The weight that fights the sway

A tuned mass damper is a heavy weight mounted high on a building and designed to fight it. Tuned right, the weight sways against the building and soaks up the shaking. Tuned to the wrong frequency, it does little — or makes the sway worse.

A weight on springs

Engineers picture two masses. The building is one mass on one spring, M and stiffness K, standing in for its dominant sway mode. On top sits a smaller mass m on its own spring k, with a dashpot c beside it: anything that turns motion into heat. Mazzon, Frappa and Pauletta describe it that way; in practice it can be a pendulum or a concrete slab on rubber isolators.

The trick is phase

The paper calls a tuned mass damper a passive supplemental energy dissipation device, and the trick is phase. Its frequency is set to that of the building's main sway mode, so the weight moves opposite the building. Its inertia pushes back on the structure instead of being dragged along, so the shaking's energy goes into the damper rather than into beams and walls. The paper's reason for caring is blunt: less inertial force in the members.

A slab on sliders, six storeys up

The case study is a six-storey reinforced-concrete building in Tolmezzo, Italy — the abstract says five, section 2 says six — its plan 18.15 m by 11.90 m, storeys 3.00 m, the ground floor 3.30 m. The damper is the slab on top: reinforced concrete, cast to overlook the roof, on flat surface sliders anchored to a steel frame, free to sway out of phase with the building below. Stiffness and damping come from lead rubber isolators, which supply both, or from low-damping ones with viscous linear dampers. The aim: shift it from a dissipative response, where its own members yield to survive, to a non-dissipative one.

What it bought, and what it cost

Untouched, its first modes sway with periods of 0.35 s to 0.41 s: stiff and quick, on the worst part of the site's spectrum, where it sees 5.07 to 6.60 m/s². With the damper they split: one group stretches to between 0.62 s and 0.47 s, off the worst of the spectrum, and the rest stay on the plateau and gain only through damping. Its first scheme, TMD1, puts the whole building under lower accelerations. The damper raises the total input energy yet halves what the building sheds in its own damping, because the devices take most of it. One caveat they give against themselves: the walls carrying most of the seismic force are still over capacity.

The catch: a damper can be detuned

Its most useful finding is aimed at its own proposal: tuning holds only while the building keeps the frequency it was measured at. Neglect something in the design — soil and structure interacting, say — and members can plasticise: a building that yields gets softer and slower. The resonance between the damper and the building's main modes then vanishes and the device is detuned. Lose a fifth of the stiffness, in our own worked example, and the frequencies fall by the square root of 0.8: about 11 per cent. Detuning also arrives quietly, through added mass or stiffness lost as its isolators degrade. A device tuned to one frequency cannot follow the building once the building moves off it.

In short

A tuned mass damper is a weight on springs and dampers, mounted high and timed to move against the building's main sway. It stiffens nothing; it takes the motion and pays for it in its own damper, and the modes it helps most are the ones near the frequency it was tuned to. Hence the room the paper gives detuning.