best practice for CFD simulation rendering

[fraggle]

Still trying to make some progress on this matter and to find a way to get 0.2l/s incoming water:

The only thing you need is just the surface of your emitter - A * (voxels/sec) * voxel_size will give you the volume of emitted liquid which you can easily convert into liters...

To be able to used this approach to define the discharge in l/s I will need to check how many liters I got in one minute. But Again, It seems that the digital wall vs real word arise... timeframe depends on the number of frame per second the animation will be played, so the link with real word is impossible.

Another approach if to find out why the default value of water velocity is 14.346 cell/s. Is this value a reflect of the real water flow, under with properties (pressure...)
I can't imagine that the default value are just based on visual approach, Chaos Group must have base the value from a scientific approach somehow, and scientist don't use cell or voxel!

Thanks

[Svetlin.Nikolov]

Nope, you don't need the FPS - the voxels/sec value accounts for that, so don't worry about it. The only thing you are missing at this point is the surface of the emitter you use in your setup.

Not sure what you mean by default value of 14.3 vox/sec - maybe you used some of the toolbar presets where the discharge is automatically adjusted depending on the emitter size.

If there is anything else I can help with and you can't find it in the docs, please do ask

[fraggle]

The only thing you need is just the surface of your emitter - A * (voxels/sec) * voxel_size will give you the volume of emitted liquid which you can easily convert into liters...

1 cell = 0.000001 liter = 0.001 cm3
Emitter Diam 20mm = surface 3.14x1?=3,14cm?
Default Velocity from water preset = 14.346 cell/s = 0.014346 cm3
Voxel size = 0.1mm

3,14x(0.014346)*0.1 = 0.004504644 cm3/s (4,504644e-6 liter/s)

For 200cm3 (0.2l/s), the velocity need to be 44398.62506337904 time bigger!!! (200/0.004504644)
Velocity = 14.346*44398 = 636942 cell/s, that even worst that my initial test with a velocity of 200 000 Cell/s.
Either I make an error somewhere during the conversion, or this approach is not the good one.

Thanks
Thanks

[Paul Oblomov]

Try in CFD and then match with phx. Import mesh and just make phx sim the same Then, u'll have some math.

[fraggle]

I do it in Cell/s instead of cm3/s
1 cell = 0.000001 liter = 0.001 cm3
Emitter Diam 20mm = surface 3.14x1?=3,14cm?
Default Velocity from water preset = 14.346 cell/s = 0.014346 cm3
Voxel size = 0.1mm = 0.01cm
3.14*14.346*0.01=0.4504644 cell cubic/s

200 cm3 = 200/0.001=200 000 cell.
So I probably make a mistake on the conversion in the previous post.
But anyway, that's exactly what I tried 2 weeks ago (one of the first post), but this doesn't looks right as the water if looking like a laser.

Thanks for trying to help.
If the doc is right, voxel = cell. https://docs.chaosgroup.com/display/...+FD+in+3ds+Max


So, a 10x10x10cm grid with a cell size of 0.1mm give 1 000 000 (1M) voxels, that's a beginning.
Considering that the quick water setup give a value of velocity of 14.346 cell/second, we can conclude that this value correspond to 0.000014346 liter/second.
0.2 liter/second is then equal to a velocity of 20 0000 cell per second.
But None of the value change when the size of the emitter change, so this shortcut won't give the right result I guess.

Looking at the simulation with 20 000 Cell per second, it's definitively not the right approach![ATTACH=CONFIG]36698[/ATTACH]

[fraggle]

Right, I use the preset Tape Water, and the value change with the object emitter size, so it is not a default value.

maybe you used some of the toolbar presets where the discharge is automatically adjusted depending on the emitter size.

If there is anything else I can help with and you can't find it in the docs, please do ask

[Raven]

Not using the maths Svetlin suggested to calculate it, but do what I just did.
Create a box 10 x 10 x 10 cm. Create emitter, radius 1 cm (20 mm diameter).
Played with settings on outgoing velocity, 70 fills the box over 5 seconds. ie 0.2 l/s.
It wont be 0.2 exactly, but visually it is pretty close, and it does vary slightly depending on your cell size. I found between 65 and 70 worked most times.

Attached my scene to demonstrate.

So if you just recreate that setup in your scene you should be able to calculate the rate you need pretty quickly.

[fraggle]

Thanks Raven,
That's very nice to help on a Sunday.
I still don't get the idea behind this approach.
The bucket is overflowed around frame 120. The more FPS your video card is capable of, the shorter the animation playback will be. Same with the rendering, a codec with 10 frame/s won't give the same result than 30 frame/s.

For sure I'm not into math, but I try to be logical.

Thanks

Not using the maths Svetlin suggested to calculate it, but do what I just did.
Create a box 10 x 10 x 10 cm. Create emitter, radius 1 cm (20 mm diameter).
Played with settings on outgoing velocity, 70 fills the box over 5 seconds. ie 0.2 l/s.
It wont be 0.2 exactly, but visually it is pretty close, and it does vary slightly depending on your cell size. I found between 65 and 70 worked most times.

Attached my scene to demonstrate.

So if you just recreate that setup in your scene you should be able to calculate the rate you need pretty quickly.

[Raven]

Video card playback should have nothing to do with it.
Phoenix is tied to your scenes frame rate.
In SA we are on PAL. So that is why my animation was set to 25 frames per second rate.
I have tested it on 30 fps as well.
I did have to adjust the velocity to 80, but I think it is because it basically becomes more accurate, as it has more frames and with it being 18 steps per frame we now have 2700 steps as apposed to 2250 in the pal version.
So if you have less steps per frame it may fill up quicker.

[Svetlin.Nikolov]

Hey,

Just dropping in on the math...

Your emitter area should be more like 12.57 cm^2.

If you are using the preset, you need to go to your Dynamics rollout and reset your unit scale to 1.0, because this gets automatically adjusted with the Tap Water preset.

So, with a discharge of 1600 (only if you are using Surface Force in your source) you should get 0.2L/s = 200cm^3/sec with an emitter that has an area of pi*2cm*2cm.

Hope this helps

[fraggle]

Thanks Svetlin for the math.

Surface = pi x R?
R=1cm (diam 2cm), like a real tape water
Surface = 3.14*1*1 = 3.14cm?.
If 0.2l/s is obtain with a 12.57cm2 surface with a discharge of 1600, then I will need 4 time more discharge for a surface 4 time less.
An emitter of 2cm with a discharge of 6400 = 0.2l/s.

But are you sure about your initial statement of 0.2l/s = 12.57cm2 at 1600 discharge?

Hey,

Just dropping in on the math...

Your emitter area should be more like 12.57 cm^2.

If you are using the preset, you need to go to your Dynamics rollout and reset your unit scale to 1.0, because this gets automatically adjusted with the Tap Water preset.

So, with a discharge of 1600 (only if you are using Surface Force in your source) you should get 0.2L/s = 200cm^3/sec with an emitter that has an area of pi*2cm*2cm.

Hope this helps

[fraggle]

I'm afraid it is still not the right value.
This is the result after 2 frame, water is bumping all the way, that didn't match the real word at all.

[Raven]

Hey Svetlin

I must admit the math side still has me stumped as well.
1600 appears to put out way to much.
In a test on my 1 liter box scene, with a 2cm radius emitter it appears that 1600 puts out 0.4 l per frame.

[Raven]

Fraggle

Not sure if you would be able to send me your scene, to do a few tests?

[Raven]

Not sure if it makes any sense, but dividing the 1600 by my frame rate of 25, gives me the 0.2l/s but out of the smaller 20mm diameter tap.