[Updated to V1.3 - now MacOS support] FLUIDOS: a fluids simulator plugin for Daz Studio (commercial)

Veneris

Congratulations. I had been waiting for this, now I only need one simulator for the fire and one for physics.

Alberto

Veneris said:

Congratulations. I had been waiting for this, now I only need one simulator for the fire and one for physics.

Thank you!

It's planed fire and smoke in a future version of FLUIDOS.

Hurdy3D

Alberto said:

Veneris said:

Congratulations. I had been waiting for this, now I only need one simulator for the fire and one for physics.

Thank you!

It's planed fire and smoke in a future version of FLUIDOS.

And raid? Or a shower?

Alberto

 

gerster said:

And raid? Or a shower?

A shower can be done just now. Imago's example is here:

Imago said:

Glad to have been helpful!

For GLE, here a quick example. I used five sources here without any noticeable increase in the simulation's time. You can use more than that to obtain your shower!

Spray can be simulated enabling diffuse particles in Domain and the Mesher.

Alberto

There is attached the .duf file of spray.

SimonJM

Twice now DS has got into what looks like an infinite loop, doing '"FluidOS reconstruction - FluidOS Mesher" and jumping apparently randonly up and down the Timeline, frame 23, 4, 25, 15, 14, 22, 5, 22, 23, 3, 4, 21 are just the latets sequene.  The first time it happened i had to cancel DS as nothing would stop the process.  Rather annoying as I had just spent (this time) a fair bit of time fine tuning the scene.  It should not be a memory issue as the computer is running at under 20% RAM usage.

The first time it happend I am pretty sure the simulation had ended and I was adjusting the frame to see which would give a better result when it went 'bokers'.  This tiem I'm not sure (it's late and I am tired) if it had finsihed or was at frame 29 of 30, or if I was beginning to check out frames again.  From what I can see there have been no updates in the log - the last 4 or so entries deal with frame 30 and are from around 20 or 30 minutes ago.

Alberto

SimonJM said:

Twice now DS has got into what looks like an infinite loop, doing '"FluidOS reconstruction - FluidOS Mesher" and jumping apparently randonly up and down the Timeline, frame 23, 4, 25, 15, 14, 22, 5, 22, 23, 3, 4, 21 are just the latets sequene.  The first time it happened i had to cancel DS as nothing would stop the process.  Rather annoying as I had just spent (this time) a fair bit of time fine tuning the scene.  It should not be a memory issue as the computer is running at under 20% RAM usage.

The first time it happend I am pretty sure the simulation had ended and I was adjusting the frame to see which would give a better result when it went 'bokers'.  This tiem I'm not sure (it's late and I am tired) if it had finsihed or was at frame 29 of 30, or if I was beginning to check out frames again.  From what I can see there have been no updates in the log - the last 4 or so entries deal with frame 30 and are from around 20 or 30 minutes ago.

Set OFF the mesher to avoid reconstruction during simulation (in high resolution simulations reconstruction considerably slow down the process).

The simulation is adaptive stepsize to get more accurate results. That is, if the velocities of fluid particles are high, the engine subdivides the frame step in substeps, thus you see more than one entry per frame. You can increase speed of simulation by increasing the CFL condition number (maybe to 10 or 20) or disabling at all the adaptive stepsize by setting CFL number to 0.0. The price of the speed is some loss of accuracy. 

 

SimonJM

Alberto said:

SimonJM said:

Twice now DS has got into what looks like an infinite loop, doing '"FluidOS reconstruction - FluidOS Mesher" and jumping apparently randonly up and down the Timeline, frame 23, 4, 25, 15, 14, 22, 5, 22, 23, 3, 4, 21 are just the latets sequene.  The first time it happened i had to cancel DS as nothing would stop the process.  Rather annoying as I had just spent (this time) a fair bit of time fine tuning the scene.  It should not be a memory issue as the computer is running at under 20% RAM usage.

The first time it happend I am pretty sure the simulation had ended and I was adjusting the frame to see which would give a better result when it went 'bokers'.  This tiem I'm not sure (it's late and I am tired) if it had finsihed or was at frame 29 of 30, or if I was beginning to check out frames again.  From what I can see there have been no updates in the log - the last 4 or so entries deal with frame 30 and are from around 20 or 30 minutes ago.

