Wind load analysis: how to calculate wind loads on a building

ASCE 7 will give you a wind load number for almost any building you design, and for a rectangular tower on flat ground that number is good enough to build from. The problem starts when the geometry stops being rectangular. Tapered facades, setbacks, podium roofs, a neighboring tower 100 ft (30 m) upwind: the code’s pressure coefficients were never derived for any of it. The further your design sits from the shapes in the tables, the less those coefficients tell you about the pressures the building will actually see, in either direction. Wind load analysis is how you find out.

This article covers both halves of the job. First the code procedure, in full, with a worked example you can check your own numbers against. Then the cases where the code hands off to physical testing, and where simulation earns its place before that.

What is wind load analysis?

Wind load analysis is the evaluation of the pressures and forces wind exerts on a structure, and of the structural response those forces produce. It has two outputs: a set of design pressures for the structural frame and the facade, and an assessment of how the building moves under fluctuating wind.

Two things drive whether you need to go beyond a code calculation. The first is height. As buildings get taller and more slender, crosswind response and occupant acceleration start to govern the design, and the static pressure check stops being the binding one. The second is location. In coastal and hurricane-prone regions, even a conventional low-rise has to be checked against design wind speeds that inland sites never see.

The work sits with architects and structural engineers together, because the two changes that reduce wind load most are both architectural: the shape of the building and the treatment of its corners. Settle them during massing and building aerodynamics stays a design conversation. Leave them and it becomes a remedial one.

How wind load acts on a building

Wind produces positive pressure on the windward face, suction on the side walls, the leeward face and most roof surfaces, and a fluctuating crosswind force from the vortices shedding off the building’s sides. Design has to answer for all three.

Pressure loads on the structure and facade

Pressure analysis, mean and peak, identifies the positive and negative pressures on the envelope, which sets facade and cladding specification and shows where reinforcement is needed. Peaks concentrate at corners, parapets, roof edges and any discontinuity in the facade, and they can run several times higher than the pressure averaged over a wall. Peak suction is a gust quantity, so it comes either from the code’s gust-inclusive coefficients or from an unsteady analysis.

For simple rectangular forms the code tables give these pressures directly. For complex shapes, wind tunnel testing or numerical analysis is the only way to get them with any confidence, because the tabulated coefficients come from a small family of test geometries that your building probably is not part of.

Dynamic wind loads and vortex shedding

For slender towers with low damping, crosswind response from vortex shedding often governs. Vortices peeling alternately off each side of the building induce an oscillating crosswind force at a shedding frequency set by the wind speed and the building width. If that frequency approaches the structure’s natural frequency, the response amplifies, and the outcome ranges from occupant discomfort to structural damage.

This is why slender towers get a lock-in check on top of the peak pressure check. ASCE 7-16 makes this explicit: Section 27.1.2 sends buildings whose response is subject to across-wind loading, vortex shedding, galloping or flutter to Chapter 31 rather than the analytical procedures. A tower can satisfy every static pressure requirement in the code and still generate motion complaints on every windy afternoon.

How to calculate wind load on a building to ASCE 7

ASCE 7-16 calculates design wind pressure in two stages: convert the site wind speed into a velocity pressure at height, then multiply by pressure coefficients for the surface and system you are designing. The full procedure is five steps.

A note on editions before the formulas. Everything below is written to ASCE 7-16, the edition referenced by the 2018 and 2021 IBC. The 2024 IBC references ASCE 7-22, which moved the directionality factor K_d out of the velocity pressure equation and into the design pressure equations, revised the wind speed maps in hurricane-prone regions, and revised the terrain exposure constants. Check which edition your jurisdiction has adopted before you use these expressions.

Step 1: Risk category and basic wind speed

Assign a risk category (I to IV) from the building’s occupancy and consequence of failure, then read the basic wind speed V for your site from the map for that risk category. ASCE 7-16 publishes four maps, one per risk category, at mean recurrence intervals of 300, 700, 1,700 and 3,000 years respectively. The 3,000-year Risk Category IV map is new in ASCE 7-16; through ASCE 7-10, Risk Categories III and IV shared the 1,700-year map.

