1. A funnel for a central field
A ball moves under an attractive force towards the centre. At the same time, its sideways velocity carries it past the centre. These two effects combine to produce orbits.
DR.Orbit calculates motion in a plane and displays the balls on a silver funnel. Its height visualises the scaled gravitational potential: it gets deeper towards the centre. Tilting and rotating change only your view, never a force or a measurement.
A real funnel provides an intuitive analogy, but even without energy losses it does not generally produce the same orbits. As a ball on its surface changes its distance from the centre, it also moves up or down. Vertical motion therefore accompanies its motion across the plane. A rolling ball also rotates about its own axis. Both contribute to kinetic energy and change the equations of motion. Having the same potential shape alone is therefore not enough to produce identical orbits.
DR.Orbit omits these additional surface effects and calculates the Newtonian central field directly. This lets you explore its ideal circular and elliptical orbits. The funnel makes the potential visible, but its slope introduces no additional downhill force into the calculation.
The origin is at the centre. r is the distance from it and v is speed. Acceleration follows a Newtonian central field with fixed gravitational parameter μ.
- Gravitational parameter μ
- 0.65 m³/s²
- Mass per ball m
- 0.01 kg
- Ball radius
- 0.035 m
- Opening radius
- 0.13 m
These values give the experiment a convenient scale in space and time. They do not represent a specific planetary system. Up to ten balls can be explored at once. All balls experience the same central field. Mutual gravity is not calculated. On contact, they can exchange momentum through ideally elastic collisions.
2. Give it a sideways push
Place a ball away from the centre and push it sideways. A push straight towards the centre produces little angular momentum. A sideways push can produce an orbit. For a circular orbit, velocity must be exactly perpendicular to the direction towards the centre and match the distance:
Tcircle = 2π · √r³μAt r = 1 m this is about 0.806 m/s, with an orbital period of 7.79 s.
A different velocity often produces an elliptical orbit. The ball moves faster near the centre and more slowly farther away. An unbounded Newtonian central field also allows open trajectories. Here the central opening and outer rim limit the visible experiment.
Orbit zoom in the 3D view unlocks only after one complete undisturbed orbit. The ball must experience no resistance, collisions or intervention by the hand. Position and velocity are also used to check that the entire corresponding ellipse stays safely inside the walls and outside the opening. Negative mechanical energy alone does not prove a safe orbit. A later disturbance ends the zoom. This magnification changes only the view.
3. Energy and angular momentum
Without friction, intervention, ball collisions or wall contact, mechanical energy and angular momentum of a freely orbiting ball remain constant apart from small numerical errors. Kinetic energy is always nonnegative. Potential energy is negative because its zero lies at infinite distance.
Epot = −m · μr
E = Ekin + Epot
L = m · (x · vy − y · vx)
Angular momentum L describes motion around the centre. Its sign distinguishes the direction of orbit. The central field changes the ball’s linear momentum but produces no torque about the centre. Thus, without friction, ball collisions or wall contact, L stays constant until the hand intervenes. A circular orbit is especially easy to recognise by its constant radius and speed.
Negative total energy means a bound orbit in the unbounded field. It does not guarantee a lasting orbit in this limited funnel: even a bound orbit can reach the opening or rim.
4. Rolling resistance, air drag and collisions
Rolling resistance and air drag can be switched on independently, even during motion. Both start off. Switching them changes neither positions nor velocities or measurement time. The forces act during the motion that follows.
Rolling resistance represents an approximately constant force opposing motion. Near rest, it smoothly approaches zero. Air drag grows quadratically with speed:
F⃗air = −k · |v⃗| · v⃗aroll = 0.015 m/s² · v₀ = 0.005 m/s · k = 0.001 kg/m
Rolling resistance is a simplified braking effect, not a calculated contact between a rotating ball and the funnel surface. There is no static friction. Both forces convert mechanical energy to internal energy. This is tracked and retained even when you turn resistance off again. Angular momentum transferred to the environment is also accounted for.
When two equal-mass balls meet, their velocity components along the line joining their centres are exchanged. Tangential components stay unchanged. This ideally elastic collision preserves their combined kinetic energy, total linear momentum and total angular momentum. The energy and angular momentum of an individual ball may change. Contact begins when their centres are 0.07 m apart.
A held ball acts as a stationary obstacle. An approaching ball reflects elastically from it. The hand absorbs the reaction. Because the hand does not move during that contact, it does no work. When adding or dragging a ball, an overlapping target position is moved to a free location.
5. What your hand changes
The experiment starts with an empty playfield. In the overhead view, press an empty spot to hold a ball there. Releasing places it on the surface, moving before release gives it a push. In the 3D dialog, a tap places the ball at rest. To launch it, hold the empty spot briefly until the ball appears, then drag. Moving balls can also be grabbed again. In the 3D view, the shadow marks the placement point. The visible gap beneath a held ball is only an interaction aid and changes no energy. Reset clears the surface and starts a new recording.
