Let’s change the analogy. Take a cylinder with a sliding membrane (i.e. a piston). Fill one side with fluid, the other with air, connect through a filled pipe to a second piston. That is the analogue of two containers of fluid connected by a “hose” where the air pressure on the other sides of the pistons are equal. Now increase air pressure on one side, and it will push the piston to balance the air pressure in both cylinders.
But that’s not really a “siphon”, because there is no analogue to the fighting of gravity of going over the hump. That is just simple pressure distribution through an odd-shaped volume.
How can air pressure drive the siphon when the air pressure is the same on both ends; the top of the water in the container and the end of the tube?
Air pressure doesn’t drive the siphon (gravity does), but it will limit its height. That’s where cavitation comes in. The pressure in the tube cannot be negative. Liquid density x height of the column x gravity can’t be greater than the air pressure.
They are two different questions and therefore two different answers.
A most bogus answer from Cecil on this one. The only role air pressure has in his answer is to keep the water from boiling. The controlling factor is head differential (the difference in height between the water level in the bucket and water level, if any, at the discharge end of the siphon).
Here’s a fun and easy experiment anyone can do. Take a bucket full of water and sit it on a table or stool or something. Take a tube and start a siphon. Now raise and lower the end of the tube and observe the rate of flow of the water. The lower the end of the tube relative to the surface of the water in the bucket the faster the water will flow. Use a measuring cup and a timer to measure rate of flow vs. difference in height. You will find that rate of flow is proportional to head differential.
Here’s another. Take one full and one empty bucket. Sit them at the same level and start a siphon from one to the other. Using a ruler and a stopwatch record the rate of change in level. As the water in the two buckets approaches the same level the rate of flow will slow. Again, that rate will be proportional to the difference in water levels.
For laffs calculate the difference in air pressure between the surface of the water in the bucket and the end of the tube (varies about 0.0005 PSI per foot). Pretty close to zero, ain’t it? Now calculate the difference in water pressure at the same differences in height (about 0.43 PSI per foot difference). You will find that difference matches up pretty neatly with observed water flow.
admitedly it is with a non volatile organic solvent so that it doesn’t boil immediately like water would.
and there’s the key point it would boil. under most conditions if the atmospheric pressure minus the applied tension /surface area of the water column is less than the vapor pressure of water the water boils. at 90 degrees c you could only make a water column 1/3 of the height most of the time so yes, most of the time for a volatile substnance a siphon is in fact limited by atmospheric pressure .
Another point is that under conditions where boiling is impossible, you can in fact have a water siphon longer than 10m, trees are an example: the water is not pushed to the top by root pressure,or by capillary action but is pulled by the action of transpiration at pressures down to -15bar, the xylem vessels have always been sealed and are free from imperfections and bubbles, which would cause the water column to boil and break up so the column stays solid at pressures below atmospheric.
I encourage everyone to try the simple experiment I proposed above, and then the other one described below. All the talk about boiling (or vapor pressure of the fluid being siphoned, more precisely) is a distraction from why siphons work. Siphons work by differential head, period.
The “weight of the water column” theory is easily debunked. Take two pieces of tubing with different diameters (in my case I used 3’ of vinyl tubing with an I.D. of 3/4" and 6’ of tubing with an I.D. of 3/16"). Connect the two pieces end to end. I used duct tape and a smear of silicon.
Now take the tubing and drape it over some support so that the larger tubing is on one side and the smaller on the other (camera tripod and clothes pin, for example). Submerge the end of the larger tubing in a bucket of water and establish a siphon. Even though the weight of the column of water in the larger tubing is (in my case) 16 times the weight of a column of equal length in the smaller tubing, the siphon will flow whenever the end of the smaller tubing is below the level of the water in the bucket. The rate of flow will be proportional to the difference in height between the end of the smaller tubing and the water level in the bucket.
Have you read the paper linked below? Experiments 2 and 3 make a very strong case for the role of air pressure in a siphon. The conclusion is that gravity is necessary to pull the water down through the longer end of the tube, but air pressure is responsible for pulling it up through the shorter end of the tube.
