Siphons

[zipmo]

I'm somewhat new around here, so please pardon me if this has already been hashed.

I was going through the archives and ran across the Cecil post on siphons. The conclusion seemed to be "well it could be air pressure or it could be 'cohesiveness'".

I recommend two thought experiments, neither of which require a trip to the moon.

The first, verifying that air pressure is a factor: Imagine you're siphoning out of a jar. Now, while you're siphoning, put a tight lid on the jar that allows no more air in (and has a tight seal around the hose. You can easily imagine that once a vacuum forms in the jar, the water will stop flowing. This can be verified experimentally as well. Think glugging milk jugs and "stuck" ketchup.

The second, "cohesiveness" of the water is also a factor. Imagine our jar is now a reservoir of water behind a dam. Our hose is also now much larger. While almost any size hose can be imagined to work while the water is flowing, once we stop the flow, larger hoses will self-drain. After draining we won't be able to re-start the flow without pumping enough water to fill the pipe. There is probably some maximum pipe/hose diameter at which water simply will not siphon but will instead self-drain. There may be a physics/manufacturing engineer available who knows how to calculate this size for water.

So it looks like both are required factors for siphoning.

jas

[C K Dexter Haven]

Welcome to the Straight Dope Message Boards, zipmo, we're glad to have you here. For future ref, when you start a thread, it's helpful to other readers to provide a link to the column in question. Keeps us on the same page, and saves searching time, and (with luck) avoids people re-stating what's already in the column. In this case, I assume, it's: http://www.straightdope.com/columns/...-a-siphon-work

No biggie, you'll know for next time, and (as I say) welcome!

PS - There have also been some prior discussion threads (all several years old) here:
How does a siphon work? from 2007
Question about siphoning from 2006

[zipmo]

Thanks Dexter.

I did try to search before posting/joining. On my end, I could find no "search" functionality before joining. I did try Google on the site as well. I guess I didn't notice that the "search" option popped up after I joined.

Thanks for the tip. I'll definitely put in a link to the original article next time.

BTW, the second link you posted "Question about siphoning from 2006" doesn't appear to work. The first one does, however.

Thanks again,

zipmo

[C K Dexter Haven]

(Oops, sorry -- I've fixed the link.)

... and hoping that some others will come in to comment.

[Irishman]

Originally Posted by zipmo

The first, verifying that air pressure is a factor: Imagine you're siphoning out of a jar. Now, while you're siphoning, put a tight lid on the jar that allows no more air in (and has a tight seal around the hose. You can easily imagine that once a vacuum forms in the jar, the water will stop flowing. This can be verified experimentally as well. Think glugging milk jugs and "stuck" ketchup.

I think that if the tank is open topped, air pressure plays a negligible role in the process. The siphon is driven by the gravity difference. However, if you seal the tank, pressure becomes a confounding issue, because as the water level drops, the air pressure in the tank falls. As this pressure falls, it begins to create a vacuum inside the tank. But what is a vacuum? A lower pressure than outside the tank. Thus the gravity difference is trying to drive water one direction but pressure difference is trying to drive the water the other direction. That is why the siphoning stops.

The second, "cohesiveness" of the water is also a factor. Imagine our jar is now a reservoir of water behind a dam. Our hose is also now much larger. While almost any size hose can be imagined to work while the water is flowing, once we stop the flow, larger hoses will self-drain. After draining we won't be able to re-start the flow without pumping enough water to fill the pipe. There is probably some maximum pipe/hose diameter at which water simply will not siphon but will instead self-drain. There may be a physics/manufacturing engineer available who knows how to calculate this size for water.

Let me see if I understand. You are suggesting that if the diameter of the hose is sufficiently large, the water will fall out of the column instead of siphoning?

Because if I wish to stop siphoning with a hose, the typical method is to raise the outer end above the inner one. This reverses the flow into the tank. Once the flow has drained, siphoning will not restart. So you are correct that once the hose is empty, it will not restart, regardless of the size of the hose. So I am unclear on the question.

Are you speculating that you have a pipe system, with a valve at the outter end. Close the valve shuts off the flow, but leaves a column of water in the pipe. But if the pipe diameter is sufficiently large, the inlet leg up to the top height will self drain rather than remain primed?

[zipmo]

Irishman,
I agree with you about the air pressure bit. I had given it more thought and came to the same conclusion: the lid idea simply creates a vacuum that impedes the flow and as such doesn't really give any useful insight into the reasons siphons work. Further thought also revealed that the air pressure difference between the two ends of the hose/pipe are usually negligible.

That would seem to leave us with "cohesiveness" (there's *got* to be a more accurate/specific term for this. The closest term I can recall is "surface tension", which is surely related but probably not the "best" one.)

I do recall from my childhood lawn mowing days that the siphoning action could continue despite small bubbles in the line. The bubble could be large enough to create a "void" in the line (overcoming the cohesiveness forces) but flow could still continue so long as the bubbles weren't too big.

This leads me to draw a summary so far as:
- Flow begins when the weight of the water at the drain end starts the flow
- The "cohesiveness" of the fluid pulls in more fluid along the length of the line. In the case where bubbles large enough to overcome cohesiveness forces are introduced, in some cases the flow can lower the pressure of the bubble sufficiently to allow the flow to continue. (Does this require air pressure outside the pipe? Argh!)


As to my "huge pipe" conjecture: Not the inlet leg, but the outlet leg. The flow must be sufficient to overcome the water's propensity to form a bubble at the outlet end and simply drain out of the pipe. If the valve (located anywhere along the pipe) is throttled, at some flow rate the water past the valve will allow air to flow back in and water will simply drain out rather than maintain the flow necessary to keep the siphon going.

