Siphons

So you want to use a siphon for something… unlucky…

Why Siphon

I can’t deny that siphons can bring alot of advantages to a project or something due to their lack of moving parts, but damn, there’s certainly no research out there on them. The main challenge is getting a siphon to trigger reliably at very slow rates of increase in height in them (i.e. the reservoir its attached to is filling up very slowly). I wont go into how they work here since it seems that nobody really knows, and to understand what’s happening I recommend just watching no less than 3 conflicting videos on them.

How Long?

General rule of thumb is to ensure that the end of the siphon tube is some distance below the bottom of the intake of the siphon. If they’re level then things will be quite intermittent.

As an example. The siphon I was testing had the output end of the tube 25mm below the intake and this caused absolutely no problems. I assume that as long as the body of water within the tube during siphoning perfectly fills the tube (because your tube is small and flowrate is good) then you might expect to get away with less distance, perhaps 5mm as a suggestion although this is untested.

Materials

I discovered this to be a key part of attaining reliable triggering of the siphon at low flow rates. Not so much the material, but its surface.

Hydrophobic coatings are another mystery I struggled to find any quantitative data on. The gist is that if a material is hydrophobic then water does not appear to stick or wet to the surface. Some materials are hydrophobic like PTFE, whilst more commonly there are many coating available like bike wax. For thought purposes, you might think of a sheet of paper being the complete opposite of hydrophobic since water is immediately wets into the paper and spreads out. The main metric I was using was the contact angle of the drop with the surface.

The greater the angle of contact, the more hydrophobic the surface is, and the less sticky water is to it.

Surface finish seemed of little importance honestly as it’s the chemical composition of the material that determines how strongly the water molecules are attracted to it that impacts the stickyness the most.

I tested a few materials polished surfaces for comparisons sake and settled on wax coated ABS since I 3D printed my siphon, and found this to perform about the same as PTFE (which is naturally hydrophobic).

The reason for wanting to make your siphon tube surface hydrophobic is that it reduces the tendency for the water surface to creep up the sides of the tube. If water easily wets to the surfaces, then some water remains within the tube after a single siphoning event, and this can cause false triggering of a second siphoning at a lower height of water in your reservoir (or not at all!!).

This creep effect also acts as a resistance, and you might expect the height of water within the siphoning tube, relative to the reservoir water height, to move in jumps as it repeatedly overcomes the water tension and jumps up to reattach the water surface to another point in the tube.

This yet again has another issue in that the water tension will tend to suck the water surface over the lower edge of the bend in the siphon tube, causing the siphon to just dribble instead of causing a proper seal within the downwards half of the siphoning tube, and then siphoning and emptying the reservoir properly.

All the above issues of siphoning were solved by coating the internals of the siphoning tube with canola wax (4mm diameter siphon) and I’m sure that if you just machined it out of PTFE you’d have the same effect.

In my application I was able to get it to trigger whilst collecting simulated rainfall at a rate of 0.1mm/hr. I know this is arbitrary but take my word that this is an insanely low rate (and it was repeatable!!!).

Sorry this is brief but I’m not confident on the viability of presenting the exact numbers from tests here…