Some bonded joints fail in a way that is difficult to reconcile with their own test history: they passed every acceptance criterion when they were made, carried design loads without incident for years, and then released. Nothing about the initial qualification was wrong, and nothing obvious happened at the moment of failure. This pattern points away from initial strength and toward durability, which is a distinct property that initial lap-shear testing does not measure and cannot predict. Understanding the difference is usually what separates a defensible explanation from an unsatisfying one.
Strength and durability are different properties
A lap-shear test performed on a freshly made joint measures how much load the bond carries at that moment. It says nothing about how that capacity changes after years of exposure to moisture, temperature cycling and sustained stress. A joint can be simultaneously strong when new and poorly durable, and the combination is common enough that qualification programmes for structural bonding typically require environmental conditioning precisely because the unconditioned number is known to be unrepresentative.
How moisture reaches an interface that looks sealed
Water does not need a path around a bondline to reach it. It diffuses through the adhesive itself and, in a polymer-matrix composite, through the adherend as well. The process is slow and driven by concentration, so a joint can operate for a long period before enough moisture accumulates at the interface to matter. Edges reach saturation first, which is why progressive degradation so often presents as interfacial release working inward from the perimeter.
What moisture does once it arrives
Two mechanisms operate. Water plasticises the adhesive, lowering its glass transition and softening it in a way that is substantially reversible on drying. More consequentially, water hydrolyses the chemical bonds at the interface itself — the oxide and coupling chemistry that surface treatment created — and that degradation is not reversible. The interfacial chemistry that made the joint durable is consumed, and once it is gone, drying the joint out does not restore it.
The signature is a shift in failure mode over time
This mechanism leaves a characteristic trace. A joint that failed cohesively when it was new and fails interfacially after years in service has had its interface degraded while the bulk adhesive remained substantially intact. That shift is the most direct available evidence of progressive interfacial attack, and it is why the original qualification test records — including how those coupons failed, not merely what load they carried — are worth recovering.
Temperature cycling contributes separately
Adherends bonded to one another rarely have matched thermal expansion, and every temperature excursion therefore imposes a shear stress at the bondline. Individually these stresses are small. Accumulated over thousands of cycles, and combined with an interface being progressively weakened by moisture, they drive damage that neither mechanism would produce alone. Establishing the actual thermal history a structure experienced is often more useful than the nominal environmental specification.
Separating ageing from a latent defect
The competing explanation is that the joint was defective from the outset and simply took years to manifest. The two are distinguishable. Progressive environmental degradation should correlate with exposure — worse at edges, worse on the weather side, worse on units with more service time or harsher deployment. A latent manufacturing defect should correlate with production variables instead — a date range, a shift, a tool, a material lot. Mapping failures against both sets of variables is what discriminates between them, and it requires a population rather than a single joint.
Archive samples are what make the comparison possible
The strongest form of this analysis compares the failed joint against material from the same production run that never entered service. Retained coupons, unused assemblies or archived parts establish what the bond was when it was made, so that everything measured on the failed joint can be expressed as a change rather than an absolute. Where no archive exists, low-exposure examples from the same population — an interior unit, a sheltered location, a low-hours asset — serve as an imperfect but usable substitute.
What to preserve and what the records should cover
Both fracture faces uncleaned, the surrounding structure sufficient to establish how the joint was loaded, and any exemplar or archive material from the same lot. On the documentary side, the qualification test report including failure modes, the environmental exposure the design assumed, and the actual service history of the specific asset — location, duty, and any events that would have driven unusual thermal or moisture exposure.
Why this distinction carries weight
Whether a bond failed from a defect present at manufacture or from environmental degradation over a service life bears directly on where responsibility sits and on whether the finding extends to a fleet. It also determines whether the appropriate response is an inspection programme, a design change or a material substitution — which is why the durability question is worth separating from the strength question rather than treating both as one.
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.