Pouring a bucket of water onto an umbrella - Strange results from new user

[Deflaminis]

Hello, I've got some weirdness going on in one of my scenes here with Phoenix FD. I'm a novice user, so I don't know what I don't know but I've got two movies here with symptoms and was hoping I could get some pointers for making this look better.

I've read a lot of the forums here about Phoenix, so I'm pretty sure this scenario is the hardest there is for water sim software, I have a moving object and it's convex. (Double trouble!)

My settings are generally defaults here, with the exception that I'm at SPF 6.0. I added a camera zoom that won't be in the final shot so we can see what's going on on the top of the umbrella. I know I'm in for a world of hurt on this one, but any advice would be greatly appreciated. One of them is especially comical! I know there are strengths and weaknesses to each type of advection, but I'm not sure how to approach either one here in terms of this problem.

In the forward transfer, the drops don't want to move down the shape, they like to sort of stay put. Also, the splash itself dissapates very quickly, almost... cartoonishly. However this one seems the closest to what I want. Should I spin the umbrella or something to move those drops down the shape? In the classic it's like I left the faucet on high, however the water drops for only about 10 frames then stops discharging, but the classic method is weirdly spawning more of it where there is none. (I've read about this somewhere on here, that's why I increased me SPF to 6.) The only difference in each sim is the convection method.

I've attached the movies here, I can send the scene file if needed as well. (But I'll have to email it, it's too big to post.) I'm using FD 2.2, max 2014. I've posted in another thread about rain drops, but this is a huge bucket of water and not the same shot that I was asking about here: http://forums.chaosgroup.com/showthr...and-rain/page2

Thank you.
AJ

[Ivaylo Katev]

is this the official build or it's a nightly? seems like there is an instability in the classic advection movie.
anyway, in general this task seems practically impossible to simulate in grid based solver, because the geometry is voxelized, from the simulation point of view the umbrella is like if it was made of lego blocks. in the rain thread i suggested to use particle flow for the simulation, and only the rendering to be with the splash shader of phoenix.
however, right now we are developing a particle based solver (flip). it's still in too early stage to be integrated in the nightly build, but i think it can simulate your scene much better. if it's urgent i can send you a special build with flip solver inside

[Deflaminis]

Thank you Ivaylo for your response. I didn't see how similar this was to the other scenario until you explained it. Thanks for being patient with me. No urgency here, it's all learning right now.

Now if only I could find a way to fake the wettening effect that Phoenix has I should be set.