World with friction

[d.carbonaro]

Good afternoon,
I'd like to simulate the motion of objects with air friction.
How can I do that?
Can I include all my bodies in a cube and define the friction with the
inner part of the cube?

Daniele

[Mark Williams]

> Good afternoon,
> I'd like to simulate the motion of objects with air friction.
> How can I do that?

For each body, apply a small force which is opposite and proportional
to its velocity before stepping the world. Also apply a small torque
to counter an angular velocity.

> Can I include all my bodies in a cube and define the friction with the
> inner part of the cube?

Just test each body's position against your cube when deciding whether
or not to apply the air friction forces.

Cheers,
Mark

[Bram Stolk]

On Jan 30, 1:29 pm, Mark Williams <mark.imageeng...@gmail.com> wrote:
> > Good afternoon,
> > I'd like to simulate the motion of objects with air friction.
> > How can I do that?
>
> For each body, apply a small force which is opposite and proportional
> to its velocity before stepping the world. Also apply a small torque
> to counter an angular velocity.

Proportional to the square of the velocity.
The v^2 is what keeps race-cars from reaching 400km/h.

bram

[Jon Watte]

Bram Stolk wrote:
> Proportional to the square of the velocity.
> The v^2 is what keeps race-cars from reaching 400km/h.
>

The problem with v^2 is that it has very little effect near 0, so the
object will not be properly dampened to a stop. To simulate air
resistance, you want to use a v^2 term, but to simulate damping (and get
a stable simulation) you want to use a v term. Thus, you add a force
proportional to A*v*v + B*v, where B might be in the range of 0.01 for a
timestep of 0.01, and A might be in the range 0.001 (depending on the
maximum velocity you want to achieve).

Cheers,

/ h+