Coding Adventure: Simulating Fluids

Coding Adventure: Simulating Fluids

Sebastian Lague

9 месяцев назад

1,750,535 Просмотров

Ссылки и html тэги не поддерживаются


Комментарии:

michele daffara
michele daffara - 15.10.2023 16:57

wow

Ответить
ShMokou
ShMokou - 15.10.2023 15:57

Boundary problem not so problem actually. There is a bunch of "boundary effects" in real physics.

Ответить
지하수
지하수 - 15.10.2023 14:36

The fluid simulation worked so beautifully that I almost cried while watching the video. This is a really cool video. thank you

Ответить
Lucas EASTMAN
Lucas EASTMAN - 15.10.2023 14:20

This was awesome! It is so inspiring and I want to build something like this myself now. I am really excited to see what you do next with this project! I can't wait to see your updates!

Ответить
Jeroen Commandeur
Jeroen Commandeur - 15.10.2023 13:57

Yeej, a new video! This, as all other videos, makes me smile. 🙂

Ответить
Will
Will - 15.10.2023 13:01

God that is gorgeous

Ответить
Sashy
Sashy - 15.10.2023 12:18

The next step is to add the temperature!

Ответить
Nathan Ellis
Nathan Ellis - 15.10.2023 11:28

brilliant!! looking forward to part 2

Ответить
Over-in Gaming
Over-in Gaming - 15.10.2023 10:04

I want this as a wallpaper

Ответить
Nick De Raj
Nick De Raj - 15.10.2023 09:14

I don't see myself ever using Unreal in the near future but somehow I watched the whole video, your voice is mesmerising

Ответить
Julian Taylor
Julian Taylor - 15.10.2023 06:52

Always very inspiring! And the 3D simulation was sooooo fun to watch I would love a whole game lol

Ответить
Michel Tatoute
Michel Tatoute - 15.10.2023 06:44

The most simple and efficient way to simulate damping , with no instability and pretty nice visual results is... to scale down speeds at each step (by a % exponentially dependent of the step time). The advantage is that this method always decreases energy. Methods based on damping forces introduce errors in damped force that add unwanted energy to the system, reducing the stability.

Ответить
Vincent Parker
Vincent Parker - 15.10.2023 06:30

Very Cool!

Ответить
Yolko
Yolko - 15.10.2023 04:06

The problem with particles is that an incompressible fluid will have an infinite speed of sound, but with a simulation you are limited by the update time (for a field you'd be implementing the Boussinesq equations, which just remove the sound waves a priori). In principle you want an infinite pressure multiplier, your speed of sound will be proportional to that.
There is of course the point that a smooth density doesn't really have a meaning for a Lagrangian simulation like you have here. Removing the inertia of the particles is the right call. I wouldn't be surprised if this turns out to be some approximation to a Boussinesq fluid.

Ответить
cool plup
cool plup - 15.10.2023 02:36

You have genuinely changed my life, your videos have been a landmark for my development as a programmer and also reaffirming my interest in game development. I started game development at first just through loving games, but you have a way of presenting game development that made me love it as much as a love games themselves. Keep doing it❤️

Ответить
TheIvalen
TheIvalen - 15.10.2023 02:32

Sebastian - Is Nebula a place for your videos? Would love to see them over there. You may have an expanded audience.

Ответить
Crisscross
Crisscross - 15.10.2023 02:22

Damn a human made this video?

Ответить
Pheonix1328
Pheonix1328 - 15.10.2023 01:57

I did find it a bit strange how the water would explode at the start but maybe that's because it's not acting a try incompressible fluid.
I think two (possibly related) optimizations that might help are focusing on areas that have more particles than ones with fewer. I remember watching a video on here about that and I think they visualized it with different sizes boxes or maybe that was more for the second idea. Focusing on areas with high velocities. I remember watching a "Two Minutes Papers" video on it. The calm water areas were visualised as very large blue spheres. While the high energy areas got more attention ie, small red spheres.

Ответить
Ole Peters
Ole Peters - 15.10.2023 00:57

Shut up and take my money :D Please bring it out as is. Already Perfect :)

Ответить
nickdevprod
nickdevprod - 15.10.2023 00:21

Watching second time letsgo

Ответить
LuXoR
LuXoR - 15.10.2023 00:20

This is a well earned Abonnement.
Really nice work, a lot of affort and awesome visualization.

Ответить
Joe Helton
Joe Helton - 15.10.2023 00:13

i was totally waiting fo you to post agin letss gooooo

Ответить
DetErRasmus
DetErRasmus - 15.10.2023 00:09

Looks awesome :)

Ответить
Gabriel
Gabriel - 14.10.2023 23:18

Amazing video, great work!!!

Ответить
Kenji Yip Tong
Kenji Yip Tong - 14.10.2023 21:47

Oh my god

Ответить
Kenji Yip Tong
Kenji Yip Tong - 14.10.2023 21:41

Incredible.

Ответить
Robert Wensman
Robert Wensman - 14.10.2023 21:20

Hmmm... would it be better to use a standard harsh-table instead of inserting into an array that is then sorted?

Ответить
Romain Pinguet
Romain Pinguet - 14.10.2023 21:06

One week later, amazing video ! Very beautiful simulations

Ответить
Probhakar Sarkar
Probhakar Sarkar - 14.10.2023 19:54

awesome !!

Ответить
OldWardrobe Experiment
OldWardrobe Experiment - 14.10.2023 19:21

Very nice and educational!

Ответить
zoodoz
zoodoz - 14.10.2023 19:02

i am trying to do what he is doing but drawcircle doesnt exist

HELPPP

Ответить
egg boy
egg boy - 14.10.2023 18:27

my brain do be braining

Ответить
Curious Side
Curious Side - 14.10.2023 17:52

which softwares did you use

Ответить
mMindo
mMindo - 14.10.2023 16:10

This was my bsc degree subject. Thank god i didnt do any graphical, only calculations, in fortran, that took me almost 1 week of running the code for 2k steps of simulation.

Keep on the good work ❤❤❤. Maybe next time you could do a surface inside the simulation, like a boat or any aerodynamic surface

Ответить
Danimo
Danimo - 14.10.2023 15:02

Brings function compute time down from 20s to 5s to 18ms...

... Still not fantastic... 😮

Ответить
HiggsBoson
HiggsBoson - 14.10.2023 14:48

Sebastian Lague is the 3blue1brown of coding

Ответить
Ruslan N
Ruslan N - 14.10.2023 14:31

nice 👍

Ответить
Antololo
Antololo - 14.10.2023 14:08

I saw ur geographical game its great and i wonder if u could do something like a flight simulator out of it it would be very nice to visit the whole world

Ответить
Hannes Altes Haus
Hannes Altes Haus - 14.10.2023 13:03

I would be really interested to see the double slit experiment in this simulation

Ответить
BarkatheMoon
BarkatheMoon - 14.10.2023 12:42

Love you stuff! Would love to see the simulation handle multiple liquids with different densities / color. I think watching as an "oil" liquid rises above a "water" liquid would be pretty neat. Especially since oil and water are immiscible, so you'd have to somehow account for that. But then also watching as two colored waters try and mix would be interesting as well. While I'm at it, I'll say getting a siphon to appropriately work would be neat as well. Regardless, I'm super excited to see where this goes. Have always wanted to implement my own fluid sim, so until then I shall live vicariously through you.

Ответить
Vito Vučak
Vito Vučak - 14.10.2023 12:11

Always such high quality videos from you. I am looking forward to part 2.

Ответить