1. Limited time only! Sign up for a free 30min personal tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Siphon in a vacuum?

  1. May 11, 2010 #1
    According to http://tinyurl.com/2cfghd5", an Australian physicist has discovered that The Oxford dictionary's definition of "siphon" is incorrect, attributing its effect to atmospheric pressure, rather than gravity, which is what he maintains causes it a siphon work.

    This seems to me to come down to whether a siphon would work in a vacuum (but in the presence of a reasonable gravitational field - on the surface of the Moon, say). It seems to me that it wouldn't. If you sucked some liquid past the bend in the siphon, it would surely fall down the siphon's longer arm, but it seems to me that the liquid in the shorter arm would just fall back into the main body of liquid, leaving a vacuum in the siphon as it did so.

    Does anybody agree or disagree?

    Thanks in advance,
    Steve = : ^ )
    Last edited by a moderator: Apr 25, 2017
  2. jcsd
  3. May 11, 2010 #2


    User Avatar
    Homework Helper

    How would you get a liquid to exist in a vacuum in the first place? Even if the surface pressure of such a liquid was zero, you'd have non-zero and increasing pressure with depth of the fluid, similar to the atomsphere's pressure being zero at the outer fringes of the atmosphere, which increasing presssure versus distance from the center of the earth.

    If such a liquid existed, a siphon that went above the surface of the upper tank would probably not work, since such a liquid would permit pressure to go to zero at the surface level of the upper tank.
  4. May 11, 2010 #3
    Thanks for the reply. If I had a tub of water in a room, and then evacuate the room of air, would gravity not keep the water in the tub, and surface tension not give it a surface? What about a tub of mercury?.

    Steve = : ^ )
  5. May 11, 2010 #4


    User Avatar

    Staff: Mentor

    You are correct, Steve: A siphon needs air pressure, otherwise it will lose its "prime". Water won't flow through, it will just fall out both sides because there is nothing pushing the water up toward the top of the tube.

    I read the article and his reasoning wasn't clear, so it is possible he's just splitting hairs about what gravity does in a siphon, since you do also need gravity to make the siphon work.
    Last edited: May 11, 2010
  6. May 11, 2010 #5


    User Avatar
    Homework Helper

    The low pressure would result in vaporization at the surface. This would continue until the weight of the vapor generated sufficient pressure and saturation at the remaining surface of the fluid (if any was left) to prevent further vaporization. In the case of most solids, the rate of vaporization would be negligible.

    In the case of the moon based siphon, you could use the weight of solid objects as a substitute for pressure from the air. In this case the tanks would be cylinders filled with water with piston shaped weights containing the water and maintaining pressure within the water.
    Last edited: May 11, 2010
  7. May 11, 2010 #6
    I agree with Hughes that since the work is supplied by gravity the dictionary definition is very misleading. In particular (since the air pressure acts on both reservoirs) the siphon does not normally depend on the amount of air pressure. *

    But I think the issue has to do with the notion that "there is no such force as suction" **. Indeed, in normal circumstances for a typical siphon, it is technically true that the water in the short arm rises because it is being pushed (ultimately by air pressure upon the top reservoir) rather than being pulled (by the absence of an equal force from the opposite direction). This is why it is difficult to say the dictionary is outright incorrect.

    ** Actually, there is such a force as suction. Tall trees in fact use negative pressure to suck water up to their leaves. The problem is that the siphon tube has to be carefully tuned for the liquid to remain liquid under negative pressures. Normally, any liquid would start to boil and form an embolism, disconnecting two regions of liquid and interrupting the transfer of force (and the transport of liquid).

    * So, if a siphon tube is very carefully made, it can in fact siphon a liquid even with literally zero atmospheric pressure (by means of the gravitational force in the long arm producing negative pressure in the bend as it exerts an upward pull on the liquid in the long arm). Very similar to a "chain siphon". But you're right in that a typical siphon tube (at least with a typical liquid) will not work in a vacuum, exactly as you described (an embolism would form at the peak of the siphon tube). Now, actually a typical liquid (like water or mercury) cannot be kept in a vacuum in the first place (it boils until a certain amount of vapour pressure is restored), but nonetheless we can still say that a typical siphon fails every time the peak height of the tube (above the surface of the top reservoir) is too large for the atmospheric pressure (after accounting for the strength of gravity and density of the liquid), so indeed most practical siphons would fail if the atmospheric pressure becomes sufficiently low.

