What-If 0147: "Niagara Straw"

[ps.02]

What would happen if one tried to funnel Niagara Falls through a straw?

I enjoyed this one. An obvious approach would be to freeze the water into slugs and fire them, to avoid the fluid dynamics problems? (Which is to say, the exact opposite of sending a barrel down the Falls.)

Ultimately I guess that's not going to make the problem any easier. Nor would using a larger straw, like the 12 to 15mm ones used with slushy drinks.

[CanadianNomad]

Nor would using a larger straw, like the 12 to 15mm ones used with slushy drinks.

Well if we remember the context of the question (don't think about this part too hard), then I think we can at least get up to a 20-30mm straw diameter....

[ijuin]

I find it highly amusing that it is White Beret Guy who is trying to go over the "straw falls" in a barrel.

[benmoreassynt]

Sometime in the next few years, the falls may be turned off for maintanence[sic]. And probably to see what cool stuff they can find.

As a Canadian I cannot restrain myself from pompously noting that it's only the American Falls which will be switched off, not the whole of Niagara Falls. As all Canadians know, and are forced to swear on oath before they are allowed out unaccompanied by a responsible adult, the American Falls don't really count.

[rhetoricdevine]

The barrel over the falls quip reminded me of the first xkcd comic. His barrel has had quite the journey!

[cellocgw]

ok, so a standard straw requires V = c/4 . Randall posited this straw has 38.5 mm^2 cross-sectional area. We want to reduce the speed from 73 600 000 m/s to the speed of sound in water, which is 1482 m/s.

That ratio is approximately 49700:1,
so now we need a straw whose diameter is sqrt(49700) * 7 mm , or 1.56 m if I got all my decimal points correct. That shouldn't be too hard to build. We will need to build a funnel the width of the Niagra Falls and which constricts slowly enough not to cause turbulence, but heck, that's just a fun project for some AutoCAD jockey.

[cellocgw]

Not up yet on the forum,
but it would appear this what-if spawned today's regular xkcd.

[Zinho]

GREAT. YOU GOT THE HIGGS BOSON WET AND IT DISSOLVED.

Does the Higgs boson get wet, or does the water get it instead?

Props to TMBG + Particle Man.

[Tyndmyr]

What would happen if one tried to funnel Niagara Falls through a straw?

I enjoyed this one. An obvious approach would be to freeze the water into slugs and fire them, to avoid the fluid dynamics problems? (Which is to say, the exact opposite of sending a barrel down the Falls.)

Ultimately I guess that's not going to make the problem any easier. Nor would using a larger straw, like the 12 to 15mm ones used with slushy drinks.

Unfortunately, you are correct. Water is not strong enough to stay together even at normal rifle speeds, and we need rather a lot more. I considered larger straws pretty rapidly, but...I believe we're still limited to the size of something that'll fit in a human mouth. Anything larger than maybe a garden hose is probably stretching the limit, and that doesn't get us down to anything reasonable, I'm afraid.

Maybe we could build humans with bigger mouths, first. That should work.

[HES]

I believe we're still limited to the size of something that'll fit in a human mouth.

Why? What if you're trying to feed a hippo? Is that not still a straw?

[rmsgrey]

I believe we're still limited to the size of something that'll fit in a human mouth.

Why? What if you're trying to feed a hippo? Is that not still a straw?

Wikipedia tells me that straw is an agricultural by-product of cereal production. I'm not sure where I'm going with this, but I'm sure someone can come up with something...

[HES]

Funnelling filtering Niagra falls through a straw doesn't sound that difficult. You'd need a decent mesh to keep the dam in place, but beyond that a few truckloads of by-product would be all you need.

[RCT Bob]

ok, so a standard straw requires V = c/4 . Randall posited this straw has 38.5 mm^2 cross-sectional area. We want to reduce the speed from 73 600 000 m/s to the speed of sound in water, which is 1482 m/s.

That ratio is approximately 49700:1,
so now we need a straw whose diameter is sqrt(49700) * 7 mm , or 1.56 m if I got all my decimal points correct. That shouldn't be too hard to build. We will need to build a funnel the width of the Niagra Falls and which constricts slowly enough not to cause turbulence, but heck, that's just a fun project for some AutoCAD jockey.