Similar to nolaman’s experiment, I did one last night. Using two glasses and ~6 ft of rubber surgical tubing, I made a siphon. For my siphon, I made a coil (helix, whatever) on the left side so there was a bit over 3 times the length on the left than the right. With the left coil above the level of the glass on the right, I lowered the right glass, and water flowed into the right glass. It was slow, because my tubing had a long tube with a narrow passage, but it flowed. I repeated several times and could drive the water either direction simply by which container had the higher surface level.
In short, the “weight of the water in the longer tube” explanation is bogus. The weight of the water is irrelevant.
The driver is head, i.e. the pressure in the fluid caused by height. That is a gravity effect. From the Bernoulli equation
P + pgh + (1/2)pv[sup]2[/sup] = const
where
P = fluid pressure
p = density of fluid
g = acceleration of gravity
h = elevation of fluid vs a reference point
v = speed of fluid motion
pgh is called the head term and is internal fluid pressure due to altitude, i.e. a gravity driven term. It manifests in the Bernoulli equation as fluid pressure. Similarly, the velocity (i.e. speed) of the fluid manifests as a pressure.
A siphon works by having a closed passage between two points. The closed passage is to prevent interruption of the flow. By putting a hole in the top of the tube, it allowed air in to separate the molecules, and the flow split. But if the passage is closed, there is no break to the liquid flow.
An air bubble within the tube does not break the flow, it works as part of the flow. It is a trapped bubble, as long as there are no leaks on the passage.
It does not matter the shape of the passage, the tube can go below both end points or above both end points or wander back and forth. It can be long over the inlet and short over the outlet. The fluid in the tube is not a driving factor for the flow.
Flow is created by the altitude difference between the inlet and outlet, i.e. the head.
Did you see siphon in a vacuum video posted by Cockroachlurcher? They used a special ionic fluid so the fluid would not evaporate in the vacuum, and then operated the siphon. As long as the fluid stays coherent, air pressure is not required. It is a gravity effect.
“Water seeks its own level.” If there is a passage between two reserviors, then gravity will drive the water to even the level in both reservoirs.
This is the same reason why water spreads out flat in a reservoir rather than bunching up. Sure, external effects can pile water up (wind, currents, boat motors, etc), but in the absense of other factors, gravity spreads the water at equal potential, i.e. flat. This works with two containers in connection, and as long as the passage between the connections is sealed, it does not matter if the passage goes above the water level.
In the absence of air pressure, the only thing stopping vacuum-filled cavities opening up (and breaking the process) in the fluid-filled tube is the tensile strength of the fluid itself. Air pressure keeps the column of liquid together as a unit.
If the size of the “air bubble” in the tube is greater than a certain threshold there will be no flow. Certainly if the tube contains only air that will be the case. Even though the two reservoirs are at different levels and there is a connection between them with fluid (air) in the tube.
This suggests to me that the weight of fluid in each side of the tube must be important to start the flow.
I will say again that air pressure is a distraction. Air pressure keeps the water from boiling, but that’s not why a siphon works. I’ve suggested two experiments to try, you should try them before “theorizing” any more. The first experiment shows that the rate of flow is proportional to the head pressure (difference in height between supply and destination). The second shows that the weight of the column of water is irrelevant. Here’s another that proves the air pressure differential is irrelevant. Following that is a last experiment that proves that “fluid cohesion” is not necessarily required either.
I made a U-tube manometer (piece of clear tubing taped to a scrap board) with one end at the water surface in the “supply” bucket, the other end in the “receiving” bucket. Then I set up my shop fan blowing into the “receive” bucket. I adjusted the fan until the air pressure in the “receive” bucket was higher that in the “supply” bucket (as indicated by the manometer). I started a siphon and it worked just fine with the “discharge end” air pressure higher than the “supply end” air pressure.
This last experiment is really neat.
You will need a small thermometer for this one. Wrap an empty bucket with old towels or a blanket or something. Put a couple of trays of ice (or a couple of freezer packs) in the bucket. Tape a piece of tubing in the bucket so that the end is just above the ice level, the other end draping about 3’ below the bucket. Cover the bucket with anything (piece of cardboard? NOT an air tight cover) and wait about half an hour or so. After the wait record the room temperature indicated by the thermometer. Now suck on the end of the tube to start a siphon. Put the thermometer at the discharge end of the tubing. You will find that the temperature falls at the end of the tube. You are siphoning cold air.
