← Civil & Structural Studio
Concept Explainer · Civil & Structural

Rigid vs. Flexible Diaphragm — How Lateral Load Actually Splits Between Shear Walls

Two shear walls of very different stiffness can end up carrying an even 50/50 split of lateral load, or a heavily lopsided share — depending entirely on how stiff the floor connecting them is.

A diaphragm is the horizontal floor or roof plane that collects lateral load (wind or seismic) and delivers it to the vertical elements below — shear walls, braced frames, or moment frames. It seems reasonable to assume that a stiffer shear wall should simply attract more load, but that's only true if the diaphragm itself is stiff enough to force every wall to deflect together. If the diaphragm is comparatively flexible, it can't enforce that — it just spans between the walls like an ordinary beam, and each wall only ever sees the load tributary to it, no matter how stiff it is relative to its neighbors.

Flexible diaphragm — untopped metal deck / wood roof

Distributes by tributary area
PLAN VIEWuniform lateral load w (wind / seismic)Wall Ashort, k = 1×Wall Blong, k = 3×BEAM ANALOGY — diaphragm spans like a simply-supported beamR_A = 50%R_B = 50%
Load split
50% / 50%
Even split by tributary length, even though Wall B is 3× stiffer than Wall A.
Torsion considered?
No
A flexible diaphragm can't transfer torsional shear the way a rigid one can.

Rigid diaphragm — same building, cast-in-place concrete slab

Distributes by relative stiffness + torsion
PLAN VIEWuniform lateral load w (wind / seismic)Wall Ak = 1×Wall Bk = 3×CMCReccentricity etorsional rotation about CRSPRING ANALOGY — rigid bar forces equal deflection at both wallsR_A ≈ 25% + torsionR_B ≈ 75% + torsion
Load split
≈ 25% / 75%
Split in proportion to relative stiffness — the 3×-stiffer Wall B attracts 3× the direct share.
Torsion considered?
Yes — required
Center of rigidity ≠ center of mass here, so the rigid diaphragm also rotates, adding torsional shear on top of direct shear.
Property
Flexible
Rigid
Distribution method
Tributary area (like a simple-span beam)
Relative stiffness (like rigid bar on springs)
Torsion / accidental eccentricity
Generally not transferred through the diaphragm
Must be checked — rotates about center of rigidity
Typical construction
Untopped metal deck, plywood/OSB wood sheathing
Concrete slab, concrete-filled metal deck
Code classification basis
ASCE 7 §12.3.1 — computed or permitted-idealization
ASCE 7 §12.3.1 — max Δdiaphragm ≤ 2× avg story drift
Deflects relative to walls
Significantly — diaphragm is the flexible element
Negligibly — walls are the flexible elements
Why this works

The diaphragm can only enforce equal deflection if it's stiff enough to act as a rigid body.

A rigid diaphragm distributes load by relative stiffness because it behaves like a rigid bar sitting on top of a row of springs (the shear walls) — since the bar doesn't bend, every spring is forced to compress the same amount, and a stiffer spring pushes back harder for that same deflection, so it ends up carrying proportionally more of the load: V_wall = (k_wall / Σk) × V_total. A flexible diaphragm can't do that — it has too little in-plane stiffness of its own to force the walls beneath it to move together, so it simply sags between them like an ordinary simply supported beam, and each wall only ever picks up the load tributary to its own portion of that span, regardless of how stiff it happens to be relative to its neighbors. Which behavior actually governs isn't a property of the diaphragm alone — it's a comparison between the diaphragm's in-plane stiffness and the stiffness of the vertical elements it's spanning between.

Common misconception
"Concrete diaphragms are always rigid, and wood or untopped metal deck diaphragms are always flexible."

That correlation holds often enough in practice that it becomes a mental shortcut, but it isn't what the classification actually depends on. ASCE 7 §12.3.1 defines diaphragm flexibility as a relative comparison: a diaphragm is flexible if its computed maximum in-plane deflection exceeds twice the average story drift of the shear walls or frames it's delivering load to. That means rigidity isn't a fixed property of the diaphragm material alone — it's a comparison between two stiffnesses. A concrete slab spanning between very stiff, closely spaced shear walls could, in principle, deflect enough relative to those walls to fail the rigid test, while the code separately permits certain wood-sheathed or untopped metal deck diaphragms to be idealized as flexible without calculationunder specific conditions — a permitted simplification, not proof that the underlying physics only ever goes one way by material type. When wall stiffnesses are similar, or the exact split matters for a governing design case, the material alone isn't enough to answer the question — the relative deflection comparison is.

