The Haunted Door: Hinge Creak Resonance (3D)

A real door hinge in a candle-lit room: stick-slip friction at the pin rings the hinge bracket. Drag the door or let it swing.

Switch to 2D

Presets

Playback

Parameters

Camera

Layers

Quality

Model check

The science

What is modelled

Two things happen at a dry hinge. First, the pin grips: the door is pushed through a slightly springy contact, the pin sticks while torque builds, then breaks free and slips until the contact grips again (stick-slip). Second, every sudden change of friction torque rings the hinge bracket like a struck bell, and that ringing is the sound you hear. The creak therefore has two clocks: the slip rate sets the roughness, the bracket sets the pitch.

Equations

  • Friction limits: Ts = μs N r and Tk = μk N r, with pin radius r = 6.5 mm and hinge load N = 0.3 m g (door weight times the lever arm of its centre of mass over the hinge spacing).
  • Stick: the pin does not move while |kc(θd − θ)| ≤ Ts. The torque ramps up at kc ω, so the first slip comes after t = Ts / (kc ω).
  • Slip: I θ'' = kc(θd − θ) − Tk sgn(θ'); the pin re-sticks when its velocity returns to zero and the spring torque is below Ts.
  • Bracket resonator: Ib y'' + c y' + kb y = Tf(t) with natural frequency f = (1/2π) √(kb/Ib), damping ratio ζ = c / (2 √(kb Ib)) and damped pitch fd = f √(1 − ζ2). Heavier doors (bigger Ib) creak lower, stiffer brackets higher, rust raises ζ and so shortens the ring.
  • Friction coefficients: μs = (0.05 + 0.85(1 − lube)) (0.5 + 1.1 rust), μk = μs (0.30 + 0.55 lube). These are plausible fits, not measured values.

Checks run in the page

  • The bracket alone is hit with a torque step; the period of the ringing and the ratio of successive peaks are compared with the closed-form damped-oscillator values.
  • The time of the first slip under a steady drive is compared with Ts/(kcω).
  • Energy balance: work done through the contact spring equals stored spring energy plus kinetic energy plus heat generated in the pin. A live Fourier transform of the bracket motion also shows the peak at the predicted pitch (graph 3).

Honest simplifications

  • One hinge pin stands for three; the door is lumped into one effective pin inertia and one contact spring, so the real mode shapes of a door are not modelled.
  • Friction is Coulomb with a static and a kinetic value plus small random variation from rust; real surfaces also have velocity-dependent (Stribeck) friction and wear.
  • The bracket is a single mode; real hinges ring at several frequencies. Sound is a sawtooth at the damped pitch whose level follows the computed bracket speed; it is not a recording.
  • Door angle changes are minute during one slip, so the visible shake is exaggerated (switchable).

What to try

  • Raise lubrication to 95 percent: mu_s falls, the torque drop per slip shrinks and the creak fades.
  • Take the Heavy oak door preset: the pitch drops because the effective inertia rises.
  • Slow the swing to 3 deg/s with a soft contact: the slips become fewer and louder pulses.
  • Slow time down (time scale 0.02) and watch the stick-slip sawtooth in graph 1 and the ringing in graph 2.
Loading model...
Loading Blender model...
Pin friction torque and spring torque (stick-slip sawtooth)
Hinge bracket twist speed (ringing)
Spectrum of the creak with the predicted pitch
Made with 🟠 Blender · Three.js