Is it possible to emit particle fluids without the “Use Stream Emission”
and have a good result.

My first try is to set my birth rate per impulse. The problem with that
is as soon as a collision occurs, my emission become unstable. I think
what is happening is that the number of sub-step is calculated
automatically and it needs more sub-step with a collision and then a lot
more particles are born at this time.

The other solution is with constant birth rate but I cannot have a nice
distribution of all the points at the same time. For example, half of
them are created on a different sub-step then the rest.

Any ideas? I tried this question on the mailing without sucess.

Thanks

François

Hi Francois,

I saw your post on the mailing list but didn't manage to reply. Try turning on “Density Test Activation” under the Density Test tab. The Density Threshold prevents particles from birthing where it will exceed the Rest Density thus causing your sim to explode.

Usually, I channel reference the Rest Density from the Particle Fluid object & multiply it by 0.9, just in case.

Hope the above helps!

Cheers!
steven

Hi Steven,

I tried what you said but it doesn't work with my rest density * 0.9. Instead I had to write directly 0.2 without the multiplication. I think they've changed the way this parameter works because in the help they say they internally multiply it by the rest density.

However with 0.2 it still behaves differently when the first particles hit the ground. With stream emission I don't feel any change in the flow.

The other thing I like with stream emission is the speed. It looks like it generates let's say 3 rows in 1 timestep if it needs to fit with the speed which is great.

I guess now my question should be, how can I use the stream emission with a different direction per point based on the normal maybe.

That would be also great to be able to do that in Pop too to avoid oversampling too much just to have a nice distribution.

If it's not possible I would make a RFE. What do you think about that?

François

Hi Francois,

I just re-read your initial post again and remember that the Impulse birth is based on the Maximum substeps of the Particle Fluid Solver. If you need collisions to work better, you need higher substeps so you can take the Impulse Rate and divide it by the Maximum Substeps to ensure the number of particles birthed is more consistent. I think that might help.

However, I agree it would be very nice to have the Stream Emission work based on per point. Please submit the RFE if you can.

Thanks!

Cheers!
steven