Physics simulation: golf ball on surface

[jdelavaulx]

Hi there,

I’m trying to figure out a way to simulate how a golf ball rolls on a mesh according to physics variables such as gravity, slope, friction, etc.
The ball might start stationary, or it might have an initial speed and direction.

Is there an easy way to do this either in Rhino or Grasshopper?

-Jeremy

[martinsiegrist]

Seen this topic yet?

Simulating a marble ramp

Hi @reginald.raye The goal from here should be useful Here’s an example marblerun.3dm (4.0 MB) marblerun.gh (16.1 KB) (and here if you want to include multiple balls with collisions between them marblerun_multiball.gh (101.3 KB) ) You should be able to just replace the surface inputs with a new ramp design. Only include the stuff the ball will actually touch to avoid extra calculation. This doesn’t consider surface friction or angular momentum, and I’m sure you’d need to make ph…

[jdelavaulx]

This is great, but surface friction is critical for this application. I need to know whether the ball will accelerate or decelerate for given slopes. The mesh is quads at 0.2m size.

-Jeremy

[jdelavaulx]

Do you know if this may be able to be adapted to include Friction maps?

-Jeremy

[martinsiegrist]

Bouncing Decohedron Problem (Kangaroo)

Here’s less idle, more lively version This generates the mesh differently, and uses pressure to keep the shape. It uses the floor with friction, which allows it to roll instead of just slide, and I reduced the damping in the bouncy solver. bouncyball.gh (16.7 KB)

This should have friction

[jdelavaulx]

Thanks so much. Do you think that it would also take the mass and shape of a golf ball?

-Jeremy

[martinsiegrist]

As a comparison I’m sharing a simple example of a spring and correct force units. Unfortunately I don’t know how to add properties to a ball

Kangaroo basics

Hi Martin, In your definition it looks like there is no load applied, only a Length goal and an Anchor. In this case the Strength parameter is actually irrelevant, since all the goals can be satisfied exactly with no conflict. The relative strengths only really matter when you have goals fighting each other, as it controls in a precise way which one will be more satisfied. A better example might be have one end fixed, and a load applied to the other end. Like this: spring_example.gh (10.2 …

[jdelavaulx]

Thanks Martin.

I’m trying to get a ball to interact with the mesh surface in the attached rhino file and gh file.

Ideally, I want a grid of golf balls at 2.5m spacing which can start stationary, and then start rolling on the mesh. I want the mesh to have a friction coefficient variable, which can be from 0.0000 - 1.0000. The goal is to analyse how far the balls will roll on the surface given an input friction on the mesh.

rollingballfriction (1).gh (2.2 MB)
Friction on Mesh.3dm (18.3 MB)

Not sure if this is easy or not.

-Jeremy

[Joseph_Oster]

@jdelavaulx

Not sure if this is easy or not.

If you must, assume it’s not

[martinsiegrist]

A few things.

1. The Rhino file is obsolete when you internalise the base mesh, your landscape.

2. The landscape must be a closed mesh for the collision to work.

3. The ball itself can be defined as a rigid point set. I would only do additional subdivision for display purpose after the solver once everything else runs smoothly.

This way the simulation works. I cannot judge the physical behaviour.

rollingballfriction mrtn.gh (2.0 MB)