Fasteners and joints

In brief

A wooden chair can be built entirely from one continuous piece of bent wood, and it would never come apart at a joint, because it has none; it can also fail catastrophically the moment one glued leg joint lets go, precisely because that joint was the one place the load path had to cross from leg into seat. Almost nothing useful is built from a single continuous piece of material. Parts have to be joined, and every method of joining, a bolt, a weld, a rivet, a bead of adhesive, a press fit, a snap fit, is a specific answer to the same underlying question: how does the load, which was flowing continuously through one member, actually get from that member into the next one across a seam that is not solid material at all. Choosing among these methods is not a matter of which is strongest in the abstract; it is a matter of which one can carry the particular load the joint will see, be manufactured correctly, be checked afterward to confirm it was made correctly, and, if something goes wrong later, be taken apart and fixed.

The full treatment

First look: why a joint is where structures actually fail

Structural failures disproportionately start at joints rather than in the middle of an otherwise uniform member, for a simple reason: a joint is precisely where the smooth, continuous cross-section of a member is interrupted, by a drilled hole, a change in thickness, a bead of weld metal with different properties than the parent metal, or a thin layer of adhesive instead of solid material. Any of these interruptions concentrates stress locally even before considering how well the joint was actually made, and manufacturing quality varies far more at a joint than through the bulk of a plain member. A joint is where a load path is asked to survive a seam, and seams are where surprises live.

Bolts and screws: clamping force and reversible tension

A bolted joint most often works by clamping two parts together with enough tension in the bolt shank, the bolt preload, that friction between the clamped surfaces resists the external load before the bolt itself ever has to carry that load directly in shear. This is subtler than it looks: a properly preloaded joint is designed so the external load mostly reduces the clamping force slightly rather than pulling directly on the bolt, which is why torque specifications matter so much, an under-torqued bolt loses its grip and lets the joined parts slip, while an over-torqued one can yield or strip its threads before ever seeing service load. Bolts and screws are, in exchange for this preload requirement, the one joining method designed from the start to be taken apart and put back together, the default choice wherever maintenance or replacement is expected.

Welding: fusing material into continuity, at a cost

Welding melts the parent material at the joint, along with added filler material, so the two pieces solidify together as one continuous piece rather than two pieces held by a third element like a bolt or rivet. Done well, a weld can approach the strength of the base material itself, since there is, in principle, no discontinuity left. The cost is that welding heat alters the material's structure near the joint, in a region called the heat-affected zone, often weaker or more brittle than either the base or weld metal, and a weld cannot be undone without destroying it, so a welded joint trades reversibility for strength and continuity. Weld quality also cannot be judged by eye; internal defects like porosity or incomplete fusion require dedicated inspection methods, which is why welded joints in critical structures carry inspection requirements bolted joints of similar load do not.

Rivets and adhesives: shear-carrying pins and load-spreading films

A rivet works much like a bolt without preload: a solid pin, installed through aligned holes and deformed to fill the hole and form a head on each side, primarily carries load in shear across its shank as the two joined plates try to slide past each other. Riveting was, for over a century, the dominant method of joining structural steel and aircraft aluminum before welding and high-strength bolting largely displaced it. Adhesive joints work differently, spreading load as shear across the entire bonded area rather than concentrating it at a few discrete points the way bolts or rivets do, avoiding the stress concentration a drilled hole creates, but adhesive strength is far more sensitive to surface preparation and moisture than a mechanical fastener, and a bonded joint is hard to inspect for hidden defects once cured.

Press fits and snap fits: joining by geometry and elastic strain alone

A press fit joins two parts using only a deliberate, slight oversize of one part relative to the hole or socket it is forced into, so the elastic squeeze of the surrounding material grips the inserted part by friction, with no separate fastener at all. A snap fit uses a shaped, flexible feature, a small hook or bead, that elastically deflects during assembly and springs back once past a matching undercut, locking the parts by geometry rather than friction or a foreign fastener. Both methods are prized in high-volume manufacturing because they need no separate part to install and often no tool, but both depend entirely on the material's elastic properties staying within range: a press fit that is too tight can crack a brittle housing on assembly, and a snap fit made from a material that creeps or fatigues can quietly lose its grip years later, failing silently rather than announcing an obviously loose fastener.

Lineage

Riveting is ancient in metalwork and became the dominant method for large iron and steel structures, from ships to bridges to early aircraft, prized for its reliability under vibration and its visually inspectable head. Arc and gas welding matured through the early twentieth century and progressively displaced riveting wherever a continuous, lighter joint was worth the loss of reversibility and the added inspection burden. High-strength bolting developed in parallel as the reversible alternative for structures needing disassembly or field assembly. Adhesive bonding, press fits, and snap fits grew from niche uses into mainstream methods as polymer materials and high-volume manufacturing rewarded joints needing no separate fastener part and no assembly tool.

The strongest case for it

Treating joint selection as a distinct engineering decision, rather than an afterthought once the parts themselves are designed, works because each method's strengths map cleanly onto a recurring need: bolts where disassembly matters, welds where continuous strength matters more than reversibility, rivets where vibration resistance and visual inspectability matter, adhesives where distributing load over a wide area matters, and press or snap fits where eliminating a separate part matters. This framework holds across industries, from aerospace to consumer electronics to civil infrastructure, because the underlying trade-offs, strength versus reversibility versus inspectability versus cost, are the same in every one of them.

The strongest case against it

No joining method escapes its own honest limitations, and conflating "strong" with "correct choice" is the central mistake this entry warns against. A weld can be stronger than a bolted joint yet be the wrong choice precisely because it cannot be inspected without destructive testing, and cannot be undone if a design later needs to change. A press fit or snap fit can work perfectly on assembly day and fail years later through slow material creep a one-time inspection would never catch. A bolted joint depends entirely on correct preload, and a bolt torqued incorrectly can be structurally weaker than a much simpler adhesive joint, despite bolts having the higher theoretical strength on a data sheet. A common misconception is assuming a joint's rated strength, taken from a single fastener's data sheet, transfers directly to the assembled joint, when the true governing factor is often the surrounding material, the hole it is drilled or pressed into, rather than the fastener itself.

Where it stands now

The broad classification of joining methods and their trade-offs between strength, reversibility, inspectability, and cost is settled, standard engineering knowledge across mechanical, civil, and aerospace curricula. Active development continues in adhesive chemistry, which keeps narrowing the strength and durability gap with mechanical fasteners, and in structural health monitoring methods that aim to give bonded, welded, and press-fit joints some of the ongoing inspectability a visible bolt or rivet head has always offered for free.

Test yourself

You are choosing how to join an aluminum bracket to a steel frame in a piece of outdoor equipment that must be field-serviceable, meaning a technician with basic tools must be able to remove and replace the bracket if it is damaged, and the joint will be visually inspected once a year but never disassembled for detailed testing. Rule out at least two of the joining methods discussed here, stating specifically which requirement each one fails, and select the method you judge correct, explaining how your choice satisfies field serviceability, adequate strength, and realistic inspectability all at once, not just the strongest possible connection.