Page 1 of 1

Multipath Crazy Straw

Posted: Tue Dec 26, 2017 10:32 am UTC
by jewish_scientist
If I made this what path would the fluid actually take? And something bad has got to happen with the 'g's, right?

Re: Multipath Crazy Straw

Posted: Tue Dec 26, 2017 4:56 pm UTC
by measure
jewish_scientist wrote:If I made this what path would the fluid actually take? And something bad has got to happen with the 'g's, right?

If you assign a "resistance" to each branch, you could calculate the fluid flow similarly to how you calculate electrical current flow in a resistor network. The flow is driven by the pressure differential and impeded by the small size of the tube (because of wall friction and stuff). In the case of parallel routes, the fluid will flow along both branches, but more so in the shorter/wider ones because of their lower resistance. The dead-end branches won't do much while you're drinking, but they might trap fluid/air, and will definitely make it harder to clean.

Re: Multipath Crazy Straw

Posted: Tue Dec 26, 2017 6:43 pm UTC
by Pfhorrest
I would imagine that since there's actually no such thing as suction force, you would end up with the water continuing to find its own level as it's pushed up the straw by the greater water pressure below it. Basically, imagine you constructed your crazy straw at 100X scale, but the bottom end was capped off, so it's just an enormous weirdly-shaped bucket. Then you put a water spigot in that bottom end. You turn on the spigot and the "straw" construction slowly fills with water. Pretty sure it'd just look something like that, but faster and smaller.

Re: Multipath Crazy Straw

Posted: Tue Dec 26, 2017 9:48 pm UTC
by gmalivuk
Pfhorrest wrote:
I would imagine that since there's actually no such thing as suction force, you would end up with the water continuing to find its own level as it's pushed up the straw by the greater water pressure below it. Basically, imagine you constructed your crazy straw at 100X scale, but the bottom end was capped off, so it's just an enormous weirdly-shaped bucket. Then you put a water spigot in that bottom end. You turn on the spigot and the "straw" construction slowly fills with water. Pretty sure it'd just look something like that, but faster and smaller.

Except that's not how crazy straws (or straws in general) work, on account of the massive surface tension water has at the scale of a drinking straw.

Re: Multipath Crazy Straw

Posted: Wed Dec 27, 2017 2:19 pm UTC
by jewish_scientist
measure wrote:If you assign a "resistance" to each branch, you could calculate the fluid flow similarly to how you calculate electrical current flow in a resistor network.

What about gravity?

Re: Multipath Crazy Straw

Posted: Sun Dec 31, 2017 8:50 pm UTC
by measure
jewish_scientist wrote:
measure wrote:If you assign a "resistance" to each branch, you could calculate the fluid flow similarly to how you calculate electrical current flow in a resistor network.

What about gravity?

Gravity will affect the suction force required to operate the straw, but I don't know if it would affect the relative flow rates at steady state. I think it would be analogous to running an electric circuit in a constant external field, or something like that.

I suspect that with a high suction force on a small straw the effect of gravity would be negligible, and that if the flow rate is very large, second-order effects would break down the current analogy (although maybe electrical circuits also start behaving non-linearly at high enough currents). I'm not an expert on either electricity or fluid dynamics, though; this is just my intuition.

Re: Multipath Crazy Straw

Posted: Mon Jan 01, 2018 7:21 pm UTC
by Soupspoon
If you could engineer a tuned reed-flap at each divergent junction (and/or each convergent one), that could ease the need to engineer identical effective flow resistances via both(/every) parallel branch. And for all fluid conditions (air and liquid) at all rates. You just need to get them close enough to let resonances buff and debuff the various passageways.


However, the easiest way to deal with it (plus the 'dead ends') is to twist round in the third dimension. Double back over (or beneath) the necessary paths, in a standard unbifuracted tube, without even any crimped twists that defeat the laminar flow and leave nigh-on-impossible-to-clean gunk trapped in relatively dead zones.

Re: Multipath Crazy Straw

Posted: Mon Jan 01, 2018 8:49 pm UTC
by Zamfir
measure wrote:
jewish_scientist wrote:
measure wrote:If you assign a "resistance" to each branch, you could calculate the fluid flow similarly to how you calculate electrical current flow in a resistor network.

What about gravity?

Gravity will affect the suction force required to operate the straw, but I don't know if it would affect the relative flow rates at steady state. I think it would be analogous to running an electric circuit in a constant external field, or something like that.

