Water flow in corridor

[TSR_trix]

Heres a test to fill a corridor with water.

https://vimeo.com/43815771

[Ivaylo Katev]

good job! what about the grid resolution and the sim/rendering time?

[TSR_trix]

Grid total cells was 20 million, and hmm, maby it took about 8 hours to simulate. Water pass render time (with camera motion blur) was 4-8 minutes per render node.
Grid resolution could be even smaller, but I didnt have nerves to re-sim

[TSR_trix]

Another flood simulation uploaded in my profile. https://vimeo.com/45489909

This one takes approx 4-6 minutes per frame to simulate (0,512cm cell size) and have around 400 000 foam particles at the end.
Render time with 1280x720 resolution took 20 minutes - 2 hours per frame. It seems that having so much refractive foams with motion blur, it takes quite much time to render. Maybe I have to adjust my render setup a bit.

ps. How do I edit this thread title? I try to update it from my first post, but it didnt updates it in the main title.

[Ivaylo Katev]

excellent! what is the resolution of the grid?

[Ivaylo Katev]

the only annoying thing is the wetting of the toilet. i think we have to introduce per-object wetting instead the global one.
i suppose the wetting helps for the floor (is the thin layer on the floor due to the wetting?) but in many cases it looks bad. or we have to replace it with more complicated technique

btw, what is the reason for this simulation? is it just an exercise or you need it for a production?

[TSR_trix]

Thanks Ivaylo for your comments! The resolution of the grid was 40 000 000.

Yes, im absolutely with you about the wetting. Per object wetting would be nice. I did make a simulation without wetting, but it didnt work cause the floor looses water after leaking and it looks unnatural. Now with thin layer of dirty water and caustics doing some tint variation, it looks cool. But those wettings in walls are not working, cause it shouldnt reflect so much and caustics makes some weird flickering on it.

Second thing is again those artifacts (or fault behavior) on that toilet cover. I think its because of the round shape of that object. I should cut it off with gizmo when rendering, but its still always annoying to avoid diagonal or round shapes with proxy objects in simulations.

This was just an experimental job, mainly to test how to use right SPF to get realistic result in measure like this toilet. Im going to use PHX to make a massive inside flood in feature film so this was also a practise for that one.

[TSR_trix]

They catch my sims on http://www.maxunderground.com/
Third on news feed.