The underlying cause
An electronic component and the board it is soldered to have different coefficients of thermal expansion. When the temperature changes, each one tries to expand by a different amount. Because they are joined by the solder, the difference in movement becomes deformation of the joint. This repeats with every cycle, and the solder accumulates fatigue damage.
Why the solder is the weak point
Solder alloy is relatively soft and, even at room temperature, already works at a high fraction of its melting temperature. This means that, in addition to deforming elastically, it flows over time, a phenomenon called creep. The deformation ends in cracks that start in the zone of highest stress and advance cycle after cycle.
What factors make the problem worse
- Large or rigid components, with greater distance to the center and more mismatch.
- Ceramics on organic laminate, because of the large difference in coefficients.
- Low-standoff joints, which absorb less deformation.
- Temperature amplitude and rate of change, as well as dwell at the extremes, which promotes creep.
- Process defects, such as voids, poor wetting or brittle intermetallic layers.
How it shows up
The crack starts at the edge of the joint, usually grows and eventually opens the circuit. Before complete fracture there may be intermittent contact that appears with heat or cold, or with vibration. That is why continuous resistance monitoring during the test detects the onset better than a check at the end.
How it is confirmed
Metallographic cross-sections, microscopy, X-ray inspection for hidden joints and dye-and-pry testing are used. The orientation of the crack and its path help distinguish thermal fatigue from mechanical impact or an initial defect.
What the design changes
Mismatch is reduced with compatible materials, components with compliant leads are used, joints are reinforced with underfill, or pad dimensions are modified. The result is verified with temperature cycling and statistical life analysis.