Loads: wind, dead, live & seismic
Before you size a single mullion you must know the forces pushing on the skin — and on a tall building in India, wind almost always wins.
A 1.5 x 3.5 m glass panel on a Mumbai tower can be asked to resist over half a tonne of wind suction — and the code tells you exactly how much.
Stand on a high floor in a coastal Indian city when the pre-monsoon winds build and you can feel the glass flex. That movement is the facade doing its first structural job: catching wind and handing it to the frame. To the public the skin looks static. To an engineer it is a pressure vessel turned sideways, loaded by gusts that push on the windward face and _suck_ harder on the leeward face and at the corners. Get the number wrong and the glass cracks, the gaskets blow out, or a panel departs the building. So before any glass thickness, any mullion, any bracket, we ask the foundational question of Module 4: how big are the loads? On a tall building, the honest answer is almost always - wind decides.
Four load families, and the one that usually governs
Dead, live, wind and seismic - the four families acting on the skin
A facade carries four families of load, and they are not equal.
Dead load is gravity on the facade's own mass - the self-weight of glass, aluminium, gaskets and panels - acting straight down. It is steady and easy to predict. Live load on a facade is modest: it is mainly the barrier / imposed horizontal load at guarding height (a person leaning on the glass), plus maintenance and cleaning forces; facades carry no floor live load because they hold up no floor. Wind load is the lateral pressure of moving air - pushing on the windward face, sucking on the leeward face and, viciously, at corners and parapets. Seismic load is the inertial force when the building shakes the facade's mass in an earthquake.
For a tall, lightweight clad facade the ranking is almost universal: wind governs the panel and mullion, dead load governs the brackets and anchors, and seismic governs the movement joints. A glass curtain wall is so light that the earthquake force on its own mass is small - but the drift the earthquake imposes on the building (Lesson 4.2) is what hurts it. So we size the skin for wind, hang it for gravity, and detail it for movement.
Wind sizes the panel. Gravity sizes the anchor. The earthquake doesn't crush the skin - it shakes the frame, and the skin has to keep up.
IS 875 (Part 3): basic wind speed Vb, factors k1-k4, then the design pressure pz
Indian wind loading runs on IS 875 (Part 3): 2015. You start from the basic wind speed Vb - a 3-second gust at 10 m height, read off the code's wind map for your city. It ranges from 33 m/s (much of the interior) to 50 m/s on the cyclone-prone east and west coasts. Mumbai and Chennai sit at 44 and 50 m/s; Delhi and Bengaluru at 47 and 33.
You then modify Vb by four factors to get the design wind speed Vz at height z:
Vz = Vb x k1 x k2 x k3 x k4.
k1 is the risk/return-period factor (1.0 for normal buildings, higher for important ones). k2 is the terrain-and-height factor - it grows with height and with how open the terrain is, which is why the top of a tower sees far more wind than its base. k3 is the topography factor (1.0 on flat ground, up to ~1.36 on a hill crest). k4 is the importance factor for cyclonic regions (up to 1.30 near the coast). Finally the design wind pressure is pz = 0.6 x Vz-squared (Vz in m/s gives pz in N/m-squared, i.e. Pascals). For cladding you multiply pz by pressure coefficients (Cpe - Cpi) and a gust factor to get the net pressure the panel actually feels - and the suction at a corner can be the worst load on the whole building.
Load combinations and the IS 1893 seismic check
Loads rarely act alone, and IS 875 (Part 5) gives the combinations. For facade cladding the governing cases are usually 1.5 x wind (ultimate, for strength) and 1.0 x wind (serviceability, for deflection - Lesson 4.2). Dead load combines with wind for the brackets. You design the glass and mullion for the worst net pressure, remembering that the suction case (wind pulling the panel off) is frequently larger than the pushing case and pulls the structural silicone or the captive gasket in tension.
For seismic, IS 1893 (Part 1): 2016 treats the facade as a non-structural component or appendage. The horizontal seismic force is roughly Fp = (Z/2) x (Sa/g) x (ap/Rp) x Ip x Wp - the building's zone factor Z, the response amplification, and the component's own weight Wp. Because a curtain wall is light (Wp is small) this force is modest and seldom governs the cladding itself. The seismic action that matters for facades is not the force on the skin's mass but the inter-storey drift the earthquake imposes on the frame - the building leans, and the facade must slide or flex to follow it without cracking glass. We carry that straight into the next lesson.