    So you're right about typical siphons, but wrong about some specialised siphons. And still, the dictionary has chosen a very poor way of describing what a siphon is.
    Last edited by a moderator: Apr 25, 2017
  8. May 11, 2010 #7
    Thanks for the extremely educational response. I'm now grappling with the concent of "negative pressure".

    Steve :confused:
  9. May 12, 2010 #8


    User Avatar
    Homework Helper

    I assume he meant a negative pressure gradient, where pressure at the base of a tree is higher than pressure at the top of a tree. There are a huge number of very small tubes (xylem) that transport fluid upwards via capillary (surface tension) and chemical actions (osmosis) that draws in fluid through the roots, and evaporation through the leaves.
  10. May 12, 2010 #9
    No. :smile:

    Not gauge pressure either. I mean the sort that relies on the existence of attractive forces between real molecules (unlike ideal gas particles).
    For trees, the pressure generally comes from surface tension in the leaves (the air-water meniscus in the stoma or wherever, curved by evaporation). The issue arises after tree height exceeds the 10mH2O of normal atmospheric pressure.
    Last edited: May 12, 2010
  11. May 12, 2010 #10
    There are actually two aspects to this discussion. Firstly, the force that drives the transfer of liquid from the short leg of the pipe to the long leg of the pipe and secondly the force that prevents the short leg of the pipe from emptying backwards, and forming a vacuum at the top of the pipe. They are separate.

    To demonstrate that hydrostatic pressure alone drives the transfer, consider this experiment:

    Simply fill a hose with water and insert each end into one of two level containers of water. Nothing will happen. The water will remain in the hose and will not flow in either direction.

    Then lower one container and water will start to move from the higher container to the lower one. The only change is in hydrostatic pressure - atmospheric pressure is unchanged at both ends (in fact, to be pedantic, it is sightly now slightly higher at the lower end!). Therefore it is hydrostatic pressure that is driving the transfer, not atmospheric pressure.

    What atmospheric pressure does do is prevent the short leg of the pipe from emptying, at least up to a height where the hydrostatic pressure in the short leg of the pipe is equal to 1 atmosphere (about 10m or 33 feet), if the short leg is higher than 10m then a vacuum will start to form at the top of the pipe and the transfer of liquid will cease. So, therefore a siphon couldn't work in a vacuum because a second vacuum would form at the top of the pipe ceasing the flow.
  12. May 13, 2010 #11
    How could we explain the frictionless fountain of liquid helium that flows as long as the temperature is cold enough , in terms of siphoning ,
  13. May 13, 2010 #12
    But wouldn't a siphon still work where the effect of gravity was nil but where there is air pressure, for instance the space station? Suppose you had two plastic bags, one filled with liquid and one empty, connected with a tube. If you started the liquid flowing from the full bag to the empty one, wouldn't the liquid continue moving by its own inertia? Isn't it really the inertia of the liquid that makes any siphon work?
  14. May 13, 2010 #13
    What about energy loses (friction/turbulence)?
  15. May 13, 2010 #14
    Exactly. Siphons do eventually slow to a stop.
  16. May 13, 2010 #15
    Uh, no they don't? Not unless the reservoir surface levels equalise, or one empties.
  17. May 13, 2010 #16
    My apologies, I was speaking from personal experience rather from established theory.
  18. May 13, 2010 #17
    Nothing wrong with that, evidence trumps dogma. But I'm not quite sure what experience you're referring to? Maybe what you're accustomed to is the siphon slowing as the difference between reservoir levels falls (while the resistance to current through the hose stays the same)? Maybe you could settle this if you tried pinching off the flow in a running siphon, since there would be no inertia to restart the flow when you remove the kink. (In my experience, one prefers to start a siphon not by sucking on the bottom of the tube but by statically pre-filling the tube, with thumb sealing the lower end of the tube.) Or use two buckets (add some kind of particulates in the tube so you can see the flow) and try to find out how long inertia lasts if you eliminate the height differential?
    Last edited: May 13, 2010
  19. May 13, 2010 #18
    Liquids do have some tensile strength, which allows them to pull some negative pressure. As long as the column of liquid in the pipe doesn't break, the siphon can run, just like a rope or chain can pull itself over a pulley.

