Pipe flow sensor - non invasive?

Is this to be DIY "for interest"? seems cheap enough to just buy one otherwise ...

Nothing "non invasive" there :(

Invasive is easy - and the cheapest would be to stick a small commodity water meter in - these are very cheap and can be had with a magnetic coupling that can drive a front mounted pickup sensor.

No - I'm looking at a scenario for a mate where the pipe has very little "visibility" before it goes off on all its branches - ie pops up under the sink and then goes everywhere. No room to work to insert anything without destruction of a kitchen unit.

Ah sorry, a quick search for clamp-on liquid flow sensors seems to show plenty ... but they're in the 3 or 4 digit price range ...

Indeed! I'd like a non-invasive clip-on meter - if such a thing existed

- to assist with balancing central heating systems. If you could set the flow rate through each radiator to be proportional to its heat output, you could set a system up very quickly without needing to wait for the temperatures to stabilise.

Yep - those are the doppler effect ones - very "industrial" in their price range...

Getting very DIY now - but I was wondering in your case about 2 temp sensors (Dallas 1-wire jobbies) and a heating resistor at the flow source side.

Without the resistor power up, the sensors should read the same temperature - especially on copper pipe, say 6-12" apart.

Power the resistor up with no flow and the 1st sensor gets hotter with only conduction to move heat to the second sensor.

With water flow, the heat should be removed from the 1st sensor. Or maybe the resistor should be on the 2nd sensor?

This is the sort of thing it would be really fun to do an experiment with - bit of copper pipe on a hose on a tap (hot and cold), couple of temp sensors (multimeter type) and a resistor with a variable PSU.

Simple enough, could probably have a rough and ready stab at a simple physics calc for the system assuming all the heat from the resistor is transferred into the moving water - bit of SHC vs flow calculation should give a rough idea of the possible sensitivity.

Readings would be very dependant on thermal coupling to the pipe. Casting a blob of silicone packed with al or cu powder could help.

NT

Invasive, but look at car engine MAF sensors - manifold air flow sensors. A +ve, a -ve and an output wire reflecting the flow.

This of any use?

Yes - that does suggest listening to the "hiss" has merit.

With copper plumbing, any noise anywhere will get all around the pipes - but as I am interested *any* flow at the source (mains) so that should not be a problem.

Interesting:

I would think a slight leak would only involve movement of water in the pipe of less than half a centimetre per second. As the pipe is metal, any cooling would be conducted over a long length of the pipe and so too small to be detected. I don't think there would be any audible hiss either.

See here for descriptions of various ultrasonic flowmeters:

Since they do invasive ones designed for water, why bother?

I really have to break out my storage box of electronics and start experimenting...