Detailing Moving Joints & Seals

To detail a moving joint well you have to feel why it is genuinely hard, because the difficulty is easy to underestimate from a rendering. A building envelope has one non-negotiable job: keep the outside out - rain, wind-driven water, air, heat, cold, noise, insects. At a fixed joint this is achievable with real confidence. Two parts that never move relative to each other can be sealed permanently: a bead of sealant, a bonded gasket, a lapped and taped membrane, set once and undisturbed, will keep water out for decades because nothing ever works the seal loose. Almost the entire craft of the building envelope is built on that premise of stillness.

A moving joint removes the premise. Now the two sides of the joint slide, pivot or fold past each other, and they do it repeatedly - a shading panel that adjusts several times a day, an operable vent, a sliding wall, a retractable roof. The seal can no longer be a permanent bond, because the two sides must be free to move; it has to be a seal that touches, releases and re-touches, or wipes across a surface, or compresses and relaxes, every single cycle. And every cycle is a small assault: the seal is deformed, dragged, abraded and worked, and materials that are worked eventually take a set, lose their springiness, wear thin, tear or harden and crack. The joint also cannot be tight, because as the previous lesson established it needs clearance and tolerance to move at all - and clearance is a gap, and a gap is a path for water and air.

So the moving joint holds two demands that pull directly against each other. Weathertightness wants intimate, continuous, pressured contact and no gaps. Movement wants freedom, clearance and the ability to break contact. You cannot fully satisfy both with a single element, and the beginner's instinct - to specify a bigger, better, tighter seal and trust it to do everything - is exactly the instinct that fails, because it asks one worked, wearing component to be the sole barrier against the weather across tens of thousands of cycles. The rest of this lesson is about the better answer: stop asking any single seal to be perfect, and design the joint so that when the outer seal inevitably leaks, what gets past it is caught, drained and vented away before it ever reaches the inside.

The moving joint's front line is the seal, and a designer should know the main families and, more importantly, how each one dies, because the failure mode is the design driver. A compression seal or gasket works by being squeezed between two surfaces; it seals beautifully while compressed, but on a moving joint it is compressed and released every cycle, and over time it takes a compression set - it stops springing back, the contact pressure falls, and the seal quietly stops sealing long before it looks worn. A wiper or brush seal works by dragging across a surface as the element moves, sweeping water and dirt; it tolerates motion well, which is why it is common on sliding and pivoting elements, but the wiping action abrades both the seal and the surface, and grit - abundant on a dusty Indian facade - turns the wipe into sandpaper that wears the seal thin and scores the track. A sliding or lip seal rides along a groove or track; it depends on the track staying clean, straight and undamaged, and any debris, deflection or thermal distortion in the track lifts the seal off its line.

The common thread is a concept engineers call duty cycle: a seal on a moving joint is not rated by how well it seals when new but by how many cycles it can seal for before it degrades, and how gracefully it degrades. A brilliant seal that fails suddenly at cycle twenty thousand is worse than a modest seal that fades slowly and predictably, because the sudden failure is the leak nobody sees coming. So the design questions are practical and unglamorous. How many cycles will this joint really see over its life, and is the seal rated well beyond that. Is grit going to reach it, and can the detail keep the worst of the dust and water off it. Can the seal be inspected, and crucially can it be replaced - a seal is a wearing consumable, like a tyre, and a moving joint whose seal cannot be renewed without dismantling the facade has designed in its own eventual leak. None of the binding numbers - the rated cycle life, the material, the contact pressure - are yours to set; they are the facade engineer's and the manufacturer's for the specific tested system. What you own is the detailing that keeps grit off the seal, protects it from the worst of the weather, and lets it be replaced when it wears, because it will.

