How does water flow from a faucet?

[richengle]

water flow from a faucet will neck down its radius as it flows to the bottom of the sink. It is asserted that this is due to the fact that gravity increases the velocity of the fluid. This is shown at . I am trying to simulate this, and am not having much luck. shows this for a flat surface. Most normal faucets aren't like that. Has anyone seen a good simulation, with explanation of methods for a standard faucet, preferably with flow in the sink as well?

 

[Dale]

What is described in the first video is a good physics problem to work through, but it does not represent the actual behavior of real water from real faucets. If you look at that, you will see that water rather quickly forms into drops.

So, do you want to describe the "toy" model or something more realistic. A more realistic model probably doesn't have a closed form solution.

 

[richengle]

id like to model water from a faucet. my simulations have it just running straight down, and not necking. it must neck because outside pressure is greater than inside pressure, no?

 

[anorlunda]

Don't forget surface tension. If the stream breaks into drops, surface tension probably dominates.

 

[Janus]

id like to model water from a faucet. my simulations have it just running straight down, and not necking. it must neck because outside pressure is greater than inside pressure, no?

What are using to model it?

If I do a quick default setting fluid sim with Blender, I get necking:

This is one frame of the simulation.

 

[richengle]

What are using to model it?

If I do a quick default setting fluid sim with Blender, I get necking:

View attachment 276037
This is one frame of the simulation.

cool. blender with manta, right? how long of baking to make 5 sec of simulation?

 

[Arjan82]

id like to model water from a faucet. my simulations have it just running straight down, and not necking. it must neck because outside pressure is greater than inside pressure, no?

No. Pressure at both sides is equal (except for surface tension effect). The flow necks because it accelerates due to gravity and needs to confirm to the continuity equation (no water is lost).

 

[Janus]

cool. blender with manta, right? how long of baking to make 5 sec of simulation?

On my machine it takes about 3 min total bake time (data and mesh), without particles ( foam, spray or bubbles) for 5 sec. This is with a fluid resolution set at 96.*
This sim is about 3 1/3 sec long, and includes some bubbles and spray. Most of the time is taken in rendering the frames (Over an hour for 80 frames)

It is somewhat grainy as I converted it to a compressed GIF in order to get the file down to a size I could directly upload here.

*Your bake times may vary.

 

[Janus]

Here are a couple of more examples produced by playing around with the viscosity of the the fluid.

A bit thicker and with the "gravity" turned down a bit.

That bit where it kind of hangs there and spurts out was not planned, but this kind of thing can happen with Blender, where a tweak in the settings can produce unintended results.