Ductility and brittleness get discussed as though they were stamped into a material: this alloy is tough, that one is brittle. Fracture evidence does not support that framing. The same plate of structural steel, from the same heat, can tear apart with visible stretching in one circumstance and shatter with almost no deformation in another. Temperature, loading rate, section thickness and notch sharpness decide which occurs. The fracture surface records the outcome, and reading it is where a structural failure examination begins.
The same material, two behaviours
Ductile fracture is a process of plastic work. Voids nucleate at inclusions and second-phase particles, grow as the surrounding metal stretches, and coalesce until the remaining section can no longer carry the load. It absorbs a great deal of energy.
Brittle fracture splits the metal along crystallographic planes or along grain boundaries, absorbs very little energy, and runs fast enough to leave no opportunity for warning. The two are separated by orders of magnitude in energy, which is why the distinction carries weight.
What the surface records at arm's length
Macroscopically, a ductile fracture is dull, fibrous and grey, with the section visibly thinned where it drew down before parting. A brittle fracture is flatter, brighter and granular, and the halves fit back together with the original dimensions essentially intact.
Whether the part still measures what it did before it broke is among the quickest discriminators available in the field. ASTM E1823 supplies the terminology for the rest of the work, and reports using terms like cleavage loosely invite avoidable challenge.
Side geometry and the shear lip
Look at the edges of the fracture, not only its face. A ductile fracture in a flat section usually shows a flat central region bordered by slanted shear lips where the free surface was able to deform. As conditions shift toward brittle behaviour those lips shrink and finally disappear.
The proportion of the surface that broke by shear rather than cleavage is a recognised measure of that shift, and percent shear is among the quantities recorded in Charpy impact testing under ASTM E23.
Thickness, constraint and why gauge matters
Thick sections behave more brittlely than thin ones of identical material. Material at a crack tip in a thick section is restrained from contracting sideways, producing a triaxial stress state that suppresses the plastic flow ductile fracture depends on. Thin sections deform freely, and tend to tear.
This is why a detail that performed for years in one gauge can behave differently in another, and why testing at the wrong thickness misleads. ASTM E399 imposes explicit specimen size requirements so a measured toughness reflects constrained conditions rather than the specimen's own geometry.
Rate, notches and where stress concentrates
Loading rate matters for the same reason temperature does: plastic deformation needs time. A load delivered as an impact can produce cleavage in material that would have torn under the same load applied slowly.
Sharp notches, machining marks, weld toes, keyways and abrupt section changes raise local stress and add constraint at once, a combination that pushes material toward brittle behaviour. Cold, fast and notched is a different service condition from warm, slow and smooth, though the drawing is unchanged.
Finding the origin and the direction of travel
Brittle fractures carry their own navigation marks. Chevrons form V-shapes pointing back toward the origin, river patterns on cleavage facets converge in the propagation direction, and radial marks fan outward from the initiation site. Ductile fractures show direction in the fibrous texture.
Locating the origin turns a description of behaviour into an investigation. What sat at that point, whether an inclusion, porosity, a forging lap or a sharp corner, explains why the fracture began there rather than anywhere else on a nominally identical part.
What the electron microscope settles
A stereo microscope establishes the macro picture; scanning electron microscopy settles the mechanism. Ductile fracture shows dimples, the cross-sections of coalesced voids, often with the originating particle still lodged in the bottom of one. Cleavage shows flat facets carrying river patterns.
Intergranular fracture shows the rounded three-dimensional form of individual grains, as though the metal came apart like dry masonry. That is a distinct finding rather than a variant of cleavage, and points toward grain-boundary condition rather than a simple temperature or rate effect.
What the mechanism does not establish
A brittle fracture surface is not evidence that the material was defective. It is evidence that the material behaved brittlely under the conditions present, which may reflect a service temperature nobody specified, an impact nobody anticipated, a thickness the design assumptions did not cover, or a metallurgical problem.
The converse deserves equal emphasis. A ductile fracture does not establish that everything was in order. It shows only that ductility was available and that the section separated by overload, leaving open the question of why the load exceeded what the part could carry.
Fatigue, exclusion and preservation
Progressive cracking under cyclic loading is a separate subject, treated separately. It matters here because many examinations begin as one question and become the other when a progressively grown region appears within the surface, so excluding it is a step in the work, not an assumption.
All of this depends on the surface surviving. Mating faces pressed back together, wire-brushed, handled bare-handed or left to corrode lose the features the analysis reads. The pieces should be kept apart, dry and unprocessed until a laboratory has them.
This article is general technical orientation, not a failure analysis, an engineering opinion, or advice on any specific matter. Determining the cause of a particular incident requires hands-on examination by a credentialed expert.