Presets
Playback
Running
Speed of the train and playback. Wire deflections are centimetres, so the picture is exaggerated.
Catenary
Tension and mass set the wave speed; dropper count shapes the span stiffness.
Pantograph
Lift force, head mass, head suspension and frame damping.
Layers
Model checks
The science
The contact wire is a string
The overhead contact wire is a long, heavily tensioned metal string. A pantograph sliding under it pushes up with a force of tens of newtons and launches transverse waves that travel at
c = sqrt(T / rho)
where T is the tension and rho the mass per metre. For a contact wire of 20 kN and 1.35 kg/m this is about 120 m/s (440 km/h). The messenger wire above it, the droppers and the mast supports make the real wave speed a little lower: this page uses c = sqrt((Tc + Tm) / (rho_c + rho_m + droppers)). A train cannot sensibly run near c: standards ask for v < 0.7 c.
Doppler factor
The pantograph moves relative to the wire at speed v, so a wave reflected from a mast or dropper is Doppler shifted. Waves ahead of the train are compressed and waves behind it stretched; the standard measure is
alpha = (c - v) / (c + v)
which falls from 1 at standstill to 0 at v = c, where the pantograph keeps pace with its own wave and the force swings become violent.
Why the force oscillates once per span
Droppers hang the contact wire from the sagging messenger. Between two masts the wire is softest in the middle of the span and stiffest near the masts and steady arms, so a constant push lifts the wire by different amounts and the contact force oscillates at the span-passing frequency f = v / L (60 m span: 1.4 Hz at 300 km/h). A dropper can only pull; when the wire is pushed up it slackens, a nonlinearity that this model includes.
The model
Two strings (contact wire and messenger, 0.5 m grid, 720 nodes, a 360 m ring of six spans) obey rho y_tt = T y_xx + forces. Droppers (stiffness 80 kN/m, tension-only) join them, the messenger is hung from masts through stiff springs, and the contact wire is held at the masts by steady arms. The pantograph is two masses: a head (mass m_h, suspension k_h) on a frame lifted by a constant force F0 through a damper. Contact is a one-sided spring of 60 kN/m: F = k_c max(0, y_head - y_wire). By default the lift follows the European standard EN 50367 mean force F = 70 + 0.00097 v^2 newtons.
Numerical checks on this page
(1) A pulse is sent along a clean numerical string and its arrival time gives the wave speed, compared with sqrt(T/rho). (2) A pinned string loaded at mid-span is relaxed with damping and its deflection compared with the analytic F L / (4 T). (3) Over three laps of the ring the mean contact force must equal the lift set-point, since a wire cannot hold a net force on average.
Honest simplifications
This is a 1D vertical teaching model of one pantograph on a periodic ring, not a validated catenary code: no bending stiffness or damping measured on a real line, no lateral (stagger) dynamics, no aerodynamic force on the pantograph beyond the standard mean-force law, no multiple pantographs, no wire irregularities. The wire is shown with a visual exaggeration (slider) because real deflections are centimetres. Because the ring repeats every 360 m, waves launched in one lap are met again later, like a second pantograph; sweep statistics discard the first third of each run. Numbers are plausible orders of magnitude, not data for a particular line.
What to try
- Load Express, 250 km/h and watch graph 1: the force oscillates once per span. Then Near the wave speed: swings grow and the force touches zero (arc).
- Compare Slack wires at 300 km/h (low tension, small c) with High-speed line, 320 km/h: the same speed is calm on the high-tension wire.
- Switch the EN 50367 rule off and set a low uplift force: the contact is lost earlier. Raise it and the mean force climbs but wear increases.
- Use Sparse droppers: fewer droppers make the span softer in the middle and increase the force variation.
- Read graph 3: the standard deviation and the min/max band as speed approaches the red c line.