Wind test 2

Comment

Comment by djordje on August 8, 2011 at 4:23am

fantastic!

Is there any chance of increasing the stiffness of the surface, adding a few more supports, so that it can act like a roof membrane?

Then from a pressure of the wind, we will get a pressure distribution on the membrane. Then get the nodal (vertices) forces for each triangle part of the membrane.

Is this possible?

Comment by djordje on June 14, 2011 at 12:48pm

Great work Daniel!

I find this interesting also. Especially in the field of fluid dynamics. For example - influence (pressure) of the wind on the free shaped surfaces. Of course the surface will respond as a stiff contruction, not like in this example, light like a fabric.

Do you think this is possible?

Comment by Erick Katzenstein on June 13, 2011 at 6:51am

On second thought, I may be wrong in thinking that a feedback loop is required for air resistance. Sorry for the confusion!

Comment by Astrid G on June 12, 2011 at 7:53am

Great! This can be very useful for what I'm trying to do with Kangaroo. I'm trying to simulate some fluid dynamics. More specifically, multiple water drops on a water pond. Can anyone recommend a good starting point?

Comment by Erick Katzenstein on June 12, 2011 at 7:29am

Thanks Daniel.  Fred, my loop question was regarding air resistance, where the value would have to be updated with the changing cross sectional area of the material. As far as I know, this would require a feedback loop.

Can't wait for the new release!

Comment by Ángel Linares on June 12, 2011 at 6:27am

Thanks for the explanation :)

Comment by Daniel Piker on June 12, 2011 at 6:14am

Sorry my verbal description of how it works was unclear - I'm projecting the wind velocity onto the triangle normal. A diagram probably makes it clearer:

 

The red vector is the triangle normal, and the blue vector is the wind velocity.

 

I project the wind onto the normal - to give the green vector, which I then multiply by the triangle's area and apply to each of its corners.

(and that all simplifies down amazingly neatly to nothing more than a cross product and a dot product !, so it's pretty quick to calculate)

 

You are quite right that when the mesh is perfectly parallel with the wind direction there will be no force. That's why in this video I applied a very tiny vertical force to one point on the edge to start it off. As soon as the wind catches it, the whole thing starts flapping.

 

 

I do like boats! In fact I wanted to make a little sailing boat as a test, but realised that it would need buoyancy to make it float. That should be possible by using the triangle normal pressure, but multiplying it by depth.  Maybe I'll add that next.

 

I'm also thinking of adding air resistance - so like wind but relative to the motion of the triangle. Then you could model flying machines with actual lift.

 

It probably is possible to use Kangaroo in combination with Hoopsnake, though I've not tried it yet. There are certainly lots of cases where it would be nice to be able to make the input dependent on the output in some way, and one plan is to make a scriptable interface to Kangaroo.

 

about pseudo-physical materials - see this video and my post here

Comment by Garcia del Castillo on June 12, 2011 at 4:22am

Oh my, that is just sooo cool.

 

Just a quick question: is the mesh in this video perfectly parallel to the wind direction? And if so, how does it happen in the first place that any face has a force perpendicular to it (by projection)?

Comment by Fred Becquelin on June 12, 2011 at 3:43am

Hi Daniel!

This is awesome!

What is it that pseudo physical material? something new too?

I don't get it... the force is coplanar with the face? You say it's projected onto it.

Erick,

I've been thinking about pressure and loops too, to make things float. I'm not sure if it's in Daniel's plans to build it inside kangaroo...

I would not try to use hoopsnake with kangaroo before it has triggering functionality, and even then I don't know if it would work. It doesnt makes much sense anyway, since kangaroo is already iteratng.

 

Do you like boats, Daniel? Everybody like boats, you should too!

Comment by Daniel Piker on June 11, 2011 at 6:55pm

Thanks Erick.

This is a new component I've just added in the upcoming release of Kangaroo.

It works similarly to the pressure force, but instead of just applying a force normal to each triangle, it projects the wind vector onto it and applies that force (multiplied by area).

• ‹ Previous
• 1
• 2
• Next ›
• Page