Module 16 · Lesson 16.2

Lateral stability: four systems, and why the best one loses

A braced bay is 17 times more efficient than a moment frame and is regularly rejected. Understanding why is most of what a multi-storey stability decision involves.

Why this matters

Every multi-storey building needs something to stop it swaying, and there are only four common answers. On pure engineering terms one of them wins overwhelmingly — a braced bay is nearly seven times stiffer than a moment frame for a fraction of the cost. It is also the one that puts a diagonal member through a floor that somebody has to let, walk through and furnish. The decision is a genuine negotiation, and being able to state what each option costs in the other's currency is what makes a structural engineer useful in it.

By the end of this lesson you should be able to

Describe the four common lateral systems and how each resists sway
Compare them on stiffness, cost and plan freedom
Say why a braced bay is so much more efficient
Say when a moment frame or an eccentric brace is nevertheless right

What you should already know

Frame stability, α_cr and the amplified sway method (Module 9)
Joint stiffness and classification (Modules 3 and 13)
Base fixity (Module 14)

Four ways to stop a building swaying

System	Relative stiffness	Relative cost	Obstructs plan	Ductile
Concrete core	2.5	0.6	no	no
Braced bay	1.0	1.0	yes	no
Eccentric bracing	0.6	1.4	yes	yes
Moment frame	0.15	2.5	no	yes

The figures are indicative rather than design values, but the ranking is real and the reason for it is mechanical.

A braced bay resists sway axially. The diagonal simply stretches or shortens, and axial stiffness is EA/L — enormous.

A moment frame resists sway in bending. The beams and columns bend in double curvature, and bending stiffness is 12EI/h³ — much smaller, and it falls with the cube of the storey height. It also needs moment-resisting connections at every joint, and Module 13 showed what those cost to make and how far short of rigid they still are.

A braced bay is 6.7 times stiffer than a moment frame at 0.4 times the cost — about 17 times the stiffness per unit cost.

And it is regularly rejected, because of one column in that table.

Take care

The best engineering answer is not always the right answer

A braced bay puts a diagonal member across a bay of the floor plate. That bay cannot then be a lettable open office, a glazed façade, or a route from one side of the building to the other.

Which is why braced bays cluster around lift cores and stair wells — places where the plan is already obstructed and nobody minds. When the core is big enough and in the right place, that is the whole stability system and it is nearly free.

When it is not — a long thin building, a core at one end, a client who wants an unobstructed floor — the choice becomes a real one:

Concrete core if there is a core and the programme can take it being built first. Stiffest and cheapest, and it fixes the stability system's position at the design's earliest stage.
Moment frame if the plan must stay clear. Expensive and much less stiff, and Module 9's α_cr will need watching.
Eccentric bracing if energy dissipation is needed — a short link beam yields in shear as a designed fuse before the braces buckle. That matters in seismic design and rarely otherwise.

The structural engineer's job here is not to win the argument for the braced bay. It is to say what the alternatives cost, in stiffness, in steel and in α_cr, so that the decision is made knowingly.

Try it

Four systems, compared on the terms that decide

Relative lateral stiffness for each system, with the two you are comparing highlighted. The readout gives what each costs in the other's currency — including what it does to α_cr.

System A

System B

α_cr on a braced bay12.5

A relative stiffness: 1.00
B relative stiffness: 0.15
Stiffness ratio: 6.67 ×
Cost ratio: 0.40 ×
Stiffness per unit cost: 16.7 ×
A obstructs the plan: yes
B obstructs the plan: no
A is ductile: no
B is ductile: yes
α_cr with A: 12.5
α_cr with B: 1.9
Second-order needed with A: no
Second-order needed with B: YES

A braced bay is 6.7 times stiffer than a moment frame and 2.5 times cheaper — 16.7 times the stiffness per unit cost. But one obstructs the floor plate and the other does not, and on a letting plan that is frequently the whole argument.

Things worth trying

Start with braced bay against moment frame. 6.7 times the stiffness at 0.4 times the cost — nearly 17 times the stiffness per unit cost, which is not a marginal preference.
Now look at the α_cr rows. A building at 12.5 on a braced bay would be at 1.9 on a moment frame: not merely 'needs second-order analysis' but close to having consumed its stiffness entirely.
That factor of six is why the stability system is chosen before anything else. Every member depends on α_cr, and α_cr depends almost entirely on this choice.
Compare the concrete core with the braced bay. The core is stiffer AND cheaper, because it is already being built for the lifts — 4.2 times the stiffness per unit cost.
But look at the note: it fixes the stability system's position at the design's earliest stage and has to be built ahead of the steel. The cost is programme, not material.
Compare eccentric with concentric bracing. The eccentric option is less stiff and dearer, and it is chosen anyway where energy dissipation matters — the link beam is a designed fuse.
Look at the 'obstructs the plan' rows for the braced bay. That single yes is why the best engineering answer is regularly rejected, and why braced bays end up round lift cores.
Drop the base α_cr to 6 and compare again. On a marginal building the choice stops being economic and becomes a question of whether the frame stands up at all.

Worked example

Choosing a stability system for the ten-storey office

Given

The building from the last lesson: ten storeys, 1200 m² per floor, 3.5 m storey height
A central core containing lifts and stairs, 8 m × 6 m in plan
The client requires clear, column-free letting floors outside the core zone
Total vertical load at the lowest storey 21 000 kN; wind 900 kN at that level
Storey drift under that wind, on a trial arrangement: 12 mm

Find

Which stability system to adopt, and whether the frame is stable.

Assumptions

Relative stiffness and cost figures are engineering recommendations, not design values
The α_cr threshold is calibrated and unverified here
The core is assumed available for stability — a programme assumption as much as a structural one

Practice

A braced bay has relative stiffness 1.0 at relative cost 1.0; a moment frame has 0.15 at 2.5. How many times more stiffness per unit cost does the braced bay give?

Practice

A storey carries V = 21 000 kN, is 3.5 m high, and drifts 12 mm under a horizontal load of 900 kN. What is α_cr?

Practice

That building is rebuilt on a moment frame with 0.15 times the lateral stiffness. What is α_cr now?

Practice

A concrete core has relative stiffness 2.5 at relative cost 0.6. How many times the stiffness per unit cost of a braced bay is that?

Check yourself

Why is a braced bay so much stiffer than a moment frame for the same steel?

Summary

Four systems: concrete core, braced bay, eccentric bracing, moment frame
A braced bay resists sway axially; a moment frame bends — hence the order of magnitude
Braced bay against moment frame: 6.7× the stiffness at 0.4× the cost, 17× the efficiency
And it is rejected whenever the floor plate must stay clear
So braced bays cluster round cores and stairs, where the plan is obstructed anyway
A concrete core is stiffer and cheaper still — 4.2× a braced bay per unit cost
The worked building had α_cr = 12.5 on a core and would have had 1.9 on a moment frame
So the stability system is chosen FIRST, because everything else follows from α_cr

Progress is kept in this browser only.

This is educational material. It uses simplified examples to teach principles, and must not be relied on for real design or safety-critical decisions. Module overview and checkpoint