    Also, note that evaporation is highly dependent on surface thermodynamics. Depending on the liquid and pipe materials (which will affect the conditions at the interface between pipe and liquid), the liquid in the pipe might effectively have a much lower, even negative vapor pressure, as long as no vapor or nucleation sites are present. Keep the siphon very clean and it'll work better.
  20. May 15, 2010 #19
    if you want to siphon the fluid out of the bath tub into the vacuumed room, it’s not a problem. But if you try to siphon the fluid out of the room it’s impossible because there is no space to fill the void. The vacuum is proportionate to the volume of the room. if you take something out you increase the volume. if you increase the volume whilst in vacuum the room would implode.
  21. May 15, 2010 #20
    That's nonsense. Vacuum is vacuum, space empty of matter, it has no pressure, tension, or preferred volume. There is no force on the walls of an evacuated vessel in vacuum, no matter what changes you make to the volume of that vessel. The volume of vacuum on either side of the siphon only becomes relevant when the fluid being siphoned fills that entire volume.
  22. May 15, 2010 #21
    Sorry , decrease the volume of the room whilst in vacuum . But I am pretty sure that the walls of a vacuumed room would be affected by the fluid exerting from the vacuumed room considering you had the power to pump it out. It would be limited by the room’s elasticity.
  23. May 15, 2010 #22
    I’m no expert, but fluid dynamics is not affected by air pressure, only gravity. Vacuuming the room of air will not affect fluid dynamics in a siphon. No air would suggest an inability to move the fluid out of a vacuum without compromising the vacuum state in capsule. But the dictionary definition stats to convey liquid through a tube with suction or ((immersion)), how you immerse that tube is that of many ways. Reverse airlock is one way, fill a portion of the room in liquid and seeling it off from the vacuum and releasing it. To continue close the portion and re-vacuum the area before exposing it for submergence again in the vacuum. The dictionary states immersion this implies you can use a reverse airlock.
    Last edited: May 16, 2010
  24. May 16, 2010 #23
    Ok I just read the article at the start of the question. I should read things closer before running my mouth off. I’m probably wrong because I don’t have the stature of a senior physicist. But the definition states convey meaning transfer in a tube, which a tube is” a long, hollow cylinder used for conveying or holding liquids or gases>a long flexible metal or plastic sealed at one end and a cap at the other end “. The method of conveying liquid can be achieved with reverse airlock. Submerging the cup like tube in liquid and conveying it through the reverse airlock. The note stated in the dictionary under siphon being “maintained by different fluid pressures at the tube openings” would suggest if I fill a cup with water, the bottom of the cup is maintaining the fluid pressure equal to rXpi^2Xhight of the cup. So where does this definition suggest water is affected by air pressure? The dictionary states “or immersion”. correct me please!!
  25. May 16, 2010 #24

    Increase or decrease, it makes no difference. Vacuum is vacuum, there's nothing there to push or pull on the walls of the vessel. Obscure things like the Casimir effect aside, vacuum exerts no force, positive or negative, no matter what the volume.

    As I mentioned, liquids have tensile strength which allows them to exert negative pressure and "pull" on surfaces in contact with them that they adhere to. The forces involved are relatively weak, though, and present in atmosphere as well. It is quite common to drain liquids from containers without the containers imploding, there's no reason to think this will change in vacuum.

    The first statement is just wrong. Or "not even wrong". Fluid dynamics is the dynamics of fluids, and air is certainly a fluid...air pressure is of great importance in a wide variety of situations in fluid dynamics. Including siphons. They can work in vacuum with liquids that don't boil, but air pressure affects how high the siphon tube can rise above the higher liquid level without breaking the fluid column.

    Just how does removing a liquid from a container that otherwise only contains vacuum "compromise" that vacuum? What, there's not enough nothing to replace the liquid?

    It's not at all clear what you're proposing or why you think it's necessary.
  26. May 16, 2010 #25
    Fluid pressure at the bottom of the cup is proportional to the depth of the liquid and on air pressure, it does not depend on the area. The article was correct in saying that the siphon was maintained by the pressure differential across the tube: if the liquid column in the tube does not break, the lower leg will have a higher pressure from the liquid inside the tube, and the liquid will exit that side as long as the pressure outside is lower.

    This can work in vacuum because liquids have tensile strength. They have very finite tensile strength, however, and the liquid column will break if the middle of the siphon goes too far above the two containers...the liquid will fall down both legs of the siphon, leaving a vacuum in the tube. External air pressure will support part of the weight of the fluid column, allowing the middle of the siphon to rise higher before the column breaks.

    I have no idea where you're going with "reverse airlocks".
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook