6.1 Cracks and Fracture Modes

6.1 Cracks and Fracture Modes

Quick Answer: A crack is a planar discontinuity with little volume. Hot cracks form near solidification temperatures; cold cracks form after cooling (often hydrogen-related); fatigue cracks grow under cyclic stress. Orientation, opening, and surface access decide whether MT, PT, UT, or RT will find them. Level III judgment starts with mechanism → morphology → method, not with a method name alone.

Crack detection and evaluation cut across Domain 2 (application of NDT methods) and Domain 4 (materials and process technology) on the ASNT NDT Level III Basic exam. You are expected to know why a crack formed, how it is oriented in the part, and which method family can reasonably reveal it under a written procedure.

What Makes a Discontinuity "Crack-Like"

A crack is a separation of material along a surface that is essentially two-dimensional: length and depth dominate; width (opening) is small. That planarity is the single most important NDT property:

• Surface methods (VT, PT, MT) only work if the crack breaks the accessible surface (and for MT, if the material is ferromagnetic and flux leakage can form).
• Ultrasonics respond strongly when the beam strikes the crack face near-normal, producing a specular reflection; glancing incidence or a tightly closed face can reduce amplitude.
• Radiography needs a path-length difference (gap) along the beam. A tight crack with faces pressed together and a beam not aligned with the plane often produces little or no image contrast—hence the classic exam contrast between crack-like (planar) and volumetric flaws for RT.

Do not confuse a crack with a lack of fusion, lamination, or seam. Those may also be planar, but their origin (process incomplete join, mid-plane rolling separation, elongated surface defect from billet) and preferred locations differ. Section 6.2 and 6.3 develop those distinctions.

Hot Cracks

Hot cracks (solidification cracks, hot tears in welds/castings) form while metal is still in or near the mushy temperature range—above or near the solidus—when ductility is very low and liquid films can separate grains.

Typical drivers:

Weld metal solidification cracks often run along the weld centerline or follow columnar grain boundaries. HAZ liquation cracks form in partially melted grain boundaries of the base metal beside the fusion line, especially in alloys susceptible to grain-boundary melting. Hot cracks tend to be intergranular, irregular, and open while hot—though they may close somewhat on cool-down.

NDT implications: many hot cracks break the surface of the weld face or root and are candidates for VT, PT, or MT after cleaning. Subsurface segments or incomplete openings may require UT or, if sufficiently open and aligned, RT.

Cold Cracks and HAZ Cracking

Cold cracks form after the weld or heat-treated region has cooled—sometimes hours later (delayed cracking). In carbon and low-alloy steels, the classic triad is:

1. Hydrogen (from moisture, oil, electrodes, or atmosphere),
2. Susceptible microstructure (hard martensite in HAZ or weld metal),
3. Tensile residual stress (restraint, joint design, thickness).

Underbead and toe cracks in the HAZ are exam favorites. They may be subsurface when underbead, or surface-breaking at the weld toe. Hardness and cooling rate control susceptibility; preheat, low-hydrogen practice, and post-weld heat treatment are process controls—not NDT methods, but Level III procedure writers must know that process risk drives inspection timing (e.g., delayed inspection after hydrogen diffusion).

Cold cracks in steels are often transgranular through hard microstructures (though paths vary). They can be very tight. MT (wet fluorescent on cleaned surfaces) is a workhorse for ferromagnetic welds; PT applies when MT is not suitable or as a complementary surface check; UT addresses subsurface HAZ/weld cracks when surface methods cannot see them.

Fatigue Cracks

Fatigue cracks initiate under cyclic stress, almost always at a stress raiser: weld toe, undercut, sharp corner, corrosion pit, inclusion, thread root, or fretting site. Growth proceeds roughly perpendicular to the maximum principal tensile stress (Mode I dominated in many components).

Fracture-face clues when a part fails:

• Beach marks (clamshell marks) indicate intermittent growth under varying amplitude or environment.
• Ratchet marks suggest multiple initiation sites.
• Final overload zone may be ductile or brittle depending on remaining ligament and material toughness.

In-service NDT programs target initiation-prone locations and known growth planes. Surface methods catch early surface initiation; UT (including phased array and TOFD where procedures allow) sizes or screens deeper growth. Fatigue is a service discontinuity class (see Chapter 5 themes) with crack morphology—so method selection still follows planarity and access.

Brittle vs Ductile Fracture

Level III Basic expects you to read fracture mode language and connect it to inspection urgency and material state.

BCC steels can show a ductile-to-brittle transition with temperature and thickness; FCC alloys generally remain more ductile. A Level III does not perform metallography in the Basic exam, but must know that a brittle-prone material plus crack-like flaw plus tensile stress is a high-consequence combination that drives sensitive surface inspection, fracture-mechanics-based acceptance where codes require it, and conservative method selection.

Orientation: The Exam Lever for MT, PT, and UT

Orientation is how the crack plane sits relative to the free surface, the principal stress, and the interrogating energy.

Magnetic particle testing (MT)

Flux leakage is maximum when the crack is perpendicular to the magnetic field. Longitudinal magnetization finds transverse cracks; circular (head-shot / prod / yoke orientations as applicable) finds longitudinal cracks. Multi-directional or two-shot techniques exist because unknown crack orientation is common. Only surface and slightly subsurface openings in ferromagnetic materials produce useful indications.

Liquid penetrant testing (PT)

PT needs a clean, surface-breaking opening. Orientation to the surface matters only insofar as the crack must intersect the inspected face. Very tight or contaminated openings reduce bleed-out. PT works on non-magnetics where MT cannot.

Ultrasonic testing (UT)

A crack acts as a mirror. 0° (normal-beam) finds planar flaws parallel to the scanning surface (laminations more than typical transverse weld cracks). Angle-beam shear waves are the classic setup for weld cracks oriented roughly vertical to the plate surface: the beam hits the face at a favorable angle and returns corner traps or specular echoes. Wrong angle, wrong index, or a crack tilted out of the plane of incidence drops amplitude—procedure design must match expected orientation from process knowledge.

Radiographic testing (RT)

RT contrast for cracks requires the beam to travel a path with measurable material loss (open gap) along a sufficient fraction of thickness. Beam parallel to the crack plane maximizes chance of imaging; beam perpendicular to a tight closed crack often misses it. Hence exam language: RT is strong for volumetric flaws (porosity, some inclusions, slag) and weak for tight crack-like flaws unless geometry and technique cooperate.

Crack-Like vs Volumetric: RT Decision Frame

When a procedure relies on RT alone for crack-sensitive welds, a Level III should recognize the limitation and specify complementary MT/PT/UT as codes and risk require.

Level III Procedure and Evaluation Mindset

Writing or approving a procedure for crack-sensitive hardware means stating material and process (weld process, PWHT, hydrogen control), defining surfaces and volumes, selecting method(s) matched to expected orientation, controlling surface prep, and tying acceptance to the applicable code—not to personal preference. Evaluation must separate relevant crack indications from non-relevant geometry or process marks (Section 6.4).

Exam trap: equating "we radiographed it" with "there are no cracks." Radiography can clear volumetric quality while missing tight cracks. Another trap: running only longitudinal MT magnetization when toe cracks may be transverse to the weld axis—orientation must be covered by technique.

Master mechanism, morphology, orientation, and method physics together. That combination is what Basic items test when they describe a delayed HAZ crack, a centerline hot crack, or a fatigue starter at an undercut.