 showed some Widmanstätten ferrite, but it was rated as Grade 2 per GB/T13299-2022, implying no coarse structures that significantly reduce toughness. The absence of martensite across all joints confirms that the selected processes prevent quench hardening, a common issue in steel casting repairs when cooling is too rapid. The microstructural homogeneity can be described using phase transformation kinetics, where the cooling time from 800°C to 500°C (Δt8/5) influences phase formation. For steel casting, a balanced Δt8/5 promotes bainitic structures over martensite, as approximated by:
$$\Delta t_{8/5} = k \cdot \left( \frac{Q}{v} \right)^n$$
where \(k\) and \(n\) are material constants, \(Q\) is heat input, and \(v\) is travel speed. In these repairs, moderate heat inputs (e.g., 1.0-1.5 kJ/mm) likely resulted in Δt8/5 values conducive to bainite, ensuring good weldability for the steel casting.
Mechanical properties were comprehensively evaluated. Tensile test results, summarized in Table 4, show that all repaired joints exceeded the base steel casting requirement of 480 MPa tensile strength, with Process 3 achieving the highest average of 547.3 MPa. Elongation values were all above 20%, meeting specifications. Notably, all tensile specimens fractured in the base material, indicating that the weld and HAZ had higher strength—a testament to the high-strength matching wire. The strength-ductility product (\(U\)), calculated as \(U = R_m \times A\), where \(R_m\) is tensile strength and \(A\) is elongation, serves as a toughness index; Process 2 had the highest \(U\) at 13,181.51 MPa·%, suggesting optimal balance for steel casting applications.
| Process ID | Tensile Strength, Rm (MPa) | Elongation, A (%) | Strength-Ductility Product, U (MPa·%) | Fracture Location |
|---|---|---|---|---|
| Process 1 | 505.3 | 21.25 | 10,373.63 | Base Material |
| Process 2 | 514.3 | 25.63 | 13,181.51 | Base Material |
| Process 3 | 547.3 | 23.18 | 12,686.41 | Base Material |
| Process 4 | 480.3 | 25.63 | 12,310.09 | Base Material |
| Process 5 | 514.0 | 24.79 | 12,742.06 | Base Material |
Hardness profiles, measured along lines 2 mm (L1) and 6 mm (L2) from the top surface, revealed consistent trends: hardness peaked in the CGHAZ, with values below 250 HV3 for all processes, confirming no martensitic hardening. Processes involving annealing (1 and 4) showed lower overall hardness, indicating softening due to recovery and recrystallization. This softening can be modeled using the Hollomon equation for work hardening, but in annealing, it relates to dislocation annihilation. For steel casting repairs, maintaining hardness below thresholds is crucial to prevent brittleness; the data affirm that all processes achieve this.
Impact toughness, assessed via Charpy V-notch tests at room temperature, demonstrated that weld zones had the highest impact energy (152-171 J), followed by HAZ (102-116 J) and base steel casting (94-109 J). Process 4, with annealing, yielded the best toughness, likely due to stress relief and grain refinement. The superior toughness of weld metal underscores the effectiveness of the selected wire for steel casting repair. The impact energy (\(CVN\)) can be correlated to microstructure using empirical relations like:
$$CVN = \alpha \cdot (d^{-1/2}) + \beta$$
where \(d\) is grain size and \(\alpha\), \(\beta\) are constants. Finer microstructures in the weld, due to rapid solidification, enhance toughness, vital for dynamic loading in rail components.
