eetu's lab

[eetu]

I wanted to try a melting setup too. This one has a twist, it has a pyro blowtorch doing the melting.

 

https://www.dropbox.com/s/gxq6luymr3chhq9/pyromelt.mov?dl=0

Both sims are running lockstep in one dopnet, with the flip sim fetching the temperature field from the pyro sim.

The flip particles have an additional temperature attribute, which gets added by the pyro temp field and then dissipates slowly.

This temperature attribute is mapped to viscosity, this way the wax first melts and then gradually hardens again.

I also use the temperature attribute to modulate a couple of the shader parameters to get a more "wet" melted look.

 

edit: Oh, and the different parts have different heat response, the outer ridges and the top do not melt as easily as the middle.

pyromelt_v016.hip

pyromelt_ogl.mov

Edited November 29, 2020 by eetu

[Netvudu]

yep....still a wizard.

Awesome scene.

 

Edit: it makes a lot of sense, but I got two questions. First, what do you need the gas surface tension for?

And secondly, I see that for the wax material, the scattering is disabled, yet you are multiplying your temperature at shader level by sss attenuation...what´s going on there?

Edited May 6, 2014 by Netvudu

[Skybar]

Cool, almost the same setup I did for mine. Gj

[tomwyn]

Amazing stuff.... Very inspirational!

[eetu]

I got two questions. First, what do you need the gas surface tension for?

And secondly, I see that for the wax material, the scattering is disabled, yet you are multiplying your temperature at shader level by sss attenuation...what´s going on there?

 

I wanted the molten wax to flow freely but not splash around - I thought surface tension would work better than viscosity for that.

 

I was using sss for the surface mesh at first, but then switched to using the interior volume approach. That binding is just an obsolete leftover from that.

[pelos]

very impressive,  thanks for the hip files they are great examples.

[abvfx]

Great work, the last few entries have especially been excellent .

[eetu]

Just consolidating some smaller old things here this time:

 

- a pyro sim affecting a flip sim

- a simple pbr diffraction shader

- and.. a braid

 

ee_braid_v003.hip

ee_pbr_dispersion2.hip

ee_pyroflip.hip

Edited May 12, 2014 by eetu

[eetu]

I got inspired by the quaddel stuff and ended up with 100% more cancer:

(mov here)

 

First I got a more sensible result, but after I added curl to the displace direction all these small hairy squigglies burst out. Scary stuff

 

The process:

- measure curvature

- displace high-curvature areas in normal direction, mix in some curl noise.

- check sdf for self-intersection

- remesh

- repeat.

 

I will keep on screwing around, I'll share the hip when I'm done

 

 

Edited November 29, 2020 by eetu

[sanostol]

scary, but looks great

[slamfunk]

awesome! that quaddel growth process link you shared was nothing short of inspiring for sure!! please keep us posted on updates 

[eetu]

A couple more renders and a turntable:

 

cancer_turntable.mov

[JaydenDP]

Thats horrible!

[Netvudu]

Yes. It´s horrible in a good way

[slamfunk]

disgusting 

[eetu]

And now for something serious: The Egg Equations.

 

I once needed an egg for a project, and being a dutiful Houdnik I naturally started by googling for an egg equation. Luckily, I found a set of those from Japanese mathematician Tadao Itau, and promptly recreated a parametric egg in Houdini.

 

 

egg_equation5.hip

Edited May 28, 2014 by eetu

[Matt_K]

Thanks Eetu!

Really enjoying following your work and looking at your hip files!

 

Matt.

[ntkirby]

eetu,

 

For your flip stickiness, do you have a solution for when the viscosity is turned up above 0?

[dgoubitz]

eetu,

 

For your flip stickiness, do you have a solution for when the viscosity is turned up above 0?

 I had the same issue, you can remove the collision field in the Gas Viscosity node inside the flipsolver.

[eetu]

From the archives: cellular automata. For an acquired taste

 

It's been over a year since I did these, so I can't remember everything. I had just read "A New Kind of Science" by Stephen Wolfram, and wanted to replicate things in Houdini.

 

First I did the "basic system", a binary 2d cellular automaton with successive generations stacked on top of each other, with the initial condition on top and time going down.

Here are a couple of stills and an animation of all the basic rules #1 - #512. Most of them are quite boring..

 

 

 

After that I experimented with continuous cellular automata, here the cell values are floating point, and instead of counting active neighbours, the combined value of neighbours is remapped to get the next generation value for this cell.

The visual setup is the same, except I feed existing volume data to the successive generations for added visual interest. These are a bit more boring, but they are at least a little bit controllable and can work in combination with other data, so maybe, just maybe could be useful somewhere (I won't hold my breath..)

 

 

Attached are two scenes and corresponding python HDAs.

 

ee_2d_ca.otl

ee_2d_cca.otl

vol_2d_ca_v003.hip

vol_2d_cca_v006.hip

Edited December 9, 2020 by eetu