Gravity causes the air bubbles, and through thm how many paths are available for the flow. Think of the top of the letter A. Water flow in through the righthand leg. If suction is strong enough, the water level rises to the horizontal bar, and flows out through the left leg. The roof triangle of the A becomes a trapped volume of air. If flow velocities go up, the pressure drop over the horizontal bar goes up as well. This causes the water level to rise in the right side of the roof triangle, pushign out air onthe other side.

If flow velocity becomes high enough, the water level reaches the top of the A and spills over, and all the air is now pushed out. If the flow rate goes down, the whole A stays full of water and the flow keeps going through both the bar and the triangle.
gmalivuk wrote:Except that's not how crazy straws (or straws in general) work, on account of the massive surface tension water has at the scale of a drinking straw.

Drinking straws are just one step too large for significant surface tension effects. Otherwise you couldn't drink - PP is hydrophobic, and the capillary action pushes water out. A quick calculation says that in a 4mm straw, the water level is 1 mm lower due to surface tension. Its an inverse relation, so below one-tenth of an mm or so, the capillary action would become a major resistance in the straw.

Re: Multipath Crazy Straw

Posted: Tue Jan 02, 2018 12:37 pm UTC
by gmalivuk
The fact that you can hold water in a straw by covering only the top end tells me surface tension is doing something important.

Re: Multipath Crazy Straw

Posted: Tue Jan 02, 2018 1:25 pm UTC
by Eebster the Great
The question is if you could hold water in a straw more than about 10 m long. If you could, then friction would play a significant role. But I'm pretty sure you can't.

Re: Multipath Crazy Straw

Posted: Tue Jan 02, 2018 1:52 pm UTC
by gmalivuk
Who said anything about friction?

Re: Multipath Crazy Straw

Posted: Tue Jan 02, 2018 5:17 pm UTC
by sardia
gmalivuk wrote:The fact that you can hold water in a straw by covering only the top end tells me surface tension is doing something important.

Would that work in a vacuum? Hold a straw like cylinder full of water, pump out the air, and then rotate the cylinder. Does the water come out? Assuming it doesn't boil or something.

Re: Multipath Crazy Straw

Posted: Tue Jan 02, 2018 6:26 pm UTC
by Soupspoon
It's the pressure of the atmosphere, against the weight of the water (and the tending-to-zero pressure in the other direction, that only 'feels' like a negative pressure of suction) that counts somewhat.

With vacuum outside, there's no effective-suction holding the weight back. You'd have only meniscus-based forces holding the water from falling out the bottom (plus roughly equal evaporative pushes at both ends of the column?) and if turned open-end-up it'd be glass(-top)-half-empty without the inrushing atmosphere. And a battle between evaporation and freezing (then subliming away) depending on ambient liquid and container temperatures.

Which might even form a decent plug, in inverted mode (or if allowed to form before being inverted) behind which liquid may survive to be held back for as long as the solid plug remains solid and pressing up against thr cylinder sides. I suspect more of the details need to be nailed down before a definitive answer, though.

Re: Multipath Crazy Straw

Posted: Tue Jan 02, 2018 8:14 pm UTC
by gmalivuk
sardia wrote:
gmalivuk wrote:The fact that you can hold water in a straw by covering only the top end tells me surface tension is doing something important.

Would that work in a vacuum? Hold a straw like cylinder full of water, pump out the air, and then rotate the cylinder. Does the water come out? Assuming it doesn't boil or something.
Pressure is *also* doing something important, of course, but pressure plays basically the same role even if you blow everything up to the size of a pipeline.

Re: Multipath Crazy Straw

Posted: Tue Jan 02, 2018 10:58 pm UTC
by Zamfir
gmalivuk wrote:The fact that you can hold water in a straw by covering only the top end tells me surface tension is doing something important.

Yes, but its more subtle than lifting, and only relevant in the upside-down case. The surface tension is clearly not doing the lifting, or you wouldn't need the finger on top. Instead, it seems to stabilize the water-air interface and prevents Rayleigh-Taylor instabilities from growing. In a larger pipe, random fluctuations in the interface will grow into plumes that mix the fluids.

You might know the trick where you put a card on glass with water, turn it over, and the water plus card stay hanging. That's the same idea as the straw - without interface mixing, the water can act as its own suction cup.

Re: Multipath Crazy Straw

Posted: Tue Jan 02, 2018 11:06 pm UTC
by gmalivuk
I never said the surface tension was doing the lifting, though.