Wind is invisible on your elevation but it sets the panel's thickness and the mullion's depth. Two moves you control change the load dramatically: building **height** (k2 grows with it) and **corners**. The corners and the top two storeys of a tower see the worst suction, so a fully glazed corner detail or a thin parapet coping is where the engineer will push back hardest. Ask early what your site's basic wind speed is - a Chennai project (50 m/s) carries roughly twice the pressure of a Bengaluru one (33 m/s), and that ratio shows up as thicker, costlier glass.
Own the wind analysis end to end: read Vb off the IS 875-3 map, build k2 from the right terrain category and the panel's actual height, apply k3/k4 honestly for hill sites and cyclonic coasts, and never forget the gust and the cladding pressure coefficients - mid-panel Cpe and the far worse local corner/edge Cpe. For supertall or aerodynamically sensitive buildings the code map is not enough; you commission a **boundary-layer wind-tunnel study** and design the cladding to the measured peak local pressures. Keep the suction case visible in every check - it governs the structural silicone bite and the anchor pull-out.
The number that drives everything you install is the **design wind pressure** in kPa - it is on the facade drawings and it is why the glass on the 40th floor is thicker than on the 4th, and why the corner panels differ from the field panels. When you see a panel marked for a higher zone or a corner location, it is not an error: that piece resists more wind. The structural silicone bite, the gasket and the bracket bolts were all sized to a specific pressure - which is why substituting a thinner glass or a smaller bolt 'to save cost' is never a site decision.
IS 875 (Part 3): 2015
Wind loads on buildings
India's wind code - gives basic wind speed Vb, the k1-k4 factors, gust factor and pressure coefficients. Its city wind map is coarse: for tall or unusual buildings it explicitly defers to wind-tunnel testing.
IS 1893 (Part 1): 2016
Seismic design
Sets zone factor Z and the method for forces on non-structural components like facades. For light cladding the force on its own mass rarely governs; the imposed inter-storey drift does.
IS 875 (Part 5)
Load combinations
Defines how dead, live, wind and seismic combine. For facades 1.5x wind (strength) and 1.0x wind (deflection) usually govern; it does not itself give the wind magnitude.
ASCE 7 / EN 1991-1-4 (global)
Wind loads, components and cladding
The US and European equivalents; both carry explicit 'components and cladding' pressures with elevated edge/corner zones. Used as cross-checks on premium Indian projects; their gust/zone definitions differ from IS 875-3, so never mix codes mid-calculation.
“Wind just pushes on a building, so the worst load is the wind blowing straight onto the glass.”
The push (positive pressure) on the windward face is often _not_ the worst case. Moving air separates at corners, edges and the parapet, creating intense local **suction** that can exceed the windward push by a large margin - and suction is the load that pulls a panel off the building, tensions the silicone and yanks the anchors. IS 875-3 captures this with high local (corner/edge) negative pressure coefficients. Designing only for the 'push' is how corner panels fail first.
Worked example - design wind pressure on a curtain-wall panel
Let us put a real number on a panel: the net design wind suction on a corner curtain-wall unit near the top of a Mumbai office tower, straight from IS 875 (Part 3).
The IS 875 (Part 3): 2015 wind map and k-factor tables, a calculator, and the panel geometry from the facade grid.
GIVEN (Mumbai office tower, corner panel at z = 80 m):
Basic wind speed Vb = 44 m/s (IS 875-3 map, Mumbai)
Risk factor k1 = 1.00 (ordinary building, 50-yr)
Terrain/height k2 = 1.12 (Category 3, z = 80 m)
Topography k3 = 1.00 (flat coastal site)
Cyclonic factor k4 = 1.15 (coastal, important contents)
Panel size 1.5 m x 3.5 m = 5.25 m2
Vz = Vb x k1 x k2 x k3 x k4
pz = 0.6 x Vz^2 (Vz in m/s -> pz in N/m2 = Pa)
Net cladding pressure p = pz x (Cpe - Cpi) x Gust
corner suction: Cpe = -1.20 , Cpi = +0.20 , Gust Gf = 1.00- 1Design wind speed. Vz = 44 x 1.00 x 1.12 x 1.00 x 1.15 = 56.7 m/s. The four factors lift the 44 m/s map value to nearly 57 m/s at this height on this coast.
- 2Dynamic (velocity) pressure. pz = 0.6 x Vz-squared = 0.6 x 56.7-squared = 0.6 x 3215 = 1929 N/m2 ~ 1.93 kPa. This is the bare wind pressure before cladding coefficients.
- 3Net pressure coefficient. At the corner the external coefficient is suction, Cpe = -1.20, while internal pressure pushes outward, Cpi = +0.20. Net = (Cpe - Cpi) = (-1.20 - 0.20) = -1.40 (suction, pulling the panel off).
