FNWI --- IMAPP Department of Astrophysics
Radboud University > Faculty of Science > Department of Astrophysics

This page describes the contents, setup and possible demos for the gravity well demo kit.

Setting up

The GW demo setup consists of:

  • a trampoline frame (9 large steel-tube components: 6 ring segments and 3 U-shaped supports, plus 6 smaller connecting pieces for attaching the supports to the ring)
  • 1 3x3m square of spandex (we also have 1 spare piece of fabric in the kit, same size)
  • 32 clamps for holding the fabric to the trampoline frame
  • 1 'skirt' that wraps around the perimeter and hangs to the floor to make it look nice (and to hide the clamps from inquisitive kids).
  • 1 orange bowling ball used as a massive central object.
  • 6 petanque/jeu-de-boules balls, also suitable for use as the massive central object.
  • Many marbles and golf balls to use for our moving particles.

The frame is the first thing to set up. This is most practical to do with at least 2 people, as the frame parts will tend to become unhinged again if you try it by yourself. So far it has been the most practical option to connect the 6 ring segments first, with the leg inserts pointing up (so, with the frame upside down). The leg pieces can then be inserted and the whole thing can be flipped to its standing configuration.

When you attach the fabric, first note the side of the fabric on which the seam sticks out: this side should be facing downward. The safest thing to do is to first connect 2 opposing sides of the square piece to 2 opposite points on the frame with clamps (1 clamp per side) while letting the fabric hang somewhat loose in the frame. At the ring, pull the fabric over and around the frame tubing and clamp it in place from below. Then, the next 2 points on the circumference of the frame can be used to clamp the fabric to. After this, the next 4 clamps can be connected in the spaces between the first 4 clamps and so on. With 32 clamps, the fabric is supported nice and evenly. When clamping the fabric to the frame, be sure to use only the 'front' bit of the clamps (the part covered in black rubber): that way the fabric will not get damaged so quickly. We found that if you let the clamps bite as deep as possible (i.e., when you have the red part biting into the fabric), the metal of the clamps easily tears holes in the fabric.

After the fabric has been clamped to the frame, it can be gradually stretched at each clamp position so as to get a surface that is taut. In the test setup at the department, we simply tested the tautness with the bowling ball until it no longer touched the ground when resting in the middle of the surface. When the fabric is in place with all clamps attached, all clamp handles should be pointing down (they are grabbing the fabric from below). This is done because then the skirt can nicely hide the clamps from view when attached around the outside.

Preparing for a demo

For most demonstrations, we first put the bowling ball or a petanque ('jeu de boules') ball in the middle. This should be done carefully, as the fabric (and the seam down the middle) is not infinitely strong! The bowling ball can be placed by holding the bowling ball, leaning over the surface towards the middle, and carefully lowering it onto the fabric to guide it down into its equilibrium position. Note that the bowling ball and the petanque balls should not be used as moving objects! This because they are so massive that they risk damaging the fabric when they roll around.

