The reliability of core components in heavy industrial equipment, such as roller mills, is paramount for continuous operation. Our team has extensive experience in addressing field failures, which often stem from inherent material issues or operational stresses. Among these, casting defect problems in large, critical parts present a unique challenge. Repairing them on-site is frequently the only viable option due to the prohibitive cost and time associated with replacement, especially for oversized components. The following discourse details a proven, systematic approach developed through practical field application for the successful remediation of a significant casting defect in a mill bearing housing. This methodology integrates thorough inspection, metallurgical principles, controlled welding techniques, and innovative machining compensation strategies.
Prior to delving into the repair procedure, it is instructive to understand the context of the equipment. Roller mills integrate grinding, drying, classification, and conveying, offering high efficiency and compact design. Their dynamic components, like classifier rotors, require robust sealing. A common issue involves the failure of lip seals on rotating shafts, leading to dust ingress and bearing damage. Our team redesigned such a system, replacing two of three lip seals with labyrinth-style mechanical seals and modifying the chute for easier maintenance of the remaining seal. This improvement enhanced longevity and reduced lifecycle costs. However, a more profound challenge involves structural integrity failures, such as those arising from a casting defect in primary load-bearing structures.
In one particular project, a critical casting defect was discovered during the final inspection of a large roller mill’s rocker arm bearing housing. The housing, a substantial steel casting, exhibited visible surface cracks indicative of subsurface flaws. Transporting this massive component for factory repair was impractical. Therefore, a comprehensive on-site repair protocol was developed and executed in collaboration with metallurgical experts. The entire process, validated by a decade of subsequent trouble-free operation, can be broken down into five key phases, as illustrated in the following workflow:
1. Inspection & Marking
2. Preheating
3. Welding & Deposition
4. Straightening & Finishing
5. Measurement & Verification
Phase 1: Inspection, Marking, and Defect Removal
The initial step is a thorough non-destructive testing (NDT) assessment to characterize the casting defect. Ultrasonic testing (UT) was employed to map the approximate depth and extent of the flaw beneath the visible surface crack. Based on the UT results, the defective area was outlined on the casting surface. The objective of marking is to define the boundary for removal, ensuring all compromised material is excised.
The removal strategy is crucial to prevent further crack propagation and minimize heat-affected zones. For a transverse crack, drill “stop holes” at both ends to arrest its growth. The material is then removed using a handheld grinder—avoiding thermal cutting methods like carbon arc gouging to prevent excessive heat input and distortion. The cutting profile is trapezoidal: a horizontal cut is made at a calculated distance (H1, e.g., 10mm) below the suspected defect bottom, and an angled cut (approximately 30°) is initiated from a point above the defect (H2, e.g., 10mm above). This angled approach reduces the volume of material removed while providing good access. After each cutting pass, dye penetrant inspection (DPI) is performed on the newly exposed surface. If indications persist, cutting continues incrementally (to a new depth Hx) until a completely sound metal structure is revealed, confirming the total elimination of the casting defect.
Phase 2: Theoretical Foundation for Preheating
The bearing housing material was identified as a normalized cast steel, G20Mn5, known for reasonable weldability but prone to hardening in the heat-affected zone (HAZ). The chemical composition and mechanical properties are summarized below:
| Grade | Code | C | Si (Max) | Mn | P (Max) | S (Max) | Ni |
|---|---|---|---|---|---|---|---|
| G20Mn5 | 1.622 | 0.17–0.23% | 0.60% | 1.0–1.6% | 0.020% | 0.020% | 0.8% Max |
| Grade | Heat Treatment | Thickness (mm) | Yield Strength (Min, MPa) | Tensile Strength (MPa) | Elongation (Min, %) |
|---|---|---|---|---|---|
| G20Mn5 | Normalized (+N) | t ≤ 30 | 300 | 480–620 | 20 |
| G20Mn5 | Quenched & Tempered (+QT) | t ≤ 100 | 300 | 500–650 | 22 |
Welding on such a thick section creates a small, rapidly cooling weld pool, leading to a quench-like effect. This can result in high hardness and hydrogen-induced cracking (HIC) in the HAZ. Preheating is the primary countermeasure. It slows the cooling rate, allowing hydrogen to diffuse out and reducing the martensite hardness, thereby improving crack resistance. The optimal preheat temperature (Tp) can be estimated using the Seferain method, which considers carbon equivalent (CE) and plate thickness (t).
First, the carbon equivalent [C]c is calculated:
$$ [C]_c = C + \frac{40(Mn + Cr)}{360} + \frac{20Ni}{360} + \frac{28Mo}{360} $$
For G20Mn5, neglecting Cr and Mo, and using average values (C=0.20%, Mn=1.30%, Ni=0.4%):
$$ [C]_c \approx 0.20 + \frac{40(1.30)}{360} + \frac{20(0.4)}{360} \approx 0.20 + 0.144 + 0.022 \approx 0.366 $$
Next, a thickness factor is introduced, where t is the thickness in mm (approximately 150mm for this housing):
$$ [C]_t = 0.005 \times t \times [C]_c = 0.005 \times 150 \times 0.366 \approx 0.2745 $$
The total equivalent carbon content [C] is:
$$ [C] = [C]_c + [C]_t = 0.366 + 0.2745 \approx 0.6405 $$
Finally, the preheat temperature is:
$$ T_p = 350 \times [C] – 0.25 \approx 350 \times 0.6405 – 0.25 \approx 224^\circ\text{C} – 0.25 \approx 224^\circ\text{C} $$
Based on this calculation, a preheat temperature of approximately 200–225°C was targeted. To prevent distortion during heating, the bearing cap was bolted securely to the base, and end covers were installed to maintain structural rigidity. Heating was performed uniformly around the repair zone using heating blankets or torches, with temperature monitored continuously by infrared thermometers to avoid localized overheating.
