Presets
Playback
Jupiter and the belt
Camera
Layers
Quality
Model check
The science
What is modelled
The main asteroid belt lies between Mars (1.52 AU) and Jupiter (5.20 AU). Every rock here moves on an exact Kepler ellipse around the Sun, solved from M = E - e sin E with x = a (cos E - e), y = a sqrt(1 - e^2) sin E, rotated by inclination, node and argument of perihelion. Jupiter, Mars, Ceres and Vesta use their real orbital elements. Rock sizes, planet sizes and the probe are exaggerated so that they can be seen; positions are to scale.
Equations
- Kepler's third law:
T = a^(3/2)years forain AU (Sun = 1 solar mass), i.e.T^2 = 4 pi^2 a^3 / (G M). - Mean-motion resonance
p:q(p rock orbits per q Jupiter orbits):T_J / T_a = p/q, soa_res = a_J (q/p)^(2/3). This gives 2.50 AU for 3:1, 2.82 AU for 5:2, 2.95 AU for 7:3 and 3.28 AU for 2:1 - the observed Kirkwood gaps (Daniel Kirkwood, 1866). - Speed on an ellipse (vis-viva):
v^2 = G M (2/r - 1/a), with 29.78 km/s the circular speed at 1 AU. - Resonance half-width (scaled model):
w = 0.05 AU · s · sqrt(m_J), with a strengthsper resonance (3:1 and 2:1 strongest). The true width depends on eccentricity and libration physics. - Eccentricity pumping inside a zone:
de/d(tau) = lambda · s · m_J · (1 - |a - a_res| / w) · e, integrated exactly over each step. The clocktauis the orbital clock multiplied by the time-lapse factor - real resonant clearing takes 10^5 to 10^6 years, far slower than the orbital motion you see. - Removal: a rock whose perihelion
q = a (1 - e)falls below Mars's aphelion (1.666 AU) becomes a Mars-crosser and is removed from the belt (in reality it is later scattered, or hits a planet).
Checks run in the page
- A velocity-Verlet integration of a test body in the Sun's inverse-square field reproduces
T = a^1.5within 0.5 % and conserves energy to better than 1e-4. - The positions produced by the renderer's own orbit function sweep exactly p/q revolutions per Jupiter period at every resonance.
- The stepped pumping integrator reproduces the closed-form clearing time
ln(e_cross / e0) / (lambda s m_J)for two Jupiter masses.
Honest simplifications
This is a teaching model, not an N-body integration: gravity of Jupiter on each rock is not integrated. Instead, Jupiter's effect is reduced to resonance zones in which eccentricity grows exponentially with a rate that scales with Jupiter's mass; the semi-major axis of a pumped rock is kept fixed (in reality resonant orbits also exchange energy and chaotic zones overlap). Real asteroids show additional structure (families, secular resonances, Yarkovsky drift) not modelled. Hildas and the Trojan clouds are decorative: their libration is drawn, not computed. The Great Red Spot, Mars, Ceres and Vesta textures were baked in Blender from procedural noise and are not photographs. The probe is a generic survey spacecraft, not a particular mission.
What to try
- Run "Solar system today" and watch the orange 3:1 zone empty; the orange histogram in graph 1 develops notches at 2.50, 2.82 and 3.28 AU.
- Switch to "No Jupiter" - no gaps appear; then "Heavy Jupiter (3 M_J)" and compare the clearing speed (graph 3) and the zone width.
- In graph 4 the gaps sit at integer period ratios (3, 5/2, 7/3, 2) - the signature of a resonance.
- Use "Follow tracer" (press "New tracer asteroid" to choose another) and read its eccentricity grow in the a-e plot until it crosses the Mars limit.
- "Inside the belt" and "Follow Ceres" show the Blender-built rocks, Ceres, Vesta and the survey probe up close.