So the fluid (or gas) has to be denser than the density of the surrounding atmosphere, there has to be a difference in head (height) between the supply and the destination, and the flow is proportional to the difference in head. Now go out to the garage and try these simple experiments before theorizing any more.
That’s accurate. Similarly, in the absense of air pressure, most liquids don’t stay liquids.
Air pressure differential can drive fluids to move through tubes. It is not the operational force for a siphon, though may be involved in starting the siphon. Something has to prime the tube. One way to do it is suck fluid into the tube, i.e. lower the air pressure at one end vs the other end, and let air pressure push fluid into the tube. But once the tube is primed, the driving force is gravity.
Yes, you have to get enough of a flow to start the siphon. Let me think on this one.
Yes, this is true, for the simple reason that if the fluid is less dense than the surrounding atmosphere, then buoyancy will move that fluid up out of the bucket and air down into the bucket. Ergo, gravity will not pull the “fluid” down, it will pull the surrounding air down and that will push the “fluid” up.
Wow, how is this so hard? Cecil had it right: except in the case of not-really-fluids with crazy tensile strength, you need both gravity and air pressure for a siphon. The physics are pretty simple: the pressure from gravity pulling the liquid in the long tube down MINUS the pressure from the atmosphere pushing on the low container is greater than the pressure from gravity pulling the liquid in the short tube down MINUS the pressure from the atmosphere pushing on the high container. Therefore liquid runs down the long tube, and (because of atmospheric pressure) is pushed up the short tube and over, while more liquid is pushed from the upper container into the short tube. Continue.
Clearly, hydrogen bonds or whatever in the liquid aren’t necessary, as gasoline is notable for not really having hydrogen bonds like water does, and yet gasoline is probably the second-most often siphoned liquid. As noted you can also put an air bubble somewhere in the siphon and everything still works, so liquid cohesion is clearly not a factor.
It seems kind of obvious to me that if air pressure isn’t necessary, siphon tubes wouldn’t need to be airtight. But you can’t siphon using a tube with a hole at the top.
nolaman – what was the actual air pressure difference between the upper and lower reservoirs? Note that air pressure increases as you go down, so the lower container already had a theoretically higher air pressure anyway. But the question is was the air pressure difference greater than the pressure difference from the weight of the two different columns of water? If it was, then yes, the fluid indeed should have been blown backwards through the siphon. Heck, that’s what a pump does. But I really, really, doubt that a fan can produce much of an air pressure difference, maybe a few hundredths of an inch of water or something, but nothing close to the pressure from the different levels of water.
And, nolaman, I’m not sure what the experiment in siphoning cold air proves – other than demonstrating that liquid cohesion isn’t important. There’s still air pressure and gravity doing their things in the same way as if you were siphoning water. Though actually, I’d be really wary of using cold air in siphoning experiments, since the net forces are going to be so small that lots of things could screw up the results; with a nice strong breath to start you could probably get cold air coming uphill for a while just by momentum.
Honestly, if someone doesn’t understand what’s going on, I’m happy to try and get you through it – just explain where the reasoning here doesn’t make sense to you, and I’ll try and rephrase it to have it make more sense to you.
I’m comfortable with saying that both can affect the operation of a siphon. Under most circumstances, the air pressure differential between the inlet and outlet is sufficiently small to be overwhelmed by the gravity potential difference. Special cases, such as the experiment 3 case with the inlet bottle sealed, shows that a significant pressure differential can change things.
Think about this alternate set up: have two buckets with different water levels, and a sealed tube between them that runs underneath the buckets. The tube must be sealed to prevent the water from leaking out the bottom rather than flowing to the other bucket, but air pressure is not required to push the water from the higher bucket to the lower bucket. Gravity all by itself will do it.
Just like gravity is all it takes to keep the surface of the bucket flat, and not heaped up to one side.
Under most circumstances, air pressure pushing down on the surface of the water in the buckets is not a key element of what is driving the siphon. It is the fluid pressure differential based upon elevation.
