What you’ll explore
How does a giant ball keep a skyscraper still? In this lab you stop a skyscraper from swaying with a tuned mass damper and test your intuition against a real physics simulation.
Challenge levels
- No damper. Gusts push the tower in step with its own 6.8 s sway. On its own, that push would only lean the top 0.8 cm. The tower’s own damping is 2%. After 5 minutes, how far does the top sway, in cm?
- Add the ball. Now hang a damper near the top: a mass 2% of the tower’s moving mass (like Taipei 101’s 660-metric ton ball), tuned to the tower and linked through a damper. Same wind. How far does the top sway now, in cm?
- Tune it. Winds come in every rhythm. Tune the damper’s own rhythm (as a fraction of the tower’s) so the worst sway, over every wind rhythm, is at most 6.5 cm.
- What if it’s mistuned?. What if the damper is tuned 20% too fast (its rhythm 1.2× the tower’s)? Same resonant wind. How far does the top sway now, in cm? (Tuned: 5.2 cm.)
- After the gust. A typhoon gust leaves the tower leaning 30 cm and then stops. With the tuned damper, how many seconds until the sway stays under 3 cm? (Without it: about 123 s.)
Key ideas
Show the explanations (spoilers: try the levels first)
- At resonance the steady sway is the static lean ÷ (2ζ) = 0.8 cm ÷ 0.04 = 20 cm (19.9 cm after 5 minutes, still settling). Pushes in step with the sway keep adding energy.
- The tuned ball swings a quarter-cycle behind the tower, so its pull opposes the wind and its damper turns the sway into heat. The top sways about 5.2 cm instead of 20: about 4 times less.
- Den Hartog’s rule: tune the damper slightly slower than the tower, f = 1/(1 + μ) = 1/1.02 ≈ 0.98. Then the two response peaks come out equal and as low as possible.
- Out of tune, the ball no longer lags the tower by a quarter-cycle, so it cancels much less of the push. The sway rises to about 9.9 cm, almost twice the tuned value.
- Without a damper the sway envelope shrinks by e^(−ζωt) with ζ = 2%: ln(10)/(0.02 × 0.924/s) ≈ 125 s to drop tenfold (the last swing above 3 cm ends at 123 s). The tuned ball pulls energy out every cycle: about 28 s.
About the model
- Each tower is its first sway mode only: a moving (modal) mass m1 = 33,000 t on a spring, period 6.8 s, with 2% structural damping (real tall buildings are about 1–2%). Taipei 101’s real damper is 660 t, so μ = m2/m1 = 2% is used; the tower’s true modal mass is not published here, so treat the numbers as representative.
- The damper is modeled as a mass on a linear spring and a linear viscous damper. Taipei 101’s ball is a pendulum on cables with hydraulic dampers; for small swings a pendulum acts exactly like this spring.
- Wind is a steady sinusoidal push at the top (static lean 0.8 cm), switched on at t = 0. Real wind is random gusts plus vortex shedding; the steady rhythm isolates resonance.
- Motion is solved exactly: steady forced response from the complex transfer function plus the free response via a matrix exponential. Both towers are checked against an independent RK4 in newtons.
- Sway is drawn about 150× exaggerated (20 cm on a 508 m tower would be invisible). The ball’s own swing is drawn at a smaller scale (≈ 1.2 units per meter, kept inside the tower); the HUD and graph show real values. Ball size is not to scale.
- Den Hartog’s tuning rule (1956) assumes an undamped tower; with 2% tower damping it is still near-optimal.
Every number on screen comes from the simulation, and each lab passes automated checks against the physics formulas. Spotted a mistake? Tell us.