> OK, so how about this?  A while ago we were discussing a non-invasive fluid
> flow measuring technique here on the PICLIST.  The idea was that ultrasonic
> sound was injected into a pipe or hose, and upline and downline sensors
> measured the intensity of that signal a short distance from the source.  By
> comparing the relative intensities from the two pickup sensors, the flow
> direction and speed can be accurately determined since sound transfer will
> improve in the direction of the flow, and decrease in the opposite
> direction.


I would think accurate correlation between signal strength and fluid
velocity would be difficult.  Would it not be better to measure travel
time through the liquid?  The difference between upstream and downstream
travel times should be easily detectable and should easily correlate with
the speed of the liquid.