Set OFF the mesher to avoid reconstruction during simulation (in high resolution simulations reconstruction considerably slow down the process).

The simulation is adaptive stepsize to get more accurate results. That is, if the velocities of fluid particles are high, the engine subdivides the frame step in substeps, thus you see more than one entry per frame. You can increase speed of simulation by increasing the CFL condition number (maybe to 10 or 20) or disabling at all the adaptive stepsize by setting CFL number to 0.0. The price of the speed is some loss of accuracy. 

 

I really should read the documentation - so it's best practice to have the Mesher Enable option set to Off whilst running the simulation?

Alberto

SimonJM said:

Alberto said:

SimonJM said:

Twice now DS has got into what looks like an infinite loop, doing '"FluidOS reconstruction - FluidOS Mesher" and jumping apparently randonly up and down the Timeline, frame 23, 4, 25, 15, 14, 22, 5, 22, 23, 3, 4, 21 are just the latets sequene.  The first time it happened i had to cancel DS as nothing would stop the process.  Rather annoying as I had just spent (this time) a fair bit of time fine tuning the scene.  It should not be a memory issue as the computer is running at under 20% RAM usage.

The first time it happend I am pretty sure the simulation had ended and I was adjusting the frame to see which would give a better result when it went 'bokers'.  This tiem I'm not sure (it's late and I am tired) if it had finsihed or was at frame 29 of 30, or if I was beginning to check out frames again.  From what I can see there have been no updates in the log - the last 4 or so entries deal with frame 30 and are from around 20 or 30 minutes ago.

Set OFF the mesher to avoid reconstruction during simulation (in high resolution simulations reconstruction considerably slow down the process).

The simulation is adaptive stepsize to get more accurate results. That is, if the velocities of fluid particles are high, the engine subdivides the frame step in substeps, thus you see more than one entry per frame. You can increase speed of simulation by increasing the CFL condition number (maybe to 10 or 20) or disabling at all the adaptive stepsize by setting CFL number to 0.0. The price of the speed is some loss of accuracy. 

 

I really should read the documentation - so it's best practice to have the Mesher Enable option set to Off whilst running the simulation?

At least for high resolution, yes.

nekyo

Fluidos is a lot of fun to play with, thanks Alberto!

The recipe PDF was very helpful to get started.

RAMWolff

WOW.. really cool!  

Alberto

nekyo said:

Fluidos is a lot of fun to play with, thanks Alberto!

The recipe PDF was very helpful to get started.

I'm glad you like it!

jamesl767

Hi ould you upload more sample files to get us started 

thanks

jamesl767

Here is what I'm getting demo file 

Alberto

jamesl767 said:

Hi ould you upload more sample files to get us started 

thanks

Yes!

jamesl767 said:

Here is what I'm getting demo file 

Is this the spray.duf?, if so, the mesher must have set ON the "Diffuse particles" property.

jamesl767

Alberto said:

jamesl767 said:

Hi ould you upload more sample files to get us started 

thanks

Yes!

jamesl767 said:

Here is what I'm getting demo file 

Is this the spray.duf?, if so, the mesher must have set ON the "Diffuse particles" property.

the spray.duf works this is Fluidos Scene 2.duf I keep getting clumping

Imago

Try this, it's the one linked in Alberto's signature: Scene 2

Alberto

jamesl767 said:

Alberto said:

jamesl767 said:

Hi ould you upload more sample files to get us started 

thanks

Yes!

jamesl767 said:

Here is what I'm getting demo file 

Is this the spray.duf?, if so, the mesher must have set ON the "Diffuse particles" property.

the spray.duf works this is Fluidos Scene 2.duf I keep getting clumping

Change the Preferred device, please, and let me know if it works.

jamesl767

Alberto said:

jamesl767 said:

Alberto said:

jamesl767 said:

Hi ould you upload more sample files to get us started 

thanks

Yes!

jamesl767 said:

Here is what I'm getting demo file 

Is this the spray.duf?, if so, the mesher must have set ON the "Diffuse particles" property.

the spray.duf works this is Fluidos Scene 2.duf I keep getting clumping

Change the Preferred device, please, and let me know if it works.

that worked! my 1080 gpu maybe the problem 

 

Alberto

This is video of a river flow simulation.

