I triple dog dare you

[Noyogi]

Hello, Good Day, nice of you to join the conversation

I have a question to pose and a topic to discuss

What is the best code to build

not a pentagonal dodecahedron,

but a soap bubble space where bubbles can be created, co exist and join other bubbles, in a gravity free zone where soap film is an ideal thickness and all bubbles seek a least surface area to volume ratio, while maintaining their individuality

 

ANY and ALL ideas WELCOMED and ENCOURAGED

*I say zero gravity to make things simpler, but there could be gravity if that is what it takes

*I say seak least surface area to volume because I see that as an option for finding the shapes that the bubbles take

*The question I am posing is more shape related than shader, material, rainbow related, if you render difference clouds in Gimp or smoke material in 3dsmax then fiddle around with the colors for a few minutes you get a pretty decent soap bubble representation, this is good enough for me, on a sphere with random rotation it'd look pretty good, i'm not sure how you'd make that work with a not round bubble

ALSO CHECK THIS OUT

http://graphics.ucsd.edu/~iman/SoapBubbles/soap_bubbles_2.pdf

Here is a story

Seeking the Least:

Hey bra I got all this gasious volume inside me an its really stretchn on my awesome surface tension, i see you too have heaps of gas up in ya, wanna work(term work is not being used in a physics sense here) together and share some surface area, we can relieve a bit of this tension, ah sure sounds epic bub... and it was good and great and certainly epic the end.... minutes later... bub and les were joind by more bubbles and they formed a happy family ("cluster" if you will).

IDEALLY the code would be something that could be ran on a website

[Pryme8]

https://en.m.wikipedia.org/wiki/Metaballs

[dbawel]

Are we to try and simutate the quantum mechanics which exist in and on the surface of a bubble causing tension and color? This is actually fact and a key element to the behavior of bubbles, and a very tough task in my opinion.

DB

[Wingnut]

Yeah!  Meatballs... delicious with spaghetti, and a fun Bill Murray movie, too!

What's that you say?

Ohhh... it's metaballs?  Ahh, ok.    Yep, metaballs... that's it.  Good link, p8, and yes DB, likely a very tough task.

"We are the C-I-T's... so pity us. The kids are brats... the food is hideous. We're gonna smoke & drink & fool-around. We are the Northstar C-I-T's"

(Counselor In Training)    (From the Meatballs movie).  What a waste of a perfectly good posting-space I did here, eh? 

[webGLmmk]

On 9/20/2016 at 7:05 PM, Noyogi said:

but a soap bubble space where bubbles can be created, co exist and join other bubbles, in a gravity free zone where soap film is an ideal thickness and all bubbles seek a least surface area to volume ratio

 

Ever the pragmatic:

With CG, we're simulating. So you have to basically decide what specifics to simulate. Bubbles exist in an atmosphere WITH gravity..they simply have next to zero mass, and are easily carried along by air movement.

(Bubblemaking 101 at community college...Can bubbles exist in the vacuum of space? I'm going with no. Got no clue what I'm talking about actually.)

So in choosing to simulate the MOVEMENT of bubbles:  using the babylon.js physics engine, you create a sphere ...so you can either set the gravity really really really low, or at zero ..and apply impulses. Collisions with varying objects either make it burst, or bounce...with another bubble stick together or get bigger....hey, good idea, there's a lot of potential with this theme. It reminds me of a game I played on steam... Osmos:   http://store.steampowered.com/app/29180/?snr=1_7_7_230_150_7

Have you used the physics in babylon yet? I'll let you start that off either way.

19 hours ago, dbawel said:

Are we to try and simutate the quantum mechanics which exist in and on the surface of a bubble causing tension and color? This is actually fact and a key element to the behavior of bubbles, and a very tough task in my opinion.

DB

As far as surface area relative to volume...I dont think thats really a necessary thing to factor in a bubble simulation, or possible, or will even be noticed by user/audience/players.  The APPEARANCE of the film is what is simulated..see DB above...obviously material with low alpha/high transparency..but those swirling colors...that does seem like a task...pretty cool one...If someone could create that kind of dynamic movement, let me know, because I'll use them for my stars. or lava flow. I'm thinking alternating gradients, based on some kind of moving black and white map...thats heavy duty...haven't done any shader stuff...yet

I mean the simplest cheap out way for the bubble material would be a mostly transparent and lightly emissive grayish-bluish color...but close up,  bubbles aren't bubbles without the swirling soapy rainbow of colors

So then you decide, what is the point. Is this a game? A slightly interactive animation? Are you seeing only a few bubbles up close, or lots of bubbles from further away?

Finishing my ramble with the idea that purpose and concept must come first.