Phase 3: Controlled Welding Deposition
With the preheat maintained, the repair welding commenced. This falls under the category of “cold repair welding” (Tp < 300°C). Shielded Metal Arc Welding (SMAW) was selected for its versatility. Electrodes with a basic (low-hydrogen) coating were chosen to minimize hydrogen intake. The electrodes were baked at 250°C for one hour and stored in a holding oven at 100–150°C until use.
A multi-layer, multi-pass technique was employed. The root pass was made using a small diameter electrode (2.5mm) at a low current (~110A) to ensure proper fusion at the base of the cavity without excessive penetration. Subsequent layers were deposited using 3.2mm electrodes with a moderate current, adopting a horizontal weaving pattern. Crucially, to mitigate residual stresses, each layer was peened immediately while the temperature was between 100–200°C. Peening plastically deforms the weld bead, relieving shrinkage stresses. Slag was meticulously removed after each pass, and visual and DPI checks were conducted intermittently to ensure soundness. The interpass temperature was kept above 150°C, requiring supplemental heating. After completing the weld fill, the entire component was insulated with ceramic wool and allowed to cool slowly to ambient temperature to prevent thermal shock and further reduce cracking risk. Post-weld heat treatment was deemed unnecessary due to the controlled process.
| Process | Electrode Type | Diameter (mm) | Current (A) | Key Technique | Purpose |
|---|---|---|---|---|---|
| SMAW (Root) | Low-Hydrogen | 2.5 | ~110 | Stringer Bead | Ensure root fusion, minimize heat |
| SMAW (Fill/Cap) | Low-Hydrogen | 3.2 | Moderate | Weaving, Interpass Peening | Deposit metal, relieve stress |
Phase 4: Geometric Correction and Finishing
Despite all precautions, welding a large cavity inevitably introduces some distortion, altering the bore’s cylindrical geometry. The challenge was to restore the bore’s diameter, roundness, and surface finish to original specifications using only field tools. Conventional wisdom suggested on-site boring with a portable machine—a precise but costly and time-intensive solution.
Our team implemented an innovative thermal straightening technique. The principle is to apply controlled, localized heat to induce compensatory plastic deformation, effectively “pulling” the bore back into alignment. The procedure is iterative and experience-based:
- Measurement: After complete cooling, the bore’s geometry is meticulously mapped using inside micrometers. Reference lines are scribed axially and circumferentially to ensure consistent measurement points.
- Analysis: The distortion pattern (e.g., ovality, localized depression) is identified from the measurements.
- Targeted Heating: Specific areas on the housing’s external surface, opposite to the internal distortion, are selected. These areas are heated to a specific temperature (typically in the 600–700°C range, below Ac1) using a torch, causing localized thermal expansion.
- Induced Deformation: As the heated zone expands, it yields plastically. Upon cooling and contracting, it creates a tensile stress that pulls the distorted internal surface towards the desired shape. In some cases, strategic relief cuts are made near the heated area to guide the deformation.
- Repetition: Steps 1-4 are repeated, with minor adjustments, until the bore measurements fall within tolerance. Usually, 2-3 iterations suffice.
Following successful straightening, the final step is surface finishing. The welded area is first rough-ground flush with the original bore surface using coarse grinding wheels. Final finishing is achieved with finer abrasive tools and polishing stones, moving in alternating transverse and longitudinal directions until the surface roughness matches the surrounding area and no visible transition remains.
Phase 5: Final Verification and Quality Assurance
The culmination of the repair is rigorous verification. The bore is measured exhaustively at all previously marked locations. Key parameters include:
$$ \text{Diameter Tolerance} = D_{\text{nominal}} \pm \delta $$
$$ \text{Ovality} = |D_{\text{max}} – D_{\text{min}}| \leq \epsilon $$
$$ \text{Surface Roughness } (R_a) \leq \text{Specified Value} $$
Only when all dimensional and surface finish criteria are satisfied is the component deemed repaired. A final comprehensive NDT check (UT and DPI) of the welded and heat-affected zones is conducted to confirm the absence of any new discontinuities or residual casting defect indications. The bearing housing is then ready for assembly, having had the original casting defect completely eradicated and its functional geometry restored.
Discussion and Preventative Perspectives
The successful on-site repair of this major casting defect demonstrates that with proper methodology, even critical flaws can be remediated without component replacement. The core principles—thorough defect removal, scientific preheating, stress-managed welding, and corrective straightening—form a robust framework applicable to various heavy steel castings. This approach saves significant cost and downtime.
However, prevention remains superior to repair. The occurrence of a casting defect highlights the importance of stringent quality control during the foundry process. Modern techniques like real-time mold monitoring, controlled pouring, and advanced feeding systems can significantly reduce the incidence of shrinkage cavities, cracks, and inclusions. Implementing more rigorous NDT protocols (e.g., radiographic testing for critical sections) at the manufacturing stage is essential for early casting defect detection.
Furthermore, design optimization can mitigate stress concentrations that might exacerbate a latent casting defect under cyclic loading. Finite element analysis (FEA) during the design phase can identify high-risk areas, allowing for design modifications such as smoother fillet radii or strategic rib placement to improve stress distribution and casting soundness.
In conclusion, the battle against casting defect issues is fought on two fronts: proactive prevention through advanced manufacturing and design, and reactive remediation through skilled, theory-informed field repair. The methodology detailed herein provides a reliable arsenal for the latter, ensuring that when a casting defect threatens operational continuity, it can be decisively and permanently addressed where it matters most—on site.