Joints, Interfaces & Connections

Connections in a modular building fall into two broad families, and they have different characters. The first is module-to-module: modules joined to each other, side by side and stacked floor on floor, to make the building. These are the connections the manufacturer has usually designed and tested as part of their system — standard fixings at standard points, repeated many times — and they benefit from repetition and from parts that were made to match. The second family is module-to-site: the connections between the factory-made parts and the conventionally built world they land in — the foundations, the ground-floor podium, an in-situ concrete or steel core (lifts and stairs, often built on site for stability), and any site-built structure the modules tie into. This family is harder, because it is where the accurate factory meets the approximate site, and where the tolerance reconciliation of lesson 5.2 actually has to happen.

Structurally, the connections have real work to do. Modules must be tied together and down so the assembled building behaves as one structure — resisting not just gravity but lateral loads from wind and, in much of India, seismic action, which the connections and often a separate stabilising core or bracing system must carry. How a stack of boxes is made to act as a braced, tied structure is a genuine engineering problem, and the answer is specific to the system: some rely on the modules themselves, many rely on a site-built core or frame that the modules lean on. This is firmly engineering territory — the forces, the fixings, the load paths and the stability strategy are designed and signed off by the structural engineer and are intrinsic to the manufacturer's tested system.

The designer's job is not to invent the connection but to understand and accommodate it: to know that module-to-site interfaces need generous, deliberate tolerance and adjustment; to know that a core or podium is often on the critical path and must be coordinated early; and to know that the connection strategy shapes the architecture (where the bracing or core sits, how far modules can cantilever or step, where movement joints fall). A designer who treats connections as a detailing afterthought will find the engineering reaching back up to reshape the plan late and expensively. A designer who understands the two families and their demands designs a building the connections can actually hold together. The binding design stays with the specialists; the spatial and sequencing consequences are yours to own.

A connection that merely holds the building up has done a fraction of its job. Every performance line in the building — every line that must be continuous for the building to work — has to be carried across the joint, and this is where modular detailing earns its keep. Consider them in turn. Fire: a building is divided into compartments so fire cannot spread freely, and the compartment lines run through the joints between modules; the junction must re-establish the fire-resistance rating and the seal that the module wall or floor provides, or the compartmentation is only as good as its weakest joint. A penetration for a pipe, or an unsealed void at a module junction, can quietly defeat an otherwise good fire strategy.

Acoustics: the separation between dwellings or rooms depends not only on the mass of the walls and floors but on stopping sound flanking around them — travelling through gaps, rigid contacts and continuous paths at the junctions. A modular joint has to be detailed to break those flanking paths (resilient layers, discontinuous construction, sealed gaps) or the acoustic performance the modules could achieve is lost at the line where they meet. Weather: wherever the joint is on the external envelope, it must keep water and air out across a gap that varies with tolerance and moves in service — which is why external module joints rely on designed, layered defences (overlaps, baffles, drained and ventilated cavities, compressible seals) rather than a single bead of sealant hoping to bridge a moving gap. Services: pipes, ducts, cables and drainage frequently have to cross from module to module, so the joint must provide a coordinated, accessible, sealed route for them that does not compromise the fire, acoustic or weather lines it also carries.

The hard part is that these duties land at the same place, at the same time, and can conflict: the gap that weather wants (open, drained, ventilated) fights the seal that fire and acoustics want (closed, continuous, massive); the access that services want can breach both. Resolving them is a coordinated, multi-discipline design problem, and the resolution is specific to the system, the rating required and the exposure. The binding fire, acoustic and weathertightness design belongs to the fire engineer, the acoustician, the facade specialist and the manufacturer's tested and warranted details. What the designer must hold is the map: an awareness that every joint is carrying several performance lines at once, a refusal to let one duty quietly defeat another, and an insistence that the junctions are coordinated early rather than improvised late. (Weathertightness, fire and acoustics each get their own treatment in Module 8; here the point is that they all meet at the joint.)