- 4Net design pressure. p = pz x (Cpe - Cpi) x Gf = 1929 x (-1.40) x 1.00 = -2701 N/m2 ~ -2.70 kPa. The corner panel feels about 2.7 kPa of suction - notably more than a mid-field panel would.
- 5Force on the whole panel. F = p x area = 2701 N/m2 x 5.25 m2 = 14,180 N ~ 14.2 kN ~ 1.45 tonnes of suction trying to pull this one unit off the building.
- 6Apply the ULS factor. For strength design (IS 875-5) multiply by 1.5: 1.5 x 2.70 = 4.05 kPa is the pressure the glass, the structural silicone bite and the brackets must resist at ultimate. The 2.70 kPa (1.0x) value is what you then use for the deflection check in Lesson 4.2.
- 7Sanity-check the suction logic. Had we taken only the windward push (Cpe ~ +0.8), the load would have been smaller than this corner suction - confirming the misconception above: at corners, suction governs.
You’ll walk away with
A net design wind pressure of ~2.70 kPa (serviceability) / ~4.05 kPa (ultimate) and a ~14.2 kN panel suction force - the single number that will size the glass thickness, the structural silicone and every anchor for this unit, computed cleanly from IS 875 (Part 3).
Two quick ways to feel the numbers.
- 01Re-run Step 1-4 for a Bengaluru panel (Vb = 33 m/s, k4 = 1.0) at the same height and compare the net pressure to Mumbai's - you will see roughly half the load, which is exactly why coastal towers carry thicker glass.
- 02Take any tall building you know and ask which face and which corner sees the prevailing wind; that corner, near the top, is where the worst suction lives and where the panels are detailed strongest.
A facade carries dead, live, wind and seismic loads, but on a tall building wind almost always governs the skin. IS 875 (Part 3) builds the design pressure from a mapped basic wind speed and four k-factors, and the worst case is usually corner _suction_, not the windward push. Wind sizes the panel; gravity sizes the anchor; the earthquake's gift to the facade is drift, not force.
Four load families: dead (self-weight), live (barrier/maintenance), wind (lateral, governs the skin), seismic (governs movement, not the light skin's mass). Wind via IS 875-3: Vz = Vb x k1 x k2 x k3 x k4, then pz = 0.6 x Vz-squared, then net p via cladding Cpe - Cpi. Corner suction usually governs. ULS = 1.5x wind, SLS = 1.0x wind.
How do you calculate wind load on a facade in India?
Use IS 875 (Part 3): 2015. Read the basic wind speed Vb from the wind map for your city, compute the design wind speed Vz = Vb x k1 x k2 x k3 x k4 (risk, terrain-and-height, topography and cyclonic factors), then the design wind pressure pz = 0.6 x Vz-squared in Pascals. For cladding, multiply by the net pressure coefficient (external minus internal) and the gust factor to get the pressure the panel actually feels - and check the corner suction case, which often governs.
Does wind or earthquake govern a glass curtain wall?
For the skin itself, wind almost always governs, because a curtain wall is light - the seismic inertial force on its own small mass is modest. What the earthquake does matter for is inter-storey drift: the building leans, and the facade must accommodate that movement without cracking glass. So wind sizes the panel and glass, while seismic drives the movement-joint detailing covered in the next lesson.
Why is the corner of a building the most wind-critical for cladding?
Because moving air separates at sharp corners, edges and parapets, creating intense local suction zones. IS 875 (Part 3), like ASCE 7 and EN 1991-1-4, assigns much higher negative (suction) pressure coefficients to corner and edge zones than to the field of the wall. That suction pulls panels off the building and tensions the silicone and anchors, so corner and top-storey panels are routinely the most heavily engineered on a tower.
Peer-reviewed journals & authoritative standards
- 01Bedon, C. et al. Performance of structural glass facades under extreme loads - design methods, existing research, current issues and trends. Construction and Building Materials, 163. — Construction and Building Materials (Elsevier), 2018.
- 02Review on Glass Curtain Walls under Different Dynamic Mechanical Loads: Regulations, Experimental Methods and Numerical Tools. IntechOpen. — IntechOpen (peer-reviewed chapter), 2023.
- 03Wind Resistance Performance Evaluation of a Cable-Type Curtain Wall System on Reinforced Concrete High-Rise Buildings. — (peer-reviewed; University of Bath research portal), 2021.
_You now have the design pressure. The next question is not 'will the glass break?' but 'how far will it bend?' - because long before a panel fails it can deflect enough to crack glass and burst seals. Deflection and movement, next._