Whatever surface tension is doing, that effect is part of the explanation for why water in a couple-mm-wide crazy straw wouldn't behave the same way as in Pfhorrest's enormous weird bucket.

Picture a 4mm straw with a single loop in it and a 40cm pipe with the same loop. Water slowly filling up the pipe is indeed going to find its own level (and pour bit by bit from the top of the loop to the bottom before eventually trapping an air bubble in that section), whereas water sucked through the straw stays in one contiguous mass.

Re: Multipath Crazy Straw

Posted: Wed Jan 03, 2018 11:13 am UTC
by Eebster the Great
In a crazy straw, you can suck water to a precise point where the edge is vertical, or even hanging in midair. That is definitely something that can't happen in a pipe. I thought you were saying that surface tension was somehow necessary for a straw to work.

Re: Multipath Crazy Straw

Posted: Wed Jan 03, 2018 12:52 pm UTC
by Xenomortis
Eebster the Great wrote:I thought you were saying that surface tension was somehow necessary for a straw to work.

Isn't it? If there were nothing "holding the liquid together", what's stopping air from breaking the liquid apart and just rushing along the straw?

Re: Multipath Crazy Straw

Posted: Wed Jan 03, 2018 3:59 pm UTC
by gmalivuk
You can suck water up a wide pipe just like a straw if you lower the pressure above it. It won't flow the same way (due to surface tension), but the level will still rise.

What's stopping air from rushing in is the fact that the bottom end is submerged. Obviously if you take it out of the reservoir then air does rush into a wide pipe, but not into the straw, which was precisely my point about surface tension in the first place. But the lifting still happens.

Re: Multipath Crazy Straw

Posted: Wed Jan 03, 2018 5:47 pm UTC
by Zamfir
For the curious, here's positive-displacement pump that lifts a water column up through a much bigger pipe than a straw, including a bit of a fancy bend at the end. Start video at 3:50. Regular centrifugal pumps can't do this, they can't generate pressure when they are pumping air.

The water-air interface in the downward section of the pipe looks different from the interface in a straw, that's due to the relatively weaker surface tension. While the overall lifting effect is rather similar.

https://youtu.be/2FA6t7wHtF0

Re: Multipath Crazy Straw

Posted: Wed Jan 03, 2018 7:20 pm UTC
by The Great Hippo
Zamfir wrote:For the curious, here's positive-displacement pump that lifts a water column up through a much bigger pipe than a straw, including a bit of a fancy bend at the end. Start video at 3:50. Regular centrifugal pumps can't do this, they can't generate pressure when they are pumping air.

The water-air interface in the downward section of the pipe looks different from the interface in a straw, that's due to the relatively weaker surface tension. While the overall lifting effect is rather similar.

https://youtu.be/2FA6t7wHtF0
That's pretty cool, and weirdly relevant where I work (particularly the bit about centrifugal pumps; we encounter 'cavitation' -- what happens when the pump starts sucking on gas -- pretty regularly).

For the longest time, I've wondered at the benefits of positive displacement pumps; I didn't realize they could pull a vacuum like that.

Re: Multipath Crazy Straw

Posted: Wed Jan 03, 2018 9:55 pm UTC
by morriswalters
The Great Hippo wrote:For the longest time, I've wondered at the benefits of positive displacement pumps; I didn't realize they could pull a vacuum like that.
They're used in, wait for it, vacuum pumps. Also most refrigeration equipment, and a lot of water pumps. (My apologies, I couldn't help myself.) And if you were raised on a farm in earlier times a hand pump, on a well, or a cistern. The fancy bend is a trap I believe, that keeps the pump primed. And last but not least and internal combustion engine is also a pump used to remove exhaust gasses. And I know almost nothing about crazy straws, but pipe losses at the wall are maybe a part of the effect. I hope that I didn't screw up too badly.

Re: Multipath Crazy Straw

Posted: Tue Jan 09, 2018 8:11 am UTC
by Zamfir
@jewish_scientist, I put the letter "g" straw in a fluid calculation program that tracks water level, with surface tension etc.

It's an intense calculation, I would need to spend a little time of myself speeding it up, and dedicating enough processor-hours to it. If you're interested, I ll do it.

Re: Multipath Crazy Straw

Posted: Wed Jan 10, 2018 3:39 am UTC
by jewish_scientist
If it is not too much of I bother, go ahead.

Re: Multipath Crazy Straw

Posted: Wed Jan 10, 2018 10:36 am UTC
by p1t1o
All I can think about when I see the title of this thread:

Image