ODE suitability

[pete_dl]

I have a commercial application that requires physics simulation. I
would be very grateful if I could explain my simulation requirement
and receive comments on the applicability of the Open Dynamics Engine.
This would save me wasted work if I implement it just to find out that
ODE. is not suitable.

We would like to realistically simulate a tethered glider (very
similar to a kite on a string). The element of the simulation we would
provide is the resultant lift force from glider. The elements we would
like simulated by the physics engine are:-

1) The Rope. This will be tethered to a fixed position at one end. The
rope will be 100 weights attached to each other with ball and socket
joints and very stiff beams and finally to the glider. We need to
apply aerodynamic drag to these 100 weights, so we need to know their
position and velocity.

2) The Glider. This will have an aerodynamic force, weight and line
tension acting on it. The aerodynamics simulation and control would
need to know the orientation of the glider, its speed and the tension
between the rope and the glider.

I have been told that a large mass ratio between the rope and the
glider could cause instability in the physics engine. In one scenario,
each of the 100 segments would weigh roughly 20 grams, and the glider
would weigh 150KG, so this is a ratio of 7500:1. The tension in the
line would be up to 15,000 Newtons.

I will be grateful for any comments.

Regards,
Pete Leaback

[Jon Watte]

ODE does not do aerodynamics. I suggest you move up-market to an
engineering package of some sort.

Cheers,

/ h+

[Daniel K. O.]

Jon Watte escreveu:

> ODE does not do aerodynamics. I suggest you move up-market to an
> engineering package of some sort.

From his description:

>> The element of the simulation we would
>> provide is the resultant lift force from glider.

My understanding is he already has the air lift and friction simulated
elsewhere, and just want to integrate rigid body dynamics for the rope
and the glider.

I just modified demo_chain2 a bit (reversed the gravity, anchored one
end on the ground, attached a 150 Kg ball on the other end, increased
the number of links to 100) and did some experiments. From what I can
tell, a small timestep (< 0.001, hard to get real-time) combined with
"large" chain links masses (> 0.1) is needed to make the rope remain
stable at ~30 K Newtons. This with double precision.

If the masses can't be changed I would try a different simulation
approach; for example, mass + spring for the rope (in which case you
probably won't need ODE at all). Try asking on gamedev.net or Bullet
forums for hints on rope simulation.

--
Daniel K. O.
"The only way to succeed is to build success yourself"

[pete_dl]

On Feb 2, 6:01 pm, Jon Watte <jwa...@gmail.com> wrote:
> ODE does not do aerodynamics. I suggest you move up-market to an
> engineering package of some sort.
>

Sorry, I did not explain myself properly. I will be doing the
aerodynamics simulation and provide a resultant force vector to the
physics engine. We would like the physics engine to simulate the rope
and the glider, (the glider consists of a weight and the aerodynamic
force vector).

Pete.

[James Frye]

Pete,

Something similar to this, using ODE for body mechanics and a different
package to
calculate aerodynamics, was reported in Dickson et al, "An Integrative
Model of Insect
Flight Control". Doesn't use a tether, though, so I can't help you there.

There's also a new Physsym package (physsim.sourceforge.org, I think) that
is intended
to do accurate physics. I haven't tried it yet due to lack of useful
documentation, but it
might be worth a look..

James

[pete_dl]

Thanks for your comments guys.

I tihnk I will give ODE a go with smaller mass ratios.

Pete.