Pampered Degenerate
kiwifarms.net
- Joined
- Apr 10, 2020
I don't know the answer, but this looks like a good place to start:
Null's Math Problem for 6 Months
- Thread starter Null
- Start date
- Joined
- Sep 17, 2019
Null if this is for your game then I stand by the answer I had in the previous thread: just use really basic rate equations or something for this, don't worry about it being thermodynamically perfect. At best it'll just be a lot of complicated equation solving for something barely anyone is going to appreciate, and at worst you'll end up with the "Reality is Unrealistic!" case: where your users will think the equations are fucked because the system doesn't react the way they expect a video game heat system to react.
If you still want the 'proper' solution anyway, a lot of the Kiwis in that original thread did some solid work. (e.g. here, here, here (using simple rate equations, go figure), here...)
If you still want the 'proper' solution anyway, a lot of the Kiwis in that original thread did some solid work. (e.g. here, here, here (using simple rate equations, go figure), here...)
Bongocat
kiwifarms.net
- Joined
- Jun 20, 2020
IF this is for a game, and you need a somewhat more general equation, then you could use FD software to simulate many different situations and then linearly interpolate between the results as needed. Again though, I don't know what it's for, the parameter range ect...
If it's as straightforward as what you describe, then you could just simulate and take several "snapshots" as time goes on, then blend between them. That should be good enough for just about any purpose as far as presentation is concerned.
What if this is some kind of weird psychological test Null is doing on us.
If it's as straightforward as what you describe, then you could just simulate and take several "snapshots" as time goes on, then blend between them. That should be good enough for just about any purpose as far as presentation is concerned.
What if this is some kind of weird psychological test Null is doing on us.
- Joined
- Oct 21, 2018
this is just ugly math...
12mol gas with 20 °C and 13 mol with 40°
thats 0.2924 bar and 0.3384 bar
now there is a problem, i have no idea how you determine the speed of the pressure reaching equilibrium.
12mol gas with 20 °C and 13 mol with 40°
thats 0.2924 bar and 0.3384 bar
now there is a problem, i have no idea how you determine the speed of the pressure reaching equilibrium.
- Joined
- Apr 30, 2020
Math hard, make grug brain sore like when grug fall on rock.
That is unironically how these threads make me feel. Mathbros, how do you do it?
That is unironically how these threads make me feel. Mathbros, how do you do it?
- Joined
- Nov 14, 2012
that's not _a_ problem. that's _the_ problem. I have everything else sorted.
- Joined
- Jan 10, 2020
Chemical engineer here. This is an interesting mass transfer problem that's actually quite complex. You would have to use the diffusion equation because there is time dependence.
Because this is a 3D problem, an analytical solution would be somewhat difficult to get unless some dimensionality assumptions were made, so below is the numerical solution:
The diffusion coefficient of gases is very low, on the order of 1 x 10^-5, but changes with temperature. I couldn't be bothered to find the exact numbers, as it wouldn't have a big impact on the solution, especially in the timeframe asked for in the problem. There is no change in nitrogen or oxygen amount at all times. For chlorine however, there is movement from box B to box A.
0.2 sec:
1 sec.
5 sec.
Solution:
Edit: The reversed arc length refers the distance along the centerline of box B from where the boxes are joined.
Because this is a 3D problem, an analytical solution would be somewhat difficult to get unless some dimensionality assumptions were made, so below is the numerical solution:
The diffusion coefficient of gases is very low, on the order of 1 x 10^-5, but changes with temperature. I couldn't be bothered to find the exact numbers, as it wouldn't have a big impact on the solution, especially in the timeframe asked for in the problem. There is no change in nitrogen or oxygen amount at all times. For chlorine however, there is movement from box B to box A.
0.2 sec:
1 sec.
5 sec.
Solution:
Edit: The reversed arc length refers the distance along the centerline of box B from where the boxes are joined.