V is a 3-second gust speed at 10 m (33 ft) above ground in Exposure C. The maps are contour plots, so use the ASCE 7 Hazard Tool on the actual site coordinates. A published range will not do. As a rough sense of scale for Risk Category II: mid-90s mph across much of the non-hurricane West, 107 to 115 mph over most of the interior and eastern US, rising to roughly 170 to 180 mph on the coastal South Florida contours. Special wind regions need site-specific analysis.

One trap worth naming. From ASCE 7-10 onward the maps are strength-level, so the wind load factor in LRFD combinations is 1.0. The 1.6 factor belongs with the service-level maps in ASCE 7-05. For an allowable-stress-level speed, ASCE 7 gives V_asd = V_ult × √0.6.

Step 2: Exposure category and the velocity pressure exposure coefficient K_z

Pick the exposure category from the upwind surface roughness and fetch, then compute K_z to scale the velocity pressure from the 33 ft Exposure C reference condition to the height and exposure you care about.

• Exposure B: urban and suburban terrain, wooded areas, or other terrain with closely spaced obstructions the size of single-family dwellings or larger. Requires that roughness upwind for at least 2,600 ft or 20 times the building height, whichever is greater (1,500 ft for buildings 30 ft or shorter)
• Exposure C: open terrain with scattered obstructions under 30 ft, including flat open country and grassland. The default where neither B nor D applies
• Exposure D: flat unobstructed areas and water surfaces, including smooth mud flats, salt flats and unbroken ice. Requires that roughness upwind for more than 5,000 ft or 20 times the building height, whichever is greater

The 20h fetch clause is the one people miss. A 400 ft tower needs 8,000 ft of Exposure D roughness upwind, not 5,000 ft. Exposure A, for large city centers, was eliminated in ASCE 7-02; ASCE 7-98 was the last edition to carry it. If a reference still lists Exposure A, that reference is out of date.

K_z comes from ASCE 7-16 Table 26.10-1, or from the equation in its footnotes:

$$ K_z = 2.01 \left( \frac{z}{z_g} \right)^{2/\alpha} \quad \text{for } 15\ \text{ft} \le z \le z_g $$

$$ K_z = 2.01 \left( \frac{15}{z_g} \right)^{2/\alpha} \quad \text{for } z < 15\ \text{ft} $$

Terrain exposure constants come from Table 26.11-1. For ASCE 7-16: Exposure B, α = 7.0 and z_g = 1,200 ft; Exposure C, α = 9.5 and z_g = 900 ft; Exposure D, α = 11.5 and z_g = 700 ft. These constants were revised in ASCE 7-22, so do not carry them across editions.

Step 3: Velocity pressure q_z

Velocity pressure is dynamic pressure scaled by K_z for height and exposure, then corrected by three site factors. In ASCE 7-16:

$$ q_z = 0.00256 \, K_z K_{zt} K_d K_e V^2 \quad [\text{lb/ft}^2,\ V \text{ in mph}] $$

The four factors:

K_e is the newest of the four. It was carried in the Commentary of earlier editions and brought into the body of the Standard in ASCE 7-16.

Use q_z evaluated at height z for the windward wall, and q_h evaluated at mean roof height h for the side walls, leeward wall and roof.

Step 4: Design pressure for the main wind force resisting system

For the Directional Procedure in ASCE 7-16 Chapter 27 Part 1, which applies to enclosed, partially enclosed and open buildings of all heights, rigid or flexible:

$$ p = q G C_p – q_i (GC_{pi}) $$

• G, the gust effect factor, is permitted to be taken as 0.85 for rigid buildings, meaning a fundamental frequency of 1 Hz or higher
• C_p, the external pressure coefficient, comes from Figure 27.3-1: +0.8 windward wall, -0.7 side walls, and for the leeward wall -0.5 at L/B of 0 to 1, -0.3 at L/B of 2, and -0.2 at L/B of 4 or more, interpolating linearly between
• GC_pi, the internal pressure coefficient, is ±0.18 for enclosed buildings and ±0.55 for partially enclosed buildings, from Table 26.13-1. Both signs have to be checked, because internal pressure adds to the worst case on one surface and subtracts on another

Low-rise buildings have the option of the Envelope Procedure in Chapter 28, where p = q_h[(GC_{pf}) – (GC_{pi})] and the pressure coefficients are already gust-inclusive. ASCE 7 defines low-rise as mean roof height of 60 ft or less and no greater than the least horizontal dimension, so a 50 ft tall, 20 ft wide building is not low-rise.

