CLADDING TECHNOLOGY SHANXI CO., LTDScrap Criteria for Cracks in Pressure Welds and Cladding Layers
Literature Overview
This topic establishes the most stringent quality control principle in pressure equipment fabrication: the zero-tolerance policy for cracks that persist after two compliant repair attempts, or that exist in critical locations where safe repair is not feasible. This scrap criterion represents the safety bottom line — the point beyond which engineering judgment must yield to the fundamental requirement of pressure boundary integrity. The topic applies to all pressure-bearing welds, weld overlay corrosion-resistant layers, and clad interfaces in bimetal pressure vessels and components.
The Zero-Tolerance Principle for Cracks
Why Cracks Are Different from Other Defects
Cracks occupy a unique position in the taxonomy of welding defects. Unlike porosity, slag inclusions, or undercut — which are volumetric or geometric discontinuities that can be characterized by size and spacing — cracks are planar, stress-concentrating discontinuities that can propagate under cyclic or sustained loading. A crack, even if initially small, represents a potential failure initiation site. The energy release rate at a crack tip is governed by the stress intensity factor K, which increases with crack length. Once a crack exceeds a critical length (determined by the material's fracture toughness K_IC and the applied stress), catastrophic fracture can occur without warning.
This fundamental difference justifies the zero-tolerance approach:
- Porosity can be tolerated within limits because it does not provide a crack-like stress concentration.
- A crack, regardless of size, provides a stress concentration factor that approaches infinity at the tip.
- A crack can propagate under conditions that would be benign for other defect types.
The Two-Repair Rule
The literature establishes a clear rule: if a crack persists after two compliant repair attempts, the component must be scrapped. This rule is based on several engineering considerations:
- Material degradation: Each repair cycle subjects the material to additional thermal cycling, which can degrade the microstructure, reduce toughness, and introduce new residual stresses. After two repairs, the cumulative thermal damage may have rendered the material unsuitable for pressure service.
- Residual stress accumulation: Each weld introduces residual stresses. If the first repair weld introduces stresses that initiate a new crack, and the second repair weld introduces additional stresses, the residual stress field may exceed the material's yield strength in localized regions.
- Loss of confidence: If a crack persists after two properly executed repairs, it suggests that the root cause has not been identified or eliminated. Continuing to repair without understanding the root cause is not engineering — it is gambling with safety.
- Regulatory requirement: ASME VIII Div.1 (UG-99), NB/T 47002, and PED all implicitly or explicitly support the principle that repeated failure of a repair indicates an unresolvable non-conformance.
Critical Location Assessment
Even if a crack has not yet exceeded the two-repair threshold, it may be located in a region where safe repair is not feasible. The following locations are considered critical:
| Critical Location | Reason for Criticality | Repair Feasibility |
|---|---|---|
| Stress concentration zone (hole, notch, fillet) | High local stress amplifies crack driving force | Generally not feasible |
| Seal face or gasket surface | Leak path directly to process fluid | Not feasible without dimensional restoration |
| Clad interface in high-temperature service | Thermal cycling accelerates crack growth | Not feasible |
| Weld metal adjacent to HAZ | Microstructural transition zone is vulnerable | Marginal at best |
| Thin-wall section (< 6 mm) | Full penetration repair risks burn-through | Requires specialized technique |
| Multi-layer weld root | Root repair is inaccessible for inspection | Not feasible |
Decision Framework for Scrap vs. Repair
When a crack is detected, the following decision framework should be applied:
- Is the crack in a critical location? If yes → scrap (unless an alternative repair method such as hot isostatic pressing is available and qualified).
- How many previous repairs have been performed at this location? If two or more → scrap.
- Has the root cause been identified and eliminated? If no → do not proceed with repair until root cause is resolved.
- Is the repair method qualified for this specific material, thickness, and geometry? If no → qualify the method first, then repair.
- Can the repaired area be fully inspected to the original standard? If no → scrap.
- Does the repair introduce unacceptable distortion or dimensional change? If yes → scrap.
Common Crack Types and Their Implications
| Crack Type | Typical Location | Root Cause | Repair Difficulty | Scrap Likelihood |
|---|---|---|---|---|
| Cold crack | HAZ of high-carbon or high-strength steel | Hydrogen embrittlement, high restraint, inadequate preheat | High (requires preheat optimization) | High if persistent |
| Hot crack | Weld metal of austenitic SS or Ni-alloy | Low melting point eutectics, high sulfur/phosphorus | High (requires filler metal change) | High if persistent |
| Reheat crack | HAZ of Cr-Mo steel after PWHT | Stress relief cracking in coarse-grained HAZ | Very high (may require PWHT procedure change) | Very high |
| Stress corrosion crack | Clad interface or overlay layer | Chloride SCC, HIC, or SSC in susceptible microstructure | Not feasible (material incompatibility) | Mandatory scrap |
| Fatigue crack | Pressure weld under cyclic loading | Cyclic stress exceeding fatigue limit | Not feasible (design issue) | Mandatory scrap |
Engineering Practice Case
In the fabrication of a high-pressure hydrogenation reactor with a 316L weld overlay on SA-516 Gr.70, a 12 mm long crack was detected in the overlay layer during final PT inspection. The crack was located in a region of high restraint — adjacent to a machined groove for a bolted flange connection. The first repair involved grinding out the crack, applying a TIG overlay weld, and re-inspecting by PT and UT. The repair was successful, and no indications were found.
However, during the final hydrostatic test, a new crack was detected in the same general area, 20 mm from the first repair. Investigation revealed that the root cause was excessive residual stress in the overlay layer due to the high restraint of the adjacent machined groove. The second repair involved grinding out both the new crack and the first repair weld, applying a modified overlay procedure with reduced heat input and increased preheat (from 100 °C to 200 °C), and re-inspecting by PT, UT, and MT. The second repair was also successful.
The engineering team considered whether a third crack could occur. Given that the root cause (high restraint) had been identified and addressed through preheat optimization, and that the second repair had been successful, the component was accepted. However, the entire overlay area was subjected to an additional UT scan using phased array ultrasonic testing (PAUT) to ensure no subsurface cracks were present. The component was accepted with the additional UT report included in the quality file.
Had a third crack appeared, the component would have been scrapped without further consideration, in accordance with the zero-tolerance principle.
Study Insights and Reflections
The scrap criterion for cracks is the most consequential decision in pressure equipment fabrication. Unlike a repair that can be inspected and accepted, a scrap decision is irreversible — the material is lost, the schedule is delayed, and the cost is incurred. This makes the decision emotionally and commercially difficult. However, the engineering principle is clear: a pressure boundary with a persistent crack is a ticking time bomb. The two-repair rule is not arbitrary; it is a practical implementation of the principle that a component that cannot be reliably repaired should not be used.
I have seen cases where manufacturers were tempted to perform a third repair to avoid the cost of scrapping. In one instance, a third repair was performed on a cracked overlay layer, and the component passed all inspections. Six months later, in service, the overlay layer failed at the repair location, resulting in a plant shutdown and a full replacement of the component. The cost of the third repair and the subsequent failure investigation far exceeded the cost of scrapping the component at the second repair stage. This case reinforces the importance of adhering to the zero-tolerance principle without exception.
The scrap criterion is not a failure of engineering; it is a success of engineering. It represents the culmination of decades of experience, failure analysis, and code development that have established the boundaries of what is safe and what is not. Respecting these boundaries is the responsibility of every engineer involved in pressure equipment fabrication, and it is the foundation upon which the safety of the entire industry rests.