6.4 Fracture Types Related to Weld Defects
Key Takeaways
- WT2.6 requires a working overview of ductile versus brittle fracture so inspectors can communicate risk when defects interact with material behaviour and loading
- Cracks, lack of fusion, sharp undercut, and lack of penetration act as stress raisers that elevate local stress and can initiate fracture or fatigue
- Fatigue cracks commonly initiate at weld toes and other geometric or metallurgical imperfections under cyclic load
- The IWI-S links observed imperfections to fitness risk and acceptance rules (e.g. ISO 5817) without performing full engineering critical assessment (ECA)—that depth is IWI-C / specialist territory
- Reporting must describe defect type, location, size, and orientation so engineering can decide accept, repair, or further analysis
6.4 Fracture Types Related to Weld Defects
Quick Answer: Ductile fracture absorbs energy with plastic deformation; brittle fracture can run rapidly with little plasticity. Weld imperfections—cracks, lack of fusion (LOF), sharp undercut, lack of penetration—act as stress raisers. Under cyclic load, fatigue often starts at weld toes. IWI-S links defects to acceptance and risk; full ECA is beyond Standard level.
Module WT2.6 sits between cracking metallurgy (WT2.5) and the later design/fatigue topics (WT3). You need enough fracture vocabulary to interpret why a code treats a crack as zero-tolerance while a smooth undercut has a height limit—and to write reports that engineering can use.
Ductile vs Brittle Fracture — Inspector Overview
Ductile fracture
- Preceded by noticeable plastic deformation (necking, shear lips, dimpled microvoid coalescence on the fracture surface in lab terms)
- Absorbs relatively more energy; failure is often less sudden in the structural sense
- Favoured by tough microstructures, higher temperature (above transition for ferritic steels), smooth geometry, and low strain rate
Brittle fracture
- Little macroscopic plastic deformation; fracture surfaces may appear flat/crystalline (cleavage) in ferritic steels
- Can propagate rapidly once initiated if the driving force and toughness are unfavourable
- Favoured by low toughness (e.g. coarse grain, high hardness, low temperature below the ductile-to-brittle transition), high constraint, high strain rate, and sharp notches
For inspectors, the practical message is: the same defect size is more dangerous when the material is brittle (cold service, hard HAZ, low-toughness plate) or when the defect is sharp and highly constrained. That is why toughness testing, PWHT, and temperature limits appear in fabrication standards—and why cracks are almost never “blended and ignored.”
You are not expected at IWI-S to compute KIc or run finite-element ECA. You are expected to know that toughness + stress + defect size form the fitness triangle, and that removing sharpness (repair to smooth contour within acceptance) and avoiding hard, crack-prone HAZ microstructures reduces brittle risk.
Stress Raisers: How Defects Concentrate Stress
A stress raiser (stress concentrator) elevates local stress above the nominal far-field stress. Sharp geometric discontinuities produce high stress concentration factors. Weld-related examples:
| Imperfection | Why it concentrates stress |
|---|---|
| Crack (any origin) | Extremely sharp tip; highest severity; often rejectable under ISO 5817 and most structural codes |
| Lack of fusion (LOF) | Planar, crack-like interface between weld metal and parent or between passes |
| Lack of penetration | Unfused root ligament acts as a built-in notch, often along the joint centreline |
| Sharp undercut | Notch at the weld toe in the HAZ/parent surface—critical for fatigue and for brittle initiation |
| Slag lines / planar inclusions | Elongated planar reflectors that behave like partial cracks under tension |
| Misalignment / angular distortion | Raises secondary bending stress at the joint |
| Excessive convexity / poor toe blend | Geometric concentration even without a “defect” in the ISO 6520 crack sense |
Planar defects oriented perpendicular to principal tensile stress are more critical than rounded porosity of similar length. That is why acceptance standards treat cracks and LOF more severely than small rounded pores.
Cracks Already Present in the Weldment
Sections 6.1–6.3 covered how cracks form. From a fracture viewpoint, an existing crack is already a critical starter:
- Under static load, it may extend by ductile tearing or brittle pop-in depending on toughness and constraint
- Under residual stress alone, some cracks remain dormant; others grow if hydrogen or corrosion assists
- Under service load, residual stress + applied stress interact
Inspector action: any crack-like indication is typically non-conforming until engineering dispositions it. Do not “dress over” a crack without a qualified repair procedure and re-inspection.