Grabbing a ball stops it. Dragging determines its position. On release, it takes the direction and speed of the most recent hand movement. Up to 120 ms are considered and release speed is limited to 3 m/s. Holding still briefly before releasing places the ball at rest. Subsequent acceleration by the field can produce higher speeds.
Stopping, moving and launching change energy and angular momentum. The model records these signed changes as intervention by the hand. Adding a ball also introduces its energy and angular momentum into the system. Removing or capturing it accounts for the quantities carried away. Resetting starts a new measurement and balance.
Lballs + Lenvironment + Lremoved = Linitial + ΔLhandEnergy carried away can also be negative because of the chosen zero.
6. Opening and rectangular playfield
When the centre of a ball reaches a distance of 0.13 m, the ball is captured. Its orbit ends at the opening. The calculation therefore never reaches the singularity at r = 0. Capture is a fixed model boundary, not a calculated fall into a black hole.
The rectangular playfield uses the available panel space. Its aspect ratio follows the main panel. The shorter side is at least 2.80 m long. Both directions use the same scale. Circles and balls therefore remain round, and the corners are usable too. The central field continues to act there unchanged.
The four walls reflect the wall-directed velocity component ideally elastically. The component parallel to the wall stays unchanged. Both components can reverse at a corner. Kinetic energy is preserved. Unlike a circular rim, a rectangular wall can change angular momentum about the centre. The opposite change is assigned to the environment. The ball centre stays 0.035 m away from the wall, accounting for its radius.
When the window is resized, the boundaries adapt. The playfield expands if necessary so existing balls keep their positions. Velocities and measurement time are not reset. An active grab is safely cancelled and the simulation pauses. Opening or closing the 3D view does not change the boundaries. Tilting, rotating and the gently decaying camera motion also affect the viewpoint only.
With rolling and air resistance switched off, a free orbit does not deliberately slow down. When resistance is enabled, mechanical energy becomes internal energy. The orbit may tighten and eventually reach the opening.
7. Calculation and animation
A fourth-order Runge–Kutta method jointly calculates position, velocity, internal energy and angular momentum transferred to the environment. Internal steps are at most 0.5 ms long and are reduced further near the centre. Contact times are located within the numerical trajectory. After a collision, the remaining part of the step is calculated using the new velocities.
A ball you hold with the mouse, a finger or the keyboard can support another ball. If the field pushes the second ball against it and motion perpendicular to the contact surface has almost stopped, the simulation calculates a supporting contact force. Sideways motion remains possible. The numerical resting threshold perpendicular to the contact is 0.1 mm/s. The remaining tiny energy error stays visible in the balance and is not counted as heat. Very small, repeated collisions of the same pair are grouped as sustained contact in the chart.
Physics runs independently of display refresh rate. Rendering may interpolate between calculated states to keep motion smooth. These intermediate images do not change measurements. Numerical energy error is not hidden by artificially rescaling velocity.
8. Selected ball and recorded history
The readouts and charts belong to the selected ball. Explore radius, speed, energy or angular momentum. Total energy in the chart includes kinetic energy, potential energy and the internal energy assigned to this ball. Ball collisions can transfer energy between balls. Recording captures 60 measurement points per simulated second and retains at most 60 seconds of history per ball. Older points are discarded. The short trail in the funnel is a visual aid, not an additional measurement.
The live chart uses cached, reduced measurement points while preserving minima and maxima. It does not process the entire history in every animation frame. CSV contains the retained original values for the selected ball. Export important data before resetting or before older measurements leave the window.
A captured ball initially remains available for analysis. At most ten ball histories are retained in total. Adding more balls may replace the oldest completed history. All calculations and measurements stay locally in your browser.
9. Explore a snapshot
“Snapshot” pauses the simulation and opens an immutable copy of the measurements in a wide modal. Use the mouse wheel or two fingers to zoom and drag to pan. Click or tap a curve to see an actual recorded value pair with time and unit.
The legend sits quietly inside the chart. Magenta markers identify interventions and boundary events. Their labels are grouped according to the visible range. Zooming, panning and selecting points change neither original data nor CSV. Closing resumes an experiment that was running before opening. A previous pause is preserved.
10. Model limits
The central source remains fixed and receives no back reaction. Balls do not attract each other gravitationally. Their collisions are idealised and perfectly elastic. Rolling and air resistance are simplified drag models. Ball rotation and real surface contact with the funnel are not calculated. The field is Newtonian, with no relativistic effects. The visible height is a scaled potential representation, not spacetime geometry.
The model is for exploring orbits, energy exchange and angular momentum. It is not a tool for predicting real satellite trajectories. Comparisons with ideal circles or ellipses require the ball to stay away from the opening and outer rim.