Roofs, Openings & Debris

The roof is the most exposed and the most vulnerable surface in a cyclone, and understanding why is the key to defending it. Three things conspire. First, the aerodynamics: as wind flows over a building it accelerates most sharply at the roof, creating the strongest suctions of anywhere on the skin - and those suctions peak viciously at the eaves, ridges, corners and edges where the flow separates. Second, the roof is held down mainly by gravity and its connections - unlike a wall, which has its own weight bearing down and the floors bracing it, a roof's dead weight is often modest and it relies on being positively tied down to resist being lifted. Third, if the envelope is breached (Lesson 4.2), internal pressure adds an upward push from inside, and the combined uplift can roughly double. A roof designed only for its weight, or tied with weak or corroded connections, simply lifts.

And the consequence cascades. The roof is not just shelter; structurally it usually acts as a diaphragm and a tie that holds the tops of the walls together and stops them splaying or buckling outward under lateral load. When the roof peels away, that bracing vanishes: the walls become tall free-standing cantilevers with nothing restraining their tops, and they topple or are blown out soon after. This is why post-cyclone surveys so often show the roof gone and then the walls collapsed - the roof failure was the first domino. Saving the roof often saves the whole building; losing it usually dooms the rest.

The design implication is blunt: the roof and its anchorage deserve the most careful attention of any part of a wind design, and the worst loads are at the edges and corners, not the middle. Eaves that the wind can get under, large unbraced overhangs, and light sheet roofing fixed with too few or corroded fasteners are the classic failure points. Heavy roofs resist uplift better by sheer weight but bring their own seismic penalties (a theme this course keeps balancing); light roofs are kinder in an earthquake but must be tied down hard against wind. Which roof, how heavy, and every connection capacity is an engineered, code-based decision - IS 875 Part 3 for the uplift, the engineer for the connections - but the principle the architect owns is that the roof is the battleground, its edges fail first, and it must be positively, continuously tied down.

The defence against uplift is a single idea repeated all the way down the building: a continuous load path for uplift, from the roof covering to the ground, with every link able to carry the force and pass it on. Trace it from the top. The roof covering (tiles, sheets, membrane) must be fixed to the structure below it well enough to resist suction - this is where peeling starts, so fixings are denser at edges and corners. The rafters or trusses must be tied to the walls or ring beam with positive connections - the straps, clips and anchors often called hold-downs or, in hurricane country, 'hurricane straps' - not merely resting under their own weight. The walls must carry the uplift (and the overturning) down their height without a weak storey or a broken joint. And the whole assembly must be anchored to the foundation and the foundation to the ground, so the uplift and overturning cannot simply lift or slide the building off its base. Each link is a place the chain can break, and the wind finds the weakest.

The governing truth is that the chain is only as strong as its weakest link. A superb roof-to-wall strap is worthless if the roof covering peels off above it, or if the wall-to-foundation anchorage lets go below it. This is why wind resilience is a continuity problem, not a component problem: you cannot buy safety by over-building one connection while another is an afterthought. The designer's job is to make sure the path is complete and consistent - that uplift has a carried route from the ridge to the soil with no missing or under-designed link.

Continuity also explains some classic mistakes. Additions and alterations that tap into an existing roof often break the original tie-down path. Gable ends and large overhangs are frequently the weak link because they are poorly tied and heavily loaded. Verandah and porch roofs tacked onto the main structure can be lifted and, in lifting, damage what they are attached to. In traditional and owner-built construction, roofs are often simply laid on and weighted rather than tied, which is exactly why they fly. The modern answer is positive, engineered connections forming an unbroken chain. The architect must design for that chain to exist - detailing connections at every junction, avoiding configurations that break the path - while the connection capacities, strap sizes and anchorage details come from the structural engineer to IS 875 Part 3. Draw the chain; let the engineer size each link.

Openings are the other front in the cyclone battle, and they matter for two linked reasons. First, as Lesson 4.2 showed, a breached windward opening pressurises the interior and magnifies uplift on the roof and the outward load on the walls - so protecting openings is directly protecting the roof we just fought to tie down. Second, openings are intrinsically the weakest parts of the skin: a large pane of glass or a light door resists far less than the wall around it, and it is exactly what the storm and its debris attack first. Defending the openings keeps the envelope sealed and the pressure vessel intact.

There are a few complementary strategies. Shutters - storm shutters, rolling shutters, or even robust temporary boarding - are the classic protection: a strong, well-fixed cover over glazing that takes the wind pressure and the debris impact so the glass behind survives. In cyclone regions, permanent or quickly deployable shutters on the vulnerable openings are among the most cost-effective resilience measures there is. Impact-resistant glazing - laminated assemblies designed to resist penetration and to hold together even when cracked - is the alternative or complement where shutters are impractical, keeping the envelope sealed even after an impact. Strong, well-anchored doors - especially large doors like garage or workshop shutters, which are notorious first-failure points - must be rated and fixed for the pressures, because a failed large door is a guaranteed windward breach. And the frames and their fixings matter as much as the panel: a good shutter on a frame that tears out of the wall has failed.

The design discipline is to identify the openings that matter and defend them deliberately, not to treat glazing as purely a daylight-and-view decision in cyclone country. Large glazed areas facing the likely storm direction are liabilities unless protected; the most exposed openings should be the best defended; and a household should be able to close the building down before a storm. This interlocks with the safe-room idea in the next lesson: the most protected opening-free or shutter-protected space becomes the refuge. As always, the design pressures, the impact-resistance requirements and the rating of shutters, glazing and doors come from IS 875 Part 3, product standards and the engineer - the architect's job is to plan for protectable openings, place the vulnerable glass away from the worst exposure, and make sure every opening can be defended.

The last piece is what the wind carries. A cyclone does not attack with wind alone; it picks up and hurls debris - sheets of roofing torn from other buildings, branches and whole trees, signboards, loose site materials, unsecured objects from balconies and yards - and flings them at lethal speed. This wind-borne debris is a major cause of envelope breaches: a flying roof sheet or timber that smashes a window starts exactly the internal-pressurisation failure we have been fighting, and debris injures and kills people directly. So debris is both a structural threat (it breaches the envelope) and a life-safety threat.

Designing against debris works on two levels. At the building level, the openings most likely to be hit - those facing the storm and at lower levels where debris flies densest - should be protected with shutters or impact-resistant glazing so that a strike does not become a breach. This is the same measure as opening protection above, now justified by the missile rather than the pressure: impact-resistant assemblies are specifically tested against representative 'missiles'. Solid, well-anchored walls resist debris far better than glazing, which is another reason to keep large vulnerable glass away from the most exposed faces. At the site and detailing level, resilience means not creating the missiles: securing or removing loose objects, designing roofs and cladding so they do not themselves become the debris that destroys the neighbour, anchoring rooftop equipment, and choosing robust, well-fixed external elements rather than light ones that detach. A community where every building's roof stays on is a community with far less flying debris.

There is an order-of-magnitude point worth keeping. The worst debris tends to be other people's failed buildings - which is why building-by-building resilience is also collective resilience, and why codes and the safe-room strategy of the next lesson matter so much. For the individual designer, the takeaway is concrete: in cyclone country, assume the storm will throw things; protect the openings that a missile would turn into a breach; keep vulnerable glazing away from the worst exposure; and detail your own building so it does not shed the debris that harms others. The specific missile criteria, impact ratings and test standards are the engineer's and the product standards' to apply under IS 875 Part 3 and the governing code - the architect's job is to take debris seriously as a design load, not an afterthought.