A solder joint is a structural member made of a material that creeps at room temperature. Every thermal cycle strains it, because the component, the joint and the board expand by different amounts, and that strain accumulates. Joints therefore have a fatigue life rather than an indefinite one — which means the central question after a field failure is not whether the joint wore out but whether it wore out sooner than the design was entitled to expect.
The mechanism
Mismatch in thermal expansion between a component and the board imposes a displacement on the joint each time the assembly changes temperature. Solder accommodates it by deforming inelastically, and the accumulated damage initiates a crack, usually at a corner joint where displacement is greatest, and usually at an interface rather than mid-bulk.
Crack growth is progressive, and its intermediate stage is the intermittent behaviour that produces field returns passing every bench test. A crack that closes at room temperature and opens at temperature or under vibration is electrically continuous exactly when it is being tested.
Intermetallics and embrittlement
A solder joint is not a homogeneous lump. At each interface an intermetallic compound layer forms during reflow and continues to grow, diffusion-driven, throughout service — faster at elevated temperature. Those layers are hard and brittle relative to the bulk solder.
A thin intermetallic layer is necessary for a sound metallurgical bond. An overgrown one shifts the crack path into brittle material and lowers the joint's tolerance for both fatigue and mechanical shock. Kirkendall voiding at the copper interface, where diffusion rates differ across the boundary, is a related and well-documented degradation that appears as a line of voids weakening the interface.
Because growth is time-at-temperature dependent, the layer thickness is a rough record of thermal history, which is occasionally useful for testing a claim about how equipment was actually operated.
Lead-free changed the margins
The transition to tin-silver-copper alloys altered nearly every relevant property: higher melting point, different creep behaviour, different intermetallic growth, and a smaller process window. It also raised reflow temperatures, which stresses components and boards more during assembly.
None of this makes lead-free joints inadequate — they are the industry standard and perform well within their envelope. It does mean qualification data and design margins derived from tin-lead experience do not transfer, and a product whose thermal design was inherited unchanged across that transition is worth examining on exactly that point.
Mixed assemblies
Where a lead-free ball meets tin-lead paste, or the reverse, the resulting joint has a composition and melting behaviour matching neither, and a profile adequate for one may not fully coalesce the other. These joints are a recognised reliability concern rather than a theoretical one.
Mixing usually enters through the supply chain or through rework, not through a design decision, which makes it a documentary question about what was actually fitted and what was actually done during any repair.
Was it wear-out or a defect?
This is the question the whole analysis serves, and it is answerable. Fatigue life can be estimated from the actual thermal cycle amplitude, frequency and joint geometry, and compared against the service the product actually saw and the life the design claimed.
A joint that failed after far fewer cycles than its geometry and materials predict points at something specific: a manufacturing deficiency such as voiding or incomplete wetting, a design deficiency such as an inadequate pad or an unmitigated expansion mismatch, or service conditions harsher than specified. A joint that failed at roughly its predicted life points at the life prediction itself, and at whether the product was represented as lasting longer.
Establishing the real duty cycle
The design assumption and the service reality frequently diverge, and the divergence is often the finding. Equipment specified for two cycles a day and operated through twenty, or installed somewhere with a daily ambient swing the specification never contemplated, will reach its fatigue life early without anything being defective.
Reconstructing that means data rather than assertion: operating logs, controller and telemetry records, installed-location temperature history, and the duty the equipment was actually put to.
Tin whiskers, briefly
One lead-free side effect deserves separate mention because it is a different mechanism with a similar outcome. Pure tin finishes can grow whiskers — single crystals a few microns across and long enough to bridge adjacent conductors, driven by internal compressive stress rather than by any voltage bias.
They are not fatigue and not electrochemical migration, and they are distinguishable under a microscope. They matter here because both the mitigations and the responsibility differ: whisker risk is controlled by finish selection and by a nickel underlayer, which is a component and specification question rather than a thermal design one.
Examination sequence
Imaging before cutting, as always. Then cross-sectioning through the joints most likely to have failed first — outermost, largest expansion mismatch, highest local temperature — with metallography to characterise crack path, intermetallic thickness and any interface voiding.
The crack path is the most informative single observation. A crack through bulk solder reads as fatigue; one following a brittle intermetallic layer reads as embrittlement; one through a pad or laminate reads as mechanical overload rather than thermal cycling at all.
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.