Why welds are inspected and glued joints are not
Why one joining method carries an inspection regime and the other does not.
Welds are inspected and glued joints generally are not because a weld's most dangerous defects tend to hide entirely below the surface, invisible on a joint that can look flawless from outside, while a well-designed adhesive joint's failure modes more often leave evidence on the surfaces available for a simple visual or mechanical check. The two methods carry different risks of a hidden flaw passing straight through into a finished product.
Defects sealed inside a fast melt
A weld can contain porosity, incomplete fusion or a crack sitting entirely within the depth of the material. These defects form during a brief, fast-moving melt-and-solidify process, and a visual check of the finished surface has essentially no chance of catching them, since the outer face can look smooth and continuous while the interior tells a different story. Gas that had no time to escape before the pool froze, a shielding gas flow disturbed for a moment by a stray draught, or two adjacent passes that never fused beneath a tidy cap can all sit sealed inside a weld that would satisfy any inspector working from appearance alone.
Adhesive bonding can hide flaws too, such as a poorly wetted surface or a void in the bond line. The process is slower, cooler and more forgiving, though, and many of its common failures (poor surface preparation, a thin or uneven bead, a joint that clearly never closed) leave visible evidence on the surfaces still open to inspection after the joint is made.
Testing the ice before walking on it
Judging whether a frozen lake is safe from its surface alone is notoriously unreliable. Ice that looks uniformly thick from above can conceal thin patches, air pockets or weak regions within its depth, so anyone taking the question seriously tests the ice directly, drilling or striking it at intervals. A finished weld presents almost the same problem, an unbroken, solid-looking surface over whatever porosity or incomplete fusion formed during a rapid melt-and-freeze cycle, and the industry treats its appearance with the same suspicion.
Drilling at a few points builds confidence only in the spots actually tested, leaving the rest of the lake unknown. Weld inspection has the same limitation, so for a critical weld the method has to survey the whole length of the joint, since sampling a convenient handful of points leaves the rest to hope.
Seeing below the surface, and the paperwork that follows
Dedicated non-destructive testing methods, among them ultrasonic, radiographic and dye penetrant, exist to reveal what a visual inspection cannot, and structural welds on demanding applications are routinely required to pass one or more of them. Each method sees some flaws well and others poorly. Ultrasonic inspection sends sound through the weld's depth and finds a buried void well but a fine surface-breaking crack poorly, while dye penetrant finds that surface crack cleanly and sees nothing beneath the surface it is applied to. Demanding applications therefore often specify both, so that neither has to stand in for the other.
Because inspection has to reach below the visible surface, a welding procedure for anything structurally important specifies how the weld will be checked as well as how it is made. Skipping that check carries a different weight from skipping an equivalent check on many adhesive joints, since a weld's hidden defects have no other realistic route to being caught before the joint is loaded in service. This asymmetry explains why welding procedures, welder qualifications and inspection regimes are far more heavily codified than their equivalents for structural adhesive bonding. A team new to welding sometimes reads the heavier paperwork as bureaucracy, when it is a proportionate response to how little a clean-looking weld surface promises on its own.
Bonded joints that are inspected just as hard
Adhesive bonding is not exempt from rigorous inspection everywhere it is used. Safety-critical bonded joints in aerospace and similarly demanding fields are checked with their own non-destructive methods for the same reason welds are, because a hidden void or a poorly bonded region can sit beneath a convincing surface in an adhesive joint too when the application is demanding enough. The difference described here is one of typical practice in ordinary structural work. The underlying principle applies to both methods equally: the seriousness of an inspection regime should track what a hidden defect in that particular joint would cost if it failed, whichever joining method was used. The next article turns to a particular weld geometry, the fillet weld, whose appearance is a similarly poor guide to how much load it carries.