Fatigue performance, critical for cyclically loaded steel casting parts, was evaluated through S-N curves. Fatigue limits (\(\sigma_0\)) at R=0 ranged from 171.55 MPa (Process 1) to 241.10 MPa (Process 3), as shown in Table 5. All fatigue failures originated in the base material, with crack initiation at micro-shrinkage pores inherent to the steel casting process. This highlights that for repaired steel casting joints, fatigue life is often governed by casting defects rather than weld quality, emphasizing the need for stringent quality control in initial steel casting production. The fatigue limit ratio \(\sigma_0 / R_m\) varied from 0.34 to 0.44, within typical ranges for low-alloy steels. The Basquin equation describes the high-cycle fatigue behavior:
$$\sigma_a = \sigma_f’ \cdot (2N_f)^b$$
where \(\sigma_a\) is stress amplitude, \(N_f\) is cycles to failure, \(\sigma_f’\) is fatigue strength coefficient, and \(b\) is exponent. For steel casting, defects act as stress concentrators, reducing \(\sigma_f’\); thus, post-repair non-destructive testing is recommended.
| Process ID | Fatigue Limit, σ0 (MPa) | Tensile Strength, Rm (MPa) | Ratio σ0/Rm | Fatigue Crack Origin |
|---|---|---|---|---|
| Process 1 | 171.55 | 505.3 | 0.34 | Base Material (Micro-shrinkage) |
| Process 2 | 224.65 | 514.3 | 0.44 | Base Material (Micro-shrinkage) |
| Process 3 | 241.10 | 547.3 | 0.44 | Base Material (Micro-shrinkage) |
| Process 4 | 178.36 | 480.3 | 0.37 | Base Material (Micro-shrinkage) |
| Process 5 | 213.01 | 514.0 | 0.41 | Base Material (Micro-shrinkage) |
Discussion of the results centers on optimizing the repair process for steel casting. While all five processes met mechanical requirements, Process 5 (no preheating, post-weld insulation for slow cooling, no heat treatment) emerged as optimal based on a holistic view of efficiency, energy consumption, and performance. Preheating, as in Processes 1-3, can mitigate cracking but adds complexity and time; annealing in Processes 1 and 4 improves toughness but reduces strength and fatigue limits, besides requiring extended cycles. For industrial steel casting repair, simplicity and cost-effectiveness are key, making Process 5 favorable—it avoids preheating, uses simple insulation for controlled cooling, and omits heat treatment, yet delivers excellent tensile strength, adequate toughness, and good fatigue resistance.
The microstructural observations align with mechanical data: the absence of martensite ensures good hardness and toughness, while the presence of fine bainitic and ferritic structures in weld and HAZ contributes to strength. In steel casting repairs, controlling heat input and cooling rate is paramount; Process 5 achieves this through insulation, slowing cooling to prevent quench phases while minimizing residual stresses. The superiority of this process can be quantified using a performance index \(PI\) that combines key properties:
$$PI = w_1 \cdot \frac{R_m}{R_{m,ref}} + w_2 \cdot \frac{A}{A_{ref}} + w_3 \cdot \frac{CVN}{CVN_{ref}} – w_4 \cdot \frac{H_{max}}{H_{ref}}$$
where \(w_i\) are weights, and ref values are base material targets. For steel casting, assigning higher weights to strength and toughness might yield the highest \(PI\) for Process 5.
Furthermore, the study underscores the role of casting defects in limiting fatigue life. Even with optimal repair, inherent micro-shrinkage in steel casting can initiate cracks under cyclic loads, suggesting that repair welding should be complemented by techniques like hot isostatic pressing to densify castings. Future work could explore hybrid processes for steel casting, such as laser-arc welding, to further refine microstructures and enhance fatigue performance.
In conclusion, this comprehensive investigation demonstrates that G20Mn5 steel casting can be effectively repaired using MAG welding with AG 46 4 M21 4Si1 wire. Five different processes were evaluated, all yielding joints without deleterious martensite and with mechanical properties surpassing base material specifications. Based on microstructural analysis, tensile strength, hardness, impact toughness, and fatigue resistance, the process involving no preheating, post-weld insulation for slow cooling, and no heat treatment is recommended as the optimal repair strategy for steel casting in rail transit applications. This approach balances performance with practical considerations, supporting the sustainable use of steel casting components through reliable repair methodologies.
The implications extend beyond G20Mn5 to other low-alloy steel casting grades, emphasizing the importance of tailored thermal management in welding repairs. As steel casting continues to be integral to heavy industries, optimizing repair processes will enhance component longevity, reduce waste, and improve economic efficiency, contributing to advancements in manufacturing and materials engineering.