Here is the single most important idea in the whole lesson, and it is the one that separates facades that stay dry from facades that leak: do not rely on one line of defence. A face-sealed joint puts everything on a single outer seal - if that seal is perfect, no water gets in; the instant it is not perfect, water goes straight to the inside. On a fixed joint you can sometimes get away with this, because the seal is undisturbed. On a moving joint, where the seal is worked and wearing every cycle, betting the building on one perfect seal is a bet you will lose. So the mature envelope stops trying to make the outer line perfect and instead assumes it will leak, and plans for the water that gets past it.

This is rain-screen or drained-and-vented thinking, and it is belt-and-braces by design. Behind the outer seal or screen you leave a deliberate cavity, and behind that cavity sits a second, inner line - typically the real air and water seal, kept in the sheltered position where it is not directly attacked by the weather. Water that defeats the outer line does not reach the inside; it lands in the cavity and is drained back out through weep holes at the bottom, harmlessly, before it can do harm. The cavity is also vented so that the air pressure inside it roughly equals the pressure outside - pressure equalisation - which matters because a great deal of water is not gently falling but driven through gaps by a pressure difference across the joint; take away the pressure difference and you take away much of the force pushing water in. The outer element becomes a screen that sheds the bulk of the water and takes the weather, while the sheltered inner seal, doing far less work, does the final sealing.

For a moving joint this thinking is not a refinement, it is the whole game. It lets the hard-working, wearing, exposed outer seal be imperfect - which it will be - without the building leaking, because there is a drained cavity and a protected inner line behind it. It moves the critical seal out of the firing line to where it is worked less and lasts longer. And it turns the inevitable degradation of a moving seal from a catastrophe into a slow, drainable nuisance. The specific geometry - cavity depth, weep sizing and spacing, where each line sits, how pressure equalisation is achieved for the particular movement - is binding facade engineering and belongs to the facade engineer and the tested system. The principle you carry into every moving joint is simple and non-negotiable: assume the outer seal leaks, and make sure what gets past it is caught, drained and vented away.

Keeping water out is the loudest requirement of a moving joint, but it is not the only one - the envelope makes several promises at once, and a moving element must keep all of them across every position it takes. It helps to think of the envelope as carrying four continuous barriers that a fixed wall never breaks: the water barrier (keeps liquid water out), the air barrier (stops uncontrolled air leakage in and out), the thermal barrier (the insulation and the thermal break that stop heat short-circuiting through the structure), and often a vapour control layer. In a plain wall these run unbroken. A moving element punches a hole through all four and then asks them to stay continuous even as the hole opens and closes.

Air-tightness deserves special respect because it is invisible and routinely underestimated. A joint can shed rain well and still leak air badly, and uncontrolled air leakage carries away heating and cooling, drives moisture into the construction where it can condense, and undermines comfort and acoustics. On a moving joint the air seal is usually the sheltered inner line described above, and it has to maintain contact across the element's movement and its wear - which is hard, and is a large part of why moving joints so often underperform on air-tightness even when they keep the rain out. The thermal break is the other quiet trap: a moving element is often a run of metal, and metal is a superb conductor, so an aluminium panel or frame bridging from outside to inside can short-circuit the insulation and create a cold line that wastes energy and, in the wrong climate, sweats with condensation. The detail has to interrupt that metal path with a non-conducting break that stays continuous with the wall's insulation on both sides of the moving joint.

The honest summary is that a moving joint is asked to keep the same whole-envelope contract as a fixed wall - water out, air controlled, heat held, vapour managed - while doing the one thing a fixed wall never does, which is move. That is why it is the hardest detail in adaptive architecture, why it is where these buildings most often fail, and why its binding design - the weathertightness, the air and water performance, the thermal and condensation behaviour of the specific assembly - must be entrusted to a qualified facade engineer and a manufacturer's tested, warrantied system rather than improvised on a drawing. What you own as the designer is the strategy: keep every barrier continuous across the joint, shelter the working seals, drain and vent for the leak you assume will happen, keep the metal broken thermally, and make the wearing parts replaceable. Detail the joint as if it will be attacked - because it will be - and defer the numbers that prove it holds to the people qualified to guarantee them.