Possible demos to perform

  • Circular/elliptical/hyperbolic orbits: release the marbles/golf balls one by one with different orbital speeds. See how they follow a different orbit according to their imparted orbital velocity. This is analogous to the different classes of orbits that we see in nature (with the big difference that these orbits decay rather quickly, and that the pericenter of the elliptical orbits precesses a lot. Both of these effects occur in reality too, but at a much more subtle level and for different reasons).
  • Relation between orbital radius and orbital period: release numerous marbles over a short time interval, so that multiple marbles are orbiting at the same time. The ones closer to the bowling ball will orbit more quickly - not just because they have less distance to travel, but also because they move more quickly.
  • Tidal disruption: place multiple (~5 or more) marbles in one wide shared orbit by putting them in together at once. If done carefully, they will orbit while clumped together for a while, and will later separate as their orbit gets tighter. This is a nice approximate way to show tidal disruption: the 'gravity' of the marble group is no longer strong enough to keep the clump together against the different gravitational acceleration they feel.
  • Planet + moon around Sun (difficult!): put a golf ball and a marble in orbit together. If done right, the marble will orbit the golf ball for a while, while the golf ball itself will be orbiting the center.
  • Formation of Solar system: place a large number of marbles in a clockwise orbit, and another large number in counterclockwise orbit. Many of the marbles will bump into each other, but typically a few will remain orbiting because of a statistical advantage. This illustrates how the Solar system formed where collisions within an initially amorphous gas cloud resulted in the remnants having aligned angular momenta.
  • Effect of planets on star (wobble): illustration of how an orbiting planet affects its host star. Put a golf ball in orbit, and notice how the bowling ball is tugged a little as the golf ball moves around it. This nicely illustrates that planets and their host stars collectively orbit a common center of mass that happens to lie very close to (or within) the host star. This also shows one way of detecting planets around other stars: by observing the small motion of the star (through doppler shifts in its spectrum).
  • Gravitational waves (recording playing on TV only, will not work well with live setup): we need to record a slow-mo clip of a rotating rig that presses into the fabric at two closely spaced points, rotating rapidly. the waves that propagate out from this rotating system can be likened to gravitational waves. Unfortunately this effect is rather quick and hard to see live, but with a recording we can demonstrate it on one of the TVs in the tent.
  • Double potential well: With some help, two large groups of marbles can be positioned on the fabric in such a way that they remain in position (i.e., they don't merge together by themselves). This gives a double potential well, which can be used to show figure-of-eight orbits that pass around one marble group and then around the other group. This has some analogy to the 'free return trajectory' that was used for the Apollo Moon missions, where the spacecraft would come back to Earth if no orbit injection manoeuvres would be made at the Moon. This is a fun demo for the audience to participate in, because it can get quite tricky to get the orbit right - and interesting behaviour is often observed.
  • Warp drive: This is a somewhat more playful demonstration that is less closely connected to modern human engineering capabilities. When a marble is orbiting the central object, you can make it speed up or slow down by pushing down on the fabric ahead or behind the marble respectively (make sure you match the marble's speed while you do this). In this way it is possible to have a marble stay in orbit indefinitely, periodically boosting its energy in this way. This effect can be related to a 'warp drive', where local spacetime curvature is generated to alter the path of a spacecraft.

Resetting a demo

When a demo is finished (all the marbles are spent and lying next to the bowling ball), the bowling ball can be carefully rolled closer towards the edge (you may need to reach a bit to get it away from the center) and picked up. This also makes it easier to retrieve all the marbles/golf balls as they will follow the bowling ball when it is moved. For the petanque balls, there is a magnet on a string included in the petanque ball set which makes it very easy to retrieve it.

Tips for upgrades

Mail from Eric Hennes @ Nikhef

Ik heb het lycra doek van ons model op de Albert Cuypstraat 146 gekocht bij de firme N&N:

Let op: kies materiaal met minimaal 15% elastiek/spandex). Kost ~ 10 euro per m. Breedte is 145 cm, dus voor ~15 euro heb je een vierkant doek waarmee je een opgespannen model van 150 cm diameter kan maken. Niet teveel spanning want anders kan een simpele schroefboormachine het niet bijhouden (~ 1300 toeren/minuut). Ons frame bestaat uit handig verbonden stukken elektriciteits-pijpen, opgespannen met 4 stukken touw. alles bij elkaar voor een paar tientjes. Het geheel weegt hooguit 1 of 2 kilo. Belichting geschiedt per stroboscoop, in mijn geval een pulsbestuurde LEDlamp met regelbare frequentie en een vaste duty-cycle van ~ 30% door Jan Koopstra (ET, Nikhef) voor mij in elkaar gesoldeerd, gevoed met ~ 40 Volt DC.

Zie ook bijgevoegde demo filmpje (attachment, gemaakt door Marco Kraan) Overigens ben ik (ook) copycat; de credits gaan naar Steve Mould ( (the orbiting pair appears in min 4:30 of this 10 min video)

Mijn verbeteringen zijn:

  1. Gebruik stroboscopische lamp maakt een live slow-motion demo mogelijk. Bij de demo van Steve Mould is het stroboscopisch effect bereikt door de golf-beweging op de nemen met een camera met zekere frame rate. Maar dan moet je het dus eerst opnemen, en pas bij het afspelen zie je de slow-motion.
  2. Lichtgewicht frame, snel in- en uitelkaar te halen..
  3. Transparante en veilige circelvormige plaat ipv dwarsbalk.
  4. Gebruik van veel geluidsarmere en makkelijker te hanteren accu-schroefboormachine.