Attachments
Bongocat
kiwifarms.net
- Joined
- Jun 20, 2020
Well, don't feel bad about anything involving fluid dynamics at least. That's a highly specialized and difficult field that people spend years in school for and make bank for doing successfully. I've learned that's the case with a lot of things, and often solving it 'properly' is not worth the effort. You can get equally good results by just asking an oracle and cheesing the rest. It's not cheating if it's not against the rules.
Thank you for clarifying that.
Last edited:
- Joined
- Oct 19, 2019
The biggest problem I see is you're using chlorine instead of phosgene. Most of the feds and bankers will survive this.
Alexander Clamilton
kiwifarms.net
- Joined
- Aug 22, 2020
- Joined
- Jan 10, 2020
- Joined
- Nov 14, 2012
Really? It's that slow?
Why does it feel like if you're standing in a big room and open a jar containing a ton of some gas (like Chlorine), you'd probably see it spread around pretty fast? Is that just the energy in the air or what?
Bongocat
kiwifarms.net
- Joined
- Jun 20, 2020
I'm not an expert, but in any realistic situation, you'd probably have a door or curtain opening and that would probably create a lot of air disturbance, causing it to spread much quicker. There also isn't that much of a pressure differential in this case.
- Joined
- Jan 10, 2020
Two things:
1. The simulation assumes that only diffusion is present. When gas moves, it can move in two different ways: convection, which is the gas being carried by another fluid (think mixing cream into coffee) and diffusion, which is the movement of gas based on the random movement of gas atoms by Brownian motion. In a real room, you would likely have air moving within it.
2. The plot is based on the average concentration in the entire box. In other words, it would take ~35 hours for the chlorine concentration at every point in Box B to equal the chlorine concentration of every point in Box A.
Edit: Also, in this case, you can see that as you get closer to equilibrium, the equilibration rate is slower. 15-20 hours is probably enough for all practical purposes.
- Joined
- Oct 21, 2018
that shouldnt be to complex, but i have no idea how to find the right formula...
thats more complex than needed, the change in pressure should be the main factor.
what? im pretty sure thats wrong. or you mean chlorine concentration, not pressure...
we dont need that, the pressure difference is the main factor.
Navier-stokes is the right equation to look for, but i have no idea how to get the right formula for our problem.
Bongocat
kiwifarms.net
- Joined
- Jun 20, 2020
Is there a fluid dynamics software that you'd recommend which is reasonably beginner friendly + doesn't cost multiple arms and legs?
- Joined
- Jun 24, 2020
That gas is most likely under pressure. Also, if you cracked an oxygen tank open in a chamber with slightly lower pressure you wouldn’t see the same effect.
- Joined
- Nov 14, 2012
In my simulation I want things to move enough that people are able to work with gasses in real time. Before I had it so that the adjacent cubes would interact with each other as quickly as possible (i.e. a cube and its 6 adjacent neighbors would equalize 30 times per second, or as fast as the framerate could allow), which works, but then you have an issue where something is added to the air and then suddenly the room is just completely contaminated before you can even react.
I'm really trying to keep things simple, so I'm reluctant to say "okay I need to keep track of kinetic energy in each cube as well" - especially because there are going to be tens of thousands of these if things get bigger.
In SS13, I think gas just spreads at a fixed rate, which creates odd situations where a hull breach leaves only the neighboring tiles in an 8 block radius low pressure and things like 1 tile wide doorways effectively halt depressurization for 10 minutes before it spreads out of that room.
- Joined
- Dec 1, 2020
So it was a SS13 thing, what a fucking gay thread.
Bongocat
kiwifarms.net
- Joined
- Jun 20, 2020
Could you use a fluid solver library and adjust the viscosity of the air/gas? It could be more resistant to disturbance while maintaining it's ability to spread at a slow-ish rate through doorways. For dealing with the amount of cells in the scene, could you partition the scene into larger blocks that turn on/off their simulation as needed? Maybe level of detail or a simplified algorithm could also apply if complex geometry was not involved in that room.
So for instance, if a room is entirely filled with gas, there's no need to simulate it anymore right? Similarly, if the gas has not reached the room yet, you also don't need to simulate it. So that just leaves the rooms that are currently being filled to be simulated, which should be significantly less, if I had to guess.