Step 5: Components and cladding

Components and cladding use higher pressure coefficients than the MWFRS. C&C coefficients are keyed to effective wind area, so a window or a single fastener picks up the full local peak, while MWFRS coefficients are whole-surface values that already carry the spatial averaging across the frame. Chapter 30 covers this:

$$ p = q_h[(GC_p) – (GC_{pi})] \quad \text{for low-rise buildings, Chapter 30 Part 1} $$

$$ p = q(GC_p) – q_i(GC_{pi}) \quad \text{for buildings over 60 ft, Chapter 30 Part 3} $$

GC_p depends on the effective wind area and on which roof or wall zone the element sits in, with corner and edge zones carrying the highest suction. Skipping the C&C check is how facade failures happen on buildings whose frames were designed correctly.

Worked example: 120 ft office tower, Exposure C, 115 mph

Take an enclosed office building 120 ft tall with a 40 ft by 80 ft plan, on flat open terrain at sea level, wind normal to the 80 ft face.

Inputs: Risk Category II, V = 115 mph, Exposure C, K_zt = 1.0, K_e = 1.0, K_d = 0.85, G = 0.85, GC_pi = ±0.18.

K_z at mean roof height (120 ft, Exposure C):

$$ K_z = 2.01 \left( \frac{120}{900} \right)^{2/9.5} = 1.315 $$

Table 26.10-1 gives 1.31 at 120 ft, so the formula and the table agree to within a rounding step.

Velocity pressure:

$$ q_h = 0.00256 \times 1.315 \times 1.0 \times 0.85 \times 1.0 \times 115^2 = 37.8\ \text{lb/ft}^2 $$

Windward wall at z = h, C_p = +0.8:

$$ p = 37.8 \times 0.85 \times 0.8 = 25.7\ \text{lb/ft}^2 \ \text{(external)} $$

With internal pressure at ±0.18 × 37.8 = ±6.8 lb/ft², the windward wall design pressure at roof height is 32.5 lb/ft² with internal suction, or 18.9 lb/ft² with internal pressure. This is the top-of-wall value. K_z and therefore q_z are smaller at every level below, so applying 32.5 lb/ft² over the full height is conservative.

Leeward wall, L/B = 40/80 = 0.5, so C_p = -0.5:

$$ p = 37.8 \times 0.85 \times (-0.5) = -16.1\ \text{lb/ft}^2 \ \text{(external)} $$

Net along-wind pressure on the MWFRS: internal pressure cancels between windward and leeward faces, so the net is 25.7 + 16.1 = 41.8 lb/ft², giving a total base shear of 41.8 × 80 × 120 ≈ 401 kip if the pressure is taken uniformly over the full height. In practice windward pressure varies with height through K_z, so the frame is loaded floor by floor with q_z at each level.

Wind load calculation in SI units

The same ASCE 7-16 equation in SI, with q in N/m² and V in m/s:

$$ q_z = 0.613 \, K_z K_{zt} K_d K_e V^2 $$

Running the worked example in SI: V = 115 mph = 51.4 m/s, z = 36.6 m, K_z = 1.315, so q_h = 0.613 × 1.315 × 0.85 × 51.4² = 1,811 N/m², which is 1.81 kPa. That matches 37.8 lb/ft² to the third digit.

Wind load coefficients at a glance

Wind load calculation to Eurocode EN 1991-1-4

Eurocode reaches design pressures by a different route: it folds the gust into the pressure itself through a peak velocity pressure, and handles dynamic response separately through the structural factor. The numbers the two codes produce for the same building are not the same, because the reference speeds, averaging times, roughness models and coefficients all differ.