I don't think you can prevent turbulence just by decreasing the constricting rate, unless the walls of your straw are ridiculously slippery and smooth. Fluids flowing along walls will pretty much always turn turbulent in their boundary layers after a certain distance Reynolds number is achieved (density times velocity times distance travelled divided by dynamic viscosity), which depends on the roughness and on turbulence levels among other things. If you only very slowly constrict the water you will undoubtedly reach turbulence in the boundary layer simply because of distance travelled. I'm mostly familiar with boundary layers of gases, but liquids are fairly similar at least at subsonic speeds.

Another interesting thing about gas flows through pipes and straws is the Fanno-flow principle, which Randall didn't really touch upon, I assume because straws are generally too short to really make it matter. But an adiabatic (no energy transfer to/from the walls) flow through pipes with constant diameter that has skin friction always tends to go to sonic velocities, supersonic flows slow down by the drag, while subsonic flows speed up (talking about mean velocity over the cross-section). I think it's actually really cool that drag can actually speed up flows even though they apply a force in the opposite direction. If the walls aren't adiabatic, the Rayleigh flow principle gets added, which drives the flow towards sonic in case of heat addition (hot walls, cold gas) and away from sonic flow in case of heat removal (cold walls, hot gas).

[Eternal Density]

https://www.google.com/search?q=dam+cavitation
That doesn't look good!

[Neil_Boekend]

ok, so a standard straw requires V = c/4 . Randall posited this straw has 38.5 mm^2 cross-sectional area. We want to reduce the speed from 73 600 000 m/s to the speed of sound in water, which is 1482 m/s.

That ratio is approximately 49700:1,
so now we need a straw whose diameter is sqrt(49700) * 7 mm , or 1.56 m if I got all my decimal points correct. That shouldn't be too hard to build. We will need to build a funnel the width of the Niagra Falls and which constricts slowly enough not to cause turbulence, but heck, that's just a fun project for some AutoCAD jockey.

I don't think you can prevent turbulence just by decreasing the constricting rate, unless the walls of your straw are ridiculously slippery and smooth. Fluids flowing along walls will pretty much always turn turbulent in their boundary layers after a certain distance Reynolds number is achieved (density times velocity times distance travelled divided by dynamic viscosity), which depends on the roughness and on turbulence levels among other things. If you only very slowly constrict the water you will undoubtedly reach turbulence in the boundary layer simply because of distance travelled. I'm mostly familiar with boundary layers of gases, but liquids are fairly similar at least at subsonic speeds.

Another interesting thing about gas flows through pipes and straws is the Fanno-flow principle, which Randall didn't really touch upon, I assume because straws are generally too short to really make it matter. But an adiabatic (no energy transfer to/from the walls) flow through pipes with constant diameter that has skin friction always tends to go to sonic velocities, supersonic flows slow down by the drag, while subsonic flows speed up (talking about mean velocity over the cross-section). I think it's actually really cool that drag can actually speed up flows even though they apply a force in the opposite direction. If the walls aren't adiabatic, the Rayleigh flow principle gets added, which drives the flow towards sonic in case of heat addition (hot walls, cold gas) and away from sonic flow in case of heat removal (cold walls, hot gas).

There may be some possible gains by using additions to decrease the viscosity, increasing the Reynolds number.

[Soupspoon]

There may be some possible gains by using additions to decrease the viscosity, increasing the Reynolds number.

I was thinking of coating the inside of the pipe with a suitably hydrophobic coating. (Problems with that solution, I shall leave as an exercise to the disinterested reader. Because all interested readers clearly do not have enough spare time...)

[RCT Bob]

Higher Reynolds number flows actually turn turbulent quicker than lower Reynolds numbers. Sufficient viscosity damps out fluctuations, keeping the flow stable and laminar. On high Reynolds numbers inertia of the fluid is much more influential than viscous forces, and instabilities get amplified, rather than damped out. High Reynolds number flows are inherently unstable, which is part of the reason that turbulence is so difficult to analyse.

[Neil_Boekend]

So we should add something like cellulose or gelatin to our flow?

