Dyverso Granular experiment - Drop heavy object into sand?

Discussions about the use of Realflow
rich0550a1
Posts: 19
Joined: Mon Apr 16, 2018 2:23 pm

Dyverso Granular experiment - Drop heavy object into sand?

Postby rich0550a1 » Wed Nov 07, 2018 3:52 pm

simple enough a pile of sand and then an dynamic rigid body dropping into it.

so the interaction is not right, too heavy mass it sinks to bottom, to light it flies off. It feels more like liquid not splashing sand.


density of the DY domain 1000, resolution 50.

mass of rock 10-100

the biggest problem is the rock flies away or bounces all around. upping sim substeps 1/300 or DY substeps 5/5 20 interations, does not help, a bit of boost by taking carbonite to 100% accuracy.

tested a bunch of things like friction .5 - 1 on DY particles, cohesion up and down, roughness up and down, damping up and down.

is this too much to expect from a Granualar sim?


User avatar
ChristianZ
Posts: 3416
Joined: Sat Sep 09, 2006 11:37 am
Location: CH
Contact:

Re: Dyverso Granular experiment - Drop heavy object into sand?

Postby ChristianZ » Thu Nov 15, 2018 8:31 am

My test
https://youtu.be/NPINrT9AN64

Indeed it is a challenge to adapt the rigid body parameters here. On the side of the granular solver we have here a block of heavy snow or mud (kept in place by invisible frame around it) with cohesion. Without cohesion it would be sand.
We see here that the rigid body is not loosing energy fast enough. Instead it is "digging" its way inside the granular material, although it obviously reacts to the particles.
I will do another test with less or more interaction of the Dyverso particles with the Rigid body.
RF10 standard + RFC4D

User avatar
ChristianZ
Posts: 3416
Joined: Sat Sep 09, 2006 11:37 am
Location: CH
Contact:

Re: Dyverso Granular experiment - Drop heavy object into sand?

Postby ChristianZ » Fri Nov 16, 2018 4:54 pm

https://youtu.be/Zb0h919dPTY

This example is with DY interaction factor on the rigid body=0.001
As you can see, the rigid body sinks in, but continues to spin around and to mess up the pile of mud/snow.
RF10 standard + RFC4D

User avatar
ChristianZ
Posts: 3416
Joined: Sat Sep 09, 2006 11:37 am
Location: CH
Contact:

Re: Dyverso Granular experiment - Drop heavy object into sand?

Postby ChristianZ » Sun Nov 25, 2018 12:07 pm

As a summary of my tests: indeed it is difficult to get the rigid body to behave as expected inside the granular mud/heavy snow. To be fair, the granular material as such looks realistic and reacts to the object in the expected way. However, the forces on the object don't seem to slow down/reduce the energy of the object as expected. So, the interaction is the issue, but it is mathematically speaking a different problem than the granular particles.
What to do in a project? You have to fake it, for example by using keyframes on the "air drag" parameter of the rigid body in order to slow it down and freeze it at some point in the process.
RF10 standard + RFC4D


Return to “Realflow”

Who is online

Users browsing this forum: No registered users and 5 guests