The chain runs basic wind velocity, mean wind velocity, peak velocity pressure, surface pressure:

$$ v_b = c_{dir} \cdot c_{season} \cdot v_{b,0} $$

$$ v_m(z) = c_r(z) \cdot c_o(z) \cdot v_b, \quad c_r(z) = k_r \ln\left(\frac{z}{z_0}\right) $$

$$ q_p(z) = \left[1 + 7 I_v(z)\right] \cdot \tfrac{1}{2} \rho \, v_m^2(z) $$

$$ w_e = q_p(z_e) \cdot c_{pe}, \qquad w_i = q_p(z_i) \cdot c_{pi} $$

Air density ρ is recommended as 1.25 kg/m³, and the terrain factor is k_r = 0.19 (z_0 / z_{0,II})^{0.07} with z_0,II = 0.05 m. The roughness expression holds for z_min ≤ z ≤ 200 m; below z_min, take c_r(z) = c_r(z_min). Terrain roughness and z_min come from Table 4.1:

Two mapping notes for anyone moving between the codes. Eurocode’s c_pe,10 for loaded areas of 10 m² or more is the rough equivalent of ASCE’s MWFRS coefficients, and c_pe,1 for areas of 1 m² or less corresponds to components and cladding. Where internal volume and permeability cannot be assessed reliably, take c_pi as the more onerous of +0.2 and -0.3. Vertical walls of rectangular-plan buildings are divided into pressure zones A to E, and dynamic response is carried by the structural factor c_s c_d.

National Annexes override several of these recommended values, so confirm against the annex for the country you are building in.

Where code calculations stop: wind tunnel and CFD

The code procedures are calibrated on a limited set of building shapes, and ASCE 7 recognizes one alternative when your building falls outside them: the Wind Tunnel Procedure in Chapter 31, with test conditions governed by ASCE 49.

Be clear about what that means for simulation. ASCE 7 does not list computational methods among its permitted procedures for determining design wind loads, and CFD results on their own are not a code-compliant substitute for Chapters 26 to 30 or for a Chapter 31 wind tunnel study. SimScale has written about this in detail: for code submittal, the calculation or the tunnel is what counts.

What CFD is for is the design decision that happens before submittal. The situations where code coefficients are least reliable are exactly the ones where the design levers are still open:

• Tapered, twisted or setback towers, where no tabulated C_p applies
• Interference from neighboring buildings, which ASCE 7 has no explicit provisions for and which EN 1991-1-4 treats only coarsely in Annex A.4
• Crosswind and torsional response on slender forms
• Local peak suction at corners, parapets, balconies and canopies
• Comparing corner treatments or shape options before committing to a tunnel campaign

Wind tunnel testing answers the same questions with code standing, at a cost and lead time that means you get one or two geometries, late. CFD answers them for as many geometries as you want, early, and the tunnel then confirms the option you chose. The CAARC facade pressure validation case shows how close the two methods sit on surface pressure for a standard tall-building geometry, and SimScale’s pedestrian wind comfort validation study does the same for the velocity field against wind tunnel and field measurements.

Wind load analysis with CFD

CFD models a virtual wind tunnel numerically, and returns pressure, force and velocity fields across the whole domain. A physical model gives you data only at the pressure taps you instrumented. Separation and recirculation are captured in the flow field, and the atmospheric boundary layer is imposed directly as an inlet profile, which makes multi-direction runs driven by a wind rose a setup choice rather than a new test campaign.

Solver choice matters here. Steady RANS gives you mean surface pressures cheaply and is what you want for a shape comparison. Unsteady shedding, crosswind response and peak suction need a transient run, and the lattice-Boltzmann method is what makes those tractable at building scale on cloud hardware.

For a working engineer the practical difference is iteration count. A virtual wind tunnel running in the browser puts wind load analysis inside a design cycle instead of at the end of one.

Thornton Tomasetti, a global engineering consultancy with more than 1,500 staff, built exactly that into their own workflow. Their CORE Studio team wrapped the SimScale API in an in-house digital wind tunnel app that runs from Rhino, so wind loading is available to designers who are not CFD specialists.

Project spotlight: vortex shedding on a 50-story tower

We ran a 50-story tower at 45 m/s to see how much corner geometry changes the crosswind load. The tower is 150 m tall on a fixed 20 m by 20 m square base. Two designs: an initial version with sharp corners, and a second with the corners rounded. Both use a transient incompressible turbulent flow analysis, because the quantity of interest is unsteady.

The sharp-cornered design produces strong, organized vortex shedding, and with it high-amplitude periodic crosswind forces. The shedding frequency works out at roughly 0.23 Hz, uncomfortably close to the 0.2 Hz typical natural frequency of a 50-story building [2]. That is a lock-in risk.

