Stonehenge Haul and Trench Lab

2D companion of the Stonehenge explorer: a stone hauled up a slope with a Monte-Carlo crew estimate, and a trench section with radiocarbon dates and a Harris matrix.

Switch to 3D
Blender-rendered sarsen circle with trilithons at sunrise.
Scene plate rendered in Blender (Cycles). Illustration only: the charts and checks run live in the page's own model.

Presets

Playback

Haul or trench

Model checks

The science

What is modelled

Two parts of the 3D Stonehenge lab in cross-section. First, hauling a stone on a sledge, rollers or a raft up a slope with a crew whose pull is a hypothesis range, plus a Monte-Carlo estimate of how many people that really needs. Second, a trench section with five layers and dated finds: radioactive decay gives raw ages, a model calibration curve turns them into calendar years, and the layer order (a Harris matrix) is used to flag a find that contradicts it. All coefficients are hypothesis ranges, not measurements; the calibration curve is invented for teaching and is not IntCal.

Equations

  • Pull needed up a slope: F = m g (c cos θ + sin θ); a sledge slides back when tan θ > μ, i.e. at θ = arctan μ.
  • Crew: N = F / f with f = 150 to 300 N per person; steady speed v = vmax(1 − F/(N f)); work W = F d.
  • Monte-Carlo: draw c, f and the daily walking rate uniformly from their ranges, 2000 times, and report the 5, 50 and 95 percentiles of the crew and of the days.
  • Radiocarbon: N = N0 e−t/τ with τ = 5730 / ln 2 = 8267 y; age t = −τ ln(N/N0); error σt = τ σN / N.
  • Stratigraphy: a later context must not give an older date than an earlier one it overlies.

Checks run in the page

  • Slip angle by stepping the slope against arctan(mu); work by time-stepping a haul and integrating F dx.
  • Monte-Carlo mean force against the centre of the ranges; the 25 t worked example (49.03 kN, 197 people).
  • Radiocarbon age by bisection against the logarithm; half-life; the Harris matrix has no cycle and a planted cycle is caught; the contaminated find is flagged.

Honest simplifications

How the stones were really moved is debated; this page only shows what each hypothesis would demand. Friction is a single coefficient, the crew pulls in a steady tow, and the finds and their N/N0 values are model data for the exercise, not published results. Not affiliated with any heritage body.

What to try
  • Press Sarsen on the level and read about 200 people; then Sarsen up a 3 degree rise and watch the crew needed grow.
  • Press Slope steeper than mu: tan(slope) beats the coefficient, so the load slides back without stop-logs however large the crew.
  • Press Antler pick dating, then step through the finds: the rootlet charcoal R5 is red because it is younger than the layer above it.
  • Raise the counting precision slider and see the calibrated range of one find widen.
Crew needed against slope
Monte-Carlo crew
Speed
Days
Made with 🟠 Blender