Evacuation Corridor: Convoy Flow, Checkpoint Bottleneck and Stop-and-Go Waves

A convoy of trucks, medical vehicles and buses on a 2 km road through a damaged town: car-following physics, a checkpoint that limits speed, weather and a braking lead vehicle.

Switch to 2D

Presets

Playback

Parameters

Camera

Layers

Quality

Model check

The science

What is modelled

A convoy of trucks, medical vehicles and buses drives a 2 km road (shown folded into four legs and three hairpin turns) to a safe zone. Each driver follows the vehicle in front with the Intelligent Driver Model (IDM), a standard car-following equation from traffic engineering. The road is a ring: a vehicle that reaches the safe zone is counted and re-enters at the start, so the density stays constant. A checkpoint limits the speed to 6 m/s in a zone of 230 m, and the lead vehicle can be braked to a stop to create a shock wave.

Equations

  • Acceleration of a vehicle with speed v, bumper gap s to the leader and closing speed Δv: a = amax [1 − (v/v0)4 − (s*/s)2] with the desired gap s* = s0 + v T + v Δv / (2 √(amax b)). Here s0 = 3 m, amax = 1.5 m/s², b = 2 m/s² (comfortable braking), T = 1.2 s (time headway), and every vehicle is 9 m long.
  • Equilibrium (all gaps equal, no relative speed): 1 − (v/v0)4 = ((s0 + vT)/s)², giving a speed ve(s) and a flow Q = ve / (s + length) per lane. This curve is the fundamental diagram in the first graph.
  • Free road: dv/dt = a (1 − (v/v0)4) has the exact solution t = v0/(2a) (atanh u + atan u) with u = v/v0.
  • Checkpoint: for positions from 80 m before the checkpoint to its end the desired speed is min(v0, 6 m/s). The best flow the zone can carry is the maximum of ve(s)/(s + length) with v0 = 6 m/s, which is the capacity shown in the readout.
  • Weather multiplies the desired speed (1, 0.72, 0.55), lengthens the headway (1.2, 1.6, 2.0 s) and softens acceleration and braking (1, 0.8, 0.6). Integration uses a 0.05 s step with a trapezoidal position update.

Checks run in the page

  • Free-road acceleration time against the exact integral.
  • An evenly spaced ring at the equilibrium speed stays there and carries the analytic flow.
  • No collisions and no overtaking through four hard stops of the lead vehicle.
  • Queue discharge through the checkpoint against the equilibrium capacity of the 6 m/s zone.
  • Length of the Blender vehicles and of the road polyline against the numbers the model uses.

Honest simplifications

  • The parameters are textbook values for ordinary traffic. Real evacuation convoys, army trucks, buses and ambulances behave differently, and drivers in an emergency do not follow one law; the weather factors are plausible multipliers, not measurements.
  • One lane, no overtaking, no curves slowing the vehicles (the hairpins are only drawn: speed does not depend on the bend), no hills, no breakdowns and no reaction delay. The ring road is an idealisation; the real destination would take vehicles out of the system.
  • The checkpoint is a speed limit, not a queue of vehicles being inspected one by one, so the capacity reported is that of a moving zone.
  • The model has a known weakness: the IDM is collision-free but can produce unrealistically strong braking (limited here to 9 m/s²) when a vehicle is very close.
  • This is an educational traffic-flow demonstration, not a planning tool.

What to try

  • Raise the number of vehicles until the average speed collapses and a red stop-and-go band appears in the space-time diagram.
  • Press "Brake lead vehicle" in a dense column and watch the wave travel backwards while the cars drive forwards.
  • Compare Sunny with Fog at the same column speed: the equilibrium flow drops because the headway grows.
Loading model...
Loading Blender model...
Fundamental diagram: flow against density
Space-time diagram of the column (last 300 s)
Mean speed and checkpoint flow against time
Made with 🟠 Blender · Three.js