Pressure within the tube is important. If you break the seal, you prevent operation. If the inlet opening is not sufficiently submerged, air will leak in, gravity will sort the more dense water from the less dense air and preferentially pull the water down and allow the air to flow. The reason why the outlet does not need to be submerged in fluid to work is that the gravity is already pulling the water down, and there is no room for the air to sneak past it. The water in the end of the tube functions as the second reservoir seal. Since the inlet is submerged, it does not require air pressure to push fluid into the inlet - the weight of the fluid in the first reservoir is sufficient.
Description: guy takes bucket of water with a tube in it, fills tube, pulls one end out and over to start siphon, emptying into a bucket below. Then slowly raises the exit end of the tube, letting it drain and fall to the bucket on the floor, not submerged. As the exit end raises up, water flow continues until 00:25, when the exit is [del]level with[/del] slightly above (due to momentum) the water level in the inlet bucket. Suddenly, flow stops, and starts to reverse. He then lowers the tube end again, and reestablishes flow in the original direction.
Are you telling me that the air pressure difference between the top of the first bucket and the end of the hose is substantially different at ~ 1 ft lateral offset and ~4 inches vertical displacement?
You are trying to use the wrong parts of physics. You should be thinking of this as an hydraulic system. Siphons will work just fine in a vacuum as long as the siphoned fluid doesn’t vaporize.
Take two buckets each half full of water. Start a siphon with a nice, long piece of tubing from one to the other. Now put them on a table or something so they are at the same height. Wait for the levels in the buckets to equalize. Support the tubing so the top of the siphon is, say, 2’ above the water level.
Where in the siphon is the pressure the lowest? At the top of the siphon.
Where is the pressure the highest? Right at the elevation of the water level (I’m excluding the portions of the siphon tube below the water line for simplicity).
Where in the “up” tube is the pressure equal to the pressure in the “down” tube? At all points of equal elevation.
Let’s lower one bucket. The water begins to flow toward that lowered bucket. Why? For that answer you have to go back to the top of the siphon.
On the “down” side just past the apex the pressure falls slightly lower than the pressure on the “up” side just before the apex. That’s because the column of water on the “down” side is vertically longer (NOT heavier, LONGER), which creates a lower pressure at the top of that column. So the water flows from the relatively higher pressure at the top of the “up” tube toward the relatively lower pressure at the top of the “down” tube. It is a differential in pressure caused by the differential in liquid head that makes the siphon work. I can’t emphasize strongly enough that it’s the pressure difference in the fluid, not a weight difference.
You can calculate the rate of flow through the tube using the physical properties of water, the physical characteristics of the tubing, and the differential head. I will emphasize here that the source of the energy to move the water is identically the differential head. The weights of the two columns are irrelevant.
Bingo. You said this in fewer words than I’d have taken, so kudos for that too, and saving me a bit of typing.
Irishman, the air pressure differential between the ends of the siphon is not significant – you’re right about that. But the air pressure is required.
Take your example of a tube between two buckets, connected to the bottom of the buckets. This is not a siphon, and will work with no air pressure: gravity does everything we need.
But if we grab that tube and raise the middle of it above the water level for either bucket, we now have a siphon.
If the liquid is one that doesn’t tend to cavitate, then we’re fine. But if it’s one that does (like water), as soon as the height over the water is enough to cause cavitation, it won’t work in a vacuum. That height is different for different kinds of liquids. I don’t remember the name for this “cohesiveniss” property; hopefully someone does. But I suspect that the height required for the siphon to fail in a vacuum would be lower for alcohol than for water, for example. And no doubt, for any liquid, there is some height where cavitation would occur.
Oh – also, the fact that a siphon does not work when the distance up exceeds about 30 feet is proof that air pressure is required. That’s when the weight of the water in the tube exceeds the air pressure.
This is why we can’t pump water out of a well from the top of the well, if the water level is more than about 30 feet deep. (For practical purposes, much less than that, since we want a decent flow.) For wells that are deep, the pump has to be near the water level.
Grravity makes the water move down the “down” part of the hose. But ambient air pressure is what makes it move up the “up” part of the hose.
These are easily testable hypotheses, btw. Just get 80 feet of garden hose and a way to raise the middle well over 30 feet, if you don’t believe me.
