Lets talk about Mantaflow

General Forums Blender and CG Discussions
341

I believe the goal is merge the Mantaflow branch into Blender master for version 2.8. The Mantaflow branch is unifying the current smoke and liquid simulations into a single simulation type (fluid).

There is a GUI interface within Blender to set up Mantaflow simulations that is similar to how fluid simulations were set up before. The last build I tried, the Blender UI freezes while the simulation is baking though, which makes it difficult to test simulations.

Supposedly the Mantaflow branch does have viscosity implemented now in this commit, but I have not tested it yet.

342

I have build the latest from git hub, but i can’t see any viscosity and surface tension. But I am a noob when it comes to building.
Is this “commit” a part of the files that I am cloning?

BTW Rlguy, I had the same problem with freeze or crash, I had to choose the cpu for rendering.
even when baking.

343

Sebbas remove viscosity in the next commit, maybe it was a test only. The viscosity commit only allow viscosity but you cannot configure nothing. Only you can make a simulation and see how fluid have a different viscosity.

344

Yeah I was comparing the codes, this commit are in the github. So you are right

345

Hi! I see the viscosity commit caused a lot of confusion … in that commit I was just experimenting a bit. @RLGUY Yes, so it turns out since v0.11 mantaflow already supports viscosity. So it’s all set.

I also just pushed a bunch of commits, including new settings for “fluid diffusion” (viscosity, surface tension) and also some secondary particle updates. Lets see how this goes, as always if you find a bug, please post to the issues tracker on Github.

346

With High viscosity (0.1 or more) simulation make strange things.

347

Yes, I didn’t restrict the values in the UI. Just make sure to keep the viscosity value very small for now

348

actually forces don’t work with mantaflow?

349

They do, you just need to turn them up pretty high it seems!

350

First of all Sebbas, I want to say thank you for your effort to bring a better fluid / smoke simulation into Blender. Thanks a million.
I have learned to build now, but there are some expression I do not understand, when you say “pushed” does that mean that these are in the repository now? So I could do a update?

351

The win64 build at https://graphicall.org/1199 has been updated at request of sebbas.

352

Yep

in git, when you push a change, you upload a change. When you pull changes, you download last changes. For that reason to update git the command is git pull. But blender developers make a script that make auto all this, make update

353

when I did a “make update” the system said that the branch is up to date. When I look at me buld it says Date: 2017-11-19
should it not say 2017-11-29? Sebbas did the changes 6 hours ago

Edit: I had to use make full to get the changes. Now there are som sweet changes :slight_smile:

354

Are you sure that you don’t have checkout of a old commit or other branch? write

git checkout fluid-mantaflow

and then make update

355

Yes I am very sure, because when I did a make-full i got the fluid-mantaflow branch with the new changes.

356

Check that you write the correct command in the correct folder

Git checkout fluid-mantaflow
Make update
Make full

357

Thank again, I will do that next time when I am updating :slight_smile:

358

Double post

359

@esimacio Thanks for the nice words! Yes, @DcVertice already said it: In git, Blenders version control system, commits are local at first. Once I “push” them to the central repository everyone has access to them.

360

Thank you Sebbas! On other note, in the latest update, “plane” shape emitter does not emit liquid(fire, fire+smore, smoke still works).