[Soupspoon]

So we should add something like cellulose or gelatin to our flow?

I see absolutely no problems with this plan...

[danbert8]

The whole analysis of drag and flow profile is meaningless in this discussion. Assuming the flow of Niagara Falls could go through a sub 2 meter pipe at near the speed of sound, just putting a funnel in front of the river wouldn't make that happen. You'd need to pump it to get it to flow fast enough. Good luck with that pump that can accelerate that volume of water to near the speed of sound...

If you neglect the option for a pump, then you are restricted to gravity flow, not pressurized flow and you're going to be looking at the equivalent of how large of a hole you'd need with the depth of the Niagara river to get the same flowrate. Basically, a bell-mouth or "morning glory" spillway. It's going to be pretty darn large, well larger than 2 meters, maybe on the order of 20 meters or more.

[LtPowers]

I think we're missing something here. The question involves "funneling" the Falls through a straw, not "forcing" it. Randall answered the "forcing" question in his usual exhaustive detail, but the original question involved funneling. So it seems to me the answer is that the river would back up behind the funnel and eventually flood the surrounding land.

As a Canadian I cannot restrain myself from pompously noting that it's only the American Falls which will be switched off, not the whole of Niagara Falls. As all Canadians know, and are forced to swear on oath before they are allowed out unaccompanied by a responsible adult, the American Falls don't really count.

Don't forget the Horseshoe Falls are partially in the U.S.!


Powers &8^]

[SDK]

I think we're missing something here. The question involves "funneling" the Falls through a straw, not "forcing" it. Randall answered the "forcing" question in his usual exhaustive detail, but the original question involved funneling. So it seems to me the answer is that the river would back up behind the funnel and eventually flood the surrounding land.

At the risk of stating the obvious, the real answer is, "Niagara Falls wouldn't fit through a straw."

[danbert8]

I guess the discussion we are digressing into is how big of a straw would you need to handle the flow over Niagara Falls? Based on the actual falls, I would say the upper bound of the straw (one big enough that we know it would work) would be a straw with the same perimeter as the length of the existing falls... The existing falls have an edge that's somewhere around 3000' long. Let's round it up to 3142' because I am lazy and want simple math giving us a straw that needs to be no larger than 1000' in diameter. I'll let someone who knows how to design emergency spillways of the bell-mouth type firm up the size estimate...

Actually, being less lazy, some engineers probably already figured this out... The new Niagara Tunnel Project is a 42' diameter tunnel that has a maximum flow capacity of 500 cubic meters per second (collaborations between the US and Canada are a pain for units). The average flow over the falls is around 6000 cubic meters per second, so you'd need 12 of these 42' diameter "straws". There is less resistance to flow in one big pipe vs multiple small pipes, but I've reduced our upper estimate of straw size from 1000' in diameter to one with the same area as 12 42' pipes. A pipe with 145' diameter has the same area as 12 42' pipes, so I'd say the maximum straw diameter is probably somewhere between 120-140' in diameter.

[hbarsquared]

I would like to point out that a "coffee straw" (as described in the original Amazon review article) is distinctly smaller than a normal straw. A coffee straw is NOT 7mm in diameter, closer to 2-3.

How does this effect the end results? I demand a rewrite.

[Soupspoon]

I would like to point out that a "coffee straw" (as described in the original Amazon review article) is distinctly smaller than a normal straw. A coffee straw is NOT 7mm in diameter, closer to 2-3.

How does this effect the end results? I demand a rewrite.

Halving the diameter (which it would, and slightly more) would quarter the cross-section, thus quadruple the flow-rate, which (ignoring boundary-effects, in both cases) would take the velocity from the "1 quarter" to "4 quarters" of c bit in that table...

[Copper Bezel]

I would like to point out that a "coffee straw" (as described in the original Amazon review article) is distinctly smaller than a normal straw. A coffee straw is NOT 7mm in diameter, closer to 2-3.

How does this effect the beginning results? I demand a rewrite.

I have no idea how much of this message made sense before the begending. = .

[mathmannix]

Is anyone else disappointed that the item about which the linked review was written is apparently no longer available, or at least no longer explicitly mentioned? I would buy it... for science, yeah, that's it!