[Noyogi]

I have edited the post to be more specific thank you for the responses thus far

@Pryme8thank you for the metaballs that gives me plenty of reading and is in a good direction.

The joining of bubbles I was referring to was without them popping, and is more specifically oriented towards the shapes that they create when joined

@dbawel I am open to simulating quantum stuffs if that's what it takes but I think there is a much easier way to achieve results which represent how bubbles join together forming less bubbly shapes

@Wingnut While this is not MY space, I welcome you to try your best and fillet.

@webGLmmk I must apologize for my lack of specificity, what kind of place would that be, Specifi-city, ha... ha.... ha... I had assumed that after watching the video everyone would be on the same page, it's about the shapes. Thank you, I have edited the first post to have more words and have hopefully aligned some vectors.

Scincerely, OyO

 

 

[Wingnut]

Hi guys.  Thx for the extra details, Noyogi.  I don't have anything pertinent, except maybe http://paulbourke.net/geometry/implicitsurf/

I enjoy reading about these things.  Paul Bourke is an early founder of our Spherical Harmonics features, in BJS.

When vertices are plotted continuously and dynamically with math, you get a big pile of variables to play-with.  Primarily, influences and susceptibility to those influences.  Every vertex is "managed", one could say.  Each has a coach... a behavioral therapist.  Mesh merging becomes a whole new ballgame, because each vertex is under the instructions of its therapist.  Some could be reluctant to be around others, and avoid commune-ing.  Others might be very social, and feel purposeless without partners.

Where am I going with this?  I have no idea.  Maybe someone with soap bubble experience, can answer these:

Do bubbles attract each other?  Do they seem to WANT to join?  Or perhaps it takes some forcing of two bubbles together... to make them decide to share surfaces?  But maybe Nyogi doesn't care so much about that.  He seems to be interested in the shapes formed when hemisphere bubbles share surfaces, and thus make the inner shapes.

So... hmm.  Noyogi... have you done some experiments with using math... to positionally overlap (many) semi-transparent spheres in various symmetries, and see the resulting overlap areas?  If we don't need to consider mesh merges, then we (you) are ready to do some fun experiments, even in the BJS playground (so we all can turn knobs/change variables).

Would you be hoping to "harvest" those in-the-center geometric shapes... into being stand-alone mesh?  Would you be wanting to apply materials to those inner shapes, possibly without applying materials to the 48 overlapping spheres that CAUSE that inner shape?  *shrug*  I'm not really qualified to talk about any of this, but... I think there might be some fun... in creating a math-driven sphere positioner.    I'm a simpleton yooper boy... what can I say?  Fibonacci is my kick, because... I grew up with a backyard full of pine cones (Fibonacci-infested fun shapes that we kids examined early and often.  Snail shells, too.)  (We're related to the Conch tribes who are native to coastal beaches.)  (seashell pickers)

All in all, math-derived geometry... is better than any Blender model, 8-to-1... in my opinion.  Let's try some math-based sphere positioners... and get some ooohs and ahhhs going, here, huh?  Maybe it starts... with hacking-up the spherical harmonics code that @jerome was kind enough to convert-into BJS goods.

[jerome]

Unfortunately, I've currently no time to play with such interesting things :-(   (new univ year here, new students everywhere asking for stuff, etc... I hope it will be quiet back in some weeks... months, arg)

[Wingnut]

Yeah @jerome, I wasn't implying that YOU should do any hacking. Just giving you credit for the coolest dynamic mesh on the planet.    ALL the stuff you code... just a fantastic lesson in BJS geometry.  I love it all.  I was recently studying Curve3.  What a great tutorial and demo you made for that, and what a useful class.  I'm beginning some study on building a HUD in Canvas2D, and Nockawa used Curve3 in his Lines2D Shape-class tutorial and demo.

Combining these two "features", man, I have everything I need.. to do anything... HUD-wise.  Just sweet.... but quite off-topic here in Bubbleville. 

On-topic:  After re-considering my overlapping spheres thought, I suspect that will not give the desired results.  It is more about how the bubbles attach to each other, isn't it?

All in all, I think I must excuse myself from this topic, because it is just too far outside of my intelligence.  Hell, it seems I have plenty of work myself, learning when to use position.clone()  (I have had some failures with that, lately.)

This is an interesting topic.  Only a small portion of the topic... is directly-related to Babylon.js, it seems (the displaying part).  The rest... looks like math.  (Wingnut hides under the bed... joining his very-scared dog). 

I DID derive a possible forum-ula, though.  (balloonElasticity - soapBubbleElasticity)/2 = bubbleGumElasticity. heh

[BitOfGold]

There are some nice shaders on shadertoy, with thin film interference...
Nice math
https://www.shadertoy.com/view/ld3SDl
https://www.shadertoy.com/view/XddXRj