Ask what an exhibit must survive and most people answer with show days: lights, crowds, two weeks of handling a year. The real number is worse and different. A four-show-a-year exhibit gets assembled and disassembled dozens of times across its service life — by labor crews who have never seen it before, at 7 a.m., on concrete, against a clock. Connection cycles, not show days, are the duty spec. The exhibits that look tired after one season and the ones still crisp after five are usually built from similar materials; what separates them is whether anyone designed for the cycle count.
The connection hierarchy
Reassembly design starts with one sorting rule: permanent bonds only within sub-assemblies that never come apart. Solvent-bonded acrylic joints are structural and beautiful — inside a module. A bonded hero case travels as a bonded hero case. But between modules, everything that separates for the crate must be a mechanical connection designed for cycles, and the specific hardware choices matter more than they look on a drawing:
- Threaded inserts over screws. Threads cut directly into acrylic strip — maybe not on cycle one, but across a season of torque from cordless drivers. A metal insert puts the wear interface in metal, where it belongs, and makes the fastener joint as reusable as the panel it holds.
- Captive hardware over loose. Loose hardware is lost hardware — not sometimes, always, eventually, and the missing bolt is always the one the header needs. Captive fasteners, retained washers, and hardware that stays with its panel through the crate convert a bag-of-parts problem into a fold-and-fix operation.
- Cam and toggle fittings. Show-floor labor is capable and hurried. Connections that need alignment, patience, and the right small tool get improvised around; quarter-turn cams, toggles, and latches get used correctly because using them correctly is the fast path. Design for the crew you’ll actually have, not the one in the instructions’ imagination.
Wear parts are designed, not discovered
Every exhibit has surfaces that slide, seat, and take torque on every single cycle — panel feet, track edges, the faces where modules stack. Left undesigned, those surfaces are “discovered” as wear parts by the fleet: scuffed, chipped, and photographed in an incident report by season two. The move is to make them replaceable pieces from day one. A sacrificial polycarbonate wear strip on an acrylic fixture is the canonical example — polycarbonate shrugs off the impact of repeated seating and takes the wear so the acrylic doesn’t, and when the strip finally looks like a season of shows, it unscrews and a new one goes on. Fleet damage becomes a consumable. The same thinking applies to glides, corner bumpers, and any edge that meets the crate: put the wear in a part whose replacement is a line item, not a rebuild.
Tolerance thinking for parts that mate hundreds of times
A joint that mates once can be fitted; a joint that mates a hundred times must be toleranced. Three practices carry most of the load:
Slip-fit clearances. Mating parts need designed clearance that accommodates the per-process variation in our Fabrication Tolerances standard and still assembles smoothly on cycle sixty — after the season’s dust, temperature swings, and honest wear. A snug fit that delighted everyone at the shop is a stuck fit in a cold hall; clearance is not sloppiness, it is the cycle count showing up in the drawing.
Lead-in chamfers. A chamfered entry turns alignment from a skill into a shove — the part finds its own seat under time pressure. Chamfers cost almost nothing at fabrication and pay on every one of those dozens of cycles, in assembly minutes and in edges that survive being missed by a quarter inch.
Asymmetric keying. If a panel can go in rotated or swapped with its twin, some morning it will — at 7 a.m., with the show opening at 9, by a crew meeting the exhibit for the first time. An offset pin, an asymmetric notch, a keyed bracket: physical impossibility beats labeling, because labels require reading and keys don’t.
Field-repair reality
Something will eventually fail at a show — the design question is whether the failure is a swap or a crisis. Design so the plausible failures are hand-tool-swappable on the floor: a cracked panel that unscrews from the front without dismantling the module behind it; a dead LED strip on a connector, not soldered in the dark; fasteners standardized so one driver and two bits service the whole booth. Pair the design with a spares strategy — the known-fragile panels, the strip, the hardware kit riding in the crate. Our spares-planning guide covers the kit itself; the design-side rule is that a spare only helps if the broken part was designed to come out.
The economics, said in factors
Reassembly-hardened design costs more at build one — inserts, captive hardware, keying features, and wear strips are all real fabrication content — and it pays on every show after: faster installs at billed show-floor labor rates, panels that survive the season instead of joining the reorder list, and a booth that opens on time because nothing assembled backward. The correct accounting period is the fleet’s life, not the first invoice — a fact worth saying out loud at budget time, because build one is when the money conversation happens and show twelve is when the design decision is actually experienced.
One honest closing note on dates, because seasons are schedules: a multi-show program’s build almost always lands against an immovable first show, and that is what Reserved Production exists for — the scheduling tier that holds dedicated capacity against a guaranteed ship date, which earns its keep the season your booth debuts. If you’re planning next season’s build — or auditing this season’s damage and recognizing a design signal in what keeps breaking — bring us the exhibit drawings and the show calendar, and we’ll go through them connection by connection.
