More planetary close encounters and collisions

[Greenleaf]

The most interesting part in this video is how important tidal forces are during close encounters and actual collisions. These effects can currently not be, and likely never will be, modeled properly with the gradual collision mode where planets are spheres. It can however be included in UniverseSandbox² - and it will be. The only question is the time frame.


Another interesting thing is the texture based coloring of planets, which is used to make two Earths collide as opposed to two generic blobs. The resolution, at the time of recording last year, was not that high which gives the planets a blurred appearance. You can still pick out continents though. In the final version the heating from deformation is also modeled so everything would glow, and nothing would keep the original texture coloring.

http://youtu.be/dajIf1BHq48

[dechireur77]

So it will be put in "Universe Sandbox ²" ?

[shadow6061be]

Amazing work !

I wonder if this option will be implemented in 2D or 3D ?

[smjjames]

Amazing work !

I wonder if this option will be implemented in 2D or 3D ?

Why would it be implemented in 2D when Ubox is 3D?

[Greenleaf]

Amazing work !

I wonder if this option will be implemented in 2D or 3D ?

Why would it be implemented in 2D when Ubox is 3D?

Well, its UniverseSandbox² ;-)
No, seriously, these videos are already in 3D. It was, and is, considered to perhaps have super high resolution 2D models of specific simulations which naturally lend themselves to 2D, but in general you are right. Universe Sandbox is 3D and to keep the seamless experience, SPH has to be 3D as well.


The main issue with 3D vs 2D is that to have the same spatial resolution (in 3D as in 2D) you need a lot more particles. As an example, if 1000 particles works for 2D, then you need sqrt(1000)^3=31623 in 3D to get the same resolution.


That is just something we have to deal with though

[WitheHole18]

ok,thanks,the their jobs is fantastics,hopefully it will be included in UniverseSandbox ² 

[Xriqxa]

Just amazing. Your progress is just beautiful with this.

[tuto99]

Very nice video. Lucky your computer can run the sph simulations AND record videos..
Just wondering, aren't these sort of models you see in scientific simulations seemingly a bit exaggerated? I know it's supposed to represent collision situations and formation of bodies, but it all looks too smooth and fragile, as if the bodies had no strength. It's a nice and pretty way of simulating collisions, but seems at least somewhat inaccurate. I don't know, maybe it's not meant to simulate the stature of planets.... Just wondering, I am really into these sort of things.

[Greenleaf]

Very nice video. Lucky your computer can run the sph simulations AND record videos..
Just wondering, aren't these sort of models you see in scientific simulations seemingly a bit exaggerated? I know it's supposed to represent collision situations and formation of bodies, but it all looks too smooth and fragile, as if the bodies had no strength. It's a nice and pretty way of simulating collisions, but seems at least somewhat inaccurate. I don't know, maybe it's not meant to simulate the stature of planets.... Just wondering, I am really into these sort of things.

The interesting thing is that it is not inaccurate, quite the contrary. The issue is your expectations. You are used to a rock having a solid feel to it, but if you scale things up, size and force, then this changes drastically.
If you were a giant, so the earth was the size of a football relative to you, and you gently poked at it with your finger, it would act as if it was entirely fluid. not because it is, but because the structural strength doesn't scale like the finger poking force does. Also, a gentle poke for you would perhaps be 1000 km/s, meaning that things would start heating and evaporating due to shock when you touched it.

Another fun example is the strength of a leg bone which holds a mans weight. If you scale the man to double width height and thickness his weight is now 8 times larger. The cross section area of a leg bone is however only 4 times larger, meaning the pressure inside that bone is now double what it was before. Scale enough and the leg breaks.

Same with planetary physics.

[tuto99]

Very nice video. Lucky your computer can run the sph simulations AND record videos..
Just wondering, aren't these sort of models you see in scientific simulations seemingly a bit exaggerated? I know it's supposed to represent collision situations and formation of bodies, but it all looks too smooth and fragile, as if the bodies had no strength. It's a nice and pretty way of simulating collisions, but seems at least somewhat inaccurate. I don't know, maybe it's not meant to simulate the stature of planets.... Just wondering, I am really into these sort of things.

The interesting thing is that it is not inaccurate, quite the contrary. The issue is your expectations. You are used to a rock having a solid feel to it, but if you scale things up, size and force, then this changes drastically.
If you were a giant, so the earth was the size of a football relative to you, and you gently poked at it with your finger, it would act as if it was entirely fluid. not because it is, but because the structural strength doesn't scale like the finger poking force does. Also, a gentle poke for you would perhaps be 1000 km/s, meaning that things would start heating and evaporating due to shock when you touched it.

Another fun example is the strength of a leg bone which holds a mans weight. If you scale the man to double width height and thickness his weight is now 8 times larger. The cross section area of a leg bone is however only 4 times larger, meaning the pressure inside that bone is now double what it was before. Scale enough and the leg breaks.

Same with planetary physics.

Ah, makes sense now. I am just too used to having the belief that rocky planets are super solid. Thanks for clearing things up.

[pac0master]

The idea is very good, but will it take a lot of processing power to render it all?

Will it mean we will be able to make similar impact than the videos we see all over the internet?
maybe not as precise but just something similar would be awesome.


This for example
We've all seen this vid :3
https://www.youtube.com/watch?v=PENT_hnyO-o

[Greenleaf]

The idea is very good, but will it take a lot of processing power to render it all?

Will it mean we will be able to make similar impact than the videos we see all over the internet?
maybe not as precise but just something similar would be awesome.

We are actually using that as a target for development. :-)
The hard part about Universe Sandbox ², and any sandbox really, is that the solution needs to handle any situation. If we were just to show one specific collision, it would have been done long ago, and looked as good as that video.
If we should handle all situations, then we need to actually calculate things physically correct at very high resolution. That can also be done without too much trouble, but then the issue is that it will not be at interactive speeds. Therefore we sit between two chairs, and try to hold on to thin air, while calculating as accurately as we can without making it slow.


Speaking of speed, do note that real time would be quite doable. Generally things, in the simulation, are running thousands of times faster than real time.

[pac0master]

That would be awesome, I already enjoy crashing things into planets, I can't Imagine how fun it might be.

[Retsof]

Can we get some more videos?  Also I am sad that I can't find any planetary SPH toys on the internet to play with, it looks like great fun.