Again, from my lawn mowing days, I remember that while larger diameter hoses would siphon gas faster, they were harder to start and harder to keep going. I see this as an example of conditions (large diameter, slow flow) overcoming the cohesiveness of the fluid.

So I would say there are the main factor would be the "cohesiveness" of the fluid. The diameter of the hose/pipe as well as flow rates are also factors in the effectiveness of the siphoning.

I do get the feeling that I'm trying to "reinvent the wheel" of fluid dynamics. Surely there are fluid mech-E's/phycisists who can put this whole issue to bed quickly and much more elegantly?

zipmo

[Chronos]

I think that if the tank is open topped, air pressure plays a negligible role in the process. The siphon is driven by the gravity difference.

There are two different issues, here. The outlet must, in all circumstances, be lower than the inlet, or water will not flow. This is the gravity difference. What air pressure is important for is how high you can raise the water in between. If the top of the hose is more than about 10 meters above the inlet, then water will just fall down both sides of the hose, and not flow from one to the other, because a column of water 10 meters high generates the same pressure as the atmosphere.

[Dr. Strangelove]

Cohesion isn't necessary for a siphon to work. Air pressure is enough to keep the fluid from forming bubbles of vacuum that disrupt the flow (though the flow itself is driven by gravity, since air pressure presses the same on all sides).

However, you can actually operate a siphon in a vacuum, and in that case cohesion does play a part:
http://www.youtube.com/watch?v=8F4i9M3y0ew

[zipmo]

Originally Posted by Chronos

There are two different issues, here. The outlet must, in all circumstances, be lower than the inlet, or water will not flow. This is the gravity difference. What air pressure is important for is how high you can raise the water in between. If the top of the hose is more than about 10 meters above the inlet, then water will just fall down both sides of the hose, and not flow from one to the other, because a column of water 10 meters high generates the same pressure as the atmosphere.

Ah, good point. That more clearly explains the barometer reference from the original article.

[zipmo]

Originally Posted by Dr. Strangelove

Cohesion isn't necessary for a siphon to work. Air pressure is enough to keep the fluid from forming bubbles of vacuum that disrupt the flow (though the flow itself is driven by gravity, since air pressure presses the same on all sides).

However, you can actually operate a siphon in a vacuum, and in that case cohesion does play a part:
http://www.youtube.com/watch?v=8F4i9M3y0ew

Excellent Dr. Strangelove! Thank you! So that video shows that atmospheric pressure's only role on a siphon is to prevent "cavitation" in the liquid column.

Zipmo

[Dr. Strangelove]

Originally Posted by zipmo

Excellent Dr. Strangelove! Thank you! So that video shows that atmospheric pressure's only role on a siphon is to prevent "cavitation" in the liquid column.

Zipmo

I suppose that's a reasonably accurate way of putting it. You couldn't siphon a bucket of ball bearings because they would just fall apart--unless they were either magnetic (as with cohesion), or there was some membrane on the ends that pushed the balls together via pressure (as with air pressure).

It's just a hunch, but I'll bet that the cohesive forces are far weaker than atmospheric pressure--you wouldn't get anywhere close to a 10 meter column.

[Chronos]

It's just a hunch, but I'll bet that the cohesive forces are far weaker than atmospheric pressure--you wouldn't get anywhere close to a 10 meter column.

And my hunch would be that that's correct in almost all circumstances, but that the height you can support via cohesion would depend on the properties of the tube, and you might be able to get a significant height using a very thin capillary.

[zipmo]

Originally Posted by Chronos

...but that the height you can support via cohesion would depend on the properties of the tube, and you might be able to get a significant height using a very thin capillary.

My understanding of barometers is that the required height of the liquid column is a function of the liquid used. E.g. A water-based barometer requires a much taller column than one of mercury. Based on our conversation about siphons, it would seem that the air gap formed at the top of the barometer is formed when the liquid 'cavitates'. The size of the cavity being a function of the surrounding air pressure.

Chronos, I would agree that very thin tubes might allow a slightly taller column before such cavitation occurs, but would that additional height be the result of the interaction between the liquid and the material of the tube, i.e. the force that causes a miniscus?

[newme]

I did a little experiment. I filled a tube with water and put the low end in a glass of water. I clamped the other end shut and put it in a glass of oil that was higher than the glass of water. The highest point of the tube was about half a meter above the oil.

Oil and water have no cohesion. In fact, they have "negative" cohesion; oil is hydrophobic and repels water.

When I released the clamped, the siphon worked just fine. The water flowed out of the tube and oil flowed up and in then down and out. I will let you draw your own conclusion as to how much role cohesion plays in siphoning water or oil.

[zipmo]

Originally Posted by newme

I did a little experiment. I filled a tube with water and put the low end in a glass of water. I clamped the other end shut and put it in a glass of oil that was higher than the glass of water. The highest point of the tube was about half a meter above the oil.

Oil and water have no cohesion. In fact, they have "negative" cohesion; oil is hydrophobic and repels water.

When I released the clamped, the siphon worked just fine. The water flowed out of the tube and oil flowed up and in then down and out. I will let you draw your own conclusion as to how much role cohesion plays in siphoning water or oil.

Excellent idea! So this brings us back to the "vacuum" experiment from the video. If your experiment had been done using two similar liquids in a vacuum (assuming that both liquids do not immediately boil off) would they simply separate within the tube and each flow back into their respective containers?