Lack of Fusion and Lack of Penetration as Crack Analogues
Lack of fusion is metallurgically not a solidified crack, but mechanically it is a planar discontinuity with little or no bond. Loaded in tension across the plane, it behaves like a crack of comparable size and orientation.
Lack of penetration leaves an unfused root; in single-sided butts this is a long notch along the joint. Many applications forbid incomplete penetration when full penetration is specified. Even when partial penetration is designed, the root notch must be accounted for in design—inspectors verify that the as-built joint matches the design intent and WPS (full vs partial penetration).
Undercut and Toe Geometry
Undercut is a groove melted into the base metal at the toe and left unfilled. Sharp undercut:
- Reduces section thickness locally
- Creates a notch in a region that already has residual tensile stress and HAZ microstructure
- Is a preferred fatigue initiation site under cyclic tensile or bending stress
Codes set depth and length limits (quality levels B/C/D in ISO 5817 terms). Smooth, shallow undercut within limits may be acceptable; sharp, deep undercut is not. Dressing and re-welding follow the repair WPS when required—not arbitrary grinding into thin wall.
Fatigue Initiation at Toes and Imperfections
Fatigue is progressive crack growth under cyclic stress below the static tensile strength. Welded joints are fatigue-sensitive because of:
- Toe geometry — sharp transition from weld reinforcement to plate
- Residual tensile stress near the toe
- HAZ microstructure and possible undercut or cold laps
- Embedded imperfections that act as starters if surface-connected or near-surface
Typical story: a structure sees vibration, traffic, pressure cycles, or wave loading; a crack starts at a weld toe or at a surface defect and grows until leakage, fracture, or NDT detection.
Inspector contributions to fatigue performance (within Standard scope):
- Enforce toe profile and undercut acceptance during VT
- Ensure specified toe grinding, TIG dressing, or peening is done only when designed and qualified—and done correctly
- Prevent crack-like defects from entering service
- Report misalignment that adds secondary bending
Full fatigue design (S-N curves, detail categories) appears more deeply in WT3 and design chapters; here the link is defect → initiation site.
IWI-S Role vs ECA / IWI-C Depth
| Task | IWI-S expectation | Beyond Standard (specialist / IWI-C) |
|---|---|---|
| Identify and classify imperfections | Yes (ISO 6520 language) | Advanced fractography |
| Apply code acceptance (e.g. ISO 5817 levels) | Yes, per ITP/spec | Deriving new acceptance by ECA |
| Link defect type to stress-raiser / fatigue risk in reports | Yes, qualitative | Quantitative fracture mechanics (K, CTOD, R6, BS 7910-type assessments) |
| Approve fitness-for-service of a crack left in service | No—escalate | ECA by competent engineering |
| Stop work / quarantine non-conforming welds | Yes, within authority | — |
Engineering Critical Assessment (ECA) uses fracture mechanics to decide whether a known flaw is acceptable for a given stress and toughness. IWI-S must know that ECA exists and when to call for it, not perform it as a routine inspection decision. Leaving a crack because “it looks small” without acceptance criteria or ECA is outside Standard competence.
Reporting That Supports Fitness Decisions
A useful non-conformance report includes:
- Type — crack, LOF, undercut, lack of penetration (ISO 6520 group where used)
- Location — weld metal / HAZ / parent; toe / root / centreline; joint ID
- Size and orientation — length, height/depth, surface or buried
- NDT method and procedure that found it
- Related process context — consumable, preheat, pass where relevant
- Immediate action — hold point, quarantine, temporary support if structural
That package lets welding engineers choose repair, cut-out, or formal assessment.
Putting Chapter 6 Together
- 6.1 Hydrogen cold cracking — delayed, four conditions, low-H and thermal control
- 6.2 Hot/solidification and liquation — high temperature, chemistry and bead shape
- 6.3 Lamellar tearing — through-thickness plate weakness under T/corner strain
- 6.4 Fracture vocabulary — ductile/brittle, stress raisers, fatigue toes, inspector vs ECA
Together these topics equip the IWI-S to recognise crack mechanisms, prevent them through procedure verification, and escalate defect significance without overstepping into specialist fracture analysis.
Which statement best contrasts ductile and brittle fracture for an IWI-S level overview?
Why are cracks, lack of fusion, sharp undercut, and lack of penetration treated as high-severity weld imperfections from a fracture perspective?
Where do fatigue cracks in welded fabrications commonly initiate under cyclic loading?
What is the correct IWI-S boundary regarding engineering critical assessment (ECA) of weld defects?