Integrating c4d dynamics with realflow

[Eric Nixon 20170904101515]

filter by coll appears to work, and is fairly efficient, the viewport show the rf blue particles filtered to orange particles by a large collider interaction distance, blackdots are c4d dynamic collisions, gray boxes are the clones, there is a lot of lag in the viewport feedback but that lag may be a lot less in the actual sim I hope.

filter.png

Btw. It was so easy to adapt that scene, via the online docs and a few mouse clicks, even when crippled..., nice software guys

[Eric Nixon 20170904123948]

rf granular 2-way castles filter TP.7z

try this, very interactive setup

filtertp.png

Is there a way to setup a dummy collider, just to trigger the filter? I think its not possible, It would help for high speed collisions, as i can make the dummy fly ahead of the current trajectory, via xpresso, so would need a smaller margin and so less tp/clones would be generated.

[Eric Nixon 20170904123948]

leading ball doesnt slow down enough, this is the cache of uploaded scene.

issues.png

[Thomas Schlick]

Hey, nice progress. Thanks for sharing your results!

Glad you've found the "On Particle Collision" attribute in the filter. This would have been my next suggestion, since "Pressure" is a bit difficult to control in many cases.

It's possible to connect collision objects to certain fluid containers, since every collider tag has it's own "Links" field where you can define those interactions. If you have a low-res dummy fluid, for example, just in order to trigger a few collision particles then this is the way to go.

With regards falloffs, it's definitely nice to have such a feature with every daemon, but for some it's unfortunately not possible, e.g. "Drag". Drag is simulated on the fly based on the particles' current speed and positions, but you have to know a force field's dimensions in advance in order to calculate a correct falloff.

[Eric Nixon 20170904165334]

Looking good with proxies, you can see how the mud gets carved and leading ball meets resistance. Fastest setup so far;

prxies.png

I used the link fields to limit the proxy coll to the original particles and used the sufrace offset to make the collision volume almost zero, but used a larger interaction distance, proxies are offset along trajectory via that xpresso rig.

Not sure what you mean by low-res dummy fluid, can you reword that?

[Eric Nixon 20170904232110]

I've just been looking at the particle skinner (i thought maybe you meant that..) I can't understand what the particle skinner is for, I have watched two tutorials including the balloons one, The elastic/rigid deformers already do what particle skinner does.. I must be a confused noob

In the docs it says;

This daemon transfers the position of particles to the vertices of the object(s), attached to a rigid, elastic, or fluid container, and deforms it/them.

?????

Select a fluid container to deform the objects, associated with a "Rigid" or "Elastic" deformer.

also ??????

I think these docs need work, sorry I've gone off-topic, and also thanks for helping.

Edit: I've read the rf10 docs for Pskinner, I now believe that it lets you use all fluids as deformers, rather than just the elastic/rigid ones. But deformations by a fluid may give weird reults.

[Thomas Schlick]

Not sure what you mean by low-res dummy fluid, can you reword that?

My idea is to use an extra fluid with a low "Resolution" value only to create the collision particles. These particles will then be converted into TPs. This might help to get rid of the computationally expensive filter daemon - esp. when you have really high particle counts.. This fluid can also be exclusively linked to a proxy object, for example. Not sure if it'll help to make things faster or easier, I'm just gathering suggestions you might want to try. But it seems as if you've come very far already.

I've read the rf10 docs for Pskinner, I now believe that it lets you use all fluids as deformers, rather than just the elastic/rigid ones. But deformations by a fluid may give weird reults.

Thanks for pointing us on that. You're right: the description in the docs is definitely misleading, but you've got it anyway. The skinner can be used with fluid particles as well, not just with rigid/elastic. But as you've mentioned the deformation is difficult to control.

[Eric Nixon 20170906201826]

Hi Thomas, please look at this cached file, its creating a 2-way interaction between the rigid-deformer and a pool of fluid.

https://drive.google.com/file/d/0BwwAfn ... sp=sharing

The file sims correctly occasionally but often c4d becomes corrupted, or crashes.

Have had many crashes with elastic/rigid deformers last two days, refreshing the deformers by changing the resolution often helped to fire it up again.

[Eric Nixon 20170907032703]

Heres a working file (uncached but watchable), I cant say what fixed it, setting the rigids density can trigger it, but i've had so many glitches and crashes, that I amno longer paying attention to what causes them,other than using rigid/elastic deformers and dynamics together..

float rigid2 cached physics first.7z

my rf is 2.0.0.0037, gtx 670, c4d 18.057, tfd 1.0.149

[Eric Nixon 20170907175653]

Hi i just re-opened the 'working' file and the problem returned of no interaction between the fluid-rigid, So I changed the rigid density slightly and voila, its workign again.