This is the setup of the river. There is a source in one extreme of river bed, and a sink in the other. A cube is used for the initial water. It was added a little value to Force X to force the water to flow. The smoothing is for avoid the staggering over the river bed. Diffuse particles are enabled. 

Here are the Source Properties. All velocities are set to zero. The body forces will push the water.

Here are the Sink Properties

The cube properties:

The terrain acts as a obstacle, thus, no changes were done, but the parenting to Domain.

These are the properties of the first mesh, for reconstructing the water mesh.

These are the properties of the second mesh, for reconstructing the Whitewater (Diffuse particles).

The meshes should be enabled after concluiding the simulation process.

The .duf file is attached. If you load it in Daz Studio, don't forget to adjust the preferred device.

 

 

smaker1

Hello Alvin

having fun with Fluidos :-)

What would be the best way to have a "calm" water cube with someone waking and making ripples on the surface ? I tried to liquify a cube but my results was not good

Thanks for you help 

Richard Haseltine

 

smaker1 said:

Hello Alvin

having fun with Fluidos :-)

What would be the best way to have a "calm" water cube with someone waking and making ripples on the surface ? I tried to liquify a cube but my results was not good

Thanks for you help 

You are parenting the figure to the domain, and turning the fluid body option off so that it's an obstacle? That said, I am not getting the results I would expect from a cylinder pushing into and out of a block of water - there are not ripples propagating away from the point of intersection, just a dent there. This is with force on for the cylinder prop, the last frame of a 200 frame simualtion:

Alberto

Richard Haseltine said:

 

smaker1 said:

Hello Alvin

having fun with Fluidos :-)

What would be the best way to have a "calm" water cube with someone waking and making ripples on the surface ? I tried to liquify a cube but my results was not good

Thanks for you help 

You are parenting the figure to the domain, and turning the fluid body option off so that it's an obstacle? That said, I am not getting the results I would expect from a cylinder pushing into and out of a block of water - there are not ripples propagating away from the point of intersection, just a dent there. This is with force on for the cylinder prop, the last frame of a 200 frame simualtion:

Turn ON "Enable Moving Obstacles" in Domain's Main settings. This way, the cylinder could push the water.

Alberto

smaker1 said:

Hello Alvin

having fun with Fluidos :-)

What would be the best way to have a "calm" water cube with someone waking and making ripples on the surface ? I tried to liquify a cube but my results was not good

Thanks for you help 

Try to use the shallowest water you can (to get faster simulation). Set ON "Enable Moving Obstacles" in Domain's Main settings.

This is the same I used in Genesis walking over snow in the second demo. Here, viscosity is disabled.

The cell size is 2.5. You could resize the domain (and the cube), but try to not reduce the resolution. In this scene, the Grid size is 217 X 7 X 40 = 59,360 cells. Thus, any modifications should have around  60,000 or more cells.

Here you see the Domain settings.

Cube settings:

If you use aniblocks, bake them so the plugin can recognize the keyframes.

The .duf file with the Domain and the cube water is attached.

smaker1

Many Thanks Alvin !

This look very promissing. I'm gonna study this tomorow (too late for me right now!)

Fluidos open plenty of new possibilities. Congratulations!!!

Thanks again for the support

 

Richard Haseltine

Alberto said:

Richard Haseltine said:

 

smaker1 said:

Hello Alvin

having fun with Fluidos :-)

What would be the best way to have a "calm" water cube with someone waking and making ripples on the surface ? I tried to liquify a cube but my results was not good

Thanks for you help 

You are parenting the figure to the domain, and turning the fluid body option off so that it's an obstacle? That said, I am not getting the results I would expect from a cylinder pushing into and out of a block of water - there are not ripples propagating away from the point of intersection, just a dent there. This is with force on for the cylinder prop, the last frame of a 200 frame simualtion:

Turn ON "Enable Moving Obstacles" in Domain's Main settings. This way, the cylinder could push the water.

