Walk across the atrium of a mall or an airport concourse and look down. Somewhere near the midspan there is a thin metal strip running wall to wall — and if you follow it, the same line climbs the wall, crosses the soffit and reappears on the façade. Most people read it as a trim detail. It is a lid. Underneath it is a deliberate slot cut clean through the structure, and that slot is one of the few things stopping the building from slowly tearing itself apart.
Why buildings have to move
Steel and concrete are stiff. They are not inert. Four things make a building move, and if you do not give that movement somewhere to go, it arrives as a crack.
- Temperature. Concrete's coefficient of thermal expansion is roughly 10 × 10−6 per °C — about 0.4 mm per metre for a 40 °C swing. On a 100 m long block in Delhi that is around 40 mm of end-to-end movement between a January night and a June afternoon, plus a smaller version of the same cycle every single day. It is why IS 456 asks for a joint in structures running longer than about 45 m.
- Shrinkage and creep. Fresh concrete keeps losing moisture and shortening for months, then keeps deforming under sustained load for years. Neither process stops when the contractor demobilises.
- Differential settlement. A 20-storey tower and a 2-storey podium load the ground in completely different ways and settle at different rates. The line between them has to be free to move.
- Dynamic load. Wind sways tall buildings. An earthquake moves adjacent blocks out of phase with each other, and where two blocks sit too close they collide — pounding, a documented cause of structural damage in every major urban earthquake. IS 1893:2016 derives the required separation from design storey drift for exactly this reason.
None of this is preventable. The choice is only between letting the building crack wherever it likes, and deciding in advance where it will move and giving it room. That room is the expansion joint.
The joint and the joint system are not the same thing
These two get conflated constantly — including by people who ought to know better — and it causes real trouble at tender stage.
The joint is the gap itself. Its width is a structural decision, made by the structural engineer: thermal joints under IS 3414 and IS 456, seismic separation calculated from design drift under IS 1893:2016. Most building joints land between 25 and 100 mm; seismic gaps in high-rises run considerably wider.
The expansion joint system is what covers it: aluminium frame profiles, a centre plate or slide, elastomeric seals, a waterproofing membrane and, where required, a fire barrier. It is a designed assembly whose job is to let people, trolleys, water and fire behave sensibly at a line where the building is deliberately discontinuous. That is the part we make.
What the cover actually has to do
The brief sounds simple. It is not, because every requirement fights the one next to it.
- It has to move in three axes at once — opening and closing, vertical offset from differential slab deflection, and lateral shear — without locking, buckling or lifting out.
- It has to carry load: foot traffic, trolleys, hospital beds, and in car parks and podium decks, vehicles.
- It has to be watertight, because the joint cuts straight through the building's waterproofing membrane. A joint over occupied space with no membrane installed inside it is a leak with a commissioning date.
- It has to be smoke-tight, because an unprotected joint is an open vertical shaft running through every fire compartment it crosses.
- And it has to sit flush and trip-free — now an accessibility compliance requirement, not a matter of taste.
Any fabricator can put an aluminium plate over a gap. Making one that stays flush, dry, quiet and serviceable for twenty years is a different problem entirely.
A short history
The idea is old. Roman and medieval construction used lime mortar and dry stone — weak in tension, and therefore able to crack and creep harmlessly. Movement was accommodated because the fabric itself was forgiving.
Rigidity created the problem. Nineteenth-century railway engineers understood it well enough to leave gaps between rails; long-span iron bridges sat on rockers and rollers for the same reason.
Reinforced concrete made it worse. Continuous monolithic frames give stress nowhere to go, and by the middle of the twentieth century the codes caught up — ACI in the United States, IS 456 in India from the 1960s onward. The first covers were crude: bitumen and felt, then bolted steel chequer plate that rattled, corroded, and was fixed with anchors driven straight through the waterproofing it was meant to protect.
Extruded aluminium changed the economics from the 1970s. A precision die produces a frame profile that anchors into the slab and a centre plate that slides on it, in continuous lengths, at a fraction of the weight of steel. Vexcolt Ltd was founded in the UK in 1985 into that generation of specialists, and now holds over thirty registered patents across joint, control-joint and fire-barrier products, with manufacturing in the UK, the Netherlands, China, Bahrain and India.
Where the innovation has actually happened
Less in the aluminium; more in everything around it.
Elastomeric and pre-compressed foam seals displaced metal plates in wall and façade joints. Fire barriers moved from improvised stuffing to systems tested and rated against standards such as UL 2079 — and, more usefully, tested in combination with the water seal, so one assembly does both jobs instead of three trades arguing about whose responsibility the interface is. Centring mechanisms in seismic profiles return the cover to position after a major event rather than leaving it displaced. Thermal breaks arrived in roof and external joint systems. And factory-made corners, transitions and floor-to-wall junctions removed the welded site improvisation that is where most leaks actually start.
The Indian trajectory, without the marketing gloss
The demand is real without needing a market report to prove it. The National Infrastructure Pipeline, metro construction in a dozen cities, airport expansion, and continued high-rise development across Delhi-NCR — which sits in seismic zone IV — are visible drivers on their own.
Vexcolt India operates under a royalty licence from Vexcolt UK, extruding UK profiles through Indian dies, and supplies and installs for L&T, Shapoorji Pallonji, Ahluwalia and DLF among others. Local manufacturing is not just landed-cost arbitrage. It removes the lead time that kills a specification when the material is needed next month.
Two honest limitations.
First, procurement here is still overwhelmingly lowest-bid. An anonymous fabricator beats a proven system on a comparison sheet every time. The life-cycle argument does work — but it has to be made in writing to the consultant before the tender is drafted, not to the contractor afterwards.
Second, the failure mode in this market is almost never the profile. It is the installation: an unlevel or unprepared substrate, a membrane left out, anchors driven into poor concrete. Which tells you where the defensible position actually is — in the team, not the extrusion.
There is a quieter opportunity in age. India's construction surge of the late 1990s and early 2000s is now more than twenty years old. Those original joints are failing, and the people specifying the replacements have first-hand experience of what the cheaper option cost them.