Low tech telecommunications (water)

[Vir27]

I'm a historian. I've struck on a curiosity about telecommunication technology in pre-modern times. I've read about a number of methods, and I'd like to ask some physics questions that got beyond my ken. This information is for my curiosity and creativity, not for any rigorous study, so I may be mentioning Wikipedia.

Thank you for humoring me, and I hope I'm not asking too many questions :) I've now separated them into multiple threads by subject; this one ought to be about water. To conform with PF custom, I'll be adding the other questions to the threads after the first ones have been answered. Let anyone feel free to answer only those questions he prefers to.

Water

An idea I had myself which may or may not be sensible is burying a small, very long line pipe, and filling it with water. Looking into this, I read about a British "hydraulic telegraph," but its not quite the same since it's trying to relay more information with devices that can affect the water pressure more precisely than interests me. Obviously leaks, earthquakes, freezing, expense, and thirsty creatures are possible problems of this approach, but in theory...

Q1. Suppose, as a signal, the sentry at one end of the pipeline pulls his plug. If the pipe was full, won't the fact of the drainage be communicated down the whole line fairly quickly?

 

[mfb]

"Quickly" depends on the pipe. You have a lot of water that has to start moving and you have friction all along the line. Both will slow down the process. The first one has an even worse side-effect: you can get oscillations - water levels that go up and down even without further human influence.

Let's consider an example: 5km pipe length, 10cmx10cm square for nice numbers, 1m height difference between the sides. That gives ~100N accelerating force for 50 tons of water, or an acceleration of 1/500 m/s^2 neglecting friction. Your data transmission would need 10 seconds to change the water level by 10cm. I guess friction will slow that down significantly. And then you have to wait a bit more until the water levels are equal enough to start the next bit.

 

[russ_watters]

Google "water hammer". It's a shock wave caused by a sudden change in flow (like a valve slamming shut), that travels through the water at the speed of sound (1,500m/s).

Or, if the pipe is full, but the water isn't flowing, you could literally hit it with a hammer...

 

[jbriggs444]

Q1. Suppose, as a signal, the sentry at one end of the pipeline pulls his plug. If the pipe was full, won't the fact of the drainage be communicated down the whole line fairly quickly?

With considerable refinement, this technique is called a "hydrophone".

 

[DaveC426913]

No need to drain any water. It can be a closed, non-lossy system.

As Russ hints at, water is essentially incompressible. A shock provided at one end (say by a hammer on a membrane) will travel to the other at almost a mile per second - five times faster than through air. You should, in theory, be able to set up a sort of telegraph system of ins-and-outs. Data rate would be determined by practical matters, such as the rigidity of the pipe casing. (The signal would attenuate and dissipate.)

 

[berkeman]

One hint to add to Dave's post -- you will want to "terminate" the ends of the pipe in the "characteristic impedance" that the water wave experiences in its travels through the pipe. This is to prevent reflections of the wave from the ends of the pipe, which will cause interference in reception of the waveform at both ends of the pipe.

You can research the terms that I've used in my previous paragraph to learn more about transmission lines and terminating them in their characteristic impedance.

 

[Vir27]

I think my comment about the hydraulic telegraph may have led us to believe the idea involves sending messages via the drainage. I was thinking just, "The water is going down. That's the signal." No second or nuanced communication: nothing more informative per use than a flare.

But despite my possibly confusing the issue, I understand from mfb's post that that method would communicate at maybe .25km/s (significantly more than a frictionless 10 seconds for the 5km). Whereas just striking the full pipe would transmit at ~1mi/s. Well, that's an answer I'd never have worked out on my own! Fascinating.

berkeman, if I have applied what I googled about characteristic impedance, we want the pipe to end in a shape which is suited to stopping a water wave or else we'll get reflections of the wave (coming back?). Clearly that'd matter for distinguishing the data of several transmissions in close succession (before friction ate the water's movement and calmed it back down), but as just a one-off signal, maybe not. Another thing I think I picked up while reading what characteristic impedance is is that it's length-independent for a uniform line. But practically, I think that might not be the case for this little scheme of mine because they're burying a pipe over a long distance and maybe if they bury it at the same depth that whole time it's actually curving a little bit because of the curvature of the Earth? So maybe that's not a uniform line. If the curvature of the Earth does come into it, they may not have the surveying technology (or math) to adjust. So they would be sort of stuck at the flare-level signal--which is fine, since it's the extent of my curiosity's ambition.

Oh, if the curvature of the Earth does come into it, then will this idea of striking the pipe still work, or will the curve break the wave?

 

[nasu]

The sound will go even faster through the pipe itself (unless is made from lead). Why do you need the water? Speed of longitudinal sound in steel is around 6000 m/s and in copper over 4500 m/s.

 

[Vir27]

Well, my original idea used the water without any percussion because I wanted something that would communicate over long distance (i.e., water, in my imagination, will keep falling into place and leveling off all along miles of a full pipe as space opens in one end of the pipe), but I had the impression from the previous posters that the water hammer effect was important to their even faster suggestion.But if sound carries for significant distance in just a pipe, maybe we could discuss it in my sound thread? That's the one that I find most interesting. I'm very much piqued to find out how relays of acoustic mirrors might be made to send a signal. Pipelines with a relay pipe-ringers is kind of fun, too!

 

[parkland]

I'll just throw this in the hat, not being a super smart person, just pointing out what I know from experience.

In the oilfield, down hole tools use mud pulses to send information to surface decoding equipment.
The signal rate can sometimes be around 1 pulse / second. Depths can be thousands of meters.
Because drilling is done with mud, not water, issues arise which are not part of your question, such as aeration,
noise from the pipe rotating, etc.

Looking at using purely water, and pressure spikes, for communication, the issue I see is the pipe, or containment.
The longer it is, and more volume, the slower the reaction.
Over a long distance, the material of the pipe, I think would influence the pressure more than the properties of the water itself.

Imagining a system made from primitive materials sounds like a very unreliable system. I could be way off here, but I am imagining a trough more than a pipe, something where even wind direction could change the flow of water.

 

[Vir27]

Thanks for the input, parkland. I think your experience was quite valuable. I wouldn't have known you can rapidly pulse through mud at thousands of meters distance. Very possibly you're right that pipe technology constraints would limit this one before the water physics came into it.

 

[DaveC426913]

The sound will go even faster through the pipe itself (unless is made from lead). Why do you need the water? Speed of longitudinal sound in steel is around 6000 m/s and in copper over 4500 m/s.

True. But a pipe made of metal will have significant costs related to the length. The OP is talking about "curvature of the Earth" distances, so...

Oh, if the curvature of the Earth does come into it, then will this idea of striking the pipe still work, or will the curve break the wave?

CotE will have no effect. It'll work anyway.

 

[Vir27]

Ok, thanks. I don't know how far it might take for that to matter (though I hear you that it won't) is the only I brought it up. I've been thinking for at least tens of miles, but since there is no specific purpose I have no specific length in mind. As far as the ancient costliness of pipe going dozens of miles, oh yes, quite true. We're assuming these people are highly motivated like ancient people pulling off all their costliest engineering stunts ;)

I think we answered all the questions I had thought of on this topic, so thank you all again.