Yes, I was just trying that, once with a continuing oscillation and once with only two impacts on the surface (too much water here, so it took about half-an-hour to sim with a cell-size of 1.25 and 200 frames). I'm not sure why the water starts with an inverse meniscus and finishes dead flat.

https://www.dropbox.com/s/602w7h1cc3yfqya/Fluidos ripple test.avi?dl=0
https://www.dropbox.com/s/y4h890dl1wvrew3/Fluidos Ripple Test II.avi?dl=0

Alberto

Richard Haseltine said:

Alberto said:

Richard Haseltine said:

 

smaker1 said:

Hello Alvin

having fun with Fluidos :-)

What would be the best way to have a "calm" water cube with someone waking and making ripples on the surface ? I tried to liquify a cube but my results was not good

Thanks for you help 

You are parenting the figure to the domain, and turning the fluid body option off so that it's an obstacle? That said, I am not getting the results I would expect from a cylinder pushing into and out of a block of water - there are not ripples propagating away from the point of intersection, just a dent there. This is with force on for the cylinder prop, the last frame of a 200 frame simualtion:

Turn ON "Enable Moving Obstacles" in Domain's Main settings. This way, the cylinder could push the water.

Yes, I was just trying that, once with a continuing oscillation and once with only two impacts on the surface (too much water here, so it took about half-an-hour to sim with a cell-size of 1.25 and 200 frames). I'm not sure why the water starts with an inverse meniscus and finishes dead flat.

Could I see the settings, please?

Alberto

Richard Haseltine said:

Alberto said:

Richard Haseltine said:

 

smaker1 said:

Hello Alvin

having fun with Fluidos :-)

What would be the best way to have a "calm" water cube with someone waking and making ripples on the surface ? I tried to liquify a cube but my results was not good

Thanks for you help 

You are parenting the figure to the domain, and turning the fluid body option off so that it's an obstacle? That said, I am not getting the results I would expect from a cylinder pushing into and out of a block of water - there are not ripples propagating away from the point of intersection, just a dent there. This is with force on for the cylinder prop, the last frame of a 200 frame simualtion:

Turn ON "Enable Moving Obstacles" in Domain's Main settings. This way, the cylinder could push the water.

Yes, I was just trying that, once with a continuing oscillation and once with only two impacts on the surface (too much water here, so it took about half-an-hour to sim with a cell-size of 1.25 and 200 frames). I'm not sure why the water starts with an inverse meniscus and finishes dead flat.

https://www.dropbox.com/s/602w7h1cc3yfqya/Fluidos ripple test.avi?dl=0
https://www.dropbox.com/s/y4h890dl1wvrew3/Fluidos Ripple Test II.avi?dl=0

Oh, the cube is hiding the mesher.  Set Visible to OFF. I think this is the reason of the dead flat. You don't need to run again the simulation.

Any mesh supposed to be fluid mass must be invisible for rendering.

Richard Haseltine

Alberto said:

Richard Haseltine said:

Alberto said:

Richard Haseltine said:

 

smaker1 said:

Hello Alvin

having fun with Fluidos :-)

What would be the best way to have a "calm" water cube with someone waking and making ripples on the surface ? I tried to liquify a cube but my results was not good

Thanks for you help 

You are parenting the figure to the domain, and turning the fluid body option off so that it's an obstacle? That said, I am not getting the results I would expect from a cylinder pushing into and out of a block of water - there are not ripples propagating away from the point of intersection, just a dent there. This is with force on for the cylinder prop, the last frame of a 200 frame simualtion:

Turn ON "Enable Moving Obstacles" in Domain's Main settings. This way, the cylinder could push the water.

Yes, I was just trying that, once with a continuing oscillation and once with only two impacts on the surface (too much water here, so it took about half-an-hour to sim with a cell-size of 1.25 and 200 frames). I'm not sure why the water starts with an inverse meniscus and finishes dead flat.

https://www.dropbox.com/s/602w7h1cc3yfqya/Fluidos ripple test.avi?dl=0
https://www.dropbox.com/s/y4h890dl1wvrew3/Fluidos Ripple Test II.avi?dl=0

Oh, the cube is hiding the mesher.  Set Visible to OFF. I think this is the reason of the dead flat. You don't need to run again the simulation.

Any mesh supposed to be fluid mass must be invisible for rendering.

Ah, that was it yes - thank you.

Alberto

smaker1 said:

Many Thanks Alvin !

This look very promissing. I'm gonna study this tomorow (too late for me right now!)

Fluidos open plenty of new possibilities. Congratulations!!!

Thanks again for the support

 

 

Richard Haseltine said:

Ah, that was it yes - thank you.

You're welcome!