The corner-softened design produces weaker vortices and low-amplitude crosswind forces. Rounding the corners cut the wind-induced dynamic forces in the crosswind direction substantially, and with them the risk of damage. The geometry change was free to make at concept stage, before the structural system was fixed. Recovering the same reduction later means supplemental damping, and that comes out of the structural budget and the floor plate.

Wind load design strategies that reduce wind effects

The design modifications that cut wind effects most work by breaking up or suppressing vortex formation:

• Flow spoilers or deliberate surface disturbance
• Corner softening, chamfering or recessing
• Tapering with height, or varying the cross-section shape up the building
• Porosity: open floors, bleed slots, or through-building openings

Studied during the design cycle, these modifications can reduce wind-induced forces by 25 to 60% across the cases Irwin reports [1]. That is a larger reduction in load than the structural system can usually recover afterward, and it is only on the table while the massing is still open.

How much shape actually buys you: a supertall in Shenzhen

Adrian Smith + Gordon Gill Architecture put numbers on this during a design competition for a supertall tower in Shenzhen. Working from a baseline rectangular extrusion, the team compared forms in SimScale and measured total wind force for each:

The finding that mattered was counterintuitive for that site: stepping back the mass at the top beat tapering by more than five to one. Tapering is the more common instinct, and on this site it bought almost nothing. Adding wind relief slots to mitigate vortex shedding took the total to 34.85%, and the slots doubled as housings for vertical axis wind turbines.

None of that was knowable from a code calculation. It came out of running the options.

Wind load analysis software: what to look for

Wind load software splits into two categories that solve different problems, and buying the wrong one is a common mistake.

Code calculators automate the ASCE 7 or Eurocode procedure: you enter site data and building dimensions, and they return q_z, pressure coefficients and design pressures per surface. They are cheap, fast, and submittable. What they cannot do is tell you anything the code tables do not already contain, which leaves them silent on exactly the geometries where you most need an answer.

Simulation platforms solve the flow field. They handle arbitrary geometry, neighboring buildings, transient response and local peak pressures. They are not a substitute for the code calculation on submittal, and they are the only way to compare shape options at the pace design actually moves.

If you are evaluating simulation tools for wind loading, the criteria that matter are:

• Validation evidence against wind tunnel data for building geometries, published and reproducible
• Transient capability, because vortex shedding and crosswind response are unsteady by definition
• Atmospheric boundary layer and wind profile setup, including multi-direction runs driven by a wind rose
• Parallel runs, so a five-option shape comparison finishes in the wall-clock time of one
• Import from the CAD you already use, Rhino and Revit included, without a remodeling step
• Facade-level pressure extraction as well as global forces, so the C&C conversation is supported

SimScale covers these in the browser on an AI-native cloud platform, with no hardware to provision and no license to check out. Explore wind simulation on SimScale or start a free account and run the first case today.

How to get started with wind load analysis

Start with the code calculation, then simulate the cases the code cannot see. For a rectangular building on flat terrain, Chapters 26 to 30 of ASCE 7-16 will give you defensible design pressures in an afternoon, and the worked example above is enough to check your work.

Once the geometry deviates, add simulation. CFD was once reserved for specialists with access to HPC hardware, and that is no longer true: building simulation now runs from a browser, and the SimScale Public Projects Library carries templates across wind engineering, including pedestrian wind comfort, tank farm wind loading, pollution control and natural ventilation. For a structured walkthrough of mean and peak wind load extraction, the wind loading simulation for structural design session covers the workflow end to end.

Frequently asked questions

References

1. Irwin, P. A. Wind Issues in the Design of Tall Buildings. RWDI, Los Angeles Tall Building Structural Design Council, May 7, 2010.
2. Irwin, P. A. Vortices and Tall Buildings: A Recipe for Resonance. American Institute of Physics, 2010. S-0031-9228-1009-350-6.
3. ASCE/SEI 7-16, Minimum Design Loads and Associated Criteria for Buildings and Other Structures, Chapters 26 to 31.
4. ASCE/SEI 49, Wind Tunnel Testing for Buildings and Other Structures.
5. EN 1991-1-4:2005, Eurocode 1: Actions on structures. Part 1-4: General actions. Wind actions.