Related Concept Explainers
Load Path & Tributary Area
Read it →
Moment Frame vs. Braced Frame
Read it →
Wind vs. Seismic Load Combinations
Read it →
One-Way vs. Two-Way Slab
Read it →

Rigid vs. Flexible Diaphragm — Concept Explainer

Explains how a lateral load (wind or seismic) actually splits between shear walls of unequal stiffness depending on the horizontal diaphragm connecting them — a flexible diaphragm (untopped metal deck, wood structural panels) distributes load by tributary area like a simply supported beam, while a rigid diaphragm (concrete slab, topped metal deck) distributes it in proportion to each wall's relative stiffness and must also be checked for torsion from any eccentricity between the center of mass and center of rigidity. Covers the ASCE 7 §12.3.1 quantitative flexibility criterion and why the classification is relative, not purely material-based.

Two Different Distribution Models

A flexible diaphragm has too little in-plane stiffness to force the vertical elements beneath it to deflect together, so it behaves like an ordinary simply supported beam spanning between them — each wall picks up only the load tributary to its portion of that span, R = w × (tributary length), independent of wall stiffness. A rigid diaphragm has enough in-plane stiffness to move essentially as a single rigid body, which forces every wall beneath it to deflect the same amount at any given point; because a stiffer wall resists that shared deflection with more force, it ends up carrying a proportionally larger share of the total load: V_wall = (k_wall / Σk) × V_total, the relative-rigidity method.

Torsion Is the Other Half of the Story

A rigid diaphragm doesn't just redistribute load by stiffness — if the building's center of mass (where the inertial or wind force effectively acts) doesn't coincide with its center of rigidity (the stiffness-weighted centroid of the vertical elements), the diaphragm also rotates about the center of rigidity under load. That rotation adds a torsional shear component to each wall on top of its direct-shear share, which can increase demand substantially on walls far from the center of rigidity. Flexible diaphragms generally aren't treated as capable of transferring this torsional redistribution, since they lack the in-plane stiffness to act as the rigid rotating plane the calculation assumes.

How the Classification Is Actually Made

ASCE 7 §12.3.1 defines the test quantitatively: a diaphragm is flexible if its computed maximum in-plane deflection exceeds twice the average story drift of the associated vertical elements of the seismic force-resisting system. This is a relative comparison, not a fixed material property — the same diaphragm could classify differently depending on how stiff the walls or frames beneath it are. The code also permits certain configurations (untopped steel decking or wood structural panel diaphragms in specific low-rise, wood/steel/light-frame conditions) to be idealized as flexible without an explicit calculation, and separately permits rigid idealization for others — practical simplifications layered on top of the underlying relative-stiffness physics.

Frequently asked questions

How do I actually determine if a diaphragm is rigid or flexible for design?

Per ASCE 7 §12.3.1, compute the diaphragm's maximum in-plane deflection under the applicable lateral load and compare it to the average story drift of the vertical elements it's delivering load to. If the diaphragm deflection exceeds 2× that average drift, it's classified flexible; otherwise it can be treated as rigid. The code also lists specific conditions under which certain diaphragm types can be idealized as flexible or rigid without this calculation.

Does diaphragm flexibility change which walls need to be designed stronger?

Yes, often significantly. Under a rigid-diaphragm assumption, the stiffest wall attracts the largest direct-shear share and torsional shear can add substantially more to walls farther from the center of rigidity. Under a flexible-diaphragm assumption, every wall simply gets its tributary share regardless of stiffness. Using the wrong assumption for a given diaphragm can under- or over-design specific walls relative to what the structure will actually experience.

Can untopped metal deck be assumed flexible without running the deflection calculation?

Yes, under specific conditions ASCE 7 permits certain diaphragm types — including untopped steel deck and wood structural panel diaphragms in qualifying low-rise wood, steel, and light-frame construction — to be idealized as flexible without an explicit deflection comparison. This is a permitted simplification for common, well-understood configurations, not a claim that all such diaphragms are inherently flexible in every situation.

What is a semi-rigid diaphragm, and when is it used?

A semi-rigid (or semi-flexible) diaphragm model uses the diaphragm's actual computed in-plane stiffness — via finite-element or similar analysis — rather than idealizing it as fully rigid or fully flexible. It's used when the diaphragm doesn't clearly satisfy either idealization, such as long, narrow, or irregularly shaped diaphragms, and produces a load distribution somewhere between the tributary-area and relative-stiffness extremes.

Why does accidental torsion matter even when the building looks symmetric?

Codes require an assumed accidental eccentricity (commonly 5% of the diaphragm dimension) in addition to any calculated eccentricity between center of mass and center of rigidity, to account for uncertainties like unaccounted-for mass distribution, construction variability, and non-uniform stiffness that a purely geometric symmetric layout wouldn't otherwise capture. This accidental torsion is only meaningfully transferred through the diaphragm when it's classified as rigid (or analyzed as semi-rigid).

🎓

Try our Civil & Structural Studio

More calculators, simulators, and guides for this discipline.

Related tools & guides

Shear Wall Design CalculatorSeismic ELF Design SimulatorFrame Deflection SimulatorSeismic Design Categories: What IBC and ASCE 7 Require