I'm new to using FLIP fluids and I'd like to have a check on some of what I'm finding with friction values.

I have a simple emitter going into a bathroom sink. I sorted out the initial velocity for what looks good coming out of the tap, but that caused the water to splash quite a bit out of the sink (I'm using the swirly solver instead of the splash solver). So I ticked the Use Friction box and am playing with friction values.

I would have thought friction values range from 0 = no friction to 1 = complete loss of motion on contact. But I'm finding setting the value to 1.0 on both the sink and the fluid gives almost no reduction in speed. I need to go up to something like 10 (on both) to get noticeable amounts.

Do values of 10 on both objects sound right?

Should I be changing the velocity on both, or leave one object set to 1.0?

How does Houdini calculate the “final” friction value in a collision between two objects? If it's multiplied, I would assume both objects having a value of 10.0 would be the same as one with 1.0 and one with 100.0, right? I haven't had a chance to try that scenario out, still sorting out some wedges and such.

thanks for any guidance!

In these kinds of situations a little math can help. What if you simply scale the velocity of the fluid when it's Y position is greater or less than a specific cross over point (like right before splash down).

Have you tried using the “stick on collision” option in the FLIP solver? It might give you better control.

I tried the “stick on collision” but I think it was causing my particles to stop completely rather than just slow down across the sink. I think I may have abandoned it too soon though, before experimenting with more settings.

I ended up sorting out the friction values, for me it ended up looking good with 3.5 on the fluid and 1.0 on the sink.

Enivob - I think you were going for the effect of slowing down the fluid just before it hits the sink to keep it from hitting as hard, correct? Would that be noticeable in the final animation as having an odd slowdown, or have you found that not to be a problem?