Volumetric construction creates a condition that panelised and site-built work do not, and it is worth understanding on its own: the double skin. When you stack finished 3D modules, the floor of the upper module comes to rest on the ceiling of the lower module. Where a site-built building has one floor/ceiling assembly between two storeys, a stack of modules has two — the top module's floor deck and the bottom module's ceiling, with a joint zone between them. The same doubling happens vertically at every party wall between side-by-side modules: two wall skins with a cavity, rather than one shared wall.

This has consequences, good and bad. On the good side, two skins with a gap can be acoustically excellent — discontinuous construction with an air gap is exactly how you stop sound transmission, so a well-detailed double skin can outperform a single site-built separator. It can also help with fire separation and gives a route for services and tolerance take-up in the zone between. On the bad side, it costs height and material: two floor zones and two ceiling zones stacked up consume storey height and add weight and cost that a single assembly would not, and across a tall building that lost height adds up. It also creates a concealed void at every inter-module joint — the double-skin zone — which is hard to inspect and must be detailed so that fire, acoustics and weather are properly closed *inside* a gap no one can easily see after assembly.

The design responses follow from naming the condition. Budget the height: account for the double floor/ceiling in your floor-to-floor dimension from the start, rather than discovering it has eaten your ceiling height late. Use the void: let the zone between the skins carry services, tolerance and acoustic separation deliberately, so the cost of the double skin buys performance rather than being pure waste. Close it before it is buried: ensure the fire, acoustic and weather seals within the double-skin joint are designed to be completed and, where possible, inspected during assembly, because once the modules are locked together the joint zone is effectively sealed away. The double skin is neither a flaw to be hidden nor a free gift — it is a defining characteristic of volumetric construction that a good designer plans around, turning its air gap into acoustic and service advantage while paying honestly for the height it costs. The exact build-ups, ratings and acoustic values are, once more, outputs of the manufacturer's tested system and the relevant engineers.

Here is the design discipline that separates connections that work on paper from connections that work on a crane-day site: accessibility for assembly. A connection is not finished when it is drawn; it is finished when someone on site can physically reach it, make it, and ideally check it — in the real sequence, with the real access, in the real time available, often with a module hanging from a crane and the next one waiting. A structurally perfect bolt that can only be reached from a space that no longer exists once the adjacent module is placed is not a connection; it is a trap. This is pure Design for Assembly, applied to the joint.

Several principles follow. Design the connection and the sequence together: every joint must have an access route and a moment in the assembly order when it is reachable, and that moment must come *before* it is closed off by the next piece. A connection might be made from inside the module, from an open face before the neighbour arrives, from a designed access panel, or from a position the sequence deliberately preserves — but it must be made from *somewhere that exists at that step*. Prefer few, simple, repeatable connections over many fiddly ones, because every connection is a task done at height on a critical-path day, and because a simpler joint is easier to make right and to inspect. Make joints tolerant and guiding — self-aligning features, generous lead-ins, adjustment built in — so a module guided into place by a crane lands and connects without millimetre-perfect manoeuvring. And think about inspection and the future: fire and acoustic seals that can be verified, and where possible connections that can be reached again for maintenance or eventual disassembly.

This is why connection design and the assembly programme (Module 7) are really one conversation. The sequence in which modules are craned and placed determines which joints are reachable when, and therefore which connection designs are even possible; a joint that ignores the sequence will stall the crane, blow the programme and tempt the site into an unapproved improvisation. The designer's role is to make connections that respect how a building actually goes together — accessible, simple, guiding, inspectable — and to coordinate that with the manufacturer (who knows their connection details) and the assembly and craneage specialists (who know the sequence and the reach). As with all of this module, the binding connection design, the structural capacities and the approved details belong to the engineer and the manufacturer's tested system; the designer owns the insistence that the joint can be reached, made and checked in the real world, not just resolved on the drawing.