11.2 Cracks — Causes & Prevention

11.2 Cracks — Causes & Prevention

Quick Answer: Cracks are planar ruptures with sharp tips. They may sit in weld metal, HAZ, or parent metal, and may run longitudinal, transverse, or form crater cracks at stop ends. Under ISO 5817, cracks are typically not permitted at quality levels B, C, and D. Causes link to Chapter 6 metallurgy (hydrogen cold cracking, hot/solidification cracking, lamellar tearing). Prevention is procedure-driven; detection uses VT plus surface and volumetric NDT as specified.

Module WI1.7–1.8 and exam-meta defect topics expect IWI-S candidates to classify crack locations and orientations, know that cracks are generally rejectable under common structural acceptance standards, connect causes to metallurgical mechanisms already studied, and choose appropriate detection methods.

Why Cracks Sit at the Top of Severity

Compared with rounded gas pores, a crack:

• Concentrates stress at a sharp tip.
• Can grow under static, cyclic (fatigue), or environmental loading.
• Often connects surfaces or critical ligaments in pressure and structural joints.
• Is treated as not permitted in ISO 5817 quality levels B/C/D for the usual crack categories listed.

Therefore the inspector’s first duty when a crack-like indication appears is conservative classification and hold—not “watch and wait” without engineering authority.

Location: Weld Metal, HAZ, Parent Metal

Location drives root-cause investigation and repair strategy. Repairing a HAZ hydrogen crack without fixing hydrogen control is not a lasting solution.

Orientation and Morphology Terms

Longitudinal cracks run roughly parallel to the weld axis. Classic examples include solidification centreline cracks in deep, narrow beads and some hydrogen cracks aligned with residual longitudinal stress fields.

Transverse cracks run across the weld axis, often in weld metal or HAZ under longitudinal residual tension. They may appear as a series of short cracks.

Crater cracks form in the final solidification crater when the arc is broken without proper fill or down-slope. They are often star-shaped or short longitudinal remnants in the crater. Many procedures require crater fill, back-step, or controlled termination—and VT specifically checks stop ends.

Radiating / branching cracks and chevron patterns may indicate progressive fracture or solidification patterns; report morphology carefully for metallurgy follow-up.

Microfissures and hot-crack networks may be fine; still classify as crack-type imperfections when identified.

Hot Versus Cold (Recap for Imperfection Classification)

Use Chapter 6 mechanisms when writing NCRs:

1. Hot / solidification cracking — during the last stages of solidification; associated with wide freezing range, impurities (S, P), deep narrow beads, high dilution/restraint. Prevention: chemistry control, bead shape, parameter and sequence control.
2. Hydrogen cold cracking (HICC) — after cooling; needs hydrogen + hard microstructure + tensile stress + low temperature. Prevention: low-H consumables, dry storage, cleanliness, preheat/interpass, reduced restraint, delayed NDT when required.
3. Lamellar tearing — parent plate through-thickness weakness under high strain (T-joints, stiff node welds). Prevention: material selection (through-thickness properties), joint design, buttering, sequence.
4. Liquation / reheat — alloy-specific; recognise as specialist metallurgical issues for stainless, Ni alloys, and some creep steels (Chapters 7–8).

The ISO 6520 name may simply be “crack (longitudinal) in HAZ”; the cause narrative for the quality system uses the metallurgical category.

Acceptance: ISO 5817 Practice for Cracks

For steel fusion welds assessed to ISO 5817:

• Cracks (including microcracks in the usual tables) are not permitted at B, C, and D for the categories covered.
• Do not apply “pore diameter logic” to cracks—there is no “small enough crack” under those tables for ordinary acceptance.
• Product standards or client specs may be equal or stricter; they rarely make structural cracks acceptable without formal engineering critical assessment (ECA), which is outside routine shop acceptance.

IWI-S rule of thumb for exams and practice: treat cracks as zero-tolerance under ISO 5817 B/C/D unless a documented ECA/fitness-for-purpose route is formally invoked by the contract—and that is not casual field discretion.

Causes Checklist (Shop Floor)

Consumable and hydrogen control

• Damp basic electrodes, wet flux, contaminated wire.
• Cellulosic electrodes on hardenable, restrained joints without justified procedure.
• Oil, paint, moisture, and primers left in the weld zone.

Parameters and technique

• Excessive depth-to-width ratio promoting centreline solidification cracks.
• High dilution into sulphur-bearing parent metal.
• Abrupt arc breaks creating crater cracks.
• Excessive heat input or too-low heat input depending on mechanism (hot crack vs hard HAZ cold crack).

Design and restraint

• High restraint fixtures, thick sections, rigid nodes.
• Poor joint design promoting through-thickness strain (lamellar risk).
• Repair welding in stiff structures without revised controls.

Material

• High carbon equivalent / hardenable steels without preheat.
• Segregation-prone compositions; poor through-thickness ductility plate.
• Incorrect consumable matching for solidification cracking susceptibility.

Prevention Toolbox (Inspector Verification Focus)

Detection Methods Overview

Visual testing (VT) — first line for surface-breaking cracks, crater cracks, toe cracks, and arc-strike cracks. Needs good lighting, access, and sometimes light grinding for profile assessment—without removing evidence prematurely when investigation is needed.

Liquid penetrant testing (PT) — surface-breaking cracks on non-magnetic and magnetic materials; excellent for stainless and non-ferrous. Cleanliness critical.

Magnetic particle testing (MT) — surface and slightly subsurface cracks in ferromagnetic materials; highly effective at toes and weld faces.

Radiographic testing (RT) — can show some cracks when oriented favourably to the beam, but tight, unfavourably oriented cracks may be missed. RT is stronger for volumetric rounded indications than for all planar cracks.

Ultrasonic testing (UT) — generally more sensitive to many planar crack-like flaws when properly applied and when geometry allows. Preferred volumetric method for many crack-sensitive butt welds in thick sections (method selection in Chapters 13–14).

Destructive tests — fracture faces, bend tests, and macro sections reveal cracks in procedure qualification and production cut-outs; not a production 100% tool.

Timing: for hydrogen cracking risk, respect delay before final NDT when specified—early clean NDT can be a false comfort.

Reporting and Disposition

Report:

1. Type — crack (and morphology if clear: longitudinal, transverse, crater).
2. Location — weld metal / HAZ / parent; surface / embedded; toe / root / centreline.
3. Dimensions — visible length; depth if known from NDT.
4. Method of detection and reference standard.
5. Acceptance statement — typically non-conforming to ISO 5817 level X.
6. Hold for engineering disposition / approved repair WPS.

Repairs must follow an approved repair procedure (often a repair WPS), with renewed NDT and, where applicable, renewed hydrogen controls and heat treatment.

Link Forward

Section 11.3 contrasts cracks with porosity, which is often limited rather than zero-tolerance. Section 11.4 covers other planar rejectables—lack of fusion and lack of penetration. Chapters 13–14 expand NDT capabilities and limitations. When cracks recur, reopen the Chapter 6 four-condition or solidification checklist rather than blaming only the last welder by default.